Re: ebay 7104 arrived :-)


Reginald Beardsley
 

Mel,

I've downloaded some of the PDFs, but not yet all. I have manuals for all my other gear. Many as PDF though I will probably get or print paper copies as that is much more convenient when working on the stuff.

What I *really* want is schematics and other component level repair data for the 11801. My understanding is that the museum has them, but the lawyers have not cleared them for release.

I'm collecting parts to allow me to measure bit skew for DSP pipelines on the Zynq and cyclone V FPGAs to create open source firmware for DSOs based on those chips. I've got four SD-26 heads so I just need to convert the differential 100 ohm high speed PMOD outputs on the Zybo Z7-20 to single ended 50 ohm. And find 8 cables which are close enough to the same lengths. I figure that my best shot is to buy 2 dozen SMA-M RG402 cables and pick the most closely matching pairs. Then use the delay adjustment for each head.

My goal is stackable math and filter pipelines a la LeCroy but with considerably more flexibility. Once I've written the FPGA code I'll start back to work on hijacking COTS DSOs from Siglent and GW Instek using the Zynq. I was making good progress on that until I accidentally fried the SPI flash chip while trying to sort out the factory programming header. I've not yet worked up the nerve to try to replace the chip. Hopefully I just fried that one cheap part and not the Zynq.

The motivation of this is the same as buying a 485 and a 7104. I do *not* like ringing on steps. As neither Keysight nor Rohde & Schwarz offer scopes which will do what I want even for $20K I decided to do it myself. Marketing wants to claim BW all the way to Nyquist and the unavoidable consequence is ringing on a step. For a decent edge you have to accept 50-70% of Nyquist as the -3dB corner.

I've never programmed an FPGA, but I spent my career in the oil industry doing DSP on general purpose computers. I did have a run in with an FPS-120B attached to an 11/780 once. So I'm familiar with the ugly side of DSP. And 20 years ago I wrote a parallel DSP code to run on an MPI cluster which required close to 2000 hours of work. I wrote *all* of it. So I do understand what I'm taking on.

Have Fun!
Reg

Join TekScopes@groups.io to automatically receive all group messages.