Table of contents

RoadTest: Tektronix PA1000 Power Analyzer (ASEAN / ANZ ONLY)

Author: Gough Lui

Creation date: 27 Jul 2014

Evaluation Type: Independent Products

Did you receive all parts the manufacturer stated would be included in the package?: True

What other parts do you consider comparable to this product?: None that I had direct hands-on experience with, however, a quick survey of the market was performed.

What were the biggest problems encountered?: A variety of firmware and software problems and enhancement requests which I have contacted Tektronix about and they have been committed to fix with one release delivered during the RoadTest review process, and the next release due mid-August.

Detailed Review:

Tektronix PA1000 Power Analyzer Review

by Gough Lui

Preface

I would like to apologize to the community, first and foremost, for taking this long to deliver my review of this product. This was due to a rather extraordinary circumstance which involved the need to contact Tektronix on several occasions in regards to a range of issues and feature requests, some of which is still currently "in the works". I had debated with myself numerous times whether to release the review, or to hold off until Tektronix has addressed all of the issues. Further to this, was the need to buy, build and test equipment which would help me make the most of the PA1000. Time continued to tick away, and seeing that it is two months overdue, I could not reasonably hold it off for any longer. I thank you all for your patience.

As a result, this is by no means the final story in regards to the PA1000, and will be appended to as further firmware and software updates are released (time permitting). I have been informed by Tektronix that they do have plans for "many further enhancements due this year ... to improve and exceed expectations, not just with bug fixes but with significant additions to capability."

It is with many thanks that I acknowledge the support of element14 and Tektronix in selecting me as one of the lucky few RoadTesters. I think that after reading this RoadTest, you will appreciate just how hard this thing has been "driven" and I hope it proves to be useful to fellow PA1000 users and purchasers alike. If you do have any issues with the product, please contact Tektronix as they have shown serious interest in feedback and feature requests. As usual, if you feel this review is helpful, please rate it or give it a like! If you have any questions, you can leave a comment and I'll try my best to answer it. For the latest of what I'm up to, feel free to visit my personal blog at http://goughlui.com

Introduction

Tektronix have been a trusted name in test, measurement and monitoring for over 65 years. Starting with their roots in supplying oscilloscopes, through development and acquisitions, their portfolio of test equipment is growing even more comprehensive. I think many engineers would easily recognize the Tektronix "teal" coloured bodies, build quality and reliability. They would also undoubtedly also recognize the price ...

Notably, Tektronix did not have a presence in the power analyzer market until 18th March 2013 when they announced the acquisition of Voltech's power analyzer design, intellectual property and patents.

This has directly contributed to Tektronix's power analyzer line-up, which currently consists of the PA1000 and the PA4000. The PA1000 is a single phase power analyzer with a voltage rating of 600V RMS, current rating of 20A RMS and 0.05% reading + 0.05% range basic accuracy. It features GPIB, USB and Ethernet connectivity as standard. Provides information up to 50th harmonic. It is backed by a 5-year warranty.

The PA4000 is a more advanced, four channel power analyzer with a voltage rating of 1000V RMS, current rating of 30A RMS and 0.01% reading + 0.04% range basic accuracy. It features USB, Ethernet and RS-232 interfaces as standard, with GPIB optional. Provides information up to 100th harmonic. It is backed by a 3-year warranty.

Features common to both power analyzers include:

- Bandwidth of up to 1Mhz

- Crest factor tolerance up to 10

- Fast auto-ranging for voltage and current

- 1A current shunt to improve resolution for standby current measurements

- USB host for data logging to a USB thumb drive

- PWRVIEW software, which allows for compliance testing to IEC62301 Ed. 2. available in 32 and 64-bit editions of Windows 7 and above

This review will be focusing on the PA1000, which is currently available from element14 for AU$3281 (on promotion). Unfortunately, for users looking to characterize three-phase loads, the PA4000 is not presently in the catalog of element14.

The Contenders

The market in power analysis can be quite confusing for the newcomer. The term power analyzer can have several different meanings depending on the context. There are specifically products which call themselves power analyzers, but are more like power meters which offer values for voltages, current and power factor but little information about harmonics or waveform. There are other products which are better referred to as power quality analyzers which are designed to pinpoint problems with power, such as sags, swells, dips, harmonic noise, flicker, etc.

Some of these products are designed for field usage, and are designed portable with battery packs and may implement current clamps and voltage clips as their sensing methodology, but are targeted at power consumptions of large plant and facilities where small currents and standby usage is not within their abilities and precision is often sacrificed for portability. Others are targeted at analysis of motors, motor drives and inverters specifically.

So when someone says power analyzer, the correct response is really, what sort of power analyzer?

This particular power analyzer is more of a mixture of a power meter, with harmonic and waveform analysis, which isn't (currently) targeted at logging or identifying power quality problems, or with energy efficiency of large plants or facilities. Instead, it is more concerned about the load profile and behaviour of an individual device, such as one that would be hooked up to a single general purpose outlet (GPO) or circuit. It has higher precision than most portable devices, and is more suited to an R&D role in designing appliances which connect to the mains. This might be in order to measure basic parameters (current, voltage, power factor) or standby power in order to understand efficiency, and ensure compliance to minimum energy performance standards and standards surrounding harmonic distortion, for example. With the use of external transducers, it is possible to analyze the load profile for larger currents, of a single phase, or even to analyze the efficiency and behaviour of LED current drivers and fluorescent ballasts.

One difficulty I came across is the problem of dissimilar specifications and overloaded datasheets. I've tried to choose some basic metrics common to all, to make comparison easier, however this may not be sufficient if your needs are very specific. I can't guarantee that I haven't made any mistakes, as some datasheets state different values for the same parameter, to different precisions, within the same document.

As a result, it took a little bit of sifting to find products of a comparable nature, with the other difficulty being that prices are difficult to come by, and as a result, the below products are not comparable in price. While I have done my best efforts, do note that this cannot cover all the products in the market - you are advised to do your own research as your needs might differ, and new products continually come on to the market.

In all, it appears on paper that the PA1000 has a unique position in this market. The majority lag behind on features such as bandwidth, connectivity, user display, measurement modes/shunts and software abilities. Many units on the market relegate connectivity, features such as harmonic analysis and software as extra-cost options. A large number of products also focus more on motor testing applications, which make them less than ideal choices for R&D work or compliance testing for standby power measurements or even fluorescent ballast testing. Also, it seems the PA1000 has an edge when it comes to basic accuracy and calibration cycle in some cases. Compared to the market contenders, it is a solid choice on specifications alone, provided this is the sort of power analyzer you're after!

Conversely, it also seems to show the PA1000 as less than ideal if your use case revolves around work with motors and inverter drives, as some of the software functionality and multiple-channel features of other analyzers are more suitable for analyzing the input and outputs simultaneously and for producing specific torque, current, RPM curves. It can be possible in some cases to use multiple PA1000s to achieve this, however.

Unfortunately, without firsthand experience with any of the competitors, I cannot present a side-by-side comparison, however, I can at least provide information on how the PA1000 behaves.

Unboxing the Device

The device came in a thick-walled box which was much bigger than I expected. It's even larger than a citrus fruit box, and had substantial weight. Inside, with foam stand-offs, the device was packaged in an static-dissipative bag with generous clearances from all sides. The unit itself is several kilograms and definitely exudes quality. It's built like a tank!

Removing the protective film from the LCD, the PA1000's front panel looks like this:

On the front are shrouded banana sockets for voltage measurement, as well as current measurement in two shunt ranges (1A and 20A). There is also a connection for an external current clamp transducer with up to 1.25v peak signal amplitude. Conventional transducers can also be accommodated through the regular banana plug sockets, through the use of the scale feature. The inputs are floating, which makes them suitable for analyzing the output from inverters, motor drives, or even solar panels (PV). They are rated for 600Vrms to earth (mains, and the lug on the rear of the unit), with a CAT II safety rating, which allows it to be used on customer circuits downstream of a switchboard.

In the centre, is a colour graphics LCD, flanked on the right side by four soft-buttons whose functionality is determined by icons on the right hand side of the screen (normally, page down, page up, top of list, bottom of list, decline/cancel, OK/accept, up a menu level, etc). On the right hand side of the front panel is a keyboard which allows for alphanumeric text entry, as well as invoking features such as logging, menu, graphs, reading hold, local-control override and on-line help. The keys are hard plastic buttons (likely to be more resilient than rubber keys), which have short travel and a positive tactile action.

Below this is a USB A host port, which can accommodate USB mass storage flash drives (USB Bulk-only-Transport, 512 byte sector, FAT/FAT32) for direct data logging with a current limit of 250mA (according to the manual, not tested). In the bottom corner is a hardware power switch, which actions with a satisfying click.

The rear of the unit looks like this:

Power is supplied from the left side, through a fused IEC connector equipped with a time-delay 250v 1A M205 fuse. The voltage and frequency accepted are "universal" 100-240v AC at 50/60Hz. Above that is an earthing stud, which should be used to earth the article under test for safety purposes. There is another fuse, which is situated near the ventilation slots, which protects the 1A shunt. It is a 3AG size fuse, with a fast-blow 600v 1A rating. Unfortunately, no spares are supplied with the PA1000, so it's probably a good idea to obtain some as soon as possible.

There are ventilation slots lining the top (along with the underside of the unit, not pictured). Connectivity for GPIB (IEEE 488), USB-TMC and SCPI over Ethernet is provided as standard.

From the side, the unit clearly possesses the Tektronix teal green body, that is their trademark. The unit features very substantial rubber feet, which elevate the unit from the bench about 1cm to allow the bottom-facing vents to be unobstructed.

Included is the PA-LEADSET, which consists of four thick connection cables, terminated in shrouded banana plugs along with a banana plug to alligator clip attachment for direct attachment. The set consists of one red, two black and one blue lead. The cables are marked with the MC logo and are rated to 1000v 32A CAT IV. The lead length is a generous 2m which allows for more convenient connection to the load.

Also included in a CD including documentation, this one stated as Version 1.0. I would highly recommend visiting Tektronix's website [http://www.tek.com/power-analyzer/pa1000] to download the latest versions of all documentation, software and firmware updates. Also included is a 2m black USB cable.

A regular IEC cable is included, although due to the region it was ordered for, it was fitted with an American 3-pin plug. Australian users should substitute a suitable Australian IEC cable - due to the insulation rating on this cable, it's not a great idea to just rewire the plug or use an adapter. They're so common, you probably have a few lying around that you can use.

Of note is that no Ethernet lead is supplied, so if you wish to use Ethernet, you must supply your own lead. This also applies to GPIB, where it is uncommon to see GPIB leads included with instruments.

Finally, there is also the calibration certificate, a packing list, a few regulatory notices and a Tektronix Rewards pamphlet encouraging you to register your product.

However, this isn't all that you might need to use the Power Analyzer. If you're interested in testing loads which plug into a general purpose outlet (GPO) for standby current, or for power consumption, the most convenient and safest alternative is to use a Tektronix Breakout Box. Presently, three variants are offered - BB1000-UK, BB1000-EU and BB1000-NA, which are fitted with the UK, Europe and US sockets respectively. Notably absent is the Australian configuration - one way around that is to utilize a UK break-out and a conversion plug similar to those used for overseas holidays.

Rather fortunately, a BB1000-UK was shipped with my PA1000, making the process of getting started much easier.

The breakout box is a grey plastic box. The front of it features the UK plug recepticle, along with shrouded banana connectors which are used to directly connect the break-out to the PA1000. There are two configurations for the Voltage measurement, one which places the Voltmeter before the Ammeter, and the other which places the Voltmeter after the Ammeter. This distinction is important for measurement of standby currents, where the current through the voltmeter could contribute a significant source of error.

One way around the UK plug issue is to use a travel adapter, like this one (not included), which allows you to connect Australian appliances to the UK plug. You will need to substitute your own Australian IEC lead, or have a "reverse" travel adapter to connect the UK IEC lead to Australian mains.

The top of the box features a single IEC connector which provides the power input to the box.

The bottom of the breakout features safety and regulatory information.

As this was the UK break-out box, it was supplied with a UK IEC lead with a fused plug. The fuse internally is rated at 10A. As the UK runs on the same voltages as Australia, it is safe to use this lead through an adapter, or rewire it if you possess the adequate skills to do so.

Also supplied is a stapled instruction manual which details how to connect the break-out to the power analyzer, along with a packing list and regulatory information.

Also worthy of mention is the availability of the BALLAST-CT accessory which allows for the connection of fluorescent lamp ballasts to the PA1000 for measurement of the ballast output, and the CL200 and CL1200 current clamp transducers which extends the measurement range of current to 200A and 1200A respectively.

Heritage

Based on appearance, and some research, it became clear that the PA1000 owes its heritage to the Voltech PM1000+. It appears to be a refreshed version of the PM1000+, eschewing the older RS-232 interface and parallel printer port interfaces entirely, while adding on a front USB port for data logging, Ethernet interface and an additional 1A current shunt. The majority of the other specifications are fairly similar, down to the wording of the manuals, physical enclosure design, graphical elements of the user interface, operating modes and keyboard layout.

Teardown

This is by far, one of the most expensive pieces of test equipment I have in my possession. Am I going to tear it down? Absolutely - because I know you guys enjoy ogling the insides! I know I do!

How does one get one of these beasts apart without breaking it, and without access to a service manual? Well, a little bit of luck and experience goes a long way. I do not recommend anyone attempt this for reasons of safety, the fact it voids your warranty and is likely to achieve little or nothing. I've done it so you don't have to.

If you insist, the most important first step is to disconnect all connections to the unit. Then start by removing the four corner screws on the rear plastic plate, and then removing the plate. This reveals the metal inner chassis - and explains why this thing is so heavy!

The next secret is just to remove the screws on the bottom of the unit. Once you do that, the teal green "shell" slides away from the front revealing a metal "box". Then you can remove the screws (some of the heads on mine were a little stripped!) on the top panel to remove the metal plate, revealing the juicy internals.

The internals are quite interesting, with an assortment of major chips and components:

- A filtered IEC socket has been used.

- National Instruments 1320LQL controller (U17) for GPIB

- Maxim MAX3420E USB Peripheral controller (U4) in charge of the USB

- Microchip PIC18F97J60 (U60) microcontroller fed by oscillator (X4)

- Xilinx XC9572XL (U9) CPLD fed by 10.440Mhz(?) oscillator (X3)

- DS1302Z Real Time Clock (U11), fed with a 32.768k oscillator (X1) with a Panasonic Vanadium Lithium Battery (BT1) soldered to the board

- Texas Instruments TMS320 DSP (U35) which forms the heart of the unit, fed by 48Mhz oscillator (X2).

- EPSON S1D13A05F00A1 LCD Display Controller (U42)

- Cypress CY7C1041DV33 4Mbit SRAM

- XP Power CU20-10 5v DC 4.4A switching power supply unit, with all quality Japanese electrolytics from Nippon Chemi-Con and internally fused at T2A/250v (unlikely to blow, as preceeded by a T1A/250v fuse in the IEC connector).

- Si8442BB Silicon Labs Quad Channel 2.5kV Isolator (near top silver can)

- STMicroelectronics 24M01RP Serial EEPROM (U44)

- Microchip 25LC256I Serial EEPROM (U61)

- FTDI HN976.00 YNC2 (? - didn't get a good enough shot of this mystery IC, near the front USB port)

There is also an array of undocumented connectors and jumpers which are likely for reset/service/reprogramming of the microcontrollers, CPLD, and RTC components.

There is also an internal fuse, which appears to protect the 20A shunt. I didn't investigate the size and rating of the fuse, but it appears to be a 5AG HRC 20A 600V fuse (to be confirmed). The 20A shunt appears to be the piece of folded copper on the right.

Speaking of shunts, the 1A shunt is on a PCB mounted to the side wall of the enclosure, suggesting it was an "addition" to the original design. The product literature of the PA4000 extols many benefits to their patented "spiral shunt" technology, of which this appears to be an example of. Unfortunately the 20A shunt doesn't appear to be a "spiral" shunt, which is probably why this model doesn't advertise this feature.

Just in case anyone else does this, I should just warn you not to touch the shunts. Touching them may affect their accuracy, and may represent a burn hazard if still hot from recent use.

Not wanting to harm the PA1000, I wasn't going to proceed any further in the teardown. That still leaves some mysteries underneath the cans, and on the other side of the PCB, but I'm happy to leave that unexplored for now. So there we are, the internals of the PA1000.

But why stop there? Lets take a look inside the BB1000-UK as well!

Pretty empty eh? That's because that's all it is - a simple way to connect your load to the power analyzer. There are no active components inside the breakout, just a connection of connectors, spade connectors and 600V rated wire.

The order of the wiring is quite neat and definitely correct. Spend a few minutes, and the reason behind the way it has been wired will become obvious.

Firmware and Software Update

In my case, the instrument was shipped with initial release firmware, version 1.000.000. The probability is high that new software and firmware are available, so it pays to check on Tektronix's website for them and update them accordingly. Updates normally bring improvements in stability, bug fixes and feature additions. I would recommend new users to perform these updates immediately, and periodically where new software is released.

The latest version of PWRVIEW (which is v1.1.4.110 at time of initial writing but since superseded), available free from Tektronix's webpage, installed just fine on my Windows 7 32-bit computer. It co-existed with my Keithley Model 2110's NI-VISA installation, as it seems the PA1000 also uses the NI-VISA runtime. The installation package is also available for 64-bit machines and is qualified for use with Windows 7 and 8.1 at this time. It's important to note that the software will also download and install NI-VISA runtime if not already installed, which is not packaged in with the installer, so the computer you are installing PWRVIEW on will require an internet connection. Furthermore, if NI-VISA is not already installed, the system must be restarted to complete the installation.

It is recommended to install PWRVIEW first before installing the Power Analyzer Firmware Update Utility to ensure the VISA communications layer is correctly installed. Please download the utility and the appropriate bin file which contains the firmware image for your power analyzer.

Once the utility is installed, it can be started up and the communications layer tested. The PA1000 must be connected by USB for the firmware update process. Once the communications has been verified to be functional, the tool can be pointed to the appropriate .bin file (at the time of writing, the latest version is PA1000Firmware_1_001_006.bin) and the flash process begun.

Be patient while the firmware update is in progress. During the process, there will be several stages where the process appears to stall, but it is progressing. Any interruption will ruin your instrument and possibly require return to the manufacturer to restore to functional order. The whole process takes about 3-4 minutes.

The Release Notes attached to the firmware indicates the addition of a command :SYST:ONTIME? to obtain the time in seconds since the instrument was booted, a change to the FRF query to return the number of parameters and results to be returned with the original FRF command changed to FRF_LEGACY, and a fix to Ethernet IP addresses not being retained through power cycles.

Further details on the operation of the PWRVIEW software will be detailed in a later part. The update has also updated the help which now reflects the correct URL for Tektronix.

This review was conducted using a mix of Firmware version 1.000.000 and 1.001.006 and with PWRVIEW v1.1.4.110. Users should note that a new version of PWRVIEW has been released just before the publication of this review, and another firmware and software release is slated for mid-August with more releases promised before the end of the year. As a result, many of the issues identified may have been fixed by the time you come to use it.

Standalone Use

The most obvious mode of usage for a benchtop instrument is directly using the product via the front panel. Tektronix have also have provided for download a set of whitepapers which cover basic measurement theory, as a primer for beginners, as well as a manual which details some operations and the supported SCPI command set. Reading the manual is highly advised to get the most out of your instrument.

Standalone usage revolves around several forms of display. The most basic is a list of values which have been selected for measurement, which can be scrolled up and down. There is a choice to select from one of two "zoom" levels - one of which provides large text but at the expense of only being able to show four values simultaneously, and the other being much smaller, but allows you to fit a total of up to fourteen values simultaneously.

There are also various graph modes which can be engaged to display a voltage and current waveform graph based, a voltage/current harmonic bar-graph, as well as mode-specific graphs such as those which show integrated power consumption over time.

The setting of the relevant parameters are done through the menu interface, which utilizes the four "soft-keys" next to the LCD which perform context-sensitive actions. Often, required alphanumeric values are keyed in directly using the number pad. Help is available from the LCD as well, which can be of benefit.

The five main pre-programmed modes of operation are:

• Normal Mode - The general mode used for general purpose measurements, a default mode from which you can customize and choose relevant measurement parameters.
• Inrush Current Mode - Mode dedicated for inrush current measurement. Features a reset button to reset the readings, and holds the peak current readings.
• Integration Mode - Mode dedicated for accumulating energy consumption over a period of time. Best shown on a graph, although a bug was determined with the graphing of certain longer time values resulting in an incorrect scale display (which Tektronix will address in the next firmware release).
• Ballast Mode - Mode dedicated for working with high frequency waveforms from ballast with special synchronization requirements.
• Standby Mode - Used to measure low currents directly with disabled "zero" blanking, a direct measurement technique not suitable for IEC compliance testing but good for a quick check.

The pre-programmed modes of operation allow for fast recall of measurement parameters and settings most appropriate for these types of measurements. It also limits the types of measurements available, as detailed in the manual. The pre-programmed modes of operation can be supplemented by manual selection of measurement parameters, and the configuration can be saved with a name in one of five locations.

The instrument menus are used to configure the measurements, inputs, interfaces, and more. The structure of the menu is given in a PDF attachment as element14's site seems to have issues with bullet points ...

While a menu-driven structure is fairly normal, the use of several layers of sub-menus can make navigation a bit difficult. In fact, having mapped out the options, it actually now makes me more comfortable with the menus. The provision of PWRVIEW which can perform configuration from a PC makes this less annoying than it would otherwise be.

The use of manually keyed-in values for parameters like time, and graph scale values can get tedious as it requires the manual deletion of the existing value and keying in the precise desired value. It would be nice to allow for some some unused buttons for scale increment/decrement which would save this and the ability for the PWRVIEW software to transfer the system time to the PA1000, which I have informed Tektronix about.

Application: Inverter Power Quality Comparison

Inverters are common appliances amongst those who need to run mains powered appliances from a battery, and the power quality obtained from an inverter can vary greatly. Inverters, such as those with "modified sine wave" outputs, often put out harsh waveforms with fast rise/fall times which can be challenging to measure and visualize, and others while claiming pure-sine wave can suffer from insufficient or incorrect filtering leading to harmonic distortion.

As a result, I was interested in what my regular inverters were like, so I grabbed one modified sine wave and one pure sine wave inverter and decided to see what the PA1000 thought of them.

The waveform was obviously very distorted, however, it did faithfully depict the "square" nature of it.

As it turns out, the cheap modified sine wave inverter is a bit of a mystery - it's a bit high in voltage and low in frequency but it's as dirty as can be, whereas the true-sine wave inverter fared much better than I expected. It's actually got a role to play further on ...

Front USB Port Logging

One of the more exciting features alluded to by the front panel layout is the possibility to log data to a USB flash drive without the use of a host computer. As mentioned earlier, the front USB port is compatible with USB flash drives featuring Bulk only Transport, with a 512 byte sector size and a FAT/FAT32 filesystem, which is the majority of products of 32Gb and below.

Before using the logging feature, it is advisable to set the date and time on your instrument to ensure that the logs contain valid data and are named sensibly.

To utilize the feature, you must first configure the instrument to your needs such that all the parameters you wish to be logged are displayed on the screen. Only parameters visible on the screen will be logged!

Then, you can insert your USB key - when it is correctly detected, USB RDY will appear transiently in place of the mode indicator (normally Normal, Inrush, Integrator, Standby). Logging can be initiated by pressing the 1 key. This will result in the mode indicator flashing once per second to indicate logging is in progress. Pressing the 1 key will return the mode indicator to a solid indication, ending the logging run.

It is vitally important that you do not remove the USB key during logging. If you do so, you will end up with a 0-byte csv file! Also, do not be too swift in removing the key after pressing 1 to stop logging, otherwise you will run into the same trouble. I have been able to log files with over 142,000 rows, which is about 10Mb, so it is possible to do some long term logging with it although with the risk that you could end up with nothing should the key be removed during logging or taken out too soon after stopping the log.

Files are stored as .csv (comma separated values) file with the folder/file convention of PA1000\<dd-mm-yyyy>\<hh-mm-ss>.csv, where the time represents the time the logging was initiated. The log file format consists of a short header, which identifies the instrument, date and time, followed by the displayed readings on the screen, with an index number. No timestamping is performed, and it's not explicit, however it appears that two readings are recorded per second. It would have been nice to get absolute time stamps, which would allow for more confidence in the time aspect, especially in long logs or when features such as averaging and auto-zero are in use (which can cause data interruptions or a change in data rate). I have requested this as a feature, which may be coming in future releases.

The header itself looks like this:

Unfortunately, no other logging appears to be possible, aside from this most basic form. It would have been also nice, seeing the nice graphical display, to take screen-shots of integration, harmonics bar charts and waveform graphs and record them to USB key. I have contacted Tektronix about this, and as it is a very often requested feature, it will be implemented within the next few releases.

Application: Power Quality Logging

While the PA1000 is not directly targeted at power quality monitoring, it doesn't mean that it cannot provide any insights. For one, I know the power quality at my home can be quite variable at times, being next to an industrial zone. In order to actually quantify it with any level of confidence I used a USB Key and left it logging for over 24 hours. In addition, I also seem to have a problem with high Zellweger signalling levels, so I decided to use the harmonics charting feature of the analyzer itself to investigate the amplitude of that harmonic.

Interestingly it measured a whopping 17v, whereas the normal target is 6-8v tops. No wonder I've been having flicker issues with lighting and humming when the off-peak signalling comes on. Anyway, back to the logged data. With the data, I imported it into Excel and produced some graphs and histograms which cover several power quality parameters. Of note is that the proper way to measure it involves 10-minute monitoring periods and averages - these are simply based on the raw "immediate" reported values.

Looking at the first 32,000 samples, as Excel only likes to graph that many samples per series, it is seen that my nominally "230v" line is quite a bit higher, but not unacceptably so. By looking at all samples, it is possible to build a histogram which shows the probability of a given voltage level appearing on my line:

Likewise, it is possible to do the same with frequency, crest factor, and total harmonic distortion, which I selected for recording due to their relevance to IEC standards testing for standby power.

Now that I know how variable my mains line is, it seems like standby compliance testing will be a bit of an issue - but none of this is the fault of the PA1000.

Computer Control - PWRVIEW Software and USB Connection

While standalone usage is the traditional bench-instrument style of operation, the provided connectivity and included software provide an additional dimension to using the PA1000. Computerized control of the PA1000 can be accomplished through USB (as a USB-TMC device), GPIB (or IEEE 488) and via Ethernet (more on this later). The software is provided free of charge and is available from Tektronix's website (as mentioned earlier in the Firmware Update section) for computers running Windows 7 or 8.1 in both 32 and 64 bit editions.

The PWRVIEW software has an interface which follows the Microsoft Ribbons paradigm, first introduced in Microsoft Office, which groups functions into buttons across a wide toolbar along the top (i.e. the Ribbon). Tabs are used to present different contexts/views within the application.

To start using PWRVIEW, you must first add the instrument. This is accomplished by the Add Instrument button, which will scan for directly connected instruments. Ethernet connected instruments can be added with the Network button. Of note is that several PA1000s or PA4000s can be added to a PWRVIEW project to allow them to be used to simultaneously determine input and output power (e.g. of lighting ballasts).

The first tab, labelled Setup, allows for the instrument's measurement parameters to be set up. Along the left-hand side, there is a list of pre-defined applications. All of these applications allow you to hit "Apply" to quickly pre-load the configuration enabling the most useful measurements for a given mode into the right-hand pane.

The right-hand pane is really just a collection of all the configuration options for the instrument, so you can easily amend settings (e.g. which range, current-shunt etc) or use the AC Power setting and fully customize your settings before clicking Upload to send the configuration to the instrument (which takes about 10-15 seconds). The AC Power application is the "Normal" mode, general purpose configuration which allows almost all parameters to be customized, whereas the other modes often have restrictions on parameters that can be changed to ensure the results are as expected.

As compared to the manual mode of using the instrument menus via the front panel, this proves to be a much quicker way to get your instrument configured to your needs, as everything is easily accessible. After uploading your desired settings to the instrument using the Upload button, you can always hit the # key to toggle the operating mode from Remote to Local to allow you to use the PA1000 with the uploaded configuration via the front panel.

However, that is not all for the pre-configured applications. You can also click the Wizard button which will bring up a guided diagrammatic view of the required connections based upon some simple inputted parameters (e.g. peak currents). These are different for each mode. Unfortunately, some of these diagrams do need updating as they reference an older version of break-out box.

Unfortunately, you cannot add your own applications to the applications list, however, you can save your configuration into a project file, with extension .vpm, which can be reloaded later to restore the customized configuration.

You don't have to use the Upload button - you can proceed to begin using the instrument through the Measurement tab, in which case, beginning an acquisition run will automatically result in an upload of settings.

The Measurements tab is the main tab for use with most modes. The measurement tab allows for you to configure the acquisition parameters - number of significant figures (generally, with Averaging at 1, the data returned contains 5 significant figures), the amount of averaging (Auto, recommended, applies an averaging value of 8, but configurable from 1-64) and whether Zero Blanking is enabled (which clips very small values to zero).

It also allows you to save settings, start and stop an acquisition run, take a snapshot of the data, continuously record data into the database, export recorded data, and view waveforms and harmonic bar charts.

This tab has a body which resembles a spreadsheet in more ways than one - while data appears in columns from each instrument, free cells can have formulae entered to perform real-time calculations on collected data to compute parameters not directly measured.

To begin seeing live values from the instrument, after having first configured the power analyzer to your liking, you simply need to click the Start button. Values will begin populating the columns at a rate of about two values per second which is not adjustable. I still find it a little strange, given the waveform and continuous sampling abilities of the PA1000, that it returns values only once every two seconds - in some sense, more like a multimeter than an oscilloscope. For those looking to capture very transient events, such as switching glitches, or flicker in mains waveforms, I don't think this meter is capable of it, although the hardware is possibly capable of doing so. It is noted that the PM1000+ on which this is based had some ability to measure flicker, however, no mention of this is made by Tektronix.

To record a single reading, you merely have to click the Snapshot button. To perform continuous logging, you can click on the Record button. Clicking on it again stops the recording. A list of prior recordings is available under the Data button.

Selecting any one of the recorded data sets (displayed by date-time) allows the data to be exported in a Microsoft Excel .xslx file with two sheets.

The first sheet contains the metadata, with the second sheet containing the values including absolute timestamps. Hurray!

But, unfortunately, with every hurrah comes a bit of a disappointment in the performance of the database. On a modest AMD APU based system, exporting a spreadsheet can sometimes take several minutes to complete as it combs through the database.

When the database starts getting a little big, it gets frustratingly slow to work with - even if that is to delete and clear the database. There's no bulk-delete or bulk-export features yet, meaning that you must independently deal with each data-set. In fact, if you log for an extended period of time (about 30 hours was enough for me, resulting in a database file greater than 100Mb), you can find the database data will not be able to be exported, and instead displays an error.

The database is actually in SQLite format, stored in %appdata%\Tektronix\ which, given an appropriate viewing application, you may be able to generate the requisite query to export the desired data. I haven't had the time to analyze how the tables are related, but it does open up in SQL Lite Viewer, although performance is still an issue. You can delete the database file with no ill effect, except for clearing all the recorded data.

Tektronix has been informed about the performance issues with the database, and have informed me that some improvements will be made for the next release due mid-August, with further improvements to be made over the next few releases.

Clicking on the Waveforms button allows you to display the voltage/current/power waveform as well as the harmonic bar graph. If a measurement isn't running in the main window, you will need to click Start to be able to see the graphs, and the relevant harmonics need to be selected in the Setup tab to ensure that the graphs are available.

The waveform graphs are regenerated from the harmonic amplitude and phase values reported by the PA1000, up to the 50th harmonic. As a result, the waveform displayed on the PC does differ from the actual waveform displayed on the PA1000's internal graph feature. Specifically, it is smoother and square "edges" are ringy due to the limited harmonic data available (especially if only a limited number of harmonics are selected).

A nice sine wave will have only small harmonic components, so it's probably not so bad for real sinusoidal waveforms. The scale is adjusted with the sliders on the side of the graph, but the sliders are not linear in their response. It can be a little challenge to move the sliders to scale the graph to your liking. No grid lines or cursors are provided.

The Harmonics bar graph is displayed in a similar fashion, with a slider based scale, however, it is referenced to fundamental as a percentage. You can run the scale to clip off certain bars, to provide better resolution for harmonics with lower amplitudes. The absolute value for the harmonic is available when the mouse is hovered over its respective bar.

A relatively comprehensive help-file is included, for easy reference, although it does also reference the older break-out box. It explains most of the options in sufficient depth for new users to acquaint themselves with PWRVIEW.

The Test and Results tabs are used only for Full Compliance Standby test modes which will be covered in a later section.

Application: Voltage vs. Power Analysis

One of the biggest controversies surrounding power lately involves the regular accusation that utilities are driving up energy consumption by keeping mains voltage at the point of supply on the high side of what is legally permissible, claiming that it is a requirement for the stability of the network due to "long feeders" or whatever. In some cases this is true, but there has been a big furor amongst different groups of people, both technically competent and ordinary people, about whether this is true. There is still some sort of division of opinion.

As it turns out, this instrument can help, as it gives us the ability to log the voltage and power consumption and export the values for graphing and analysis. By varying the voltage supplied to a specific load (more on this later), we can determine just how much power is being drawn and see if there's a correlation between voltage and power consumption.

The main types of appliances can be separated into a few categories. Regular resistive loads (heaters, old fashioned light globes), inductive loads (motors, fans, unloaded linear transformers) and complex loads (switching supplies). There aren't many capacitive loads in the house that spring to mind.

For a quick sampling, I've decided to graph just what happens when the voltage to a set of linear transformers is varied from 250v to 220v. This was done by using PWRVIEW to log the relevant parameters while the voltage was being varied using a Variac.

The results are not too surprising - there is some truth to the claim, mainly because the transformers may be skimping on iron, and running into saturation at higher voltages. However, it does go to show that unloaded linear transformers do consume more power in standby if the line voltage is higher - in the worst case of the five tested, about 80c/year from 230v to 240v. I sense a lot of fun potentially happening when I see what happens with a wider variety of appliances.

Computer Control - Ethernet (LAN) Connection

The PA1000 offers a 10Mbit/s half-duplex Ethernet port. You can choose for DHCP configuration, where the IP is assigned by a DHCP server or a static IP configuration where the instrument uses a fixed IP address. There is a diagnostic page which allows you to see the present assigned IP address, which is especially handy when you are using DHCP. Generally, it is most practical to set the IP manually or by setting a binding in your DHCP server, as the PWRVIEW software does not scan the network for PA1000's - you must manually give PWRVIEW the IP address your PA1000 is using for it to connect to it via Ethernet.

The device appears without a host-name on the local network. A port-scan revealed only port 5025 was open, which is for SCPI over Ethernet requests. As this port will accept requests from any source, it is vital that the product be connected to a firewalled/secured network consisting of purely trusted hosts, otherwise, there is the potential for remote-commanding of your instrument. For writing your own programs, the manual presents a full command reference.

It's probably useful to note that the Ethernet interface is not LXI capable, and does not offer the ability to control the instrument through a web browser (unlike other, more recent Ethernet-enabled test equipment). This can be considered both a blessing and a curse, depending on how you look at it, as a web-interface may encourage novice network intruders to "play" with your instrument more than an "anonymous" device with an open port they might not understand. However, it does mean that the features offered by LXI, such as the web-interface, device to device modes, high precision triggering are not available on this device, although I can't really see a need for that.

The Ethernet interface is especially useful, where existing Ethernet/IP networks exist in locations where the analyzer is to be stationed, and allows for a computer to control and log data from it without any extra connections. It also overcomes the single-run length limitation for USB of 5m (or 25m when you have chained the maximum length of hubs and cables), while potentially improving safety by isolating the ground (which USB does not). It also allows sharing of the same instrument (one at a time, of course) between workstations without plugging and unplugging, however on the downside, the instrument may be taken over by others on the same network.

When a connection is made to the port, the instrument goes into Remote mode, which causes a yellow textual indicator to appear on the right side of the screen. By pressing the # key, you can return control to the Local mode, allowing for local keypad control as long as an acquisition run is not currently in progress.

Application: IEC Compliant Standby Load Analysis

As mentioned earlier, one of the features of the PWRVIEW package is the pre-configured application for full compliance standby testing to IEC62301 Ed.2. This makes it very simple process to configure the power analyzer, collect the required test data, ensure the data meets required standards and generate reports. In fact, it's possible in the best of cases to have a product tested from start to finish within under 45 minutes!

Other power meters often have computer connectivity and the required applications as an additional cost extra, so their inclusion in the case of the PA1000 deserves particular praise.

Just Won't Pass!

It's at this point I should disclose that, while this write-up is presented as a logical progression, the actual use and experience of a product often isn't. In fact, one of the first things after updating the firmware was actually trying to get Full Compliance Standby testing operating. After all, it's my firm belief as an engineer that the best way to learn is really to get your hands dirty and get "knee deep" into it. As a result, I managed to have the PA1000 teach me some power engineering concepts and inspire me to solve several power issues. You definitely can learn something just by having your hands on such gear.

As I was rather busy and wanted to get some things over and done with, I hooked up the breakout box to mains and the PA1000 with a simple USB Charger as a load. I then ran the Full Compliance Standby process as detailed above, but I could never pass. It was clear that I was failing on Voltage, Crest Factor and sometimes Total Harmonic Content.

It will seem obvious to readers that my mains supply just won't meet the IEC requirements without some help - that's because this test actually inspired me to go back and determine just how far from compliance my mains electricity is.

As a reminder, the IEC standards require the nominal voltage (230v) +/- 1 % and frequency (50Hz) +/- 1% with a Crest Factor between 1.34 - 1.49 and a Total Harmonic Content (first 13 harmonics) under 2%.

Fixing the Voltage - Using a Variac

Given that my main trouble lay in the Voltage, and the CF/THC generally gets better outside of business hours, I decided to go about fixing it. One way to "change" the voltage is to use a variable autotransformer, commonly known by its trade name, Variac. This consists of a single winding on a core which has a surface ground away for a carbon brush wiper contact, thus allowing the output to be varied continuously from 0 to about 100% or 112% the input voltage depending on the wiring. You can think of it like a continuously variable transformer with a mutual primary and secondary coil. Virtually anyone who salvages, services or otherwise develops mains powered appliances should probably have one. I've always meant to have one, but I didn't see a good enough reason to invest in one until now.

Before I continue, I should probably quickly mention that a Variac does not isolate the input and output, and in case of certain sorts of failure, may result in the full mains voltage being applied to the output when it is not expected. Only experienced technicians and engineers who know what they're doing should operate such a thing! In addition, unqualified people should not work with mains wiring!

Of course, since I'm a bit of a penny pincher, I decided to build mine from an open frame Variac, rated at 1A, sold by element14. I already had an enclosure, and I added two fuse holders and two T1A M205 ceramic fuses to protect the variac. That was wired up to a scrapped extension cable (for a male and female mains plug). Lo and behold, I have a way to control the mains voltage. Speaking of which, for safety, one should probably also use an isolation transformer to isolate the input from the output. It's not strictly necessary, but it's a desirable form of protection which limits the shock hazard. I did manage to build one from two old toroidal transformers with their secondaries hooked together, but alas, that only makes the voltage regulation worse so I don't use it much.

What the Crest Factor!?!

As it turns out, having the Variac in line with the mains to the breakout box did allow for the voltage to be "trimmed" so as to be 230v, however, the daily variations still proved to be a little frustrating. At times, the voltage would move in such a way that it would fall out of the 1% requirement before 15 minutes had passed. The moving voltage also increased the error in the standby power, as did the occasional transients and Zellweger off-peak signalling.

Unfortunately, while it was technically possible to pass the IEC requirements with the Variac and the home supply, due to a software range issue with crest factor (they've got it as 1.39 to 1.49 for compliance when the IEC documents state 1.34 to 1.49), I still could not get pass reports from their software. I have contacted Tektronix about this, and this has been fixed in the current version of PWRVIEW. But it does point to mains testing being rather futile, maybe, unless I had a feedback loop stepper-control Variac to regulate the voltage. At one stage I thought "maybe a ferroresonant transformer might be helpful", only to find them extremely expensive and rare ... not to mention, a probable source of harmonic distortion (so thus possibly violating the THC requirement).

Finally, we get some Passes

Sick of all this, and still very seriously craving some passes, I decided to solve this problem once and for all by employing a synthesized mains source - a commercially available 300W pure sine wave inverter. I hooked this up to a fairly high powered switchmode desktop lab supply, which also solves the isolation issue. I wasn't really convinced of the ability of a so-called pure-sine-wave inverter to produce nice power, but apparently I was wrong, since the $50 inverter produced a good result (shown earlier), although the voltage was slightly off (and easily trimmed with the variac). In the end, this approach resulted in much success, with many pass reports being easily attained.

Running a Test

Running a test is fairly easily accomplished. With the PA1000 switched on and connected to an instance of PWRVIEW, one merely has to fill in the tester and product details, configure the shunt, connect the load and click on the start button.

There is a running graph of the current power consumption which allows you to visually confirm the trend and also clear details on the stability, voltage, crest factor and total harmonic content amongst others, which lets you monitor the progress of the test. In the case a value has gone out of range, it will be shown in red, and if the value has formerly gone out of range (causing the test to fail once the time is up), it will be shown in orange. Pursuant to the IEC standards, should the data sampling frequency requirement not be met, or an out of range condition exists, the meter will clearly display an error, and the test will be restarted with the current range stepped up to the next higher range.

In all, given a stable enough power source, it's mostly a plug and play operation. All tests are saved into the database automatically, and can be recalled by date and time for display on screen or export as a PDF report (sample attached to this RoadTest) or raw values as an Excel spreadsheet. It does get quite confusing that the tests are listed by date and time only, and no product information is shown unless the mouse is hovered over the listing, and as such I have requested a feature enhancement which should take place in the next few releases to address that. Another issue is that loading the result from the database resets the product displayed and exported as the first one alphabetically listed in the products database - this issue will be fixed in the next release. The lack of a bulk export facility does make it a bit annoying when working with large datasets, however, the feature will be coming after several releases.

A total of 120 of my own (relatively old) household products were tested in order to best test the instrument. Is it any wonder why it's kept me busy? All for a hobby ...

The results are colour coded with red results indicating either over-power, or results which don't meet the uncertainty requirements for the IEC standards. It seems that many loads can be quite challenging to profile, as the IEC standards require continuous sampling with no range changes, so for very peaky loads, the analyzer may need to change up to a higher range and suffer a greater uncertainty in each reading. The IEC standards, however, are very stringent on the uncertainty levels, which makes such tests difficult to pass. However, even if the results don't necessarily meet the standards requirements, they're miles better than what you can achieve using low-cost plug-in power meters when it comes to accuracy and might be enough for your requirements anyway (definitely enough for me). There is no bulk export at the moment for the database, so I had to painstakingly re-enter the above results by hand - but I have requested that this be a feature which may arrive in a future release.

A problem was identified with some loads which reach their stability requirement in-between 1h00m and 1h15m, and 2h00m and 2h15m, where the test did not terminate until 1h15m or 2h15m was reached. I hypothesized that this was a problem in the software in the coding of the termination requirement, where 15-minutes minimum test time disregarded the hour in the computation. I contacted Tektronix about this, and they believe they have found the problem with the fix due to be released in mid-August.

The difficult loads (red in uncertainty ratio/UCR column) which may not be possible to qualify using Sampling Method due to measurement uncertainty. That being said, it is doubtful that such difficult loads would be handled any better by any of the other power analyzers predominantly as the difficulty lies in the low power factor, high current crest factor values and number of current and voltage ranges which are not significantly bettered by other analyzers.

The results show a large variation in standby energy usage when considering products that perform nearly the same function, however, the vast majority seems to consume 1W or less, which is good news for the environment and your power bills!

Current Shunt Fuse

One thing that's bound to happen after some testing is that for some reason (whether it be by mistake, or due to inrush current) the 1A shunt fuse may blow and bring your compliance testing to a halt. Unfortunately, Tektronix did not supply any spare fuses, and only have the bare specification of F1A 600V printed on their device and manual. Removing the fuse reveals it is marked 'A12FA660V 1A' which is actually a Schurter Super-Quick Acting FF fuse in 3AG size (6.3 mm x 32mm).

There are many unsuitable 5AG sized fuses, so you need to pay particular attention. As this part is fairly hard to get, the best alternative I can find is a Siba 70-125-40 Ultra Rapid 1A 700V 3AG fuse. It may be pricey, but it fits the bill.

You will find regular ceramic 250v and glass 250v fuses for a lot less that will fit, and are fast-blow (as opposed to FF which was supplied), but do not do this. As the inputs to the power analyzer are rated for 600v, if anyone connects high voltage or high current sources, the safety of the operator and the analyzer is at risk if the fuse is unable to break the required fault-current and source voltages! If you do choose to replace it with a 250v fuse temporarily, you will probably void the warranty and should ensure that you do not exceed 250v input.

Build Your Own Breakout

Despite having the appropriate UK to Australian adapter so that I could use the provided Tektronix BB1000-UK breakout, I still wasn't particularly satisfied. For measuring inrush current, for example, there's no easy way to switch the appliance on and off. Plugging and unplugging the product is a rather tedious process. For standby power consumption, many new "wall-wart" supplies are made with a right-angle plug layout that means the body of the adapter clashes with the shrouded banana leads. Replacing the UK IEC lead with an unfused Australian IEC lead means that we lose the protection of the fuse in-line which could be desirable when testing appliances with "unknown" faults.

I thought, "how about we take care of all of these concerns at once, by building our own breakout?" That's exactly what I ended up doing. That allowed me to reconfirm my understanding of the wiring of the breakout box, and gave me a chance to have it "just the way I like". For safety reasons, I cannot recommend unqualified persons attempt to do this themselves - it can be a safety risk if improperly constructed!

My concept was to use a similar weatherproof plastic box, and reserve one face for a switched GPO outlet. On the other face, a fused IEC socket will be used to supply power, and shrouded banana connectors will be used to connect with the power analyzer on the same face. After a little shopping and rummaging my spares box for parts, drilling, filing, and soldering, I came up with this.

It's not suitable for use above 250v 10A, although it's not likely going to be used in that situation due to the reliance on mains voltages and the fact the input IEC socket was fitted with a quick blow 10A fuse for protection. I suppose if I had the tools, I would have used crimped spade connectors rather than soldering the wires on.

I guess that satisfies me now. Maybe Tektronix can manufacture a dedicated breakout box for the Australian/Chinese market in the future. Surprisingly, the configuration wizards which illustrate the original universal breakout from Voltech seem to illustrate something similar in concept (switched GPO).

Overall

I think it's pretty clear that the PWRVIEW software makes standby energy compliance testing very simple. It implements the Sampling Method in an almost completely automated way, making it very simple to check for standby energy compliance. The main problem was with my mains and difficult appliances!

Conclusion

I think there is much to like about the Tektronix PA1000 power analyzer. A full colour LCD with graphics ability, and data logging to the front panel USB allow for comprehensive and comfortable standalone instrument usage, while a full complement of GPIB, USB and Ethernet connectivity with the PWRVIEW software included as standard elevate the instrument above most competitors especially when price is considered.

Its DSP based architecture allows for measurements more complex than that of most basic power-meters at the same price level, such as harmonics up to the 50th order, while simultaneously boasting a wide bandwidth of 1Mhz, basic accuracy of 0.05% reading +/- 0.05% range and crest factor tolerance up to 10, which are very respectable specifications.

For basic power measurements, it is very capable both for use in AC and DC environments. It is especially useful at fluorescent and LED ballast testing, GPO appliance load profiling, general power measurements (via external transducers if necessary) and standby energy compliance, which are the PA1000's strengths.

However, the meter itself doesn't seem as ideal for work with some motors and inverters, as it lacks the inputs and calculation modes required to display motor-relevant parameters unlike some of the other power analyzers which are more motor oriented. It also does not seem suitable for cycle-by-cycle analysis for capturing flicker and transient events (even though the PM1000+ did have some ability to do so), or for power quality measurement and logging, for example. It's important that prospective buyers determine what their intended application is and then shop with that in mind.

While the package is generally quite comprehensive, it is not free of bugs, performance issues and unexpected behaviour, and I have been in contact with Tektronix who have taken the feedback seriously and have committed plans to bring fixes to the identified issues and feature enhancements over the next few software releases.

As usual, if you feel this review is helpful, please rate it or give it a like! If you have any questions, you can leave a comment and I'll try my best to answer it. For the latest of what I'm up to, feel free to visit my personal blog at http://goughlui.com

After Review Updates

I have remained in contact with Tektronix after the main body of the review was written, and they have released a new version of the PWRVIEW software, v1.1.5.109, which I have confirmed fixes the crest-factor range issue identified in the review above and comes with some improvements to display and robustness. The software also claims to have fixed problems with TekVISA (which I don't use and cannot confirm) which should be a welcome addition to loyal Tektronix users.

They have informed me of their commitment to deliver the next release of software and firmware mid-August with fixes for the identified problems of a broken hidden menu, integration time scale, product-selection difficulties in PWRVIEW, standby test run time issue, and slow performance dealing with large datasets.

According to Tektronix, other feature requests identified (mainly with screenshots, bulk exports and deletes) will see improvements on the next few releases.

While it has taken some time for this to emerge, I can very much appreciate the effort that Tektronix has put into providing me comprehensive, well considered replies to my requests, and the seriousness in which they take these reports and suggestions. I can also appreciate that developing and refining software for test equipment can be a longwinded process, especially to ensure that existing results are not compromised, and in order to ensure no regression takes place (i.e. working features getting broken). I look forward to seeing the future releases and testing them. I continue to be in contact with Tektronix and will bring you the latest in developments as they happen.

For all users of the PA1000, if you have any issues or feedback, please contact Tektronix as I'm sure they would love to hear from you.

UPDATED: Fixed a few typos and a missing image.

UPDATE 24th August 2014: Corrected incorrect information in other power analyzers comparison table upon contact by N4L.