Here’s my (quite) brief review of the oscilloscope received as part of the Forget Me Not -challenge, which includes using the scope to measure powering of the Pi B+ with a cheap DC/DC converter. The review is written as a flow of mind, so I'm fairly sure there is typos, mistakes etc. There's also some notes about using the scope I'd like input on, if someone has found a better way to do them...
As part of the kit, every selected competitor received a Tektronix TBS1052B-EDU oscilloscope. It’s an entry level scope, with 50 MHz bandwidth, 2 channels and 2.5k point record length for both channels. It has 7” color screen, two USB-ports, front port for plugging in your USB-stick and another port in the back, for plugging into a PC. It is good enough for most of the stuff you need done at a “home-lab” and best of all really, really cheap, being sold at around 400 euros.
I guess it’s good to mention at this point, that I do have some experience in using an oscilloscope, but not too much. The scopes we have at the university are mainly manufactured by Agilent (currently known as Keysight), but all scopes are more or less the same when it comes to basic functionalities.
I’ve included some basic usage instructions, but I suggest reading Tektronix Oscilloscope fundamentals as a reference. It’s really well written and goes through using an oscilloscope very well. The same “manual” is used at my uni for teaching scope usage and is proven to be very good.
The scope is very, very light. I haven’t measured it, but my guess is that it’s under 2kg. Also, it’s quite small, which is very handy for desktop use and carrying with you. The kit didn’t include any kind of cover for the front panel or the scope itself, which is a bit of a disappointment and takes away some of the advantages of having a small, light-weight scope.
Once booted up you’re greeted with a usual looking screen, with a waveform display, trigger level, timing etc information. Screencap 1 shows the main screen.
It’s quite clear, with all the necessary information shown clearly. On the bottom is the status field, which shows “Please wait…” because it’s saving the image to the USB (more about this later). It also shows the date and time (once setup), which is quite handy for reference. I’m fairly sure I’ll use this in the future by writing some basic notes about every capture on a plain old paper…
Second row from the bottom shows all the essential information for the measurement, like voltage and time dividers (how much each grid slot represents) and the trigger info.
Starting from the third row there’s the selected measurements. These are quite handy, if the signal is quite constant and clear. I’ve ran into some signals already, where the scope doesn’t know how to handle them. Fortunately there is an indicator about this, the measurement is marked with a question sign as shown for the frequency in the image. This isn’t that reliable though, I measured a signal where I knew the main frequency but the scope was picking up some distortion, even though I had the trigger level correctly set.
After the initial tests, I spent some time just browsing around the menus and this was kind of a pain. The menus are quite slow. Of course it’s worth noting that most of the time you’ll be able to just use the physical buttons instead of menus. I knew already, that most oscilloscopes are upgradable, so I started to browse around and found that, indeed, the firmware is upgradable in this one also.
The firmware is downloadable from Tektronix website (link will probably expire at some point, when a new firmware is released, you can find the new version by using the site’s internal search) after you register to the site. The firmware upgrade is quite simple, just download the file (.TEK) to an USB-stick and plug it into the scope. After the stick is scanned, press “Utilities” button on the front panel and using the buttons next to the display, go to “File utilities” -> “Update Firmware”. The needed steps are shown in the next couple of images.
The scope finds the file automatically and starts upgrading after a while. Once the upgrade is finished, the display tells you to reboot the device. Also you can format the USB-stick from this menu, which I did, as my guess it’s best to use the format the scope wants, for maximum speeds etc.
I suggest you upgrade, as the menus do feel a bit snappier and it does seem to add some functionalities as well. Some of the functions are marked as incomplete at the moment of writing, so I’m sure there will be more upgrades in the future as well, hopefully with speed optimizations.
As I don’t have the scope next to my PC at the moment (and most likely never will), I’ll use the USB port in the front panel as my “data logging” device. As the main purpose of a scope is to see the signal, I (usually) don’t need the extra stuff the scope offers, such as saving the signal to a CSV file. Images are enough for me, but of course I see the use case for CSV as well, for example making some data analysis on a PC later.
I do however see the need for saving the capture settings to a file useful, as the signal position is set in divisions rather than voltage. I can tell you that scrolling to +5V with 20 mV division is a pain. I’d be a great improvement if changing scale would “zoom” to the current position rather than just some to the same division value, instead of using the division as reference effectively jumping back down. Any tips on usage are appreciated, as I haven’t found a way to get around this yet!
But enough of this “rant”, back to saving the images. The oscilloscope allows you to set the function of “Print” button quite easily. Just press “Save / Recall” on the front of the scope and you’re presented with the menu shown in the next image.
From this menu you can select what the print button does, I’ve selected “Save Image To File”. You can also select “Prints” (which I guess would print to a printer attached to USB, not sure what it’s needed nowadays?) and “Saves All To Files”. The last function is quite useful, as it saves also the settings you used to capture the image, along with CSV and image itself. The main reason I don’t use this at the moment is once again speed. It takes a just bit longer to save all the data instead of just the image. It’d be great to have multiple selection with image, CSV and settings instead of all or just the image.
At the moment, my settings are so that the image is captured by the print button and if I need the other stuff, I use the menu itself or change the setting to match the needs of project.
After I had setup all the settings as I want them, it was time to check the calibration. Of course the scope should be calibrated correctly, being that it’s straight from the factory, but you can’t ever be too sure.
At first I decided to check the probe calibration by using the probe compensation function found in any oscilloscope. As with all the others I’ve used, this is done by attaching the probe to the pins reserved for this purpose. More about calibration process can be found in the Tektronix manual posted in the beginning. As the next image shows, the square wave is quite clean, with just a minor anomaly in the signal and requires no calibration (at least for my use).
I also ran the self calibration function found somewhere in the menu, although I’m sure there wasn’t actually need for it.
Actually measuring something
So, once everything is setup, I finally got to the point of actually measuring something! I know it sounds quite crazy as a concept, but some of the measurement devices aren’t actually used just a paperweight. An oscilloscope is a perfect example of something you can use almost every day.
My current plans for the Raspberry Pi B+ and EnOcean receiver consist of supplying 12V voltage to the board. I’m planning on extending the Pi with a custom made PCB and it should include some way to get the voltage down to 5V for the Pi. At first I planned on using some spare regulators which can handle drops like this up to 3A, but decided it was too wasteful. After all, one of the points of using home automation is to save energy. I browsed through the Farnell website to find a ready made DC/DC converter small and cheap enough (under couple of euros) to use on the board with enough power and found nothing. Of course I found IC’s to use, but I don’t really want to spend the time designing and testing a circuit for powering the Pi, at least not yet.
After awhile, I remembered that I had these El Cheapo DC/DC converters based on TI’s LM2596 from China, which I ordered couple of years ago. They say the converters are powerful enough for up to 3A. I’ve used them to create an adjustable “lab power supply” for my workspace, with three outputs. I did some testing with them ages ago, and noticed that there is a clear switching frequency “peak” shown, as there is with basically every switching converter.
As the Pi “requires” 2A for operation, these should do, as long as the supply is stable enough for operation. After double-checking the pin header and voltage coming from the converter, I plugged it in to the Pi’s GPIO +5V and ground. And the Pi booted up, success!
The image below shows the Pi running idle, being powered by the converter, with the switching frequency clearly showing. The center line is set at 5V for the remainder of measurements, so you can refer to that.
The measurements show quite well that the mean voltage is 4.97V, with maximum at 5.00 and minimum at 4.95V. Peak to peak is measured at 46.4 mV. The buyer I bought the converters from, said that the switching frequency is 150 kHz, which is also shown in the image… However, this shows the unreliability of these automatic measurements, they must have measured the frequency similarly as well
Calculating from the divisions the timing is actually about 20 microseconds. This calculates to about 50 kHz using f=1t. My guess is that the chinese manufacturer or seller used an oscilloscope to test the performance as well, leading to the misinformation. I couldn’t get the trigger to work correctly and measure 50 kHz as frequency in this case, not sure of the reason at the moment.
As the measurements above were made running with the Pi running idle, it was time to put the DC/DC converter under some load and check if it the Pi would still be stable. So, I ran the commands found here and checked the results again. The CPU was run at 100%, with loads rising up to 4.xx. The Pi stayed stable, even though I forgot the script running for 8 hours...
The voltage drops a bit, but not enough for the Pi react by freezing or something like that. The Pi actually doesn’t seem to be affected by the full load in any way. I can still run other programs etc with it and don’t have any problems. This seems like a big improvement over the last Pi, which was quite picky about the power supply (actually I couldn’t get the old Pi running with similar setup).
The oscilloscope includes a lot more functions than I’ve used here, I might make another post about some of them but most probably will not. The functions I’ve described before are the ones I’ll most likely use the scope for and I’ll leave covering the rest for other contestants. There is some functions that I’ve found useful though and I’d like to briefly introduce them.
By pressing “Measure” button on the front panel, one can find an interesting function called “SnapShot”. This takes all the possible measurements by the scope and displays them as shown in the next image. The image is showing measurements from the Pi running at full load.
These kind of measurements are really useful, if you are for example comparing two different situations (like Pi running idle vs. full load). You don’t need to take multiple images, two is enough for a reference and comparison in the future. Of course, this requires that the measurements you take are accurate and representative of the signals you’re studying.
There’s also a help function in the TBS1052B, but it’s quite hard to use, as the instructions are displayed on the screen, effectively covering the function you’re trying to find the instructions for. It’s just easier to use Google or use the Tektronix manual referenced at the start of this document.
There’s a lot this little scope can do. Some of the measurements baffle me, even if I’m supposed to be studying electrical engineering. Part of the measurement selection menu is shown in the following image.
The only thing I have some problems with is adding and removing measurements. It’s quite irritating as the menu closes after you add or remove a measurement, this makes toggling multiple measurements quite hard and time consuming.
As the TBS1052B is an educational model, it of course includes some educational functions. These seem to consist of basic lessons of using the scope, with the addition of using examples you can upload to an Arduino or MSP430 Launchpad. The lessons seem to range from “very basic” to “needs a lot of external stuff” with nothing in between them.
For example one tutorial I checked out seemed to be mainly about measuring a square wave generated by an Arduino. This is basically the same thing as using the probe comp output of the oscilloscope and doesn’t really offer anything except the instructions of how to save the results etc. At the other end, there was an example which required a signal generator as an input, which are quite expensive and rare, even at an university.
I do realize that there is a need for devices designed for educational use, but I can’t really see the point of having part of the integrated memory (or even circuitry, not sure?) designated to that function. Most of the people who will ever wind up using oscilloscopes do have computers and can view or print a PDF with instructions. Also, the tutorials seem to be more about the external circuitry than using the scope itself. The tutorials also do a lot of the data capturing, setting up the scope etc without any user interference. In my view, a tutorial embedded in an oscilloscope should be more about using the scope rather than using the external devices.
In my true, honest opinion the TBS1052B is a great product, especially when you consider the price. To be honest, I didn’t know proper oscilloscopes existed in this price range. I’ve been looking into buying one for a few years now, but they’ve always seemed too expensive. There is some “pocket scopes” out there for a very low amount of money, but they’re hardly any use even for the most basic needs (I’ve bought one...). The TBS1052B has all the functions I’ve found a use for. This even includes the measurements I've needed during my studies so far, I’m sure that I could have passed all the lab exercises etc using this scope. The scopes we currently use at the university don't even have color displays, which is quite a pain when you're measuring multiple signals...
All in all, TBS1052B is great value for money. And as with wine, it probably will get better with age with many firmware upgrades to come.