12 comments

[ 1.6 ms ] story [ 35.1 ms ] thread
Hi, I am the author. This is a hobby project. I think it could be a useful tool for DIYers. Any feedback is much appreciated.
IMHO the naming isn't great, because it would be useful for plotting slow-ish sensor data, but I'd struggle calling something with only ~3kS/s an "oscilloscope"
Thank you for the feedback. I appreciate it.

I understand that it wouldn't compare with the actual oscilloscope, but this was never meant to be a replacement for oscilloscope. It's just a low cost way to convert available electronics into a portable handy scope for basic applications. It was very useful for me while I was designing a characterization board.

Sampling rate depends upon the analog to digital converter (ADC) used with the Raspberry Pi. ADS1015 has a relatively lower sampling rate of 3.3kS/s. However, I have also used MCP3002 with Raspberry Pi. It has a sampling rate of around 75KS/s for 2.7V (and 200KS/s for 5V), which gives a very good performance.

So, just with a change of ADC, one can achieve better performance. I will probably include support for more ADCs in the library soon.

You might call it a DAQ, since that's much more in line with what you're doing.
Nice project. Thanks for posting it. In my view, any sampling rate is useful for something, and lots of low speed phenomena are worth scoping.

While looking into various ADC's, I suggest also providing as an option, a reading from the built-in audio interface.

Looks useful for data logging and general experimental work in low speed scenarios. I'm currently using a TI 83 silver edition and a TI CBL and some Vernier sensors for this sort of stuff and it's a bit sucky but was essentially skip dived.

As a scope though, forget it. You need at least 1-2MHz BW for something usable for even audio applications. At least 20MHz is probably a good idea. I have a 20MHz analogue scope (LG) and it cost me £10 (~$15) which is less than the Pi.

Sweet and totally within the spirit of RaspberryPi. Thanks for releasing it.

PS. 3K sample/sec is good enough for a lot of work on sensors and also for the youngish to get a handle on how to do ADC and process it - who knows some young person may start of experimenting with something like this and go onto design something quite awesome in the future.

That's a nice starter project. There are lots of "turn your PC into an oscilloscope" USB devices, starting around $50 for a 200KHz sampling rate.[1] For $269, you can get a 2GHz sampling rate.[2] It's hard to justify getting a dedicated scope any more, unless you need something exotic.

[1] https://www.bananarobotics.com/shop/Educational-PC-Oscillosc... [2] https://www.sparkfun.com/products/9263

Oscilloscope #2 is $269 and has a 60 MHz bandwidth. How on earth is that good value?

I would like to recommend the Rigol DS1052E in that price range, or a used HP/Tektronix analog scope for the sub-$100 market.

I second this, as the proud owner of a Rigol 1102E (and several other scopes), it's amazing what $400 gets you there. A good scope is about so much more than just the bandwidth, it's about having a good front end with low noise, and things like AC coupling, and an easily readible screen. I think if I were to only be able to own one scope and do it on a budget, Rigol 1XX2E series is really the only way to go sub-$1000.
Exotic needs like listening to any major serial bus invented in the last 20 years or peeking at full-speed IO on a hobbyist ARM chip? 60MHz of analog bandwidth doesn't go very far these days.

It's an improvement, sure, and full scopes certainly don't compete on price. I'm thankful for their efforts. But somehow I can't help but feel that it should be possible to do significantly better.

In much the same vein, if you've got a little extra cash to cough up for a cheapo $60 scope (which can be pushed to 20ish MHz bandwidth, or perhaps better), I've managed to a lot of the similar stuff (in Python) on a Raspberry Pi with a library I rolled for the Hantek 6022BE: https://github.com/rpcope1/Hantek6022API