EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: TGyuri on October 07, 2014, 10:39:22 am
-
Hey Dear All,
I'm new on this forum.
My name is Gyuri.
My english is crappy, please forgive it.
Just as my knowledge in electronics.
But that does not stopping me to build amplifiers, DACs etc.
Indeed I have repaired four TEK scopes.(2445, 2465, 2467B and 2440.)
Of course I have got some help from very nice people at TekScopes.
I have had four scopes, two of them from above, 2467B and 2440, and 2225 and a Velleman PCS500.
I have acquired a beautiful HP1745A now.
It is in very good cosmetical state inside and outside also.
Wonderful, that screen filter is glass, not some crappy plastic.
But I have some problems:
First, scale illumination, it is works only after a hour or so, after switching on.
It is uniform and stable otherwise.
Second, although waveforms are quite O.K., at all frequencies with stable trigger,
there is a strange problem.
I have built a Jim Williams pulse generator for fast rise signal, HP can not be able displayed of it's waveform.
If I press Beam Find button, the signal appears at the begining or ending of trace.
I don't know, where I can start.
I have checked LVPS voltages, seems O.K. to me.
Thanks for any help in advance!
Kind regards,
Gyuri
(https://hbpeza.dm2301.livefilestore.com/y2pg3v-Fh4qKluBUVwMFBjOKiW9-tlfpiTC33MxBAU2fmYv1RjVXVwmMpEVVpIncfLTK1mftHB9i4vHj4HsYe3zFkmRBBfKZKOfz6l8MHn5nAM/HP1745A_01SMALL.JPG?psid=1)
-
Can you see the pulse on your other scopes? What's the amplitude and duration? The HP scope is "only" 100MHz.
If you drive the HP with a simple digital signal, what is the observed risetime? The scope should have a 3.5ns risetime, but that is often limited by an inappropriate probe.
Do you have the BW limit on?
Do you have the 50ohm termination in/out?
-
Thank you for dealing with me!
I'm trying to answer your questions:
1.,
Yes of course! See picture, it was taken from 2467B.
It has a frequency about 130 kHz, amplitude about 7V into 50 ohm and 8.5V into 1 Mohm.
Even 2225 which is only 50 mHz scope, possible to view this signal, only amplitude and rising time degraded.
2.,
I have a good old Bradley 192 oscilloscope calibrator.
It has a fast rising output 200-250 mV.
According to 2467B it has a rising time 1.25 ns.
HP1745A can show this signal, I've measured rising time about 3 ns it's quite good, isn't it?
But there are some strange thing, which I discovered now.
It shows falling edge, and if I pressed down trigger slope, then it shows leading edge.
Moreover, if I playing with trigger level knob in the range, then I get twice stable trigger. One at position 9, other one at position 15.
I think there is a problem.
I have read something about it in the SM, chapter 5-21 TRIGGER SENSITIVITY ADJUSTMENT. If I understand it correctly.
3.,
Of course not.
4.,
It only changes amplitude slightly, and I get a better waveform in 50 ohm, because impedance matching.
Generators has 50 ohm output impedance.
Kind regards,
Gyuri
-
I probably wouldn't worry about it. Based on my experience with the scopes I've worked on, I generally consider the ability to trigger on an avalanche pulser nothing more than a nice bonus for anything under 150 MHz. It's not a nice waveform at all...
-
I know it's not so nice, probably I need to do some homework on my pulse generator.
So Tek 2225 a very kind scope, because it displays stable trigger with this!
Despite of it is a 50 MHz scope.
There is an other picture, it shows pulse generator output, while I pushing Beam Find button permanently.
It is quite stable, although not so nice. :)
-
There are some photographs about oscilloscope:
https://onedrive.live.com/?cid=E894CB7D50B65FBF&id=E894CB7D50B65FBF%21857 (https://onedrive.live.com/?cid=E894CB7D50B65FBF&id=E894CB7D50B65FBF%21857)
Gyuri
-
I have one of the 1741A storage scopes as a daily driver at my service bench. The 1745/46A’s came out in 1983 and were right at the end of the 10 year run of the HP 1700 series scopes. With this series in the late ‘70’s HP finally had scopes that could match Tektronix head to head in performance and reliability, and they eventually passed Tek in scope sales once digital scopes started becoming popular. Despite the Tek 2225 being a few years newer than it, the 1745A when working right should be able to trigger on any signal a Tek 2225 can trigger on, so I suspect there is an issue with your 1745A somewhere.
-Mark-
-
Yes, I think it so! :)
Gyuri
-
If your Tektronix 2225 can display the pulse, then your HP 1745A should not have any problem.
Maybe the HP 1745A is not unblanking quickly enough or the CRT is too dim.
-
Well, I have solved first problem.
I wrote it, LVPS voltages are O.K. but there was an odd-one-out:
That one, which is built around CR3-C7-Q11-Q12....
As there was no output voltage, when scale illumn. doesn't work.
Turned out, that connector between tranny and six diode bridges, has some ugly cold soldering.
It is clear that there was some repairing earlier.
So I removed LVPS board, and changed all those bridges pre-emptively.
Also I removed that connector, cleaned and soldered back its pins
And voila, scale illumination works, as new. :)
I will be unable to solve the second problem so simply.
(https://hbpeza-ch3301.files.1drv.com/y2pRXLSe5nX8hofvapzfeupE9f8uDKLltJSSeeOiUx7YqJhm9ndKymf96SdO8xzZ1WXpF6EhJ79cfCXHVIs9Lsx6BhVIOZG--sPKzDXHxgUskg/Img00250001.JPG?psid=1)
Well, I killed one of those little blue bugs, I have pulled out one of its legs.
Doesn't matter, I put back a red one.
(https://hbpeza.dm2301.livefilestore.com/y2p4Tj11ONmQD7rcIKDjOfys_LD4kLtWw-FgmuQfQtnuSvikHRUIyaZ3XHXs2_JakhHYNUL-8Myp_0zl4uXQD4YckBXkyYb8cQkw4x1Kk2wKFU/Img00300001.JPG?psid=1)
(https://hbpeza-ch3301.files.1drv.com/y2p5Z5Xu49RN7-nzoWkNzOs2VHi6wlyiCdKziwwIRaZhYoy6GIdmWBbOXvPVwaKER7VerFraVOztuLoGfi3s7BbLluRxIUlID9YqPd_Q_IaLe8/Img00340001.JPG?psid=1)
-
I think, needed a little bit more help, what direction I going to with second problem. :)
I think there is only some settings need to do.
But I don't want to turn the trimmer potmeters and caps at random.
I suspect it is not a trigger problem, because on the picture above, where beamfind button was pushed, there is a stable signal.
Also trigger led shows, signal is triggered.
Since I do not understand it, just guessing.
Gyuri
-
To me it looks like the CRT is dim which will create problems displaying fast low repetition rate signals. The vertical and horizontal deflection looks good so the acceleration voltage at the cathode is accurate but maybe the post deflection acceleration voltage which is like 19 kilovolts is missing. Can you disconnect the CRT anode or the voltage multiplier and see if that makes any difference?
You really need to find a service manual. I think the HP1746A service manual covers the most or all of the HP1745A. Artek Manuals has that one:
http://artekmanuals.com/manuals/hp-manuals/ (http://artekmanuals.com/manuals/hp-manuals/)
-
Hey David,
Thank you for your answer and questions as well!
Do you think it's too dim?
https://hbpeza-ch3301.files.1drv.com/y2pIuG0UF3x5abjKciibiCYYNU9UBEXnooDXE6wl9xT8MJ8TkXXdeaT1zcyk-ylsBn2W4bF5m42g69ikxbKpZab90_n9_tWWxABiQL_TZsAX80/Img00460001_1.JPG?psid=1 (https://hbpeza-ch3301.files.1drv.com/y2pIuG0UF3x5abjKciibiCYYNU9UBEXnooDXE6wl9xT8MJ8TkXXdeaT1zcyk-ylsBn2W4bF5m42g69ikxbKpZab90_n9_tWWxABiQL_TZsAX80/Img00460001_1.JPG?psid=1)
Yes, if we see at the beginning of the signals, certainly is.
Photograph is cheating some extent.
What can you see, it is two 10 MHz signal from my TOELLNER TOE7711 generator.
(Is there anybody who has service manual for this?)
A channel is main output triangle, B channel its ECL output.
I have the manual from Keysight:
http://literature.cdn.keysight.com/litweb/pdf/01745-90901.pdf (http://literature.cdn.keysight.com/litweb/pdf/01745-90901.pdf)
There are more pictures:
https://onedrive.live.com/?cid=E894CB7D50B65FBF&id=E894CB7D50B65FBF%21857 (https://onedrive.live.com/?cid=E894CB7D50B65FBF&id=E894CB7D50B65FBF%21857)
-
At 100 MHz things are worsened.
See attachments.
Trigger is stable, but the beginning of signal dimmed further.
First pic at 50 ns, without 10x mag., second with mag., somewhere in the middle, third one is at the beginning of signal.
So, Dear David, if you can show me the way, according to the service book, it would be good.
Many Thanks,
Gyuri
-
...Can you disconnect the CRT anode or the voltage multiplier and see if that makes any difference?
You mean which connectors?
Can you show me in the linked manual?
Otherwise, I respect the high voltage, and that 19 kV does not sound good. :)
-
Hey David,
Thank you for your answer and questions as well!
Keep in mind that I do not have any experience with HP oscilloscopes. I am a Tektronix dude but I do have a Tektronix 7603 which is roughly comparable as far as having a large CRT. The HP1745A and the others in that series look like fine oscilloscopes and based on their high PDA (post deflection acceleration voltage), I suspect they should produce a somewhat brighter and clearer display than a Tektronix 7603.
Do you think it's too dim?
It looks dim and unfocused to me but someone more familiar with the HP1745A or HP1746A and how their large CRTs should look will know better.
Yes, if we see at the beginning of the signals, certainly is.
Photograph is cheating some extent.
At 100 MHz things are worsened.
See attachments.
Trigger is stable, but the beginning of signal dimmed further.
First pic at 50 ns, without 10x mag., second with mag., somewhere in the middle, third one is at the beginning of signal.
I noticed that as well. It could indicate an unblanking problem or a problem with the z-axis amplifier which could also cause a dim display.
On analog oscilloscopes, it is usual for the trace to become fully unblanked before 1 or 2 divisions into the sweep. Yours does not seem to show this until 3 or 4 divisions but it is difficult to tell without knowing the time/div setting you used.
I have the manual from Keysight:
http://literature.cdn.keysight.com/litweb/pdf/01745-90901.pdf (http://literature.cdn.keysight.com/litweb/pdf/01745-90901.pdf)
So, Dear David, if you can show me the way, according to the service book, it would be good.
Well, that is handy. I spent some time looking for the service manual yesterday and came up with nothing. I am looking at PDF page 91 and it seems I guessed right about the post deflection acceleration voltage of 19 kilovolts.
I would disconnect the 19 kilovolt lead going to the CRT, ground it to the chassis, and then check the display again to see if it gets dimmer or otherwise changes. PDF page 90 shows that it has a high voltage plug at the end of the lead going to the CRT which plugs into the high voltage multiplier (A6 on PDF page 91 in the lower right corner). The alternative is to disconnect the high voltage multiplier from the inverter but I do not know how easy that would be.
Alternatively if you have access to a high voltage probe, the output of the high voltage multiplier could be measured directly after disconnecting the lead.
The usual precautions should be taken when doing any of the above since it *is* a high voltage circuit. Do not mess with the high voltage connections when the oscilloscope is turned on. It might be a good idea to have someone monitoring you.
I am basing the above idea on a Tektronix oscilloscope we diagnosed a couple months ago as having a bad high voltage multiplier which displayed a similarly dim and slightly unfocused display.
If doing the above is a bad idea, I am sure and hope some other EEVBlog member will say so.
If you want to start off with something safer, then I would check the waveform at test point 1 shown on PDF page 93 which is the z-axis amplifier output. PDF page 92 shows what the waveform at that point should look like. You will need another working oscilloscope to do this however.
...Can you disconnect the CRT anode or the voltage multiplier and see if that makes any difference?
You mean which connectors?
Can you show me in the linked manual?
Otherwise, I respect the high voltage, and that 19 kV does not sound good. :)
None of your photos show the area so I am not sure how to point it out to you. The description of the area in the service manual on PDF page 90 is cutoff on the left hand side.
-
Thank you for your detailed answer, David!
I have found meanwhile what you suggested.
I will be careful, because I don't want to kill this beauty, and of course not myself.
And I'm not in hurry, so I will listen anyone else opinions.
And I will try to learn this scope, read manual and try understand it, as I can, you know, my crappy english, and my knowledge in electronics, which is similar than hen's knowledge to alphabet.
So now I do not pull apart any connectors for the time being.
I can't do so high voltages in check anyway, because I do not have instrument for this.
If I get a detailed script how can I perform this test, what you suggested for me, I will do it.
Kind regards,
Gyuri
-
Thank you for your detailed answer, David!
I have found meanwhile what you suggested.
I will be careful, because I don't want to kill this beauty, and of course not myself.
And I'm not in hurry, so I will listen anyone else opinions.
As far as I know, disconnecting the high voltage at that connector which just pulls straight apart is safe.
I can't do so high voltages in check anyway, because I do not have instrument for this.
I do not have one either. I have a 6 kilovolt high voltage probe which I use to check the CRT cathode voltage which is in the -2000 to -3000 volt range but for checking the PDA voltage which is much higher, typically a 40 kilovolt probe is used.
-
I would disconnect the 19 kilovolt lead going to the CRT, ground it to the chassis, and then check the display again to see if it gets dimmer or otherwise changes.
You say, pull apart that connector on the thick red cable, and grounded it?
It doesn't kill the multiplier?
Just ask, because I never ever doing similar.
You know, I only once repaired a CRT television, when my wife flooded the flower on it.
There was some burns on the PCB, I have repaired it with some tiny copper wires.
It still works. :)
-
I would disconnect the 19 kilovolt lead going to the CRT, ground it to the chassis, and then check the display again to see if it gets dimmer or otherwise changes.
You say, pull apart that connector on the thick red cable, and grounded it?
It doesn't kill the multiplier?
Just ask, because I never ever doing similar.
The connector pulls straight apart with the plug going to the CRT and the socket going to the multiplier. The terminal inside the socket going to the multiplier is enclosed which shields it from exposure so that is pretty safe. The terminal on the plug going to the CRT is exposed and may or should be safely grounded to the chassis. Do not touch it however because the CRT may have residual charge still present.
The CRT should still work without the anode connection to the high voltage multiplier but it will be dim and not as well focused. By comparing it to your previous results, you should be able to tell if the high voltage was originally missing which I suspect may be the cause for the dim display.
The reason I suggested this test is that it is relatively easy to do and requires no special equipment.
-
Meanwhile, I was thinking about (although this is difficult for me :) ) it is a nonsense, to be grounding the wire on the side of multiplier, so I misunderstood you.
CRT side of that thick wire is, what must to be grounded.
I think, I will be waiting with it, altough I'm very curious, what is the problem.
I'm performing checks now on side 92 in the manual first.
-
I have one of the 1741A storage scopes as a daily driver at my service bench. The 1745/46A’s came out in 1983 and were right at the end of the 10 year run of the HP 1700 series scopes. With this series in the late ‘70’s HP finally had scopes that could match Tektronix head to head in performance and reliability, and they eventually passed Tek in scope sales once digital scopes started becoming popular. Despite the Tek 2225 being a few years newer than it, the 1745A when working right should be able to trigger on any signal a Tek 2225 can trigger on, so I suspect there is an issue with your 1745A somewhere.
-Mark-
Lucky. I had a 1741 for a while but the tube was weak. I ended up donating it to a local hacker space.
-
I'm searching for test points.
I have found number 4, it is almost like exactly as expected, if we doesn't count that its frequency almost double that we waiting for.
Or I'm doing something wrong.
-
Here the situation is similar:
-
Waveform TP2, almost good.
-
Waveform TP3, there is practically nothing. :( DLY'D GATE
-
Waveform TP3.
My bad.
I measured the wrong place.
The case is similar to TP4 and TP1.
-
I do not know, have not touched it yet, but I feel it is nothing wrong with 19kV. :)
I will continue to investigate.
-
I may see a problem in the gate output waveform which is TP1, waveform 4, and the first one you posted. That rising edge should be squarer and it will be affected by trimmers C11 and R12 shown on service sheet 3 which is PDF page 93. That by itself could explain the dim part at the beginning of the sweep.
Page 5-4 of the service manual tells how to adjust C11 and R12.
If that waveform cannot be made "flatter" on the top and especially after the leading edge, then there may be something wrong around that circuit.
-
Hey David, I'm just awaken now, drink my coffee, so forgive me I'm lost in my posts.
So which waveform is bad according to your opinion?
I have noticed, the frequency of the signals almost double than expected, except waveform 2, which is chopped mode.
Thank you for your valuable help!
I have now some works must to do.
Buying for weekend, dishwashing, housekeeping etc... but I hope afternoon I will work on this wonderful equipment.
Kind regards,
Gyuri
-
On the service manual sheet I have marked with lightblue the corresponding waveform.
-
So what I don't understand waveforms 1,3, and 4.
These are all of them (according to Service Sheet 3) 500 Hz signals, with 2 ms period.
In contrast, I am getting 714 Hz signals (1.4 ms) in all three position.
And waveform 1 is slightly asymmetrical.
Waveform 4 also not suitable spaced.
It is divided in 5/8 - 3/8, and I get about 27/35 - 8/35 ratio.
What a surprise its ratio is the same what I got at waveform 1.
Only it has 8/35 - 27/35 ratio.
I don't know where and how can be set it to 500 Hz.
I do not know what the consequences of this, but I do not like it.
-
I may see a problem in the gate output waveform which is TP1, waveform 4, and the first one you posted. That rising edge should be squarer and it will be affected by trimmers C11 and R12 shown on service sheet 3 which is PDF page 93. That by itself could explain the dim part at the beginning of the sweep.
Page 5-4 of the service manual tells how to adjust C11 and R12.
If that waveform cannot be made "flatter" on the top and especially after the leading edge, then there may be something wrong around that circuit.
Maybe that measure wasn't so good, so I give it another try.
I think it is not so bad.
-
The timing in these waveforms (except for transition timing which the adjustments I discussed above have to do with) depends on the sweep settings. That little short bump at the top of #4 is the brightness being increased during the intensification of the delayed sweep.
If you want to get an idea of what is going on and what you are looking at, measure #4 while changing the sweep, trigger holdoff, and intensity controls.
What #4 is really showing is the intensity control of the CRT with positive values yielding a brighter trace. The most negative portion of the waveform is the period when the trace is blanked. It even says below the graph, "amplitude varies with intensity".
What we do *not* want to see is a slow rising edge or in this specific case, a gradually rising section *after* the fast rising edge. The top of the pulses in #4 should be flat except for the short bump from the intensified sweep.
-
I may see a problem in the gate output waveform which is TP1, waveform 4, and the first one you posted. That rising edge should be squarer and it will be affected by trimmers C11 and R12 shown on service sheet 3 which is PDF page 93. That by itself could explain the dim part at the beginning of the sweep.
Page 5-4 of the service manual tells how to adjust C11 and R12.
If that waveform cannot be made "flatter" on the top and especially after the leading edge, then there may be something wrong around that circuit.
Maybe that measure wasn't so good, so I give it another try.
I think it is not so bad.
That photo looks fine. What changed it?
I have run across oscilloscopes before that had problems with the intensity control signal where it gradually rose after unblanking yielding a dim CRT trace during the beginning of the sweep which yours seems to be doing.
-
I have changed from TEK 2225 to TEK 2440.
But what most important, changed a cheap 100MHz probe to an original TEK P6137 probe. I ashamed myself, but even that cheap probe gave me a better signal shape if I calibrated it. Excuse me.
So, do you think it is an appropriate signal?
I can see an overshoot after rising edge.
So this is, what I should remove with setting of C11/R12, am I right?
Is it worth it a try?
-
Have you an idea about timing mismatch what I experienced?
Thank you for your patience!
-
Detailed view of rising edge waveform 4:
-
If you want to get an idea of what is going on and what you are looking at, measure #4 while changing the sweep, trigger holdoff, and intensity controls.
Yes, of course!
I have played with these settings, very interesting.
Even I have made an exactly similar waveform what I missed permanently, only setting TIME/DIV Vernier and Trigger Holdoff. But it doesn't have practical results.
-
I have changed from TEK 2225 to TEK 2440.
Hmm, good choice. The Tektronix 2440 is an excellent vintage from 1990.
The 2225 is probably just fast enough at 50 MHz to make the rise time measurement discussed below.
But what most important, changed a cheap 100MHz probe to an original TEK P6137 probe. I ashamed myself, but even that cheap probe gave me a better signal shape if I calibrated it. Excuse me.
Doh!
This would completely explain the observed slope on the flat part of the pulse.
So, do you think it is an appropriate signal?
I can see an overshoot after rising edge.
So this is, what I should remove with setting of C11/R12, am I right?
Is it worth it a try?
It looks fine to me now or at least the top part does but see my comment below.
The overshoot can be corrected with C11 and R12; that is what they are there for. I would certainly tweak them for the best shape.
The rise time of the pulse looks slow to me given that the oscilloscope's vertical delay line is 100 nanoseconds. I think the rise time should be less than 1/4 of that or faster than 25 nanoseconds. Maybe that is related to the apparent low CRT intensity or is a completely different problem. Adjusting C11 and R12 may help resolve it or not. Maybe someone who has one of these oscilloscopes can confirm whether a 140 nanosecond rise time is typical for this series.
This slow rise time easily explains the original problem of not being able to see the pulse from the Jim Williams avalanche pulse generator and it was next on my list of things I was going to recommend checking. Figuring out why the rise time is slow if adjusting C11 and R12 does not fix it will not be easy.
These are the kinds of problems I love to solve.
-
These are the kinds of problems I love to solve.
Yeah, just like me!
But the things are aggravating on my part to what I wrote about hen and ABC. :)
-
Interesting.
C11 here is not trimmer, it is fixed.
Maybe tomorrow I will check it.
Setting trimmer pot. R12.
I have achieved worse results.
We will see!
Kind regards,
Gyuri
-
I was wrong, again.
It was because my dimmed eye.
(I have bought this scope mainly because its nice big screen. :) )
It is a trimmer cap, just not that shape what I have known.
-
I have carried out settings described in paragraph 5-17.
I have got slightly better waveform, (see attachment) but it hasn't affected rise time.
It hasn't affected dimmed trace at 50ns neither.
So I performed that test, what David suggested me.
I disconnected 19 kV connector.
The scope was turned off at least one hour before.
I was very careful. But not enough.
As I pulled apart connectors, I have dropped that end of cable, which has a naked pin.
As David wrote it earlier, it has a residual charge for sure!
It has produced a very nice blue lightning, with some thundering.
So it certainly had some kV in it.
And now, what happened?
I was worried about, what went south because of lightning.
Because I didn't seen it, where the lightning was striked.
But I was lucky.
Be very-very careful, if you doing this!!!
Well, so this test proved, my feeling about a good PLD was right.
And what shows the screen, without PLD?
I have got a much smaller and dimmer picture.
So PLD doing its job.
-
...
As I pulled apart connectors, I have dropped that end of cable, which has a naked pin.
As David wrote it earlier, it has a residual charge for sure!
It has produced a very nice blue lightning, with some thundering.
So it certainly had some kV in it.
And now, what happened?
I was worried about, what went south because of lightning.
Because I didn't seen it, where the lightning was striked.
But I was lucky.
Be very-very careful, if you doing this!!!
This is why I suggested testing other things and leaving this for last. If the problem was solved in some other way, then messing with it would not be necessary.
Well, so this test proved, my feeling about a good PLD was right.
And what shows the screen, without PLD?
I have got a much smaller and dimmer picture.
So PLD doing its job.
This tells us some things. Since the deflection accuracy was affected by the acceleration voltage, we know that the acceleration voltage is correct, that there is nothing wrong with the power supply, and HP used a CRT design were the PDA (post deflection acceleration) voltage affects deflection. Since the screen was dimmer, we sort of know the acceleration voltage was present.
That leaves the slow rise time of the blanking pulse as the problem. Like I said, I think it should be 25 nanoseconds or faster. As shown it is slow enough that the avalanche pulse is over before the CRT beam finishing unblanking so there is nothing to see. When you press the beam finder, it disables blanking making the pulse visible.
Now the question is why is it slow and why does it have all of that overshoot? The two are likely related.
-
R12 and C11 are in an extreme position.
I don't think it is intentional, and the waveform isn't perfect. (Meant: Not designed in this way.)
It would be nice, if someone (Tekfan) would measure at incriminated TP4 one of his scope. :)
Today's day is almost equal with a victory.
At least I did not ruin anything.
I'm healing slowly.
And the weather was beautiful.
BTW, I'm wondering, that gate circuit able to make a signal, which is needed for a 25 ns leading edge?
Do not take it seriously, I have nothing to do with the electronics that I can decide the issue.
-
I'm searching for test points.
I have found number 4, it is almost like exactly as expected, if we doesn't count that its frequency almost double that we waiting for.
Or I'm doing something wrong.
So what I don't understand waveforms 1,3, and 4.
These are all of them (according to Service Sheet 3) 500 Hz signals, with 2 ms period.
In contrast, I am getting 714 Hz signals (1.4 ms) in all three position.
And waveform 1 is slightly asymmetrical.
Waveform 4 also not suitable spaced.
It is divided in 5/8 - 3/8, and I get about 27/35 - 8/35 ratio.
What a surprise its ratio is the same what I got at waveform 1.
Only it has 8/35 - 27/35 ratio.
I don't know where and how can be set it to 500 Hz.
I do not know what the consequences of this, but I do not like it.
I'm in trouble, why this signal is so asymmetrical?
Its frequency does not match with what is expected.
Lastly, it is the control signal for gate circuit, isn't it?
Any thought?
-
R12 and C11 are in an extreme position.
This indicates a problem but we already expected that. If the circuit is slow, then these adjustments are not going to work.
BTW, I'm wondering, that gate circuit able to make a signal, which is needed for a 25 ns leading edge?
The similar Tektronix circuits are about that fast. Generally they finish settling about 20 nanoseconds before the vertical signal appears and their transition times are similar. Surprisingly it is really not that difficult to do.
I'm in trouble, why this signal is so asymmetrical?
Its frequency does not match with what is expected.
Lastly, it is the control signal for gate circuit, isn't it?
Any thought?
The main gate signal which is the one that matters right now is waveform 1. Waveform 3 is the delayed gate signal which we do not need to worry about yet.
Transistors Q1 and Q2 in U1 amplify the main gate signal and their output drives the high voltage amplifier made up of Q1 through Q4 and the output is waveform 4 which is the one that needs to be fast.
I think the things to check now are the various resistor values around Q1 though Q4 to see if one is open. There could be an open capacitor somewhere but that would be difficult to find.
Take a careful look at R9 above Q2.
-
Thank you again, David!
I will follow your suggestions above.
I'm also thinking to do something similar, thanks for the confirmation!
You're an excellent teacher, and I know this because both of my parents was.
Although both of them were music teachers.
But excellence is not based on what should be taught, but how.
I think, you are one of the man, who taught me the most in this field, the other one is One and Only. (NP)
Kind regards,
Gyuri
I'm getting better slowly.
-
I think I have found it!
Resistor R7 39.2K seems to be an open circuit.
So signal just going thru C3 which is parallel with it.
I'm going to solder. :)
-
Arrrgh.... Duh and so on!!!
I have all kind of resistors.
I was sure about it I have 39.2k.
And what do you think?
Yeah, you won!!!!
I don't have.
Sometimes I would like to kill Mr. Murphy with my own naked hands.
I just wish they had strangled him in the cradle.
-
I think I have found it!
Resistor R7 39.2K seems to be an open circuit.
So signal just going thru C3 which is parallel with it.
I'm going to solder. :)
R7 should definitely not show open.
Arrrgh.... Duh and so on!!!
I have all kind of resistors.
I was sure about it I have 39.2k.
And what do you think?
Yeah, you won!!!!
I don't have.
I do not think the value of R7 is that critical. The closest standard value within 5% should at least work until you can get an exact replacement.
Sometimes I would like to kill Mr. Murphy with my own naked hands.
I just wish they had strangled him in the cradle.
If you tried that, something would just go wrong and Mr. Murphy would live.
-
You are a good man, David!
I'm just puts together things.
So we can see the result soon.
Regards,
Gyuri
-
So, that resistor was the guilty!
It works now, despite I didn't make any setting regarding R12/C11.
Only I settled them about middle position.
On the attached you can see output of Jim Williams generator with 5ns/DIV and 0.2V/DIV settings.
Its voltage peak seems low to me, so there will be some work yet, I think.
The more so as now it can't be done turn off the scale illumination.
If you remember how this thing started it is frustrating me and very funny indeed at the same time.
In this regard I'm the guilty for sure.
Main rule was violated, by myself, don't try to change more thing at once, if you are repair something.
So I have killed something around Q11 Q12 VR3 etc. on LVPS, while I implemented a socket and a heat sink for Q12.
Never mind.
I will repair it certainly, and now it is easy to change Q12, at least. :)
Dear David!
I now drinks a beer for your health!
Many thanks again!
continuation follows
Cheers!
https://hbpeza-ch3301.files.1drv.com/y2pqM36ild0SoAx9L65s2Zw2kVe1PWbfkv_JMa90kGqxU6RIcJoqbPzI-vMO02IwsUEh9qu4Q7ZOLHQsiwAaFNl3sldgo_8lCN3OYfXGT393kg/Img00820001_1.JPG?psid=1 (https://hbpeza-ch3301.files.1drv.com/y2pqM36ild0SoAx9L65s2Zw2kVe1PWbfkv_JMa90kGqxU6RIcJoqbPzI-vMO02IwsUEh9qu4Q7ZOLHQsiwAaFNl3sldgo_8lCN3OYfXGT393kg/Img00820001_1.JPG?psid=1)
Those confluent mottlings on photo are only reflections captured by camera.
(and some particles of dust)
Gyuri
-
I'm at 26.5 ns now with some settings on R12/C11.
So it is now, as you said.
-
I'm at 26.5 ns now with some settings on R12/C11.
So it is now, as you said.
Adjust R12 and C11 for best shape and declare victory!
On my Tektronix oscilloscopes, the equivalent adjustments make a noticeable difference in both the shape of the leading edge and visually on the CRT but it is not like anything will fail if they are adjusted wrong.
On the attached you can see output of Jim Williams generator with 5ns/DIV and 0.2V/DIV settings.
Its voltage peak seems low to me, so there will be some work yet, I think.
If you use a charge line on the pulse generator instead of the collector capacitor, then the pulse will have a flat or at least flatter top which will allow the height to be seen even if the oscilloscope is too slow which is the case here. It is much easier to use an edge with a flat top than a pulse to measure oscilloscope bandwidth.
-
Best shape and fastest rise time are antagonists of each other.
So you say, I must concentrate for best shape with no ringing, even if degraded risetime?
I will setting it according this.
I have repaired (again) screen illumination.
I was killed that I-dont-know-what-a-type power transistor on A16 board.
I'm just curious, is there anybody who know what type transistor is?
I know it is a NPN power darlington with about 70W TDP.
I have chosen something, let it be say TIP 142.
It works again now.
Altough probably I will never using it. :)
Regards,
Gyuri
-
Best shape and fastest rise time are antagonists of each other.
So you say, I must concentrate for best shape with no ringing, even if degraded risetime?
I will setting it according this.
It is true that you generally cannot have both at the same time but in this case, the rise time just has to be fast enough for unblanking to finish before the vertical signal reaches the CRT after traveling through the delay line. The service manual says to adjust R12 and C11 for best shape.
I have repaired (again) screen illumination.
I was killed that I-dont-know-what-a-type power transistor on A16 board.
I'm just curious, is there anybody who know what type transistor is?
I know it is a NPN power darlington with about 70W TDP.
I have chosen something, let it be say TIP 142.
It works again now.
Altough probably I will never using it. :)
I know there are indexes to convert HP part numbers to industry standard part numbers but I do not have one. I would have to study the circuit and see the original transistor.
-
I know there are indexes to convert HP part numbers to industry standard part numbers but I do not have one. I would have to study the circuit and see the original transistor.
It is not important, David.
It works excellently with TIP142 what I have choosed.
I only would like to know its type.
I have a HP parts number table, what I have found on net.
(I have attached it, thanks to Greg Davis who was the original uploader.)
But it doesn't helped.
I the SM its circuit number on A16 Service Sheet 3 is Q12, while in Replacable Parts list it is somehow confused, I suspect its number is Q14 HP P/N 1854-0553 or 0565
or something, not so readable, it is a TRANSISTOR NPN DAR. PO=70W FT=1MHz.
But it doesn't matter David, it is only an academic question.
I must have to lay back to bed now, with some ice.
(Without whiskey.) ;)
-
In this picture, readable perfectly:
https://public.dm2301.livefilestore.com/y2pVyE4gzcX4zFxMQJ2QLI_bdc0NDfFVSEckPNoScZ19jRtpvCOP10ZUkzVkKwXIPe5odSZ1wraBMFe-xSanhUmq-0z1w0c56puEjxHvNSa-bs/Img00300001_1.JPG?psid=1 (https://public.dm2301.livefilestore.com/y2pVyE4gzcX4zFxMQJ2QLI_bdc0NDfFVSEckPNoScZ19jRtpvCOP10ZUkzVkKwXIPe5odSZ1wraBMFe-xSanhUmq-0z1w0c56puEjxHvNSa-bs/Img00300001_1.JPG?psid=1)
Something Motorola, I think.
-
Yeah, and I have found a broken diode in this picture:
https://public.dm2301.livefilestore.com/y2p80Q22H6kCXXsfIP2rUYb2ENFlDDzDrnfz4mOGpnRjVkJAnL1gWcwoRlZxIgpF4OjwtaBjC4h95IsGqxp4cUyhPZrIdiwSOp9M2KDd27_MqY/Img00160001_1.JPG?psid=1 (https://public.dm2301.livefilestore.com/y2p80Q22H6kCXXsfIP2rUYb2ENFlDDzDrnfz4mOGpnRjVkJAnL1gWcwoRlZxIgpF4OjwtaBjC4h95IsGqxp4cUyhPZrIdiwSOp9M2KDd27_MqY/Img00160001_1.JPG?psid=1)
It is on the pic's left side, between that two spring steel sheet.
It is soldered on the pins of MAGx5 switch.
I'm don't know yet, what is it.
-
I think, I have found it:
-
And what does it do?
I suspect doing nothing, if we take into account the broken status..
Yes, I know it is a 2.37V zener.
But it causing what?
-
My bad, again!
it is 5.1 V zener actually.
No, it is not.
-
In this moment I really don't know. >:(
From that wrecked diode body, it is impossible to read its number.
But it was VR6 so it's a very special piece.
-
The zener diode has something to do with the vertical positioning range and also controls the vertical shift used to align the traces when x5 vertical magnification is used. A thorough checkout would have revealed a problem.
Something Motorola, I think.
It is definately a Motorola part but the number is an HP house number. "House number" is the term I learned for OEMed parts with custom numbers.
It is not important, David.
It works excellently with TIP142 what I have choosed.
I only would like to know its type.
I have a HP parts number table, what I have found on net.
(I have attached it, thanks to Greg Davis who was the original uploader.)
But it doesn't helped.
I the SM its circuit number on A16 Service Sheet 3 is Q12, while in Replacable Parts list it is somehow confused, I suspect its number is Q14 HP P/N 1854-0553 or 0565
or something, not so readable, it is a TRANSISTOR NPN DAR. PO=70W FT=1MHz.
I found the same problems. It is Q12 on the schematic but Q14 in the parts list and that is hard to read. The part number shown in the photo is obviously wrong because the wrong part was used to replace it.
The number in the parts list might be 1854-0663 which is TIP140 making your TIP142 an ideal replacement.
-
Well, it was VR6 actually, I think I will never get something similar like it was:
1902-3002 SZ10939-2 CD35526 DO7 2.37V 5% 400mW TC -0.074%
-
Thank you David,
It is good to me to know, that you are with me.
Yes, you absolutely right about Q12 part number, you have a good eye, I doesn't found it.
It is really that TIP140, so my more or less blind choice was absolutely good.
Well done, I like it!
-
In this moment I really don't know. >:(
From that wrecked diode body, it is impossible to read its number.
But it was VR6 so it's a very special piece.
The schematic and parts list say it is a 2.37 volt zener diode in a DO-7 (glass) package which is what your photo shows. The DO-7 package means that its power rating is about 500 milliwatts.
A 2.4 volt zener diode like a 1N5221, BZX55B2V4, or ZX2V4A should work fine in its place. Even 4 silicon diodes in series would work as a 2.4 volt diode.
-
Well, it was VR6 actually, I think I will never get something similar like it was
http://www.mival.ro/diode-znr-2-37v-1902-3002-18-en (http://www.mival.ro/diode-znr-2-37v-1902-3002-18-en)
-
Thanks for the reply guys!
That mival link is sensational, it is here in the neighborhood, I will contact to them! :)
Till then will be four diodes.
But it will be perfect if I can get original part.
-
Thanks for the reply guys!
That mival link is sensational, it is here in the neighborhood, I will contact to them! :)
Till then will be four diodes.
But it will be perfect if I can get original part.
Remember to reverse the 4 diodes. The zener diode operates in reverse breakdown to generate 2.4 volts but 4 silicon diodes in series will need to be forward biased to do the same job.
The 4 diodes in series will be less accurate and drift with temperature more but the oscilloscope should work fine.
-
Thanks for your clarification, David!
Maybe I have recognised it from myself, maybe not.
It can't be know, because I'm Wacky.
For example, I realized TIP142 has reverse pin order compared to that don't-know-what-is-it-darlington.
I have used only my brain! :)
BTW, I have got some 2,4V zeners.
So I have looked at once again that little part of SS4.
I have surprised in some extent, because I have recognised that simple fact:
At MAGx5 there is nothing to do with that precious 2.37V zener with special thermal coefficients.
Because in this setting it is shunted by MAGx5 switch!
May be it is only news for me. :)
It is certainly has there because of eliminate trace shifting.
I have observed some annoying trace shift on scope, soon as I have got it.
But I have thought, if it will be in a good shape, I will carry out some calibration, and it will be gone.
I have soldered a 2.4V zener to its place.
It certainly has some effect on trace shift.
But I haven't satisfied.
I have got about 2V through it in normal position.
And in the light of the exploring above, there is exactly 0V on it. (With MAGx5 pushed.)
It needs further examining.
I think, I drink now some further beers for health of David and myself and everbody else, and I will let it down a bit.
Now, I listen some music rather.
Happy weekend for all of us!
Kind regards,
Gyuri
-
As I getting more and more beer, I'm in a more and more in a philosophical mood.
I think, it is not news, most of us are a similar way.
I'm wondering, why David dealing with me?
Apart of that he is a very nice man, as I know him better and better.
(I think, at least.)
There are important things in the world.
One of them, the most important according to me, the knowledge.
(I don't know may be arts are more important, I'm hang in the balance, as always.)
And what about money and power, you may ask it?
My answer is, even a completely idiot can have money and power, while lacking any knowledge.
It is the worst and most hazardous combination I think.
These people worth nothing in my eyes, maybe double zero.
They even can share out all of the money they have.
They will be absolute nothing.
Nobody will remember of them.
BTW, you can not take money to the afterlife.
But what about knowledge?
If you have much knowledge, and you share it, you will have increasingly more and more knowledge hereby.
BTW, you can not take knowledge to the afterlife.
So if you share it, it will be remain with us, and you will be worth more and more.
And posterity will be grateful to you.
I know, all of above about me.
Perhaps not only about myself, I hope.
I don't know, people like Jim Williams, Nelson Pass, David William Hess etc., when dealing with people like me, Wacky Gyuri thinks things like above?
I do not think so, and that's the beauty of it.
Excuse me, if I seems to be muddling, please take account my crappy english.
And those beers also.
Kind regards,
Wacky Gyuri
-
I think I have got it!
What do you think, David?
Sorry for blurred picture.
-
For example, I realized TIP142 has reverse pin order compared to that don't-know-what-is-it-darlington.
Which means that the TIP140 series was not the original transistor. I forgot to mention the pinout issue which shows up a lot in these old instruments. Old transistors in the TO-127 package are ECB so TO-126 or TO-225 replacements are needed if you want to keep the same pinout but these packages are limited to about 40 watts which should be sufficient in this case. I might use a 2N6039 or similar.
At MAGx5 there is nothing to do with that precious 2.37V zener with special thermal coefficients.
Because in this setting it is shunted by MAGx5 switch!
Yes, the zener diode only matters in normal mode.
I think I have got it!
What do you think, David?
Sorry for blurred picture.
I have made worse pictures.
I looks great. Like I said earlier, the shape is not critical for proper operation.
-
TIP142 is perfect at this position I think.
I've made some progress.
I've performed calibration 5-31 vertical amplifier balance adjustment and 5-32 position and sync balance adjustment.
Annoying trace shift, when changing Volts/Div almost has gone.
I think there are much more things what need to be calibrated.
I don't have a sine generator to about 100 MHz.
Bradley 192 has only fixed outputs at 20 MHz, 50MHz and 100MHz.
-
The postman, who is usually ringing twice, now only once. :)
He haven't had more time.
So I have got the original service manual now for HP1745A.
I have ordered it from ebay seller teoutlet.
It is in very good condition.
It is much better to read, than crappy .pdf, which is downloadable.
I'm wondering I'm going to make a better copy about it,
and I will send it to Keystone, to put it up to their website.
Regards,
Gyuri
-
So I have got the original service manual now for HP1745A.
I have ordered it from ebay seller teoutlet.
It is in very good condition.
It is much better to read, than crappy .pdf, which is downloadable.
I'm wondering I'm going to make a better copy about it,
and I will send it to Keystone, to put it up to their website.
I often end up extracting pages from PDF scans, cleaning them up, and then printing them out so I can mark them up with colored pens as I work. For paper manuals that I have, I sometimes scan in pages so I can print them out and mark up the copies as well.
-
I wasn't here for some time.
Lately I've purchased a Leo Bodnar 40ps pulser.
It is a very useful little gem, I like it very much!
Of course I've tested my HP1745A scope with it.
I've got about 130MHz bandwith on CH1, it is quite good I think.
However, I've got only 70MHz on CH2.
You can say, don't worry about it, then there is a 100MHz average.
Yet, I think I will check some things at CH2, someday.
Kind regards,
Gyuri