EEVblog Electronics Community Forum
Electronics => Repair => Topic started by: BFX on July 23, 2014, 08:24:14 pm
-
Hi,
I try to repair this nice analog oscilloscope but is still there one problem.
Have anyone any idea where can be problem?
All electrolitics capacitors are changed.
-
Not triggering right and TR (Trace Rotation) needs adjustment?
-
Thank you.
Yes TR will be adjusted during calibration.
I'll check trigger circuits and trigger signal amplifier.
-
Some news.
I found bad zener diode in trigger circuit (Z400 in scheme) but after replace without change.
Here is video (sorry only from hands).
https://www.youtube.com/watch?v=8utXkW6RnCA&feature=player_detailpage (https://www.youtube.com/watch?v=8utXkW6RnCA&feature=player_detailpage)
And trigger signal screenshot from oscilloscope (IC400 pin 9)
Scheme of timebase and trigger at the end.
Any another suggestions?
-
What are you using as a signal source? It looks as if there is another asynchronous signal, perhaps mains, superimposed on the squarewave. Notice how the height of the displayed waveform varies - if it were just a trigger problem it would just show as left-right jitter.
-
My homebrew generator 8)
But here is another video.
https://www.youtube.com/watch?v=1AQNDU3poxw&feature=player_detailpage (https://www.youtube.com/watch?v=1AQNDU3poxw&feature=player_detailpage)
In this case I use calibration output of Owon oscilloscope.
-
Whooa! Took me a while to work out why the signal was apparently retracing backwards! I assume that's a trick of the camera and what you saw was two relatively steady traces?
I still think you have a problem with mains corrupting the signal. Ground loop? - does the Hameg's cal output work. Is there anything producing a magnetic field near the base of the tube? Is there ripple on the scopes power supplies?
-
No it's no trick it's real :D (I think it's depends on temperature)
First pic is ripple of 140V that's used for X and Y final amplifier.
Second pic is ripple of -2kV but I'm not sure that is correct measurement.
I used 22nF/2kV capacitor between test pin and scope probe as is recommended in the service manual.
I am using an isolation transformer to power tested scope.
Last pic is power supply scheme.
-
Ok next day without progress :-//
At all low voltages (+5V,+24V,-12V) is ripple about 10 - 20 mV and 140V, 260V about 40mV.
Some pictures again. Ramp output and output of Y final amplifier.
Any hints? Or suggestions what can I measure?
-
If you study the PSU schematic you will see a ripple spec for each of the 5, 12, and 24 V rails.
In all cases it is better than 5 mV p-p. (1, 3, 5 mV respectively)
Get the PSU to meet those specs first.
You said you replaced the Ecaps, but maybe lower ESR versions are needed.
-
Hmm but now is really problem to buy low esr axial capacitor.
There are currently used radial 105 C. I think there may be problem long capacitor wires.
-
The problem is not because of the capacitor lead length when you swap radial for axial capacitors.
Try setting the trigger to line input and sweep to about 5 ms/div to see if the noise is line related.
-
This problem seems to come from an unregulated -2Kv power supply with high ripple.
Look for T800/T801 shorted or with leakage.
A good way to test de -2KV is to vary the main voltage of+/- 10% with a variac.
Pattern would not vary at all if regulation is ok.
Remember that deflection factor (V/mm) depends of accelerating voltage.
http://www.jmargolin.com/xy/xymon.htm (http://www.jmargolin.com/xy/xymon.htm)
-
This problem seems to come from an unregulated -2Kv power supply with high ripple.
Look for T800/T801 shorted or with leakage.
I agree, the -2kV is interesting. It might explain the weird Z-modulation in the video. Bear in mind that the reference for the 2kV appears to be derived from the +5V supply - any disturbance here is going to be magnified on the -2kV output.
-
This problem seems to come from an unregulated -2Kv power supply with high ripple.
Look for T800/T801 shorted or with leakage.
I agree, the -2kV is interesting. It might explain the weird Z-modulation in the video. Bear in mind that the reference for the 2kV appears to be derived from the +5V supply - any disturbance here is going to be magnified on the -2kV output.
Exactly
The OP has stated his V+rails ripple spec is 4-10 times higher than that specified in the schematic.
Also:
All electrolitics capacitors are changed.
As the -2KV is derived from a dedicated ~470V AC winding then multiplied, that being regulated by circuitry supplied from the 12 & 5 V rails, any ripple on those supplies could be imposed on -2 KV
Not to mention any other affect the ripple might have on other CRO circuitry.
Get the LV rails ripple to spec.
-
Have a look for ripple upon C817 ! What is the frequency of this ripple ? 50 or 100 Hz?
With Hameg scopes, a very common fault is a rectifier bridge (BR802) with an open diode.
Then ripple is 50Hz instead of 100Hz and the input voltage drop too low for IC804 (5V voltage regulator) to do his job.
-
Thank you for answers T800/T801 seems to be ok also BR802.
I try to measure ripple of 5V but it's difficult because ripple when I turn on the scope and asign probes (both) to GND ripple is more than 10mV :palm:
I'll try to assign HV control voltages from battery, to inspect HV modulation ripple.
-
I try to measure ripple of 5V but it's difficult because ripple when I turn on the scope and asign probes (both) to GND ripple is more than 10mV :palm:
Are you saying that your measurement scope is creating a ground loop and introducing >10mV? Do you have a multimeter that measures AC? - If it will not directly measure the ripple on a DC supply, use a capacitor to remove the DC component - like you are doing with the scope. Be careful, neither your measurement scope, nor your DVM will like having 2kV "spikes" pushed up their inputs!
Also, remember that whatever you measure on the 5V line is multiplied by ~400 on the 2kV. - You need to get the 5V correct.
-
When I measure 5V output it's 4.961V
and -12 is -12.037
measured by HP 3478A
I don't know how to measure ripple lower than 20 mV to get right accuracy using my equipment.
I try suggestion by David Hess "Try setting the trigger to line input and sweep to about 5 ms/div to see if the noise is line related."
Yes it litle betther but I'm not sure with result.
-
measured by HP 3478A
I don't know how to measure ripple lower than 20 mV to get right accuracy using my equipment.
The HP 3478A has a 300mV AC range. Set the meter to measure AC and use the capacitor (as you did previously) to block the DC. But connect the meter and capacitor before you power-up the Hameg - otherwise you risk damaging the meter.
I try suggestion by David Hess "Try setting the trigger to line input and sweep to about 5 ms/div to see if the noise is line related."
Yes it litle betther but I'm not sure with result.
Your video shows that the trigger is working - it is locking to your test waveform. Something else is superimposed on your test waveform and it is drifting across the screen because it is at a slightly different frequency to your test waveform (or, actually, an integer frequency divisor of your test waveform). What David is suggesting is that if setting the trigger to "line" causes the problem waveform to lock and your test wave to drift, then the problem is most likely due to mains/line related issues (like excessive ripple).
Edit: It might be clearer if you move your test waveform frequency away from being an integer multiple of your mains frequency - say 1005Hz.
-
I don't know how to measure ripple lower than 20 mV to get right accuracy using my equipment.
I try suggestion by David Hess "Try setting the trigger to line input and sweep to about 5 ms/div to see if the noise is line related."
Yes it litle betther but I'm not sure with result.
Set you other scope input to AC coupling.
For a perfect DC source it will show a flat line trace irrespective of input voltage.
Turn any HF filters on. (optional)
Set V/div to 5mV or better.
Use Line triggering if available, this will set triggering to mains frequency and multiples of.
By using Line there is no need to adjust the trigger level to within the waveform.
Timebase, 5 or 10 mS/div
5, -12 and 24V share a common GND that is also common with mains earth.
This is perfect for your reference clip.
Connect to each rail in turn and note ripple.
-
Some ripple measurement of low volages by HP 3478A.
(http://www.eternal.sk/hamphoto/IMG_2403_5V.jpg)
5V
(http://www.eternal.sk/hamphoto/IMG_2405_12V.jpg)
-12V
(http://www.eternal.sk/hamphoto/IMG_2407_24V.jpg)
24V
(http://www.eternal.sk/hamphoto/IMG_2411_260V.jpg)
260V
(http://www.eternal.sk/hamphoto/IMG_2414_140V.jpg)
140V
(http://www.eternal.sk/hamphoto/IMG_2415_clips.jpg)
-
So for each rail:
5 V measured 0.591 mV and schematic spec is 1 mV :-+
-12 V " 0.890 mV " 3 mV :-+
24 V " 0.669 mV " 5 mV :-+
For the 160 and 240 V there is no schematic spec to meet however these are nice and low for non critical rails.
It will be interesting to see if the problem is gone.
But before you put her back together, see if you can duplicate your findings using your other scope to measure the ripple, but you probably can't as it is so low.
Even if your most sensitive setting is 5 mV/div and using a 1:1 probe, these ripples are ~ 1/10 division and barely discernible.
But try the 160 & 240 V rails for a hopefully quite visible result.
Good homework, this is good basic scope usage that should give confidence in use of scope functions.
-
Story continues :)
Measurement of ripple 2kV ... hmm something wrong is there >:D
From where ? :palm:
-
Is that ripple on the -1876 or -2000 volt supply?
The later is almost certainly the CRT cathode supply and must be tightly regulated because it directly affects vertical and horizontal deflection accuracy. 50 volts of AC ripple is about right to create exactly the effect shown.
It would be helpful to see the CRT schematic. I am not clear about what the -1876 volt supply is for.
As others have pointed out, the -2000 volt supply uses the +5 volt supply as a reference so ripple on the later will show up on the former and the +5 volt supply regulation is critical.
I would consider getting a high voltage probe so you can accurately measure the magnitude of the -2000 volt supply directly.
-
It's -1.9kV on check strip.
-
That is the unblanking pulse which goes to the CRT grid and has nothing to do with CRT deflection. I would expect that level of ripple on it which is really just the unblanking waveform.
-
oldway: Have a look for ripple upon C817 ! What is the frequency of this ripple ? 50 or 100 Hz?
I made this measurement on power lines and here is result (the similar is on each).
David Hess: I think, unblanking pulses is on pin 10 but I made this measurement on pin 9.
-
What voltages are on the pins of IC801 ? Is it trying to regulate?
-
Voltages on the IC801 pins are:
1 -9.4
2 2.2
3 0.84
4 -9.4
5 -9.4
6 2.14
7 4.95
8 -2.1
but here is measurement by oscilloscope on pins 3 and 6.
I tried to replace IC801, but without visible change.
I think something can be wrong in voltage multiplier circuit.
I'll try to replace diodes D801 - D808 (I think 1N4007 will be enough).
-
measured by HP 3478A
I don't know how to measure ripple lower than 20 mV to get right accuracy using my equipment.
The HP 3478A has a 300mV AC range. Set the meter to measure AC and use the capacitor (as you did previously) to block the DC. But connect the meter and capacitor before you power-up the Hameg - otherwise you risk damaging the meter.
I try suggestion by David Hess "Try setting the trigger to line input and sweep to about 5 ms/div to see if the noise is line related."
Yes it litle betther but I'm not sure with result.
Your video shows that the trigger is working - it is locking to your test waveform. Something else is superimposed on your test waveform and it is drifting across the screen because it is at a slightly different frequency to your test waveform (or, actually, an integer frequency divisor of your test waveform). What David is suggesting is that if setting the trigger to "line" causes the problem waveform to lock and your test wave to drift, then the problem is most likely due to mains/line related issues (like excessive ripple).
Edit: It might be clearer if you move your test waveform frequency away from being an integer multiple of your mains frequency - say 1005Hz.
Thank you when I set trigger to line waveform is clear :-+
But now I don't know what's the problem :palm:
-
Looks like it is trying to regulate but doesn't have the necessary voltage for part of the cycle. Either there's a fault with the quadrupler, or the EHT supply is over loaded.
Edit: this was a reply to the post - 2.
With regard to triggering on the line, what you have done is freeze the "problem" - indicating that the problem is very likely caused by something that is mains related - e.g. excess ripple.
-
I have in the past had to rework this area of a CRO.
This Hameg is unusual in that the -2kV is derived from the mains transformer and not a separate transformer.
The diodes may be similar to 1N4007 and as they are working at mains frequency they may be OK, but cross reference to the listed diode.
The caps are part of what is often called the DC restorer and I have had to replace them in the past.
That you have 4 of the same value is good for checks against each other and less types to get for replacement.
-
Looks like it is trying to regulate but doesn't have the necessary voltage for part of the cycle. Either there's a fault with the quadrupler, or the EHT supply is over loaded.
Edit: this was a reply to the post - 2.
With regard to triggering on the line, what you have done is freeze the "problem" - indicating that the problem is very likely caused by something that is mains related - e.g. excess ripple.
It's not freeze :)
Only I just set trigger level and tune my generator to waveform move slower to be possible take good photo.
Here is litle video https://www.youtube.com/watch?feature=player_detailpage&v=LuXJr5GrI8c (https://www.youtube.com/watch?feature=player_detailpage&v=LuXJr5GrI8c)
-
I tried also external trigger and the same problem as internal. :-//
Can be replaced 120pF/600V using 200pF/1000V on the position C807 of -2kV circuit? (150pF is impossible to buy)
-
Can be replaced 120pF/600V using 200pF/1000V on the position C807 of -2kV circuit? (150pF is impossible to buy)
I would not. C807 forms part of the frequency compensation network within the regulator loop - changing its value will change the loop response. In fact, unless it is known to be faulty, I would not replace.
-
Thank you Andy Watson I try to find some distributor of similar part but int difficult in central europe. (I think distributors are crazy)
I found only one in mouser but price is 61 euros :-- its impossible , its crazy dream :o
-
Only this types of cpacitors is possible to buy 120pF/500V (it's lower voltage as currently used 600V) or 150pF/3kV
http://www.tme.eu/en/details/cch-120p/500v-tht-ceramic-capacitors/sr-passives/ (http://www.tme.eu/en/details/cch-120p/500v-tht-ceramic-capacitors/sr-passives/)
http://www.tme.eu/en/details/cc3k-150-5/3kv-tht-ceramic-capacitors/sr-passives/ (http://www.tme.eu/en/details/cc3k-150-5/3kv-tht-ceramic-capacitors/sr-passives/)
Which to use?
I think better diodes to quadrupler will be http://www.tme.eu/en/details/by2000-dio/tht-universal-diodes/diotec-semiconductor/by2000/ (http://www.tme.eu/en/details/by2000-dio/tht-universal-diodes/diotec-semiconductor/by2000/)
Because original diodes is EM513 / 2kV
But I'm still not sure that ripple of 2kV output is cause of this problem, does have anyone any other hypothesis ?
Why it's happens in internal and external trigger mode and not in line mode?
-
Only this types of cpacitors is possible to buy 120pF/500V (it's lower voltage as currently used 600V) or 150pF/3kV
http://www.tme.eu/en/details/cch-120p/500v-tht-ceramic-capacitors/sr-passives/ (http://www.tme.eu/en/details/cch-120p/500v-tht-ceramic-capacitors/sr-passives/)
http://www.tme.eu/en/details/cc3k-150-5/3kv-tht-ceramic-capacitors/sr-passives/ (http://www.tme.eu/en/details/cc3k-150-5/3kv-tht-ceramic-capacitors/sr-passives/)
Which to use?
For what purpose? Unless you have evidence to the contrary, I would leave C806 and C807 alone. C802/3/4/5 could be replaced with slightly larger values - if necessary - but find the fault first.
Because original diodes is EM513 / 2kV
EM513 are still available, e.g. http://www.rapidonline.com/electronic-components/diotec-high-voltage-rectifier-diodes-526299 (http://www.rapidonline.com/electronic-components/diotec-high-voltage-rectifier-diodes-526299)
But I'm still not sure that ripple of 2kV output is cause of this problem, does have anyone any other hypothesis ?
It is difficult to be certain, however, your earlier measurements show that you have 140-200V peak-peak ripple on what should be a regulated supply. I believe somebody has already posted a link to specifications of the tube which will indicate the effects of EHT voltage variation on the geometry (sensitivity) of the CRT display. The waveforms that you measured around IC801 show that it attempting to regulate for part of the mains cycle but is being prevented from regulating for 100% - presumably because it runs out of input voltage. As I said above - either the quadrupler is faulty, or the EHT supply is overloaded. (At least that is my reading of the evidence presented to date.) Checking the quadrupler would, to my mind, be the next, easiest step to take.
Why it's happens in internal and external trigger mode and not in line mode?
It is still happening in line mode. Triggering on the line frequency has frozen the problem on the display - I bet if you put the graticule back in front of the CRT you will find that your square waves are not quite as square as the should be ;)
-
But I measured all diodes in quadrupler and all seems to be ok, also capacitors :(
Only one is suspected, under diode D409, D408 is dark BCB as overheated.
Also on another place is signs of breakdown of EHT to ground, I made a gap in PCB.
-
One problem with testing high voltage diodes and capacitors is that they may fail by having high leakage only at high voltage and most testers use a low voltage.
If they cannot be tested under realistic operating conditions and if other causes are eliminated, then it may be easier to just replace them.
-
.... then it may be easier to just replace them.
and clean the PCB whilst you're there! What is the history of this instrument - looks like it has been dipped in the sea!
-
Yes will be best to replace all diodes and capacitors.
Therefore, I asked which be better to applied.
I don't know history of that , it's looks like has been in a humid environment.
-
Ok all parts of EHT was changed ... let's go testing :-DMM
-
... let's go testing :-DMM
... and ?
Or are you waiting for the magic smoke to clear ;)
-
OK :D
Testing testing is complete and oscilloscope works again as new ;)
(http://www.eternal.sk/hamphoto/IMG_2490a.jpg)
I replace all diodes in quadrupler and also capacitors c806 and c807
(http://www.eternal.sk/hamphoto/IMG_2484b.jpg)
I used BY2000 (2kV/3A) diode as replacement of EM513
http://www.tme.eu/en/details/by2000-dio/tht-universal-diodes/diotec-semiconductor/by2000/ (http://www.tme.eu/en/details/by2000-dio/tht-universal-diodes/diotec-semiconductor/by2000/)
and those capacitors
http://www.tme.eu/en/details/cc3k-10n-7.5/3kv-tht-ceramic-capacitors/sr-passives/# (http://www.tme.eu/en/details/cc3k-10n-7.5/3kv-tht-ceramic-capacitors/sr-passives/#)
http://www.tme.eu/en/details/cch-120p/500v-tht-ceramic-capacitors/sr-passives/# (http://www.tme.eu/en/details/cch-120p/500v-tht-ceramic-capacitors/sr-passives/#)
Now is time to calibration , clean and put together.
Front panel is already clean. (img. 2)
(http://www.eternal.sk/hamphoto/IMG_2492a.jpg)
Thank you :-+
-
Well done.
Nice sharp trace.
Yes that PCB looks grubby and wanting a good clean.
Hope she is trouble free for you now.
-
Well done.
Nice sharp trace.
Yes that PCB looks grubby and wanting a good clean.
Hope she is trouble free for you now.
Thanks , it was good experience in troubleshooting of analog oscilloscope.
And also very good relax for me because I'm software developer.
:-+
-
Ok boys ,
good news first :D here is some pohotos after litle callibration process.
generator signal from owon
(http://www.eternal.sk/hamphoto/IMG_2502a.jpg)
The same signal from hameg
(http://www.eternal.sk/hamphoto/IMG_2500a.jpg)
Front panel cleaned and completed
(http://www.eternal.sk/hamphoto/IMG_2499a.jpg)
And now somethink bad :-DMM
Chanell B was broken input FET transistors and I used BF245B. Also T204 was bad and replaced by same type.
(http://www.eternal.sk/hamphoto/HM203_Edit.jpg)
But now is aplitude on the screen from channel B about 10x lower as on channel A.
Is possible change some pasive or is needed to change FET?
Better replacement of TIS69 I think like 2N5199, 2N5046, BFS21A.
Or can be something other wrong? Signal after attenuator is the same as from channel A.
-
For T201 & 204 to have failed, I would suspect a single over-voltage event.
If that is the case, expect IC201 to have failed also, along with damage to diodes and caps.
It will just take a little time to work through them.
-
I'm not sure that IC201 is wrong. Here is photo.
(http://www.eternal.sk/hamphoto/IMG_2511a.jpg)
The same signal on both chanels.
Only attu on channel 2 is set to 0.1V
-
Looking pretty!
Should be easy to find the problem with channel 2 since you've got a working channel to compare it against.
-
The leading edges do not look right to me. The overshoot might exclude a probe compensation problem.
As subtle as that is it points to a major problem.
-
Ok boys, problem solved 8)
David Hess: yes it's probe compensation problem.
Here is some updates to solve major problem
(http://www.eternal.sk/hamphoto/HM203_Edit1.jpg)
And oscilloscope works well
(http://www.eternal.sk/hamphoto/IMG_2525a.jpg)
Tomorow final calibration and testing .... and finish.
-
Here is some interim results of calibration.
I'm surprised form this, it's realy nice:)
I don't shorten parallel connection of resistors on input and it's beautiful to see to interference. :-+
(http://www.eternal.sk/hamphoto/IMG_2539a.jpg)
from owon 8202
(http://www.eternal.sk/hamphoto/IMG_2536a.jpg)
from hameg 203-1
It's measured on second corrected channel. It's same time base and same attenuator settings.
-
So the overshoot was in the pulse generator. That looks good to me.
-
yes it's my old homemade poor generator ;D
but in this case it is easy to see that the oscilloscope works very well :)
here is screenshot from scope (before I was lazy :D )