Calibrate the time base, or horizontal amplifier?


emissionlabs
 

Hello,

I need a little guidance, and don't want to make misakes while fixing a 5113 Dual Beam Scope. The 3400V high voltage was gone, and when I fixed that, the chopper circuit appeared bad, and when then was fixed, there was only one beam. Not mentioning a few smaller defects. But I have it all fixed now, and it works very good. Sharp traces, and very nice geometry.

I have adjusted the high voltage nicely at 3400V. +/-30V is as good as the single turn pot meter can do. +/-170V is allowed. Probably the HV is not exactly the same as it was before, and this affects the picture size, and if so, horizontal and vertical gain calibration needs to be done. I have two time base modules with it, which do not show the same result on the calibrator of the scope itself. So at least one of them is out of calibration. The horizontal trace is a tiny bit too slow, which can be caused by the horizontal amplifier being set wrong, or the time base being set wrong. I do not have a time mark generator, but I have an accurate hewlett packard pulse generator, which probably can do the same. And also we have 50 Hz here, and the internal calibrator shows exactly 100Hz. So a good reference is not the problem.

At this point I get stuck a little bit because I could set the horizontal gain a little bit higher, or set the time base a little bit faster. Both would "cure" the problem. But it feels wrong to set the time base faster, while in fact the horizontal amplifier would be out of calibration, or vice versa.

I have neve done this before, and I don't want to mess it up. It is such a crispy sharp CRT for the rest if it, I would like to get this scope in good condition. Perhaps someone has good advice for me. Thanks in advance!


redarlington
 

You'll need a time mark generator like a Tek 2901. But be very careful
with this. At the last second before adjusting the time base, you may
start to wonder how accurate the oscillator in the generator is. Then you
might fall down that slippery slope of time nuttery. It's a curse or a
blessing depending on your personality. Sometimes it's both.

-Bob N3XKB

On Sun, Jul 11, 2021, 3:29 PM emissionlabs <jac@406777.de> wrote:

Hello,

I need a little guidance, and don't want to make misakes while fixing a
5113 Dual Beam Scope. The 3400V high voltage was gone, and when I fixed
that, the chopper circuit appeared bad, and when then was fixed, there
was only one beam. Not mentioning a few smaller defects. But I have it all
fixed now, and it works very good. Sharp traces, and very nice geometry.

I have adjusted the high voltage nicely at 3400V. +/-30V is as good as
the single turn pot meter can do. +/-170V is allowed. Probably the HV is
not exactly the same as it was before, and this affects the picture size,
and if so, horizontal and vertical gain calibration needs to be done. I
have two time base modules with it, which do not show the same result on
the calibrator of the scope itself. So at least one of them is out of
calibration. The horizontal trace is a tiny bit too slow, which can be
caused by the horizontal amplifier being set wrong, or the time base being
set wrong. I do not have a time mark generator, but I have an accurate
hewlett packard pulse generator, which probably can do the same. And
also we have 50 Hz here, and the internal calibrator shows exactly 100Hz.
So a good reference is not the problem.

At this point I get stuck a little bit because I could set the horizontal
gain a little bit higher, or set the time base a little bit faster. Both
would "cure" the problem. But it feels wrong to set the time base
faster, while in fact the horizontal amplifier would be out of
calibration, or vice versa.

I have neve done this before, and I don't want to mess it up. It is such
a crispy sharp CRT for the rest if it, I would like to get this scope in
good condition. Perhaps someone has good advice for me. Thanks in
advance!







Zentronics42@...
 

No need to tweak the pots!! If you just repaired the HV supply this would only effect brightness in the scope it should not really mess with the time base at all or vert amp at all. Now this goes with out saying that adjusting the reference in the power supply with cause the need for a FULL recalibration / adjustment of the scope. This can be a very deep rabbit hole. I have made a few videos on You Tube about this topic. Mainly the 7000 scopes. I am currently working on a 400 series. I can post a link if there is any interest. Remember on a 400 series spec is 3%. Not sure what it is on a 5000 but it might still be "in spec"

The type 184 time base crystal is not the most accurate thing on the planet. The type 184 or TG501 would be around the time frame for a 5000 series scope. I have only seen one of the older time base unit ever. So they are even more rare. However The reference crystal in the time base 184 has about a 50 Hz swing on 10 Mhz. +25 to -25 Hz. This is mainly caused by the ovenized crystal heating and cooling as the oven is cycling. (REALLY annoying to adjust.) The tg501 is not near as bad to adjust.

If I can be of help let me know.

Zen

-----Original Message-----
From: TekScopes@groups.io <TekScopes@groups.io> On Behalf Of redarlington
Sent: Sunday, July 11, 2021 5:49 PM
To: TekScopes@groups.io
Subject: Re: [TekScopes] Calibrate the time base, or horizontal amplifier?

You'll need a time mark generator like a Tek 2901. But be very careful with this. At the last second before adjusting the time base, you may
start to wonder how accurate the oscillator in the generator is. Then you
might fall down that slippery slope of time nuttery. It's a curse or a
blessing depending on your personality. Sometimes it's both.

-Bob N3XKB

On Sun, Jul 11, 2021, 3:29 PM emissionlabs <jac@406777.de> wrote:

Hello,

I need a little guidance, and don't want to make misakes while fixing
a
5113 Dual Beam Scope. The 3400V high voltage was gone, and when I
fixed that, the chopper circuit appeared bad, and when then was
fixed, there was only one beam. Not mentioning a few smaller defects. But I have it all
fixed now, and it works very good. Sharp traces, and very nice geometry.

I have adjusted the high voltage nicely at 3400V. +/-30V is as good
as the single turn pot meter can do. +/-170V is allowed. Probably
the HV is not exactly the same as it was before, and this affects the picture size,
and if so, horizontal and vertical gain calibration needs to be done. I
have two time base modules with it, which do not show the same result on
the calibrator of the scope itself. So at least one of them is out of
calibration. The horizontal trace is a tiny bit too slow, which can
be caused by the horizontal amplifier being set wrong, or the time base being
set wrong. I do not have a time mark generator, but I have an accurate
hewlett packard pulse generator, which probably can do the same. And
also we have 50 Hz here, and the internal calibrator shows exactly 100Hz.
So a good reference is not the problem.

At this point I get stuck a little bit because I could set the
horizontal gain a little bit higher, or set the time base a little bit faster. Both
would "cure" the problem. But it feels wrong to set the time base
faster, while in fact the horizontal amplifier would be out of
calibration, or vice versa.

I have neve done this before, and I don't want to mess it up. It is
such a crispy sharp CRT for the rest if it, I would like to get this scope in
good condition. Perhaps someone has good advice for me. Thanks in
advance!







Harvey White
 

Not sure that they ever made one for the 5000 series, but an input standardizer would give you direct access to the frame, which would be fine if you had a known AC amplitude to calibrate the frame.

You'll only be able to get close, but remember that a scope is good to 3 to 5% anyway.

I'd be tempted to find an accurate AC source.  A good meter would help, square waves would be best, but you'd need to adjust for the waveform if you're using a meter.

If you could feed a signal directly into the frame, then  that allows you to calibrate both the H and V amplifiers.

If you have only the one frame, or only the one plugin, then I'd be tempted to take the mainframe's calibration on the Y channel as valid (arbitrary, I know), then adjust the vertical plugin to read the appropriate p-p value on the scope.

Then swap the H and V plugins (should be allowed, please check), but then adjust the H frame gain to the same deflection.

Then swap the plugins back to H and V.  Adjust the H timebase for the appropriate scale.

that ought to get you close.

That's what I'd do.  Adjusting the power supplies pretty much requires a recalibration.

A time mark generator can be made with some TTL logic and a TTL oscillator chip.

There are some amplitude calibrators used for a DMM that would give you at least a good 10 volt reference, so that range would be ok, but you'd find it difficult to check the other ranges.

Harvey

On 7/11/2021 5:28 PM, emissionlabs wrote:
Hello,

I need a little guidance, and don't want to make misakes while fixing a 5113 Dual Beam Scope. The 3400V high voltage was gone, and when I fixed that, the chopper circuit appeared bad, and when then was fixed, there was only one beam. Not mentioning a few smaller defects. But I have it all fixed now, and it works very good. Sharp traces, and very nice geometry.

I have adjusted the high voltage nicely at 3400V. +/-30V is as good as the single turn pot meter can do. +/-170V is allowed. Probably the HV is not exactly the same as it was before, and this affects the picture size, and if so, horizontal and vertical gain calibration needs to be done. I have two time base modules with it, which do not show the same result on the calibrator of the scope itself. So at least one of them is out of calibration. The horizontal trace is a tiny bit too slow, which can be caused by the horizontal amplifier being set wrong, or the time base being set wrong. I do not have a time mark generator, but I have an accurate hewlett packard pulse generator, which probably can do the same. And also we have 50 Hz here, and the internal calibrator shows exactly 100Hz. So a good reference is not the problem.

At this point I get stuck a little bit because I could set the horizontal gain a little bit higher, or set the time base a little bit faster. Both would "cure" the problem. But it feels wrong to set the time base faster, while in fact the horizontal amplifier would be out of calibration, or vice versa.

I have neve done this before, and I don't want to mess it up. It is such a crispy sharp CRT for the rest if it, I would like to get this scope in good condition. Perhaps someone has good advice for me. Thanks in advance!





Zentronics42@...
 

Harvey,
One of the main calibration issues that requires "exotic" signals Is the frequency calibration of the amplifiers. Having these be off can nuke the bandwidth of the scope. 465B 3db down 75-80 Mhz instead of 135 Mhz. (Yes I know 135 is above the 100 Mhz rating of the scope but that is what mine caled up to when properly adjusted.) These are done with fast edge square waves. 1ns rise time for 465 and under, ps rise time for 485's and faster for even further up the BW scale. But in terms of an AC calibrator a fluke 5200A would not be bad with its amp 5205A can go to 1500 Vac. Which is a bit spicey. If using a meter something that needs to be taken in to account most good meters work in AC RMS scopes work in PK-PK so appropriate calculations need to be taken in to account when setting up the signal source. In terms of calibration a functioning and calibrated TG501, PG506, and SG503 will cover 99% of the scopes up to 100 Mhz. Over that will need "exotic" signal sources tunnel diode pulsers and such. I will admit I have seen the bottom of that rabbit hole it is kind of on the deep side.

Zen

-----Original Message-----
From: TekScopes@groups.io <TekScopes@groups.io> On Behalf Of Harvey White
Sent: Sunday, July 11, 2021 6:14 PM
To: TekScopes@groups.io
Subject: Re: [TekScopes] Calibrate the time base, or horizontal amplifier?

Not sure that they ever made one for the 5000 series, but an input standardizer would give you direct access to the frame, which would be fine if you had a known AC amplitude to calibrate the frame.

You'll only be able to get close, but remember that a scope is good to 3 to 5% anyway.

I'd be tempted to find an accurate AC source. A good meter would help, square waves would be best, but you'd need to adjust for the waveform if you're using a meter.

If you could feed a signal directly into the frame, then that allows you to calibrate both the H and V amplifiers.

If you have only the one frame, or only the one plugin, then I'd be tempted to take the mainframe's calibration on the Y channel as valid (arbitrary, I know), then adjust the vertical plugin to read the appropriate p-p value on the scope.

Then swap the H and V plugins (should be allowed, please check), but then adjust the H frame gain to the same deflection.

Then swap the plugins back to H and V. Adjust the H timebase for the appropriate scale.

that ought to get you close.

That's what I'd do. Adjusting the power supplies pretty much requires a recalibration.

A time mark generator can be made with some TTL logic and a TTL oscillator chip.

There are some amplitude calibrators used for a DMM that would give you at least a good 10 volt reference, so that range would be ok, but you'd find it difficult to check the other ranges.

Harvey


On 7/11/2021 5:28 PM, emissionlabs wrote:
Hello,

I need a little guidance, and don't want to make misakes while fixing a 5113 Dual Beam Scope. The 3400V high voltage was gone, and when I fixed that, the chopper circuit appeared bad, and when then was fixed, there was only one beam. Not mentioning a few smaller defects. But I have it all fixed now, and it works very good. Sharp traces, and very nice geometry.

I have adjusted the high voltage nicely at 3400V. +/-30V is as good as the single turn pot meter can do. +/-170V is allowed. Probably the HV is not exactly the same as it was before, and this affects the picture size, and if so, horizontal and vertical gain calibration needs to be done. I have two time base modules with it, which do not show the same result on the calibrator of the scope itself. So at least one of them is out of calibration. The horizontal trace is a tiny bit too slow, which can be caused by the horizontal amplifier being set wrong, or the time base being set wrong. I do not have a time mark generator, but I have an accurate hewlett packard pulse generator, which probably can do the same. And also we have 50 Hz here, and the internal calibrator shows exactly 100Hz. So a good reference is not the problem.

At this point I get stuck a little bit because I could set the horizontal gain a little bit higher, or set the time base a little bit faster. Both would "cure" the problem. But it feels wrong to set the time base faster, while in fact the horizontal amplifier would be out of calibration, or vice versa.

I have neve done this before, and I don't want to mess it up. It is such a crispy sharp CRT for the rest if it, I would like to get this scope in good condition. Perhaps someone has good advice for me. Thanks in advance!







Harvey White
 

I'm not sure how well I can hear you, your voice is a bit echo-y from the bottom of that rabbit hole.

IMHO, what most people seem to worry about is horizontal and vertical accuracy, time and amplitude.  You're quite correct that checking out the frequency response is difficult.  The SG502 and SG503 are constant amplitude generators (at least the SG503 and SG504), and those can be used for frequency calibration.  There are workarounds for fast risetime squarewave generators, which can be built easily.  How much is a Fluke 5200?  Thought they needed an amplifier for the high voltage output on AC.  I personally can do DC 0-2 amps and 0-1KV.  AC is a problem.

I do wonder how many people will calibrate a scope to that extent, unless the scope simply won't work without those sources. (the DM5101 *needs* a 600 volt measured AC source, but it's not a scope).

Harvey

On 7/11/2021 6:26 PM, Zentronics42@gmail.com wrote:
Harvey,
One of the main calibration issues that requires "exotic" signals Is the frequency calibration of the amplifiers. Having these be off can nuke the bandwidth of the scope. 465B 3db down 75-80 Mhz instead of 135 Mhz. (Yes I know 135 is above the 100 Mhz rating of the scope but that is what mine caled up to when properly adjusted.) These are done with fast edge square waves. 1ns rise time for 465 and under, ps rise time for 485's and faster for even further up the BW scale. But in terms of an AC calibrator a fluke 5200A would not be bad with its amp 5205A can go to I"1500 Vac. Which is a bit spicey. If using a meter something that needs to be taken in to account most good meters work in AC RMS scopes work in PK-PK so appropriate calculations need to be taken in to account when setting up the signal source. In terms of calibration a functioning and calibrated TG501, PG506, and SG503 will cover 99% of the scopes up to 100 Mhz. Over that will need "exotic" signal sources tunnel diode pulsers and such. I will admit I have seen the bottom of that rabbit hole it is kind of on the deep side.

Zen

-----Original Message-----
From: TekScopes@groups.io <TekScopes@groups.io> On Behalf Of Harvey White
Sent: Sunday, July 11, 2021 6:14 PM
To: TekScopes@groups.io
Subject: Re: [TekScopes] Calibrate the time base, or horizontal amplifier?

Not sure that they ever made one for the 5000 series, but an input standardizer would give you direct access to the frame, which would be fine if you had a known AC amplitude to calibrate the frame.

You'll only be able to get close, but remember that a scope is good to 3 to 5% anyway.

I'd be tempted to find an accurate AC source. A good meter would help, square waves would be best, but you'd need to adjust for the waveform if you're using a meter.

If you could feed a signal directly into the frame, then that allows you to calibrate both the H and V amplifiers.

If you have only the one frame, or only the one plugin, then I'd be tempted to take the mainframe's calibration on the Y channel as valid (arbitrary, I know), then adjust the vertical plugin to read the appropriate p-p value on the scope.

Then swap the H and V plugins (should be allowed, please check), but then adjust the H frame gain to the same deflection.

Then swap the plugins back to H and V. Adjust the H timebase for the appropriate scale.

that ought to get you close.

That's what I'd do. Adjusting the power supplies pretty much requires a recalibration.

A time mark generator can be made with some TTL logic and a TTL oscillator chip.

There are some amplitude calibrators used for a DMM that would give you at least a good 10 volt reference, so that range would be ok, but you'd find it difficult to check the other ranges.

Harvey


On 7/11/2021 5:28 PM, emissionlabs wrote:
Hello,

I need a little guidance, and don't want to make misakes while fixing a 5113 Dual Beam Scope. The 3400V high voltage was gone, and when I fixed that, the chopper circuit appeared bad, and when then was fixed, there was only one beam. Not mentioning a few smaller defects. But I have it all fixed now, and it works very good. Sharp traces, and very nice geometry.

I have adjusted the high voltage nicely at 3400V. +/-30V is as good as the single turn pot meter can do. +/-170V is allowed. Probably the HV is not exactly the same as it was before, and this affects the picture size, and if so, horizontal and vertical gain calibration needs to be done. I have two time base modules with it, which do not show the same result on the calibrator of the scope itself. So at least one of them is out of calibration. The horizontal trace is a tiny bit too slow, which can be caused by the horizontal amplifier being set wrong, or the time base being set wrong. I do not have a time mark generator, but I have an accurate hewlett packard pulse generator, which probably can do the same. And also we have 50 Hz here, and the internal calibrator shows exactly 100Hz. So a good reference is not the problem.

At this point I get stuck a little bit because I could set the horizontal gain a little bit higher, or set the time base a little bit faster. Both would "cure" the problem. But it feels wrong to set the time base faster, while in fact the horizontal amplifier would be out of calibration, or vice versa.

I have neve done this before, and I don't want to mess it up. It is such a crispy sharp CRT for the rest if it, I would like to get this scope in good condition. Perhaps someone has good advice for me. Thanks in advance!















Zentronics42@...
 

Hi Harvey,
Well if there is a list of people that calibrate a scope to that extent. I will have to add my name to the list. I ended up building the lab based on the 7000 series scopes (7603 X2,7854,and 7904, and many associated plugins) but I had a chicken and egg issue. I got the scopes and plugins from eBay cheap but I did not trust them. A while later I ended up with both sets of cal standards the TM500 and the type units I still have a bit of work to do on the SG504 the high range is not working but low is. So I have done the full calibrations and everything is up to spec. It is a DEEEEEP rabbit hole that starts at a SG503 and ends at a type 284 and hits time mark and amplitude calibrators along the way. However in the service manuals the low frequency and high frequency compensations of the amps are done with square waves or various rise times and frequencies mostly around 1khz,10khz,100khz, rise times 1ns and higher for the high frequency compensations. I had to cal the calibrators before I used them so bas accuracy is against a Keithley 6500. All this was made reasonable when I called Tek for a calibration request. For a 465 they quoted me $100.00. for the 7603 frame $800 for a 485 $1200.00. I did not even ask about the 7854 and the 7904A. but given that I have done 3 7603’s a 7704A a few 465’s and 2 485’s as well as a partial on a 7854 ( need to recreate a board to complete that one unless the thermal calibration of the verticals is not that important don’t know yet. ) the gear was worth the investment. And the knowledge gained in doing the calibrations has been priceless.

In terms of AD/DC I have a fluke 343A for DC and a Fluke 5200A with the 5205A amplifier. The good news is for your needs you can get just the 5205A as it has a fixed gain and put the correct signal in the front end of it. If I remember correctly it is a gain of 100 so 60 V signal will output 600 Vac. As to for cost I don’t know. I paid $700 for both units broken. The amp was fine, however the 5200A -190V rail had burned up and the power amp in the 5200A needed to be rebuilt. Needed these to calibrate the 6 DM502’a. 6 DMMs is overkill however when working on the Tek multi rail power supplies it can be really convenient to watch them ALL at once. On thing to note on the meter cal does it need variable frequency? The 5200A with the 5205A can do 0-1200 Vac 10 Hz to 1 Mhz. Haven't needed to play with 1200Vac @ 1Mhz yet that is a bit spicey.

Zen

-----Original Message-----
From: TekScopes@groups.io <TekScopes@groups.io> On Behalf Of Harvey White
Sent: Sunday, July 11, 2021 8:56 PM
To: TekScopes@groups.io
Subject: Re: [TekScopes] Calibrate the time base, or horizontal amplifier?

I'm not sure how well I can hear you, your voice is a bit echo-y from the bottom of that rabbit hole.

IMHO, what most people seem to worry about is horizontal and vertical accuracy, time and amplitude. You're quite correct that checking out the frequency response is difficult. The SG502 and SG503 are constant amplitude generators (at least the SG503 and SG504), and those can be used for frequency calibration. There are workarounds for fast risetime squarewave generators, which can be built easily. How much is a Fluke 5200? Thought they needed an amplifier for the high voltage output on AC. I personally can do DC 0-2 amps and 0-1KV. AC is a problem.

I do wonder how many people will calibrate a scope to that extent, unless the scope simply won't work without those sources. (the DM5101
*needs* a 600 volt measured AC source, but it's not a scope).

Harvey


On 7/11/2021 6:26 PM, Zentronics42@gmail.com wrote:
Harvey,
One of the main calibration issues that requires "exotic" signals Is the frequency calibration of the amplifiers. Having these be off can nuke the bandwidth of the scope. 465B 3db down 75-80 Mhz instead of 135 Mhz. (Yes I know 135 is above the 100 Mhz rating of the scope but that is what mine caled up to when properly adjusted.) These are done with fast edge square waves. 1ns rise time for 465 and under, ps rise time for 485's and faster for even further up the BW scale. But in terms of an AC calibrator a fluke 5200A would not be bad with its amp 5205A can go to I"1500 Vac. Which is a bit spicey. If using a meter something that needs to be taken in to account most good meters work in AC RMS scopes work in PK-PK so appropriate calculations need to be taken in to account when setting up the signal source. In terms of calibration a functioning and calibrated TG501, PG506, and SG503 will cover 99% of the scopes up to 100 Mhz. Over that will need "exotic" signal sources tunnel diode pulsers and such. I will admit I have seen the bottom of that rabbit hole it is kind of on the deep side.

Zen

-----Original Message-----
From: TekScopes@groups.io <TekScopes@groups.io> On Behalf Of Harvey
White
Sent: Sunday, July 11, 2021 6:14 PM
To: TekScopes@groups.io
Subject: Re: [TekScopes] Calibrate the time base, or horizontal amplifier?

Not sure that they ever made one for the 5000 series, but an input standardizer would give you direct access to the frame, which would be fine if you had a known AC amplitude to calibrate the frame.

You'll only be able to get close, but remember that a scope is good to 3 to 5% anyway.

I'd be tempted to find an accurate AC source. A good meter would help, square waves would be best, but you'd need to adjust for the waveform if you're using a meter.

If you could feed a signal directly into the frame, then that allows you to calibrate both the H and V amplifiers.

If you have only the one frame, or only the one plugin, then I'd be tempted to take the mainframe's calibration on the Y channel as valid (arbitrary, I know), then adjust the vertical plugin to read the appropriate p-p value on the scope.

Then swap the H and V plugins (should be allowed, please check), but then adjust the H frame gain to the same deflection.

Then swap the plugins back to H and V. Adjust the H timebase for the appropriate scale.

that ought to get you close.

That's what I'd do. Adjusting the power supplies pretty much requires a recalibration.

A time mark generator can be made with some TTL logic and a TTL oscillator chip.

There are some amplitude calibrators used for a DMM that would give you at least a good 10 volt reference, so that range would be ok, but you'd find it difficult to check the other ranges.

Harvey


On 7/11/2021 5:28 PM, emissionlabs wrote:
Hello,

I need a little guidance, and don't want to make misakes while fixing a 5113 Dual Beam Scope. The 3400V high voltage was gone, and when I fixed that, the chopper circuit appeared bad, and when then was fixed, there was only one beam. Not mentioning a few smaller defects. But I have it all fixed now, and it works very good. Sharp traces, and very nice geometry.

I have adjusted the high voltage nicely at 3400V. +/-30V is as good as the single turn pot meter can do. +/-170V is allowed. Probably the HV is not exactly the same as it was before, and this affects the picture size, and if so, horizontal and vertical gain calibration needs to be done. I have two time base modules with it, which do not show the same result on the calibrator of the scope itself. So at least one of them is out of calibration. The horizontal trace is a tiny bit too slow, which can be caused by the horizontal amplifier being set wrong, or the time base being set wrong. I do not have a time mark generator, but I have an accurate hewlett packard pulse generator, which probably can do the same. And also we have 50 Hz here, and the internal calibrator shows exactly 100Hz. So a good reference is not the problem.

At this point I get stuck a little bit because I could set the horizontal gain a little bit higher, or set the time base a little bit faster. Both would "cure" the problem. But it feels wrong to set the time base faster, while in fact the horizontal amplifier would be out of calibration, or vice versa.

I have neve done this before, and I don't want to mess it up. It is such a crispy sharp CRT for the rest if it, I would like to get this scope in good condition. Perhaps someone has good advice for me. Thanks in advance!
















Harvey White
 

I'm not sure if I'll go that far.  I have some 468's, two 5000 series scopes, a 7704a, 7904, and a 7103 and 7104.  (let alone a few other scopes).  I've got a PG506, TG501, SG502 and SG503. I have the signal standardizers for up to 400 Mhz (the -01, I think).  Oddly enough, I have a fluke 343A, and I think it's a Fluke 331A voltage/current calibrator as well.

I have a very old, and very odd weston meter calibrator that was designed for calibrating moving needle meters.  The nice thing is that it will put out the 600 or so volts AC frequently needed to calibrate a DVM.  Using a "calibrated" meter allows duplicating results.  Don't think that the DM5101 needs variable frequency, but likely 400 Hz or so.  I have to completely redesign the CPU board to get any of mine to work, another project, though.

Too busy writing display drivers for LCD displays (ARM) for the moment, although that work will go into a few projects.

Harvey

On 7/11/2021 10:40 PM, Zentronics42@gmail.com wrote:
Hi Harvey,
Well if there is a list of people that calibrate a scope to that extent. I will have to add my name to the list. I ended up building the lab based on the 7000 series scopes (7603 X2,7854,and 7904, and many associated plugins) but I had a chicken and egg issue. I got the scopes and plugins from eBay cheap but I did not trust them. A while later I ended up with both sets of cal standards the TM500 and the type units I still have a bit of work to do on the SG504 the high range is not working but low is. So I have done the full calibrations and everything is up to spec. It is a DEEEEEP rabbit hole that starts at a SG503 and ends at a type 284 and hits time mark and amplitude calibrators along the way. However in the service manuals the low frequency and high frequency compensations of the amps are done with square waves or various rise times and frequencies mostly around 1khz,10khz,100khz, rise times 1ns and higher for the high frequency compensations. I had to cal the calibrators before I used them so bas accuracy is against a Keithley 6500. All this was made reasonable when I called Tek for a calibration request. For a 465 they quoted me $100.00. for the 7603 frame $800 for a 485 $1200.00. I did not even ask about the 7854 and the 7904A. but given that I have done 3 7603’s a 7704A a few 465’s and 2 485’s as well as a partial on a 7854 ( need to recreate a board to complete that one unless the thermal calibration of the verticals is not that important don’t know yet. ) the gear was worth the investment. And the knowledge gained in doing the calibrations has been priceless.

In terms of AD/DC I have a fluke 343A for DC and a Fluke 5200A with the 5205A amplifier. The good news is for your needs you can get just the 5205A as it has a fixed gain and put the correct signal in the front end of it. If I remember correctly it is a gain of 100 so 60 V signal will output 600 Vac. As to for cost I don’t know. I paid $700 for both units broken. The amp was fine, however the 5200A -190V rail had burned up and the power amp in the 5200A needed to be rebuilt. Needed these to calibrate the 6 DM502’a. 6 DMMs is overkill however when working on the Tek multi rail power supplies it can be really convenient to watch them ALL at once. On thing to note on the meter cal does it need variable frequency? The 5200A with the 5205A can do 0-1200 Vac 10 Hz to 1 Mhz. Haven't needed to play with 1200Vac @ 1Mhz yet that is a bit spicey.

Zen
-----Original Message-----
From: TekScopes@groups.io <TekScopes@groups.io> On Behalf Of Harvey White
Sent: Sunday, July 11, 2021 8:56 PM
To: TekScopes@groups.io
Subject: Re: [TekScopes] Calibrate the time base, or horizontal amplifier?

I'm not sure how well I can hear you, your voice is a bit echo-y from the bottom of that rabbit hole.

IMHO, what most people seem to worry about is horizontal and vertical accuracy, time and amplitude. You're quite correct that checking out the frequency response is difficult. The SG502 and SG503 are constant amplitude generators (at least the SG503 and SG504), and those can be used for frequency calibration. There are workarounds for fast risetime squarewave generators, which can be built easily. How much is a Fluke 5200? Thought they needed an amplifier for the high voltage output on AC. I personally can do DC 0-2 amps and 0-1KV. AC is a problem.

I do wonder how many people will calibrate a scope to that extent, unless the scope simply won't work without those sources. (the DM5101
*needs* a 600 volt measured AC source, but it's not a scope).

Harvey


On 7/11/2021 6:26 PM, Zentronics42@gmail.com wrote:
Harvey,
One of the main calibration issues that requires "exotic" signals Is the frequency calibration of the amplifiers. Having these be off can nuke the bandwidth of the scope. 465B 3db down 75-80 Mhz instead of 135 Mhz. (Yes I know 135 is above the 100 Mhz rating of the scope but that is what mine caled up to when properly adjusted.) These are done with fast edge square waves. 1ns rise time for 465 and under, ps rise time for 485's and faster for even further up the BW scale. But in terms of an AC calibrator a fluke 5200A would not be bad with its amp 5205A can go to I"1500 Vac. Which is a bit spicey. If using a meter something that needs to be taken in to account most good meters work in AC RMS scopes work in PK-PK so appropriate calculations need to be taken in to account when setting up the signal source. In terms of calibration a functioning and calibrated TG501, PG506, and SG503 will cover 99% of the scopes up to 100 Mhz. Over that will need "exotic" signal sources tunnel diode pulsers and such. I will admit I have seen the bottom of that rabbit hole it is kind of on the deep side.

Zen

-----Original Message-----
From: TekScopes@groups.io <TekScopes@groups.io> On Behalf Of Harvey
White
Sent: Sunday, July 11, 2021 6:14 PM
To: TekScopes@groups.io
Subject: Re: [TekScopes] Calibrate the time base, or horizontal amplifier?

Not sure that they ever made one for the 5000 series, but an input standardizer would give you direct access to the frame, which would be fine if you had a known AC amplitude to calibrate the frame.

You'll only be able to get close, but remember that a scope is good to 3 to 5% anyway.

I'd be tempted to find an accurate AC source. A good meter would help, square waves would be best, but you'd need to adjust for the waveform if you're using a meter.

If you could feed a signal directly into the frame, then that allows you to calibrate both the H and V amplifiers.

If you have only the one frame, or only the one plugin, then I'd be tempted to take the mainframe's calibration on the Y channel as valid (arbitrary, I know), then adjust the vertical plugin to read the appropriate p-p value on the scope.

Then swap the H and V plugins (should be allowed, please check), but then adjust the H frame gain to the same deflection.

Then swap the plugins back to H and V. Adjust the H timebase for the appropriate scale.

that ought to get you close.

That's what I'd do. Adjusting the power supplies pretty much requires a recalibration.

A time mark generator can be made with some TTL logic and a TTL oscillator chip.

There are some amplitude calibrators used for a DMM that would give you at least a good 10 volt reference, so that range would be ok, but you'd find it difficult to check the other ranges.

Harvey


On 7/11/2021 5:28 PM, emissionlabs wrote:
Hello,

I need a little guidance, and don't want to make misakes while fixing a 5113 Dual Beam Scope. The 3400V high voltage was gone, and when I fixed that, the chopper circuit appeared bad, and when then was fixed, there was only one beam. Not mentioning a few smaller defects. But I have it all fixed now, and it works very good. Sharp traces, and very nice geometry.

I have adjusted the high voltage nicely at 3400V. +/-30V is as good as the single turn pot meter can do. +/-170V is allowed. Probably the HV is not exactly the same as it was before, and this affects the picture size, and if so, horizontal and vertical gain calibration needs to be done. I have two time base modules with it, which do not show the same result on the calibrator of the scope itself. So at least one of them is out of calibration. The horizontal trace is a tiny bit too slow, which can be caused by the horizontal amplifier being set wrong, or the time base being set wrong. I do not have a time mark generator, but I have an accurate hewlett packard pulse generator, which probably can do the same. And also we have 50 Hz here, and the internal calibrator shows exactly 100Hz. So a good reference is not the problem.

At this point I get stuck a little bit because I could set the horizontal gain a little bit higher, or set the time base a little bit faster. Both would "cure" the problem. But it feels wrong to set the time base faster, while in fact the horizontal amplifier would be out of calibration, or vice versa.

I have neve done this before, and I don't want to mess it up. It is such a crispy sharp CRT for the rest if it, I would like to get this scope in good condition. Perhaps someone has good advice for me. Thanks in advance!
























Peter H
 

The "official" way is to use a 067-0680-00 calibration plug-in (https://w140.com/tekwiki/wiki/067-0680-00, the 067-0587-xx is the same for the 7000 series) to adjust the mainframe first.
Without one, you could in principle feed a signal in through the amplifier, and adjust its amplitude (as measured on the scope's backplane) to the specified V/div - but you'd have to be really sure you know how to get this right.

The '680 produces a stair-step signal that really helps getting the geometry right, of course a suitably configured arbitrary generator could be substituted.

I would still recommend the '680, maybe you can borrow one from someone on the list.


emissionlabs
 

Thank you for the many inputs.

I am looking for a way to calibrate this 5113 scope without dedicated calibrators. Otherwise it becomes too much for me, just for this one time action. I can use an Agilent 33120A digital controlled oscillator, and a 34301A 6 digit multimeter with build in frequency counte, a stabile DC power supply and an HP pulse generator. Even the build in 100Hz signal calibration signal, derived from the mains is a good time reference.

I can input signals directly on the scope boards if needed, or tap signals there. I understand with a 100MHz scope it can not be done like that. The 5113 is just a low frequency scope, and the equipment I have can do much higher frequency than the scope itself.

Having repaired the HV circuit, it is not possible to say what the HV was before, but it was over 3530V after the repair. So I set it to 3400V, by the book. Changing the HV, from what I know, does change the size of the projected picture on the CRT.

Any good hints before I start are much appreciated.


Tom Phillips
 

My understanding is that the frame calibration is to ensure the scope
frames are identically calibrated which you would want if swapping plugins
between frames.


emissionlabs
 

Here is how I did the calibration without time mark generator.

Using a 5B10B time base and a 5A14N vertical module in slot #1.

The +30V has a very small tolerance, which I used to calibrate the scope calibrator output. So I set the +30V until I had exactly 400mV on the calibrator, and the 30V was still within calibration. 400mV measured with an Agilent 34401A. Then I set the -30V to exactly the same level. So +30 and -30 are nicely balanced.

Make the calibration signal visible with 4 divisions vertical.

The sensitivity at the backplane is 50mV/cm. I soldered two wires on the PCB of the 5A14N, which lead to pins 7A and 7B. I hooked those up to the Agilent 34401A. Connect the calibrator to the vertical module input, and use the "Cal" potmeter to adjust the backplane signal to exactly 200mV. So 4 divisions on the screen. Do not touch this potmeter setting. Any deviation is now caused by the vertical amp. Adjust the vertical gain for 4 divisions.

The next step is a bit strange, but the manual sais so too.

Swap the time base module and the vertical amp.
This rotates what you see on the CRT by 90°. The vertical amp output which has a known 200mV still, is now in the time base slot (#3). This 200mV drives the horizontal amp. Adjust it for 4 divisions.

So now the vertical and horizontal amp are cablibrated.

Put back the time base and the vertical amp in the normal slots, and make the calibration signal visible. Any error is now caused by the time base being too slow or too fast. The calibration signal is not 50% duty cycle nicely. So I used a digital controlled oscillor and set it for 10kHz square wave. Now adjust the timebase to 10kHz is displayed correct.

That's it. A bit unconventional, and I hope there is no mistake in the system, but it worked very good.