HP 54502A Digitizing Oscilloscope

[drakke]

It's a 400MHz, 2ch, 400 MHz repetitive bandwidth, 400 MSa/s sampler

Are these oscilloscopes worth anything for hobby work?
There is one locally for sale that is in good cosmetic and functional condition with a little screen burn.

Hard or expensive or repair? Any ASIC's or unobtainable parts?

Any advantages over a modern 100MHz beginner-level DSO?
What is a reasonable price for one?

[tom66]

I used to have a 54501A (100MHz with 20MSa/s ADCs.) The 54502A is similar.

Main issues:
- It's dog slow to do anything more than viewing waveforms. Turning on line mode makes the display update <5Hz. Dot mode is 60Hz so not bad.  Math functions are slow.  Autoscale is slow.

- Waveform memory is one screen worth, about 512 points, pretty limited.

- The calibration memory is backed by a lithium battery which tends to fail after 15-20 years. After this fails the oscilloscope will work, but will always display "calibration failed" until it is recal'd.

- My unit had a fault which caused the trace to drift up the screen over about 30 seconds, then the oscilloscope would reset itself or just spin the fan with a frozen/blank display. I traced it down to a cold solder joint on the -12V output of the PSU causing that rail to drop out under certain conditions. Not too bad after 20 years though.

- Pre trigger is very limited only about 4 wave memories (2k points); post trigger virtually unlimited though it tends to jitter badly beyond about 1000 memories post trigger.

- No knobs, all digital entry with one encoder. After a while though you get used to it. Maybe not as quick to adjust things but reasonably fast.

Advantages:
- Powers up in <2sec, CRT takes longer to warm up.

- When used like an analog scope is very good.

- I think the only custom IC is the timing/control IC, which is unlikely to fail (it's not on the analog front end); the front end on the 54501A is implemented with discrete parts and high speed DIP opamps from Linear Tech so likely to be repairable.

- It's built like a nuclear bunker and would probably survive the nuclear holocaust too. All the ROMs are socketed. The power supply is a switch mode power supply and is very well made.

[dom0]

DSO tech of the 80s is ... not competitive today (or yesterday, for that matter). The very high end stuff of the 90s is a different thing, though. Although I personally wouldn't buy the Teks of that time(*), because most of them have the plague and probably require replacing all capacitors in the entire device. And if something isn't working chances seem to be pretty good, that leaked electrolyte damaged PCBs and parts badly.

Note that LeCroy scopes also do use ASICs (A/D converter) and hybrids (AFE) to some degree, so if you see a LeCroy failing it's input or A/D converter self-test one of those is probably fried.

(*) https://www.eevblog.com/forum/testgear/advice-on-buying-first-bench-scope-need-to-know-whats-best-for-me!/msg610950/#msg610950

It at least got 400 MS/s real time, that's something, but as tom66 mentioned memory will be likely very limited, much like the analysis functions of that old scope.

[Smith]

The NVRAM is the weak point. Its good for 20 years, after that you get all sorts or errors beccause the internal battery is empty and loses the calibration. Replace it, and do an auto calibration and your fine for another 20 years. You can place a new battery, or replace the whole module for about 10-20 bucks.

I do like this scope. Yes its different but i think its in a good way. This one is much faster than the 54501A. Drawbacks are the 6 bit resolution in single shot mode, and the main powerswitch on the back. This could be a pain if you stack more gear.