E36313A Triple Output DC P.S. - Review

Table of contents

RoadTest: E36313A Triple Output DC P.S.

Author: luislabmo

Creation date:

Evaluation Type: Power Supplies

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?: I've considered other products in similar category based on information available online but can only provide a comparison point based on my experience as this is the first time I've ever gotten my hands into a Variable Power Supply.

What were the biggest problems encountered?: Critical Bugs: the unit has some bugs, also the user manual. I consider this needs special attention as some bugs freeze the PSU completely in whatever state it is representing risk. Cosmetic damage: there was an area were the paint was rubbed off, it may have been caused by too much friction, a bad paint job, or maybe poor packaging.

Detailed Review:

The https://www.element14.com/community/view-product.jspa?fsku=&nsku=15AC9660&COM=noscriptE36313A is very powerful and brings a lot of good features with a very intuitive user interface, I think it will be great addition to any workbench. If you are a hobbyist like me, an instrument in this price segment may be a little overkill but, will serve you for many years to come as it packs very good specifications and will help you in your projects with ease. I've found few negative aspects but I don't think this will take merit from a good product like this as it performs very well in what its main purposes are: providing 3 outputs with clean and stable power and extra features.



The E36313A came held in place with custom packaging inside a sturdy box, on top of it came a small box with documents and parts. Unfortunatelly there were few scuff marks present in the PSU, the paint was was rubbed off where the scuff marks were, matching exactly the corrugation pattern of the smaller box that came on top, which led me to think that too much friction or a not so good paint job cased the issue?. Well, before I took the PSU out of the box, I noticed that it wasn't inside a plastic bag -and then held in place with the custom packaging- this practice I've seen in many electronic devices I've bought before like Monitors, Computer cases, Soldering stations, etc., maybe this was the real cause of the issue.


{gallery:width=640,height=360} Unboxing


E36313A Unboxing: Big box


E36313A Unboxing: Paint damage


E36313A Unboxing: Custom packaging


E36313A Unboxing: Box with items


E36313A Unboxing: Extra items


E36313A Unboxing: Extra Parts


Package contents

  • E36313A Triple Output Programmable DC Power SupplyE36313A Triple Output Programmable DC Power Supply (configured for 115V)
  • 6 ft. grey AC power cord for North America
  • A Connector kit
    • One 10 A, 3.5 mm female 4-pin terminal block connector
    • One 12 A, 5 mm female 4-pin terminal block connector
    • One 15 A, 5 mm female 8-pin terminal block connector
  • Keysight Instrument Control DVD with
    • IO Libraries Suite 2017 Update 1
    • Command Expert 2017 Update 1
    • BenchVue 2017
  • Certificate of calibration
  • Extra documentation
    • Packing list
    • Quick start guide
    • Safety leaflet
    • others


This PSU provides many great features. I'm going to list some that are advertised which I think are the the best features this unit may bring to you:

  • Three independent outputs (6 V/10 A, +25 V/2 A, +25 V/2 A)
  • Low output ripple and noise
  • Color-coded channels and 4.3" color LCD Display
  • Meter view to display more info on a selected channel
  • Individual voltage and current knobs with rotary encoder control
  • Auto series/parallel connection
  • Data logging
  • Front and rear output terminal


How does it feel?

It feels great overall!. Unit looks good, construction is rigid and sturdy, rubber buttons look and feel good, the rotary click switches and their rubber knobs are perfect -to my taste every click feels right: soft but accurate... and I tried them A LOT-.

One thing that got me mixed feelings was how every button press felt. I personally dislike long buttons, some times they feel wobbly and imprecise -depending how firm your hand is-, unfortunately, most of the buttons are like this.


{gallery:width=768,height=432} E36313A


E36313A: Idle


E36313A: Off


The AC power entry feels wobbly, tested this with other AC power cords, ironically, this may represent an advantage in some situations; this PSU has some software bugs that may freeze it completely and the only option available is to disconnect the mains power -more on that later-. On a serious note, a grounded power switch is highly recommended until a more stable firmware is released.


  • Power consumption while OFF: 14mA, fan works at very low speed and LCD is On -can be appreciated on a very dark room-.
  • Power Consumption while Idle: 16mA -just 2mA more than OFF, interesting-.

The fan in this unit is always ON. When the PSU is Off, the fan will be running at a very low speed.


The power supply's front panel features a high contrast 4.3-inch LED color display, a little glossy -not a problem here-, image quality is very sharp, the view angle is just enough for an instrument like this, the layout and the information displayed is on spot, displaying the necessary information for each channel matching the color-coded channels (buttons, output terminals and the paint job). The Multiple Output View displays the actual output voltage and current and the present output voltage and current settings.



Meter View

Also when needed, a meter view is available which displays more information on a selected channel:

  • Output identifier matching the channel color-code.
  • Output meter (actual output voltage and current).
  • Output settings (present voltage and current settings).
  • Output status
    • OFF
    • CV: Constant Voltage mode
    • CC: Constant Current mode
    • OV: Over-voltage protection is tripped
    • OC: Over-current protection is tripped
    • OT: Over-temperature protection is tripped
    • UR: Unregulated
  • 2wire or 4wire sensing.
  • Over-voltage Protection (OVP) setting and Over-current Protection (OCP) status
  • OCP, On and Off delays.
{gallery:width=480,height=272,autoplay=false} Meter View


Meter View Ch1: LCD detailed view of Output 1


Meter View Ch2: LCD detailed view of Output 2


Meter View Ch3: LCD detailed view of Output 3

Can be noted also that when changing a configuration affecting a single channel, the configuration color will match the the channel color-code, also the matching output selection key will light up.

{gallery:width=480,height=272,autoplay=false} Screen examples


Output 1 - Source Settings: Note the Yellow color matching the Output 1 color-code


Output 2 - LIST Properties: Note the Green color matching the Output 2 color-code


Output 3 -  Output LIST: Note the Blue color matching the Output 3 color-code


Output selection keys

Selects an output to control. The lit key indicates the selected output.


Output On key

Turns individual outputs On/Off; outputs are On when the key is lit.



The user experience is good overall, simple to operate and intuitive to use in big part thanks the color-coding scheme and the color LCD, making very easy to understand what are you doing and where. In all my tests, I did measurements with other instruments and this is one of the aspects this unit excels; provides precise voltage and current output control and high accuracy readings making the user experience hassle free as there is almost no need for separate measurement instruments.


First thing first, please pay attention to the User's Guide recommendations -I will talk about this in more detail later-.


Turning the E36313A On

When turning the E36313AE36313A on, it takes around 6 seconds to be ready to use. For a smart PSU with a micro-controller and a OS that has to manage a USB port, 3-outputs, Network capabilities, a nice presentation of the data, and all the features this unit brings I say 6 seconds is a good startup time. When the unit starts, the FAN speeds up for about a second, this is a common practice I've seen in Servers and some PCs of reputable brand.


A numeric keypad is available to accurately adjust the Voltage or Current settings. Independent Voltage and Current knobs may be used too, which use acceleration to make for higher/lower increments, feature that I personally like and find very useful most of the times, but, a little bit dangerous sometimes: when adjusting the Voltage/Current while an output is ON, if not careful, you may end up going too far, risking the integrity of whatever you have connected to the output. I would have done this feature a little different, by using rotary encoder with switch, this way the knob can be pushed to toggle on-off the acceleration feature and also adding an icon indicating such status.


It is been brought to my intention that with the outputs OFF, the PSU reads some current and voltage values, makes me wonder whether those are offset values or perhaps noise?. Should I account for those values when using the PSU?. I used a multi-meter and I didn't detect any voltage/current in any of the outputs while they are OFF.


One thing I didn't like at all is, when in other menu options, you won't have any idea of what is going on with the outputs (except the On/Off lights), this is ok BUT, if you turn the Voltage/Current knobs you may end up changing the Voltage/Current settings without any feedback as you are not in the meter view -I think this is a little risky-.


When entering any Voltage/Current using the keypad, it allows to input up to 9 digits which I find not very useful as the unit does not handle that much precision and only displays maximum 5 digits (not including decimal separator).


Data Logger

The Data Logger works very well, I think it is one of the best extra features this unit brings as there is almost no need of extra measuring equipment you may select to record Voltage, Current and Power independently for each output. The data logger seems fluid an accurate and allows to visualize in a nice graph what is happening with the outputs. The Data Logger also has the option to export the data in a very common format (CSV). On the downside the minimum sample period allowed for the data logger is just 200ms image also, a USB flash drive is required to use the data logger.


When the Data Logger is operating, if you decide to switch to another screen, a notification at the top of the screen will indicate that the data logger is actively recording.


Adding a time-stamp to the log filenames is not the default option, which I think it should be. You should activate this option every time you plan to use the data logger -and yes, every time you turn off the PSU the option doesn't get saved-. Ironically, screenshot filenames always include the time-stamp. If you prefer to save screenshots in PNG over BMP format <default>, this is another option that doesn't get saved once the instrument is turned Off image.


Testing Constant-current mode with a 20mA load

I think everybody tests their new shiny PSU with an LED first, right?. I wanted to run a first test with a standard LED (20mA), checking the current limiting features of the instrument and using the data logger at the same time, so this is how a ran the test:

  1. I've set the Current limit to 20mA
  2. Turned On the Output
  3. Then slowly dialed the Voltage knob up until the PSU switched from CV (Constant Voltage) to CC (Constant Current) mode and limited the current.

I've found that the current limiting for a 20mA LED works very well on Output 2 and Output 3, it is very quick, accurate and no LEDs were harmed in the process.


Well, things didn't go that well for me running the same test on Output 1. This output goes as far as 43mA on CV (Constant Voltage) mode before switching to CC (Constant Current) mode to limit the current to 20mA, here is are a graph showing the effect and the meter output.



I didn't find any clear explanation to this behavior on Output 1 in the specifications, so I'd love to reach my fellow readers if someone knows an answer to this.


Testing Constant-current mode with a 350mA load

Ran a similar test, this time with a 350mA LED running the same steps as the previous test:

  1. I've set the Current limit to 350mA
  2. Turned On the Output
  3. Slowly dialed the Voltage knob up, until the PSU switched from CV (Constant Voltage) to CC (Constant Current) mode.

Once more, a 350mA load works very well on Output 2 and Output 3 switching to CC (Constant Current) right on the 350mA mark, Output 1 did better this time but is not as accurate as Output 2 and Output 3 seems. Seems to me that Output 1 is not as good for small loads at the Output 2 and Output 3.


Output 1 switches to CC (Constant Current) after it hits the 371mA mark


Output 3 switches to CC (Constant Current) mode right on 350mA



Output 3 switches to CC (Constant Current) mode right on 350mA


Auto Series/Parallel connections

This is an awesome feature as will save us from using external wires, instead internal relays will do the work. I tested Auto-Parallel and Auto-Series mode with 3 different stepper drivers.


The current meter and the Logging features of the E36313AE36313A were not very useful in this test as they were not accurate -just for this scenario-. I used constant current stepper drivers -known also as chopper drivers-, to control the motor current, a chopper driver turns the output voltage to the motor on and off rapidly, due of the nature of the chopper drivers, the instrument's meters could not accurately determine the total current drained by the stepper driver in all 3 scenarios, this can be verified  by measuring the current independently in each coil with an ammeter, this is not bad at all -do not get me wrong here-, here the E36313AE36313A excels at providing enough clean juice to make the motors work with high DC voltage and/or high DC current that may be required for this kind of tests.


I used an Arduino to provide the step pulses to the stepper drivers at different speeds, the pulses speed was regulated using a potentiometer. The Arduino is then connected to a HCPL-2200 opto-couplerHCPL-2200 opto-coupler to provide isolation from the micro-controller.


Stepper1: Auto-Parallel (3A/phase Bi-Polar stepper motor)

In this scenario I used a LV8727LV8727 driving a 3A/phase Bi-Polar stepper motor with at a half-step resolution.

  • Output 1: powers the logic input of the LV8727LV8727 and the output of a HCPL-2200 opto-couplerHCPL-2200 opto-coupler.
  • Output 2 in parallel with Output 3: provides power for the stepper output.


This scenario may have triggered the Over Current protection when the stepper is at θ64 (71% power each coil, around 4.26A) but did not happen, this may be due to the fact explained before where the sensing circuit is not fast enough to detect an accurate reading of the current used at a given point of time.


Tested with 12V and 24V Vmot changing the stepper speed

{gallery:width=480,height=272,autoplay=false} LV8727 stepper driver test


LV8727 idle: Ch2 Idle at 9V, note current reading is only 854mA




Stepper2: Auto-Parallel (1.68A/phase Bi-Polar stepper motor)

Used a DRV8825 driving a 1.68A/phase Bi-Polar stepper motor with full-step resolution.

  • Output 1: powers the logic input of the DRV8825 and the output of a HCPL-2200 opto-couplerHCPL-2200 opto-coupler.
  • Output 2 in parallel with Output 3: provides power for the stepper output.


Tested with 12V and 25V Vmot changing the stepper speed

{gallery:width=480,height=272,autoplay=false} DRV8825 stepper driver test


DRV8825 idle 24V: Ch2 Idle at 24V, current reading is only 266mA

DRV8825 idle 12V: Ch2 Idle at 12V, current reading is only 454mAimage



Stepper3: Auto-Series (0.33A/phase Bi-Polar stepper motor)

Test with a DRV8824 driving a 0.33A/phase Bi-Polar stepper motor with full-step resolution.

  • Output 1: powers the logic input of the DRV8824 and the output of a HCPL-2200 opto-couplerHCPL-2200 opto-coupler.
  • Output 2 in series with Output 3: provides power for the stepper output, tested at maximum voltage supported by the DRV8824 (45V).


Test with 24V and 45V Vmot changing the stepper speed

{gallery:width=480,height=272,autoplay=false} DRV8824 stepper driver test


DRV8824 idle 45V: Ch2 Idle at 45V, current reading is only 266mA


DRV8824 idle 24V: Ch2 Idle at 24V, current reading is only 454mA



When testing the stepper drivers, sometimes the instrument's fan will speed up, also a switching noise can be heard inside the E36313AE36313A , depending of the stepper driver's decay mode selected and the speed of the stepper.

Inhibit Function

The Inhibit Input function lets an external input signal control the output state of all the output channels in the power supply. I've tested this option as I find very useful to have an external Panic Switch or E-Stop (Emergency Stop) when required, the following diagram shows how I wired the button to the Digital Port.



{gallery:width=854,height=480,autoplay=false} E36313A Digital Port


Digital port: 3.5 mm female 4-pin terminal block connector


Digital port: Panic switch wiring


First, I decided to use a Normally-Open switch, then in the front panel option Utilities > I/O Config > Digital IO changed the Pin 3 polarity accordingly:

  • Set Pin 3 polarity to Negative if a Normally-Open button will be used.
  • Set Pin 3 polarity to Positive if a Normally-Closed button will be used.


Then in the the menu Output Settings > Output Inhibit I've enabled Inhibit option Latched which will make the Inhibit signal remain latched until the inhibit input is turned Off and the output protection is cleared for all outputs.


The Output List

The Output LIST allows single or multiple outputs to run in sequence, you can configure up to 100 voltage and current steps in the Output LIST window.


I didn't test much this option as I didn't find any practical scenario that I could use to properly test this feature providing useful results. I set an example where I'm executing a LIST of 7 steps. The data logger may be used while the LIST is running; this can be appreciated on top of the meter view.


It can be appreciated in the meter view that the Voltage/Current settings do not reflect any data related to the LIST steps being executed.



Web Interface

This instrument can monitored and controlled from a Web browser. This is probably the only feature where I completely disliked the user's experience, In my opinion, this feature needs to be improved/redesigned is it renders the instrument very slow, the Web interface is also very slow, to get the results I expected when trying to change a voltage for example, I've found myself pressing a key in the web interface and then waiting until the LCD screen emulation was refreshed in the browser before pressing the next key .



The web interface is very slow and drives the overall performance of the instrument's down, the issue boils down to how the web interface was implemented, I've found that the web interface relies 100% on the screen capture of the E36313A


Every time a key is pressed in the web interface -or every certain amount of time-, the PSU will take a screen shot behind the scenes in the background, later sending a refreshed screen capture to the web browser -trying to match the web browser LCD emulation to what the PSU is displaying-. Of course, this process is slow, and uses a big part of the E36313A's CPU, making it slower overall. This can be noticed by using the instrument's keypad with the Web interface working, it will respond and will refresh the screen at a slower pace.


Not everything is bad here, I think the web interface is good for things such as:

  • Tracking the actual status of the PSU -output, voltages, Over Voltage/Current protection, and so on-.
  • Front panel identification indicator: when this option is enabled will make the PSU easier to find in a rack by blinking an icon.
  • Finding some useful information about the unit like the firmware version, telnet port, serial number.

Remote Interface

As a person with coding background, I wanted to test how easy/difficult was to elaborate some activities with few commands, I connected to the E36313A using a telnet session and tested few commands directly from the User's Guide, used the Web Interface's details to find the telnet port and started some tests.


Before going further, I recommend to set The Backspace key to Control-H in the Terminal > Keyboard options, this will our keyboard's Backspace key usable in the telnet session.


For a hobbyist like me who is getting his hands for the first time into a Bench Power supply, it was surprising to find that some commands in the User's Guide have syntax issues, to make this part of the RoadTest work, I've found myself spending a considerable time looking for answers online this syntax issues.


While testing this feature I realized that the instrument's error queue, which can be accessed from the menu Utilities > Error logs for errors in the remote interface, I've found this userful and the same time a little disappointing as you won't get any feedback in the telnet session -which I think is pretty simple to do to be honest-, instead if you have any error in the remote commands, you will need to head to the instrument's menu, and find what happen there to fix the issues, that kind of defeats a little the purpose of opening a remote session to manage the instrument IMO.


Here is one of the few errors I've found

APPL Ch1 5, 1


If you look carefully, the error indicator will tell you that something is not ok, but you won't know it in the telnet session, note that I've used a command directly from the User's Guide (Copy & Paste). When heading to the instruments error queue, I've found the error "Error-103, Invalid separator"


After digging a lot online, found the correct syntax for this command in the E36300 Series Programming Guide:

APPL Ch1,5,1


Another example:


Seems every error in the User's Guide goes down to a missing coma. The correct command is



The errors I've spotted in the manual are few and were fixed by adding a missing comma. I didn't have enough time to test every single command in the User's Guide, but managed to do some work, like measuring the voltage on Channel 1:


After figuring out how the commands work I've found that they are intuitive, simple to implement and will make some activities easier for you, but if I were Keysight, I would add directly in the remote interface some sort of feedback in case of errors, that would make the Remote interface experience more pleasant in an instrument like this.



Once got my E36313A, in less than an hour found a critical bug that rendered the instrument completely unusable in whatever state it was, I think Keysight should address this with priority as this poses risk for whatever is connected to it, here is an easy way to reproduce it:

  1. Turn the E36313A On.
  2. Set the Voltage/Current for a Channel and Turn it On -this step is just required to prove that the instrument will freeze leaving the output On-.
  3. In the front menu, press Data Logger.
  4. Press the Store/Recall key.
  5. In the front menu press Back.
  6. Press the Back key again in the front menu.

After following the previous steps, the instrument will freeze completely leaving the Output of <<step 2>> On, it wont respond to any key -including the power button-, the only option to solve this is to disconnect the mains power, after connecting the mains power again, the instrument will turn on automatically -also not good in my opinion-.

Since this instrument is software driven, I think it would be good for Keysight to implement a forced shutdown where you literally force the instrument to shut off like computers do, by holding the power button-in for few seconds. Of course first I think the critical bugs have to be solved.


Also when I was trying some remote tests with a few telnet sessions, the unit froze without any action from my part, this only happened once and I could not reproduce the issue again.

At the time of writing this RoadTest, firmware version installed in the instrument was 1.0.4-1.0.0-1.04 which matched the latest firmware available on Keysight.com

Lastly, after having some of the bugs mentioned, I expected the issues to be recorded somewhere as some kind of feedback for Keysight (or for the end user). I went to the Error queue and didn't find anything. Seems the Error queue is reserved for the self-test errors and for the remote interface only.


Voltage Spikes

Unfortunately, I don't have any instrument that allows me to demonstrate with reliable numbers what I've found here but basically found two issues:

  • When the instrument is connected to the mains power it will generate a spike in Output 1 (others RoadTesters pointed this out as well).
  • When turning off the instrument -very few times- it generated a short spike on Output 2 and Output 3.


Summary and Conclusions

The E36313AE36313A is very powerful and brings a lot of good features with a very intuitive user interface, I think it will be great addition to any workbench. If you are a hobbyist like me, an instrument in this price segment may be a little overkill but will serve you for many years to come as it packs very good specifications and will help you in your projects with ease. I've found few negative aspects but I don't think this will take merit from a good product like this as it performs very well in what its main purposes are: providing 3 outputs with clean and stable power and extra features.


What I liked

  • Provides 3 outputs with clean and reliable juice.
  • Packs a lot of useful features
  • Easy to use and very intuitive
  • Screen looks awesome and information is presented in a very clean way
  • Solid construction


What I didn't like

  • Bugs: my PSU froze in few situations -some of them easy to reproduce- which may represent risk.
  • Some important settings are not saved (or are at least I didn't find an option to do so).
  • Data logger requires a USB flash drive or it won't work.
  • When Off, the LCD screen is on and the fan works all the time -just varying the speed-.
  • User's Guide has errors in some of the remote commands.


Stay tuned as I might add some extra content in the future!. Thanks for reading.


I'd like to thank Element14 and Keysight for letting me be part of this RoadTest, it's been great learning experience for me.

  • Great review @ !


    Fine, I'm not alone - the software of your P.S.U also freeze if you press the back button twice.

    E36313A Triple Output DC P.S. - Review


    Also the voltage spikes as I showed in the video.


    The last few days, I thinking about this how this can happen, because this three cannels are analog p.s.u. units and each of they has an internal output capacitor with a bleed resistor.

    Also in the block diagram it is showing that the outputs are optical galvanic isolated as well (100% the cannel 1).

    So, if you plugs on the power cord the inrush current can switch on the regulator switches and the output voltage through this can be stored in this capacitors as long as the µcontroller doesn't shuts off the regulator switches.

    During booting the outputs can be switched on.



    Block Diagram of the E3631A as a example for the similar newer following version E36313A (the connection between +25V/-25V are switched with relays in the E36313A)






    Page 150, E3631A user manual


    In most cases, the power supply will be stable for almost any size load capacitance. Large load capacitors may cause ringing in the power supply's transient response. It is possible that certain combinations of load capacitance, equivalent series resistance, and load lead inductance will result in instability. If this occurs, the problem may often be solved by either increasing or decreasing the total load capacitance. A large load capacitor may cause the power supply to cross into CC or unregulated mode momentarily when the output voltage is reprogrammed. The slew rate of the output voltage will be limited to the current setting divided by the total load capacitance (internal and external). Table 7-2. Slew Rate Inductive loading Inductive loads present no loop stability problems in constant voltage mode. In constant current mode, inductive loads form a parallel resonance with the power supply’s output capacitor. Generally this will not affect the stability of the power supply, but it may cause ringing of the current in the load. Pulse Loading In some applications the load current varies periodically from a minimum to a maximum value. The constant current circuit limits the output current. Some peak loading exceeding the current limit can be obtained due to the output capacitor. To stay within the specifications for the output, the current limit should be set greater than the peak current expected or the supply may go into CC mode or unregulated mode for brief periods.

    AWG,             Internal Capacitance,  Internal Bleed Resistor, Slew Rate at No Load and Full Scale Current Setting

    +6V Output       1000 µF                  390 Ohm                        8 V/msec

    +25V Output       470 µF                   5 k                                 1.5 V/msec

    -25V Output        470 µF                   5 k                                 1.5 V/msec




    This is what i think about, but I'm not sure at all. We need Keysight for recommendation.


    Best Regards


  • Excellent review with a good set of tests to demonstrate the operation and use of the power supply.


    Kind regards

  • Thank you for your comment , to be honest I think I'd be rare if someone runs into the freezing fault condition under normal operation, but if you are curious to know about this in your instrument, the steps are there in my RoadTest. Good luck with your new E36313AE36313A.

    All the best,



  • Great review Luis!


    It's strange that you and others have found crashing issues with this unit yet I never ran into a problem with it freezing once when I was doing all my testing or since. Maybe I was operating things in a different way and just never triggered the fault condition in my typical usage. Hopefully Keysight will get a firmware update out which will resolve these issues soon.


    Which reminds me, I need to get my unit shipped back to them for a replacement as mine makes loud humming noises which Keysight assure me are not to be expected.


    Best Regards,