Duct-taping DSO features onto a CRO

[ramenbytes]

Hello, I came across a hackaday project that augments CROs with a webcam and some python code. While it can't replace a DSO, it looks useful for those who want to squeeze some more mileage out of their CROs before getting a DSO. Among other things, it can digitize the waveforms so that they can be analyzed, for example with FFT. It also has a single shot feature, which lets you prime the software to poll the oscillscope camera and capture the next suitable wave form it sees. Hopefully it's useful to some of you, I plan on using it myself later on if I need basic storage functionality for my CRO.

Hackaday page:
https://hackaday.io/project/171961-a-digital-camera-for-analog-oscilloscopes

The author's blog posts, to save clicking through the hackaday page:
https://josepheoff.github.io/posts/d43toc

[dzseki]

Interesting. Although most CRO have X and Y outputs as well (eg for plotting), it would have been nice to combine at least the X output with some Arduino or the like in order to get sweep speed data at least so you wouldn't have to set twice.

[ramenbytes]

Interesting. Although most CRO have X and Y outputs as well (eg for plotting), it would have been nice to combine at least the X output with some Arduino or the like in order to get sweep speed data at least so you wouldn't have to set twice.

Really? While I haven't seen many scopes yet, most of the ones I see appear to have X-Y inputs, but not outputs. Did you mean that the project author could have used a microcontroller to provide a sweeping input to the X channel, and then feed the signal of interest to the Y channel, thereby effectively replacing the sweep controls with the Arduino? Even if that it isn't what you meant, it still seems like an interesting idea to me.

[dzseki]

Interesting. Although most CRO have X and Y outputs as well (eg for plotting), it would have been nice to combine at least the X output with some Arduino or the like in order to get sweep speed data at least so you wouldn't have to set twice.

Really? While I haven't seen many scopes yet, most of the ones I see appear to have X-Y inputs, but not outputs. Did you mean that the project author could have used a microcontroller to provide a sweeping input to the X channel, and then feed the signal of interest to the Y channel, thereby effectively replacing the sweep controls with the Arduino? Even if that it isn't what you meant, it still seems like an interesting idea to me.

I really mean X output here, or gate output, so the microcontroller can measure the actual sweep speed.

[ramenbytes]

Interesting. Although most CRO have X and Y outputs as well (eg for plotting), it would have been nice to combine at least the X output with some Arduino or the like in order to get sweep speed data at least so you wouldn't have to set twice.

Really? While I haven't seen many scopes yet, most of the ones I see appear to have X-Y inputs, but not outputs. Did you mean that the project author could have used a microcontroller to provide a sweeping input to the X channel, and then feed the signal of interest to the Y channel, thereby effectively replacing the sweep controls with the Arduino? Even if that it isn't what you meant, it still seems like an interesting idea to me.

I really mean X output here, or gate output, so the microcontroller can measure the actual sweep speed.

Ah, ok. Yeah, that would be handy. It'd be cool to try that with something like a Tek 465 which has something like that, if I recall. I've also wondered about the feasibility of digitizing and recording the output.

[dzseki]

I've also wondered about the feasibility of digitizing and recording the output.

The HP 1980A/B was the pinnacle of such instrument, but as would expect its "digital" features are rather weak by today's standards, still a very complex instrument.

[Berni]

They used to make digitizer modules for analog scopes, such as this:
https://w140.com/tekwiki/wiki/5D10

Some analog scopes have this functionality built in too. They look like analog scopes with analog dials and all, but they have some digital brains in there that gives you some digital functionality like on screen cursors or capturing a waveform and downloading it digitally over GPIB. The reason for this hybrid analog+digital aproach was that digital electronics and data converters ware not fast enough to work with these fast high bandwidth analog signals that analog scopes could show.

So these scopes combined the best parts of analog and digital together. The cursors got drawn to the screen as a "3rd channel" operating in X Y mode by a CPU. The capture functionality typically only worked at very low sample rates like 0.1MSPS because that's what the common ADCs of the time could do. Or the scope would have a CCD chip that works as a analog shift register to record the waveform as a series of analog buckets of charge that then get cycled around to be read by the ADC at a slower rate, or shown on the screen. This was a big thing because it gave single shot recording functionality.

With a webcam you won't be able to get this single shot capture ability out of a analog scope because the image will be likely way too dim at any sort of high sweep speeds. While at slow sweep speeds you could use a MCU with a built in ADC to digitize it.

Electronics have progressed a lot since the 80s, so today such a digital plugin could be made for probably 50 to 100$ in as a small 5x5cm board holding a fast ADC/DAC, FPGA, MCU that you can wire in between the analog channels and the CRT electronics to capture signals and draw then to the screen. But its easier to just buy a digital scope these days.

[ramenbytes]

I've also wondered about the feasibility of digitizing and recording the output.

The HP 1980A/B was the pinnacle of such instrument, but as would expect its "digital" features are rather weak by today's standards, still a very complex instrument.

I'll have to look into it, thanks!

[ramenbytes]

They used to make digitizer modules for analog scopes, such as this:
https://w140.com/tekwiki/wiki/5D10

Some analog scopes have this functionality built in too. They look like analog scopes with analog dials and all, but they have some digital brains in there that gives you some digital functionality like on screen cursors or capturing a waveform and downloading it digitally over GPIB. The reason for this hybrid analog+digital aproach was that digital electronics and data converters ware not fast enough to work with these fast high bandwidth analog signals that analog scopes could show.

So these scopes combined the best parts of analog and digital together. The cursors got drawn to the screen as a "3rd channel" operating in X Y mode by a CPU. The capture functionality typically only worked at very low sample rates like 0.1MSPS because that's what the common ADCs of the time could do. Or the scope would have a CCD chip that works as a analog shift register to record the waveform as a series of analog buckets of charge that then get cycled around to be read by the ADC at a slower rate, or shown on the screen. This was a big thing because it gave single shot recording functionality.

With a webcam you won't be able to get this single shot capture ability out of a analog scope because the image will be likely way too dim at any sort of high sweep speeds. While at slow sweep speeds you could use a MCU with a built in ADC to digitize it.

Electronics have progressed a lot since the 80s, so today such a digital plugin could be made for probably 50 to 100$ in as a small 5x5cm board holding a fast ADC/DAC, FPGA, MCU that you can wire in between the analog channels and the CRT electronics to capture signals and draw then to the screen. But its easier to just buy a digital scope these days.

Ah right, I've heard people rave about all the modules available for those scopes. That bucket brigade mechanism sounds pretty cool too, it reminds me of the guitar pedals with similar circuits.

Electronics have progressed a lot since the 80s, so today such a digital plugin could be made for probably 50 to 100$ in as a small 5x5cm board holding a fast ADC/DAC, FPGA, MCU that you can wire in between the analog channels and the CRT electronics to capture signals and draw then to the screen. But its easier to just buy a digital scope these days.

Where's the fun in buying an off the shelf solution though?    If I'm looking for something to do maybe I'll try mangling an old cheapo analog this way.

[Berni]

Where's the fun in buying an off the shelf solution though?    If I'm looking for something to do maybe I'll try mangling an old cheapo analog this way.

True that is the fun of DIY projects.

You probably won't be able to easily find the analog CCD shift register chips for this, they are pretty rare things that soon got obsolete once they developed faster ADCs/DACs.

For doing something similar today on the cheap you could probably do it all inside a high performance MCU. There are MCUs out there that have built in 16bit ADCs that can run at 10MSPS if you sacrifice a few bits of depth. So this would handle a few MHz. They also sometimes have a built in dual DAC so you could use that to draw on the screen in XY mode.

[ramenbytes]

You probably won't be able to easily find the analog CCD shift register chips for this, they are pretty rare things that soon got obsolete once they developed faster ADCs/DACs.

For doing something similar today on the cheap you could probably do it all inside a high performance MCU. There are MCUs out there that have built in 16bit ADCs that can run at 10MSPS if you sacrifice a few bits of depth. So this would handle a few MHz. They also sometimes have a built in dual DAC so you could use that to draw on the screen in XY mode.

Yeah, I guessed ADC/DAC would be the way to go. I wonder how much of the effort in making a DSO is in digitizing and everything that follows, as opposed to actual hardware stuff. If enough of the work lies in pre-digitization stuff, maybe modding old analogs with an MCU or something would be a viable half-way point for an open source DSO. Just commandeer some hardware for your DIY'd "brain" and go.