Re: Tek 7854 intermittent issues
The Op manual p 9-43 only considers "premature termination" by the STOP
command.
After normal termination missing data are filled by linear interpolation.
(Though I think I also read filling with the last previous valid data value.)
After premature termination:
- AQR filles whole waveform with zeros
- AVR filles by interpolation (unless there were no valid data points at all).
Using AVR and nearly immediate a STOP I could generate a waveform with
interpolated sections.
Right - I don't think it knows if it's good data or bad data, only that it got filled up. I'm haven't figured out how yet, still making my way through the manual. I know it's also counting sweeps, but since it's sampling randomly, that wouldn't tell it if it was done or not. Is the CPU watching the address bus and keeping a tally of all the used locations? I still suspect that given the predictable spacing in the gaps of data he's getting, the issue is manifesting as an addressing problem during the DMA writes from the digitizer to the RAM.
|
Re: DEC PDP11/05 on a 202D cart - did one survive?
Hi Tim,
I have a P7001 Manual with dated MCI (06-17-80) for SN B100773 and above. Thus it seems thousands are sold.
Greetings,
Egge Siert
|
Re: Tek 7854 intermittent issues
On Fri, Sep 4, 2020 at 09:28 PM, Albert Otten wrote:
The Op manual p 9-43 only considers "premature termination" by the STOP
command.
After normal termination missing data are filled by linear interpolation.
(Though I think I also read filling with the last previous valid data value.)
After premature termination:
- AQR filles whole waveform with zeros
- AVR filles by interpolation (unless there were no valid data points at all).
Using AVR and nearly immediate a STOP I could generate a waveform with
interpolated sections.
Hi Albert, That's when the 'scope is healthy. In the video, it seems that acquisition proceeds to a certain stage, then sort of repeats, as can be seen in the periodicity of the pseudo-random garbling of the readout characters. This goes on for two minutes, when finally the AQR terminates. I really think PS voltages should be checked so we know whether we're chasing real ghosts. Raymond
|
Re: Tek 7854 intermittent issues
On Fri, Sep 4, 2020 at 06:00 PM, Raymond Domp Frank wrote:
@Raymond -
The scope decides the Acquisition is complete when 99% of of the memory is full. Thanks for that, Paul. Moving what I wrote earlier into this criterion: What
if the comparison is corrupted, IOW the acquisition thinks it's >= 99% done
but in reality, it isn't?
Raymond
Hi Raymond, The Op manual p 9-43 only considers "premature termination" by the STOP command. After normal termination missing data are filled by linear interpolation. (Though I think I also read filling with the last previous valid data value.) After premature termination: - AQR filles whole waveform with zeros - AVR filles by interpolation (unless there were no valid data points at all). Using AVR and nearly immediate a STOP I could generate a waveform with interpolated sections. Albert
|
Re: Tek 7854 intermittent issues
On 9/4/20 10:09 AM, Paul wrote: But yeah, check those supplies first It is best to give this step some time, doing it deliberately, since any accidental shorting out with probes can cost you lots of repair time. Each of the places just downstream of rectifiers and electrolytics needs checking for ripple or wrong DC voltage. Someone else mentioned there is a separate digital section supply to check.
|
Re: Tek 7854 intermittent issues
On Fri, Sep 4, 2020 at 08:47 AM, Nick Corvid wrote:
Raymond: You are correct, I have not gotten to check the power supply yet. For
one I've so far had some issues finding what exact values I'm looking for (ie.
acceptable ripple value) in my service manual and just haven't had time to
pull the power supply out without knowing what I'm looking for. Also, the
flare in the display (I think you were referring to my video? Or else I'm
imagining things) I didn't even notice that weird change in brightness until
you pointed it out, shouldn't have been caused by the camera anyway, all the
exposure settings were fixed.
Hi Nick, It's more than just a recommendation to check power supply (PSU) values first and make sure they are OK: You can expect virtually anything with bad power supplies, including intermittently different behavior. In my (not so humble) opinion, it makes no sense doing all the things you did, making videos, checking things etc. until you know that voltage levels are ok. The kind of problem in the PSU you're looking for is not a subtle few tens of mV above a limit or not. With a 'scope or at least a good DVM that measures AC while blocking DC, you'll see ripples of hundreds of mV at least, often many volts if a power supply voltage is bad. Until you've done that, all your other work is/may be futile. If the above sounds harsh, it's because I'm serious about it. Power supply problems are the most prominent problems in these old instruments, mostly caused by defective (electrolytic) capacitors. Re. the flare I was indeed referring to your video. It might be an interference between the frame rate of the video and the 'scope's CRT refresh rate or it's a real thing, in which case you'd also see it looking at the screen. If the latter, you (may) have a problem. Raymond
|
Re: Tek 7854 intermittent issues
On Thu, Sep 3, 2020 at 10:28 PM, Paul wrote:
@Raymond -
The scope decides the Acquisition is complete when 99% of of the memory is full.
Thanks for that, Paul. Moving what I wrote earlier into this criterion: What if the comparison is corrupted, IOW the acquisition thinks it's >= 99% done but in reality, it isn't? Raymond
|
Re: Tek 7854 intermittent issues
Nick -
The alignment pattern working would seem to indicate the CPU can read & write RAM just fine, nor are their issues from the CPU to the DACs. I'm pretty sure the digitizer writes directly to RAM, and since the pattern isn't dependent on the signal or sweep, I'm wondering those gaps are an addressing problem between the digitizer & the RAM?
But yeah, check those supplies first, and if it's still lurking, we can keep digging.
Cheers,
Paul
|
Re: Tek 7854 intermittent issues
Be aware that there is a secondary power supply for the digital stuff which lurks at the front of the scope. It's quite likely if there is a PSU problem that this is the one that is the culprit given you are having problems with the digital side of the 'scope.
David
toggle quoted message
Show quoted text
-----Original Message----- From: [email protected] [mailto: [email protected]] On Behalf Of Nick Corvid Sent: 04 September 2020 07:48 To: [email protected]Subject: Re: [TekScopes] Tek 7854 intermittent issues Raymond: You are correct, I have not gotten to check the power supply yet. For one I've so far had some issues finding what exact values I'm looking for (ie. acceptable ripple value) in my service manual and just haven't had time to pull the power supply out without knowing what I'm looking for. Also, the flare in the display (I think you were referring to my video? Or else I'm imagining things) I didn't even notice that weird change in brightness until you pointed it out, shouldn't have been caused by the camera anyway, all the exposure settings were fixed. Paul: I do get the same pattern (and similar size gaps) on pretty much any frequency (Tested from 5hz to 10Mhz) and at any sweep settings (though I haven't tested single shot acquisitions might need to get my 7B87) Also the alignment pattern appears fine both warm (powered on for a while) and cold (Off for more than 2 hours with the external ram supply active) /g/TekScopes/photo/252782/0?p=Created,,,20,2,0,0 Albert: Yeah, generally the calibrator is fine in analog. But as you noticed, and seems to be the case for any signal from any source, (until the issue disappears a while after it starts) they are recorded at half value. I'll try and get some more photos up demonstrating this. And unfortunately my proper DVM let the smoke out rather violently a few weeks ago, so I'm stuck with a multi-meter for measurements until I can source a reasonably priced replacement.
|
Re: Tek 7854 intermittent issues
Raymond: You are correct, I have not gotten to check the power supply yet. For one I've so far had some issues finding what exact values I'm looking for (ie. acceptable ripple value) in my service manual and just haven't had time to pull the power supply out without knowing what I'm looking for. Also, the flare in the display (I think you were referring to my video? Or else I'm imagining things) I didn't even notice that weird change in brightness until you pointed it out, shouldn't have been caused by the camera anyway, all the exposure settings were fixed. Paul: I do get the same pattern (and similar size gaps) on pretty much any frequency (Tested from 5hz to 10Mhz) and at any sweep settings (though I haven't tested single shot acquisitions might need to get my 7B87) Also the alignment pattern appears fine both warm (powered on for a while) and cold (Off for more than 2 hours with the external ram supply active) /g/TekScopes/photo/252782/0?p=Created,,,20,2,0,0 Albert: Yeah, generally the calibrator is fine in analog. But as you noticed, and seems to be the case for any signal from any source, (until the issue disappears a while after it starts) they are recorded at half value. I'll try and get some more photos up demonstrating this. And unfortunately my proper DVM let the smoke out rather violently a few weeks ago, so I'm stuck with a multi-meter for measurements until I can source a reasonably priced replacement.
|
Re: Tek 7854 intermittent issues
On Thu, Sep 3, 2020 at 08:38 PM, Chuck Harris wrote:
Sounds like you have a power supply ripple issue.
I never saw confirmed that you followed Dave Casey's, Chuck's and my suggestion that you may have a power supply issue. Check with DVM *and* with a 'scope. Raymond
|
Re: Tek 7854 intermittent issues
@Albert -
No big (or any?) changes to the sampling system between 'old' and 'new', so I am genuinely flummoxed about why I'm getting different results in terms of how fast a complete acquisition can happen at what P/W. It wouldn't surprise me though if the later firmware that I have is why it's at least failing gracefully instead of giving an outright error - I still get usable waveforms.
@Raymond -
The scope decides the Acquisition is complete when 99% of of the memory is full.
@Nick -
Do you get a similar pattern of filled vs non filled waveform memory at different signal frequencies? at different sweep settings?
I wouldn't rule out some of Nick's suggestions (Power supply or RAM). If you have the waveform calculator, you could punch in the test pattern generator program, which would probably show troublesome RAM.
In the first photo you posted _MG_2324.jpg, I'm seeing dots in that slope, which doesn't say to me that it's simply interpolating between two points, right? It can do that interpolation, but I think only when explicitly commanded by the waveform calculator. I believe those are captured points.
|
7854 front panel locking up
I'm slowly going through the issues with my 7854, at the moemnt I'm chasing a problem where, after being on for some time, the mode selection switches stop working, all I get is a beep from the speaker and the readout flickers whenever I press a button. From cold it seems to work OK, although I also have a fault with the readout where all the characters are squashed vertically aalmost to a line. The CPU is still running as the bus is active, I'm struggling to figure out how the mode select works as it appears to drive the data bus directly, and the beep seems to indicate that the processor is seeing the button presses. Any suggestions where to start faultfinding this?
|
Re: Tek 7854 intermittent issues
Sounds like you have a power supply ripple issue.
-Chuck Harris
Raymond Domp Frank wrote:
toggle quoted message
Show quoted text
On Thu, Sep 3, 2020 at 06:32 PM, Chuck Harris wrote:
Randomness doesn't exist in the digital logic world.
Some would argue randomness doesn't exist - full stop, nor does a digital world.
Thanks for that, Chuck. So the 7854 uses (pseudo) random sampling? If something is wrong with the pseudo random generator, i.e. repeating partial sequences a few times, then continuing, that could explain the repetition of the garbling that I thought I saw. And doesn't it show that AQR takes far too long and just repeats until finally meeting acquisition complete criteria or thinks it meets them and doesn't? That's not an acq. RAM or DAC problem. That looks more like a random address generator/acquisition-complete problem.
BTW, I just noticed a similar 3/sec or so slight brightness flare in the steady-state display. Video artifact or Z-modulation? Power supply?
Raymond
|
There¡¯s kind of a good write-up at:
Picked one of these units up new in the box about 12 or so years ago to play with. Didn¡¯t have a laptop with an external SCSI port so it got relinquished to the equipment shelves where it still gathers dust. Would like to try to get it going in my spare time but there are obviously other pressing issues to deal with.
Besides, it is a lot easier to fire up a recent model digital scope to use.
Greg
|
Re: Tek 7854 intermittent issues
On Thu, Sep 3, 2020 at 06:53 PM, Raymond Domp Frank wrote:
nor does a digital world.
... meant to say digital logic world, or digital logic, or digital electronics, anything you like, just pick at random. Raymond
|
Re: Tek 7854 intermittent issues
On Thu, Sep 3, 2020 at 06:32 PM, Chuck Harris wrote:
Randomness doesn't exist in the digital logic world.
Some would argue randomness doesn't exist - full stop, nor does a digital world. Thanks for that, Chuck. So the 7854 uses (pseudo) random sampling? If something is wrong with the pseudo random generator, i.e. repeating partial sequences a few times, then continuing, that could explain the repetition of the garbling that I thought I saw. And doesn't it show that AQR takes far too long and just repeats until finally meeting acquisition complete criteria or thinks it meets them and doesn't? That's not an acq. RAM or DAC problem. That looks more like a random address generator/acquisition-complete problem. BTW, I just noticed a similar 3/sec or so slight brightness flare in the steady-state display. Video artifact or Z-modulation? Power supply? Raymond
|
Re: Tek 7854 intermittent issues
Most humans are a lot more random than a PRNG! :)
David
toggle quoted message
Show quoted text
-----Original Message----- From: [email protected] [mailto: [email protected]] On Behalf Of Chuck Harris Sent: 03 September 2020 17:32 To: [email protected]Subject: Re: [TekScopes] Tek 7854 intermittent issues : : The randomness of the human operator relatively to the PN generator will make it appear truly random. -Chuck Harris
|
Re: Tek 7854 intermittent issues
Randomness doesn't exist in the digital logic world.
The scope knows how many pseudo random numbers it takes
before the pseudo random sequence repeats, and waits
that long...
From the operator's perspective, it is random, because
the operator has know idea where the scope will be in
the pseudo random sequence when it starts a new capture.
The randomness of the human operator relatively to the
PN generator will make it appear truly random.
-Chuck Harris
Raymond Domp Frank wrote:
toggle quoted message
Show quoted text
On Thu, Sep 3, 2020 at 08:33 AM, Nick Corvid wrote:
and since parts are missing it won't complete
How does the 'scope decide AQR is complete, if samples are taken randomly, out of order? Or are they not taken randomly, just at a slow pace, just at subsequent scans? I guess it's RTFM-time again for me.
Raymond
|
Re: Tek 7854 intermittent issues
On Thu, Sep 3, 2020 at 08:33 AM, Nick Corvid wrote:
and since parts are missing it won't complete
How does the 'scope decide AQR is complete, if samples are taken randomly, out of order? Or are they not taken randomly, just at a slow pace, just at subsequent scans? I guess it's RTFM-time again for me. Raymond
|
Raymond Domp Frank
David C. Partridge