It has long annoyed me that signature analyzer prices have really skyrocketed. I bought my first hp 5004a for $25 with a bad filter cap. I found an hp 5005b signature multimeter in a buck-a-pound box at a swap, so that was $12. Now, if you can find one, a 5004a is anywhere from 100-250, if you're lucky.
SAs are only useful for troubleshooting equipment that has signature data and that was only popular for a fairly short window of time before increasing integration made it impossible. So mostly 70s, maybe early 80s. Still, if you have to troubleshoot a piece of equipment and it has signature data, it can help track down problems very quickly. Part of the reason that prices have risen may be that some arcade game manufacturers used SA extensively to allow field techs to do component level repair.
There are a couple of current SAs available, but they are still a little pricey. So I decided to design and build a SA and release it so anybody could make one. BTW, Sigrok has a signature analysis plugin, but it has some limitations and I would not really depend on it.
I had a couple goals: cheap and easy to build for anyone. Although something like an fpga is a natural for this, not everyone can program one, so my implementation uses discreet logic with an Arduino (gasp) and a 4x20 LCD display. The displays with an I2C controller can be had for $6 or so, the Arduino Mega 2560 Pro is maybe $10, plus various bits and pieces depending on your junk box are another 10 or 20. Definitely cheap and the tool chain is free. I had initially planned on a nano, but the restricted internal RAM ended that. I use a makefile environment for the software end and avoid all that IDE crap.
Right now I have a breadboarded unit, but I will eventually get a PCB for it Everything, gerbers, datafiles, code, cad files, etc will be up on github.
This has been an on and off project over the past couple years (mostly off) but I did get back to it recently and today I had the first successful signatures. Here's a pic along with a hp5004a looking at the same data. There are still a few issues and I have yet to build the front end, but this is progress.
One thing I ran into is that the example test circuit shown in one of the HP app notes uses a degenerative situation where the stop and start lines are tied together. The HP5004a works fine with that because it uses a discrete logic state machine to control the acquisition. The Tek 308 handles things differently and it does not report the same signatures for that case. Since my design is closer to Tek's and I had made the assumption that they produced identical signatures in all situations, that had me stumped for a bit. I had to dig out my 308 and see how it behaved. The solution was to produce another test circuit that had separate start and stop signals and that worked correctly on both the 5004a and the 308 (and now mine).
Home
Help
Search
Login
Register
