Datron 1281 - Fails Full Test with unknown error (?)

[Rax]

Hi all,
A 1281 has landed on my bench due to some unlikely odds, and I'm seeing it pretty drifty on 10V (pending further tests, so please hold off on filing that), but a more distinct error is "2295" upon running the Full Test.

I can't find this error anywhere in the Manuals. The 2200s go as high as 2293 but not higher, as far as I can tell.

Has anyone seen this? Maybe one of the current owners is able to shed some light on this.

Thank you very much in advance.

[Mickle T.]

I did not find a direct link to error code 2295 in the look-up tables in the firmware source code. I'll try to assume that this error is some kind of "unexpected exception".

[Rax]

I did not find a direct link to error code 2295 in the look-up tables in the firmware source code. I'll try to assume that this error is some kind of "unexpected exception".

Didn't think that's even possible...
I'm wondering:

• If maybe they added some error codes to a very recent version of the firmware (and maybe service manual, via an addendum, for instance).
• If the firmware in memory is corrupt and a truncated code is being thrown out.

The second above is not very likely, because the instrument is working reasonably well other than that error.

In any case, the error # suggests a "DC Voltage Test" fault. If I'm going by the nearest code (2293), it seems to indicate drift. Given this was due for calibration in 2002 (in software) or 2006 (by sticker - - it may have accrued plenty of drift. Why the software reports an earlier date than the sticker is a mystery in its own right. Maybe the last cal was not registered in software? That seems highly suspicious to me.

I'm currently "baking" this out for a bit and will see if the error persists after, say, 48 or 72hrs.

Per xdevs article - https://xdevs.com/fix/d1281/ - a recap fixed some issues with two units, which is obviously a pretty clear no brainer given the age of the unit, so that's something else I may want to try. I wish it was as easy to work on as the Prema 6048, but the boards are sandwiched in parallel so they need to come off to be worked on.

[guenthert]

[..]
In any case, the error # suggests a "DC Voltage Test" fault. If I'm going by the nearest code (2293), it seems to indicate drift. Given this was due for calibration in 2002 (in software) or 2006 (by sticker - - it may have accrued plenty of drift.

It should not have drifted more than the annual specification allows times the number of years since last calibration (well, at least, if it were still under some service contract), but likely will drift much less.  In any case, I doubt that it would drift so far as to trigger a self test error.  My 1271 hasn't been calibrated or adjusted in thirty years and passes just fine.

Why the software reports an earlier date than the sticker is a mystery in its own right. Maybe the last cal was not registered in software? That seems highly suspicious to me.

Or the last (and previous) calibrations didn't involve adjustments.  One might prefer not to disturb such a precise instrument unless absolutely necessary, e.g. when drifted out of spec.

I'm currently "baking" this out for a bit and will see if the error persists after, say, 48 or 72hrs.

Per xdevs article - https://xdevs.com/fix/d1281/ - a recap fixed some issues with two units, which is obviously a pretty clear no brainer given the age of the unit, so that's something else I may want to try. I wish it was as easy to work on as the Prema 6048, but the boards are sandwiched in parallel so they need to come off to be worked on.

Dunno. TiN replaces caps early and often.  I think that mentality is born out of experience with the capacitor plague affecting many computers of the turn of the millennium.  Older and younger caps are of better quality.  I'd measure first and replace only when necessary.

[Rax]

It should not have drifted more than the annual specification allows times the number of years since last calibration (well, at least, if it were still under some service contract), but likely will drift much less.  In any case, I doubt that it would drift so far as to trigger a self test error.  My 1271 hasn't been calibrated or adjusted in thirty years and passes just fine.

I am seeing this indicating about 6-7ppm low of 10V.

Or the last (and previous) calibrations didn't involve adjustments.  One might prefer not to disturb such a precise instrument unless absolutely necessary, e.g. when drifted out of spec.

After printing out the SM and doing some reading, it looks like the error 2295 seems to indicate:

• 2 - always used
• 2 - (positive) gain measurement
• 9 - "1V Range - +1V Reference Checks"
• 5 - this seems to indicate "both the measurement number and the error number, of which several can be allocated within each step. Each error number is defined only for its own measurement" (SM Vol.1 p.2-14) which I do not understand.

I haven't gotten a chance to do anything hands on during the week, just some research in the manuals. I measured some of the rails last weekend:

• TP301: +15.230V
• TP304: -35.913V
• TP305: +4.9458V

Which all seem close enough to me to not be an issue.

I should also say that, except for this error 2295, the instrument seems to be working reasonably well and not be off by too much. That said, if SELFCAL is anywhere as critical to accuracy as ACAL is on a 3458A, the fact that it fails (due to that error) seems like something that absolutely needs to be addressed. It also fails the full self test (due to the same error), but passes the fast one.

[Kleinstein]

The error numbers are specific to the test points measured. So no real system there.
From the error number I would suspect something related to the internal CAL source, especially the +-1 V derived from the transformer. It is still a bit unclear what is wrong, as there should be a documented error code for a more gross error, like a fully missing voltage.

I would suspect an additional check added later on.  A crude guess would be a check of the 1 V also in the 10 V range. This would at least make sense and I have not found it in the list so far. This way a check if the ref. transformer was drifting, though it could also be an ADC linearity problem. The interpretation of the self tests is always tricky as there are multiple parts involved.

One may find out roughly what is measured by using a scope (at low speed) to probe the amplifier output and maybe the bootstrapped "ground" to the amplifier to see what voltages are actually measured in the self test.