Workbench Wednesday 18: Digilent Analog Discovery Review

Hi scottsto, interesting question. As my main electronics hobby, I also repair old consoles and computers. That said, I haven't used my Analog Discovery 2 yet for those repairs. However, I am working on a ZX81-related problem and just designed some PCBs to act as interposers to the Z80 and ULA, so I could connect a Logic Analyzer. (I was thinking of the digital channels on my oscilloscope at the time, but now I'm thinking about AD2 because I really do like the Waveforms software.)

That said, the documentation for the AD2 says it is compatible with 1.8-3.3 LVCMOS and 5V tolerant. Which means, the logic level is probably set around 1.17 volts, which is the lowest level for 1.8V logic. 1.17 is well below TTL's 2.0 volt level and well above the 0.4 volt for input-low. For reference, here is a chart from this TI application note on Logic Families.

In reality, I have rarely run into digital systems that were so marginal that they relied on the limits of the logic families. So I think the AD2 should be fine in logic mode. The nice thing about such a device is that if you suspect a digtial line is giving you problems, it is a quick re-connection to see the analog behavior.

In the video it is said that it is for 3.3v CMOS but is 5v tolerant. I mostly work on older electronics which are 5v TTL or 5v CMOS. (For example old video game console or pinball PCBs.) Is this logic analyser still a good choice? Are there any problems when using it on 5v devices? My ancient HP Agilent Keysight died, and I want to get something light, new and modern.

I haven't used the scripting possibility

Hi David,

I have used it. You can dowload the attached script on my RoadTest ( the link is in the suplemental content section in this page).

Thanks for putting a spotlight on this tool, James. I love my Analog Discovery 2!

I have used it many times for debugging I2C communication (I haven't used the scripting possibility, neither). When evaluating a new I2C sensor, I connect it to the AD2, use the internal power supply and run a protocol analyzer on data and clock pins. Quick and easy. Also the GUI on a PC is much more responsive than doing the same using knobs and buttons on a traditional oscilloscope.

I wish I had something like this when I was a student.

Good test report.

I had looked at the device, but decided not to buy it.

If I need a simple system like this in the future, I will reconsider.

DAB

Actually baldengineer I would love to do this but I dont have over $400 (Australian) to spend on the pro kit, otherwise I would happily order a kit and design and print an all in one case, not to mention I'm after an oscilloscope kit, as i'm currently designing a Multi Zone Amp Board for Volumio as in my Day job I do Multi Zone Audio systems for some of my clients and having to use a RPi per zone and Hi-fi HAT is annoying, My normal Job I do is I design and deploy, Security, Communication and AV systems for clients in the Domestic and Commercial sectors, but work volumes are low at the moment so unfortunately income has been tight.

I could already envisage a design, using a 5V 3A Mean well supply in the case.

From YouTube, Mohammad wrote:

Very useful instrument. I worked on a CAN bus for Automotives and needed to demonstrate the voltage quality as well as the digital messages on both CANL and CANH signals. Our Keysight oscilloscopes needed a differential probe to show both CANL and CANH. My Analog Discovery (AD )was able to show both signals on the 2 scope channels as well as the CAN messages on the logic analyzer

The differential analog inputs are something I very much glossed over in the video. Typically oscilloscopes are ground-referenced. As Mohammad mentions, this might mean you need to use a differential probe for some measurements. For example, if a signal-under-test itself is differential like for his CAN bus. Another case is measuring the voltage across a component like a resistor or a transistor. With a traditional passive probe or oscilloscope input, one side of the measurement would need to be grounded.

However, with differential input, no such reference is needed. Care does need to be taken to make sure the common-mode voltage is still within the measurement capability. So, for example, if you measured a current sense resistor on the high-side of a +12 volt rail, you have about 12 volts of common-mode in that measurement even though the difference (or differential) would probably be on the order of 100 millivolts or so.

In the case of the Analog Discovery 2, it appears the differential inputs can handle up to 50 volts of common mode. Maybe it is 100 volts? I'm not sure how to interpret the specifications. To be safe, I would recommend less than 50 volts. How do you know how much common-mode a signal has? One quick method is to measure each side of the desired component referenced to ground. In other words, measure point-A to ground and point-B to ground. Add the two voltages together and then divide by two.

Long story short, I should have made a bigger point about what "differential inputs" mean. Maybe single-ended vs differential measuerments is somethign we can cover in the future.

Interesting Idea celcius1. It would make a portable all-in-one bench.

Hmmm.

Due to ease of taking it apart, its hard to resist taking it out of its shell and making a custom all in one case with RPi 4 and Touchscreen