Can't make Digital Scope work

[CoastalMaineBird]

But that chart is in ppm, not percent.

Even if the dirty filter raises the internal temperature from 25 to 65 - your chart says that's a change of -11 ppm.

That turns a 1,000,000 uSec period into a 999,989 uSec period - hardly noticable (in my application).

I think it's far more likely that the CPU is trying to protect itself by slowing down.

In any case, thanks again.

P.S.  After 30 minutes, operation is still perfect.

[Norbert_B]

Well, this error is for the base clock...then you have the divisor circuits which can introduce additional errors... what i try to say is that temperature has a big influence on oscillators which can add up to several % of "misreadings" (or in-accuracy if you prefer that term).

But it really looks as we pinpointed the source of the issue, regardless if the CPU or the 6602 or both are creating the issue 😄

Norbert

[johnsold]

Logic and counter circuits never go out of calibration by 15%. If they are that far off, the problem is a malfunction, not calibration. Especially when intermittent. The dirty fans and temperature dependence may be important clues.  Another is the age. Is the chassis the same age as the counter?

I have seen logic circuits do really strange things if the power supply voltage dips below the minimum for the devices.  A supply running at 4.78 V (just above the 4.75 V minimum for most logic), with some ripple or noise which causes the instantaneous voltage to drop to 4.6 V for 50-100 ns may cause one or two ICs to reset, miscount, or generate incorrect outputs - and then go back to normal operation until the next glitch on the power supply.  I once had a microcontroller which was resetting 120 times per second (2 times power line frequency) due to a failing filter capacitor in the power supply.

I have fixed dozens of aging instruments with random, non-reproducible, and often thermally correlated problems by replacing all the electrolytic capacitors in the power supply or replacing the entire supply.

Lynn

[CoastalMaineBird]

Agreed. Divisor circuits will not "introduce additional errors".

The chassis is the same age, 5-6 years, I don't recall exactly.

Given that cleaning the filters has fixed the problem, I'd have to say that it's not age. I think the CPU was sweating too much and slacking off.

[johnsold]

Darn lazy CPUs!

That is about the age when electrolytic capacitors begin to degrade, especially ones which run continuously or at high temperatures for extended periods.  The dirty filters aggravate the temperature problems, so cleaning them may only be a short term fix.

Lynn