7T11 trace shifting left as time/div rotated.
|
In equivalent time, sequential mode, the trace is shifting left as I rotate the time/div knob.
All measurements were made on the 50nS sweep range with the timebase triggered from Pulse Out. Looking at the signal at TP286 which is the output of the TTH circuit, I see: Time/Div Start V Stop V Delta V 5nS -2.22V -7.82V -5.6V 2nS -2.22V -4.62V -2.3V 1nS -2.22V -3.5V -1.2V .5nS -2.22V -2.76V -560mV .2nS -2.22V -2.43mV -230mV .1nS -2.22V -2.34mV -120mV 50pS -2.22V -2.62mV -60mV (approx values) 20pS -2.22V -2.25V -30mV 10pS -2.22V -2.23V -10mV Looking at the output of U512A (TP512): Time/Div Start V Stop V Delta V 5nS -0.1V +5.5V 5.6V 2nS -0.3V +5.3V 5.6V 1nS -0.6V +5.0V 5.6V .5nS -900mV +470mV 560mV .2nS -210mV +350mV 560mV .1nS -420mV +140mV 560mV 50pS -70mV -10mV 60mV (approx values) 20pS -170mV -110mV 60mV 10pS -340mV -280mV 60mV I don't think the ramp start voltage should be moving down like that. Can anyone confirm my suspicions and suggest what the fault might be? Thanks David Partridge |
|
Joe,
toggle quoted message
Show quoted text
Thanks for the suggestion, but I'd already found Albert's posts earlier in the day, and it really doesn't match - here we're talking about the trace shifting off screen to the left not the right. C237, C292, and C313 seem to work as expected when running through the calibration procedure. Thanks Dave -----Original Message-----
From: TekScopes@... [mailto:TekScopes@...] On Behalf Of Albert Sent: 21 January 2012 19:37 To: TekScopes@... Subject: [TekScopes] Re: 7T11 trace shifting left as time/div rotated. Joe, Perhaps a good moment to tell whether your 7T11 indeed had a trim cap problem? Or was it another fault that would be good to know about? Albert
------------------------------------ Yahoo! Groups Links |
|
Albert,
toggle quoted message
Show quoted text
Thanks - you are right there are some calculation errors there - all my fault - no hardware to blame - apart from the grey mushy sort between my ears! Thinking rather more about this, the start voltage of -2.22V for the TTH Miller circuit doesn't sound right - from all my reading in the manual I think it should be starting at 0V, and the screen shots in the manual also suggest this. The waveforms shown for the Horiz. Amp. also suggest that there's no (or very little) DC offset. Comments? Thanks again Dave -----Original Message-----
From: TekScopes@... [mailto:TekScopes@...] On Behalf Of Albert Sent: 21 January 2012 23:06 To: TekScopes@... Subject: [TekScopes] Re: 7T11 trace shifting left as time/div rotated. Hi David, It takes some time to remember what all those circuits in the 7T11 are doing! The shift in output of U512A in a 5-2-1 sequence corresponds nicely (as should be) to the increase in feedback resistors 10k-25k-50k, ratio 2:3. At 1 ns/div the shift is about 0.5 V and that same shift will appear at the input R647 to the Dot Position Memory. This memory should detect that systematic error and send a correction to the Slow Ramp Inverter. That in turn would (finally) result in a shift of the THT output range to correct the error. This doesn't seem to happen, or not not enough. At the faster sweep speeds U512B and D get switched in and introduce their own offsets. Again the Dot Pos Mem should correct for this. Maybe you can check its operation/calibration. I remember it's not an easy calibration and a lot of compromise is needed. Albert (BTW your packet calculator needs some attention as well...) |
|
I also note at least one transcription error!! The Start V for .5nS in the second half should of course be 90mV.
toggle quoted message
Show quoted text
Very well spotted on the systematic nature of the offset BTW. I'd been banging my head on the wall (so to speak) with this problem for so long that I completely missed that! Regards, David Partridge -----Original Message-----
From: TekScopes@... [mailto:TekScopes@...] On Behalf Of Albert Sent: 21 January 2012 23:06 To: TekScopes@... Subject: [TekScopes] Re: 7T11 trace shifting left as time/div rotated. Hi David, It takes some time to remember what all those circuits in the 7T11 are doing! The shift in output of U512A in a 5-2-1 sequence corresponds nicely (as should be) to the increase in feedback resistors 10k-25k-50k, ratio 2:3. At 1 ns/div the shift is about 0.5 V and that same shift will appear at the input R647 to the Dot Position Memory. This memory should detect that systematic error and send a correction to the Slow Ramp Inverter. That in turn would (finally) result in a shift of the THT output range to correct the error. This doesn't seem to happen, or not not enough. At the faster sweep speeds U512B and D get switched in and introduce their own offsets. Again the Dot Pos Mem should correct for this. Maybe you can check its operation/calibration. I remember it's not an easy calibration and a lot of compromise is needed. Albert (BTW your packet calculator needs some attention as well...) In equivalent time, sequential mode, the trace is shifting left as I rotate the time/div knob. ------------------------------------ Yahoo! Groups Links |
|
Yes, the measurements ware taken with both the coarse and fine "time position" controls fully clockwise, but that gives me something extra to check on tomorrow.
toggle quoted message
Show quoted text
Regards, David Partridge -----Original Message-----
From: TekScopes@... [mailto:TekScopes@...] On Behalf Of Albert Sent: 22 January 2012 16:33 To: TekScopes@... Subject: [TekScopes] Re: 7T11 trace shifting left as time/div rotated. Hi David, The start and stop levels should differ about 5 V. With start I (and you also I think) actually mean the stop level when the trace is at the left position. On the test scope you look at vertical position of the horizontal lines. It depends of Time Position. The graph 11 is for fully CW. If you rotate to fully CCW then the whole patter shift 5 V lower, from -5 V to -10 V. Albert
------------------------------------ Yahoo! Groups Links |
|
Well - not quite fully clockwise, but almost.
toggle quoted message
Show quoted text
Now to the chase - I gave up on the calibration instructions on page 5-33 of the 7T11A manual, and went back to the original instructions in the the 7T11 manual on pages 5-10 and 5-11 - these worked a lot better for me. Now the trace is pretty much where it should be at most time/div settings and only needs a minor tweak of the FINE time position control to get a trace on screen The only problem I'm have is that all the adjustments seem to interact horribly - e.g. front panel Position and Sweep Cal are both affected by the Servo Zero and Slewing Ramp adjustments and also affect those internal adjustments - yack. Thye net result is that now the trace is pretty much where it should be at most time/div settings and only needs a minor tweak of the FINE time position control to get a trace on screen, but it's still moving around somewhat as time/div is turned :-( I don't understand how the TTH sweep voltage start point is affected by the front panel Time Position controls as they connect into the circuitry around U572D on the Analog Logic board. When Q294 in the TTH Miller circuit conducts, my dumb logic says that the TTH ramp voltage will reset to zero, but clearly I would seem to be missing something crucial, because adjusting the Time Position controls DOES appear to change the start voltage for the TTH ramp as measured at TP286, but I can't work out how :-( Regards, David Partridge -----Original Message-----
From: TekScopes@... [mailto:TekScopes@...] On Behalf Of David C. Partridge Sent: 22 January 2012 19:18 To: TekScopes@... Subject: RE: [TekScopes] Re: 7T11 trace shifting left as time/div rotated. Yes, the measurements ware taken with both the coarse and fine "time position" controls fully clockwise, but that gives me something extra to check on tomorrow. Regards, David Partridge -----Original Message----- From: TekScopes@... [mailto:TekScopes@...] On Behalf Of Albert Sent: 22 January 2012 16:33 To: TekScopes@... Subject: [TekScopes] Re: 7T11 trace shifting left as time/div rotated. Hi David, The start and stop levels should differ about 5 V. With start I (and you also I think) actually mean the stop level when the trace is at the left position. On the test scope you look at vertical position of the horizontal lines. It depends of Time Position. The graph 11 is for fully CW. If you rotate to fully CCW then the whole patter shift 5 V lower, from -5 V to -10 V. Albert
------------------------------------ Yahoo! Groups Links ------------------------------------ Yahoo! Groups Links |
|
Albert,
toggle quoted message
Show quoted text
I can only think I'm not interpreting what I'm seeing on the 'scope screen correctly. I'm probing TP286 (output of TTH) with Sweep Range set to 50nS, 5nS/div in sequential mode. I have the viewing timebase set for 5uS/div. The signal I see at high scan rates is similar to waveform capture 11 in the 7T11 and 7T11A manuals except that the top of the bright zone starts ABOVE 0V, and the bottom of it is about 5.6V below there. The picture of this signal manuals however clearly has the top of this signal at 0V with nothing above I can't reconcile what I see with the descriptions in the manual that say the TTH ramp starts at 0V and goes more negative until stopped by the slewing ramp comparator. How can the TTH ramp voltage ever be +ve in sequential mode???? I could possibly understand the bright zone moving down as you turn the Time Position controls CCW if the individual trace line in the brightened zone represents the value of the TTH ramp after it has been stopped by the slewing ramp comparator for this sample (I now think this is the case). But I cannot see how it can end up above zero! If you turn the Scan control fully CCW, you can actually see the voltage ramp down and then jump back up. D. -----Original Message-----
From: TekScopes@... [mailto:TekScopes@...] On Behalf Of Albert Sent: 23 January 2012 19:27 To: TekScopes@... Subject: [TekScopes] Re: 7T11 trace shifting left as time/div rotated. Hi David, Our start and stop terminology is confusing since both are stop levels of the THT. The "start" is the stop level corresponding to the left most dot at the trace (start of the trace), the "stop" is the stop level corresponding to the right most dot of the trace (end of the trace). The start level of the Miller circuit is always the same, like you say. About 0 V. But the stop level for first dot of the trace varies from 0 V to -5 V if you rotate Position from fully CW to fully CCW. At the 50 ns range, the stop level of the last dot in the trace is 5 V more negative (at 5 ns/div), 2 V more negative (at 2 ns/div) and so on. The Slow Ramp Inverter (which is responsible for the level shifts) sends an opposite level change to the horizontal amplifier input to cancel the shift. Also at faster time/div the horizontal gain is increased to cancel the decrease in TTH level differences. At CCW position all curves at the TTH test point have the ramp from 0 V to -5 V in common. Hence you will see this ramp very bright. Albert
------------------------------------ Yahoo! Groups Links |
|
Yes, a slight inbalance I could understand - but +2.2V is a LOT.
toggle quoted message
Show quoted text
Yes you can see the ramp very nicely at 100nS/div. But regardless of the sweep speed there's a very clear problem here. As the ramp stop voltage increases (more negative), the ramp start voltage moves from close to 0V until it reaches about +2.2V when the stabilised stop voltage (after the overshoot) gets to about 4V. I've uploaded a 3 second video clip here: <> Time to pull Q280 and put it on the curve tracer I think. 151-1011-00 seems to cross to dual 2N3822 or 2N5565, neither of which seem readily available - any suggestions on a suitable replacement dual FET? Thanks Dave -----Original Message-----
From: TekScopes@... [mailto:TekScopes@...] On Behalf Of Albert Sent: 23 January 2012 22:24 To: TekScopes@... Subject: [TekScopes] Re: 7T11 trace shifting left as time/div rotated. Hi David, At the 50 ns range, the slope of the TTH output is 5 V/ 50 ns. You have to set your viewing speed to 10 or 20 ns/div to see this nicely. At 5 us/div all you see is the stabilizing "DC" TTH output after the stop signal was received. Also the horizontal amplifier needs (long) time to stabilize. Just after the hor. output has been sampled, the TTH returns to zero. In the wait state Q294 conducts and ideally forms a shortcut. Then all depends on how good Q280A and B are balanced. The output at TP286 will differ from zero by the same amount as the "offset" A vs B. A slight difference from 0 V is no problem at all, can be corrected further on by some zero-adj. A large difference could be due to a damaged Q280. A not ideal Q294 would give negative TTH output, not positive. Albert |
|
Well now, that was interesting, swapping Q280 with Q452 (same type) initially appeared to make things slightly worse, but when I pressed on the top of the FET to straighten it up to get at one of the leads with a probe, the problem disappeared! I swapped the parts back to their original positions, and it still works as I expect (that is the voltage between T=24uS and T=34uS is now sat solidly at 0V).
toggle quoted message
Show quoted text
I can only conclude that one of the contacts in the socket had got resistive, and that removing and replacing it has fixed that (or there's a dry joint - I will check that). Either way, I am very pleased that it works correctly now, but annoyed I could not work out what was wrong. Now back to the calibration exercise which should now be a little easier (I hope). Regards, David Partridge -----Original Message-----
From: TekScopes@... [mailto:TekScopes@...] On Behalf Of Albert Sent: 24 January 2012 13:54 To: TekScopes@... Subject: [TekScopes] Re: 7T11 trace shifting left as time/div rotated. Hi David, The half period oscillation is normal. I would first test Q280 before bothering about a replacement type. They should be matched at Id about 1 mA each and Vds about 15 V. Is there any need for a curve tracer? In the absence of triggering you can measure DC levels all around, drain of Q280B (should be 15 V mine 0.7 V), common source a few volts positive, difference between gates (2.2 V?). You can even probe these points during operation; a 10x probe won't disturb the results. Maybe there is another same type dual fet in the 7T11 that can temporarily be used/exchanged? Albert Yes, a slight inbalance I could understand - but +2.2V is a LOT. ------------------------------------ Yahoo! Groups Links |
David C. Partridge