How about designing a quality scope front end?

[gb243]

With all these pocket scopes which are useless but popular Could someone with the approriate skills design an open hardware front end for such an animal. There is a lot you can do with a 5MHZ scope if it is accurate and really 5MHZ. After all that is all most people had for decades and adequate for most low frequency projects. With a solid front end the hobbyist can roll their own display and processor which these days is achievable. Designing a proper RF front end however is going to a step too far for many (at least for some time).

[HAL-42b]

So what is the upper limit with off the shelf components? 100MHz? 200MHz?

[ynfo]

This reference design from TI goes to 2GHz and has full schematics you could study or modify:

http://www.ti.com/tool/TIDA-00826

[HAL-42b]

This reference design from TI goes to 2GHz and has full schematics you could study or modify:

http://www.ti.com/tool/TIDA-00826

Brilliant!

It even has a BOM. Let's price this puppy and if we can get 4 per person and at what price.

[casinada]

That looks like a tease
They don't sell the board, just the core IC's
It is unusual for TI to do that

[Neilm]

The layout on the PCB would be the hard bit

[HAL-42b]

The layout on the PCB would be the hard bit

TI provides Gerbers and layout.  Not to say that we couldn't deviate from it if we needed to.

[julius oe.]

A pocket scope with 200MHz, 2 probes, log- function and a few clever auto-measure mode... sounds like fun !

[HAL-42b]

I created a Google spreadsheet with the TI 2 GHz frontend BOM. If we can price the components we can at least get an idea of what the final price might be.

You are welcome to join filling in the blanks.

https://docs.google.com/spreadsheets/d/1VqT09qRTmsgxbee6AHZyQpht7ohlWAWpZSCiMKgMl8c/edit?usp=sharing

[Lukas]

Some random links:

http://rigol.codenaschen.de/index.php/Schematics
https://www.mikrocontroller.net/attachment/99053/01-CH1_Input_Circuit.pdf
http://www.ti.com/lit/ds/symlink/lmh6518.pdf Appendix a

[Keysight DanielBogdanoff]

It'll be pretty tough to break 200 MHz with off the shelf components (for now).

[Kleinstein]

The TI input circuit is for a 50 Ohms input, not the more usual 1 M||25 pF found in normal low frequency scopes. Still useful but a different thing.

There is a limited use of the high input impedance input circuit for frequencies much above 100 MHz.

So one might consider a separate high frequency true 50 Ohms input, not just a switch to add 50 Ohms termination.

[xtoffer]

It'll be pretty tough to break 200 MHz with off the shelf components (for now).

I'm curious, where's the main bottleneck of off-the-shelf components? The above discussion seems to suggest that the analog frontend is not. If not there then handling of the incoming data, processing or display? Does FPGAs count as off the shelf? Surely a well picked and (not so off-the-shelf) programmed could? I'm not very knowledgable in the construction but very curious to know.

[chris_leyson]

I often wonder why nobody bothers to design under-sampling scopes. I have an HP54610B with 700MHz frontend followed by a 20MHz 8-bit sampler, 700MHz isn't a typo it was calculated from the rise time. Sitting behind me, a TEK 7854 400MHz BW with 200kHz 10-bit sampler, actually it might be 400kHz sampler, anyway it's a bit random and drifts a lot. It really is a random sampler !!

For repepitive time domain signals they're both good scopes but only 8-bit or 10-bit accurate at best, no good for RF. I use a spectrum analyzer for RF not an 8-bit or 10-bit scope. For RF, down sample to base band and use a 16-bit ADC, simples.