开云体育

--- In TekScopes@..., machine guy wrote:
>
> GREAT NEWS.

:)

> Faulty power supplies can cause a variety of misleading symptoms but I don't think that is what is happening here.?? Double check the voltages on J119 and if they are reasonably close move on for now.?? Tek gives specs for an instrument that does laboratory quality measurements so small amounts of ripple are detracting.?? But you aren't looking for a small error.?? I suggest you find that first unless voltages and ripple are far out of spec and/or changing during operation.

I checked this out again, and it's well inside spec. The ripple is "ugly" though, e.g. there are sharp spikes at ~20/40kHz. Assuming I get it working, I'll be planning to do the recap.

One surprising thing I found on a closer look; after I acknowledge the initial fault, something starts injecting about 40mV of a ~5MHz clock signal onto my 10V reference. Thankfully this doesn't propagate to the other rails though.

>
> It sounds like you may have a control board (A5) issue.?? Since the case is open you can easily do a visual check of the A5 board.?? Look for leaking or bulging caps and check the date on the memory chip and its battery.?? The 246x scopes (except for the 2465 and 2445) used a battery backed ram to save memory contents.?? These batteries are known to fail after many years and must be replaced before all memory is lost.?? There is a procedure to do this replacement.?? Its necessary to provide temporary power to the memory chip while you remove the battery and connect a new one.?? Even a momentary loss of power to the chip will lose all contents and a full calibration will be necessary.??
>
Interesting, thanks.
I'm not seeing anything obvious to my eye, though I should probably take a harder look.

> A failing A5 or its memory could explain all the problems you are seeing.?? I suggest you??focus on that for now.??
>

Thanks, this looks promising (not to mention fun:).
One thing I forgot to mention is that the scope doesn't trigger on the calibration signal, although the signal itself is fine viewed on my other scope.

Staring at the circuit diagram, I figured many of the symptoms could be explained by a DAC or analog MUX failure. It looks as though the DAC/MUX is to the center of e.g. reading all analog signals, setting trigger levels, etc.
Working through the A5 test points, I find the CPU clock and the interrupt signals are just fine, as expected.

The DAC-derived reference(?) voltages at TP2420 and TP2421 are way off, however, so I've something to look into there.

I have to admit that I can't for the life of me figure out where the -1.25V reference comes from, though. It appears to be under CPU control through the DAC/MUX, but it's also used to derive the +1.36V reference, which in turn is the centerpoint of the DAC's I->V conversion resistors (I'm a software guy, this is all alien territory to me).

Anyways, I guess working forward from the DAC's reference is something to do :).