Tektronix 7904 repair and restoration

[velik_kazakov]

Hi guys,
and sorry for my English.
My new toy now is on my desk. I buy from ebay tektronix 7904 that is not working. Oscilloscope come yesterday and because I have no patience open it for checking.
After reading some pages in the service manual I find that the PS is the same as tek 485 (I service one before 2-3 month and all is still in my head). Check it and find that have one socket is not connected right and one of the power transistor have only 2.25 h12e (need to be 10-40 in the specifications for this part). This was the problem in the tec 485 that I service and I exchange power transistors with 2 new BUX48A. I find old one transistor from 485 that have good h21e and exchange it in 7904. Now both power switching transistor have h12e ~9. Still is under the specifications but all work well. Now my new toy power up and have a good voltages from PS.
Today I think to disassembly the scop and to wash it (here was leaving small mouse , and all need to be washed). After all is is done think to recap the scope and to keep it in my lab for using and maybe will put in the old age pension my 7604.
Want to ask when recap the scope, because it is difficult for me to find good tantalum caps here can I replace all with electrolytic caps (maybe low esr caps) in the power distribution of the scope. Or to put electrolytic caps in parallel with small tantalum cap (do this in other of my old equipment and all work well).

[WastelandTek]

Hi guys,
and sorry for my English.

Today I think to disassembly the scop and to wash it (here was leaving small mouse

I am bumping just for this

well done velik!!

I too have small mouse, more in chicken feed than in scope, but pest.  Congratulations working Tek scope!!!

[velik_kazakov]

one and a half hour later it is ready to be washed .

[capt bullshot]

"here was leaving small mouse" 

"one and a half hour later it is ready to be washed" 

Watching for more pictures and posts, wishing good luck with the old boat anchor (got myself a 7603 and a 7623, without mice alas)

[velik_kazakov]

All washed, now will wait for the sun to dry the parts for 2-3 hours.
Here have some pictures from washing process. After cleaning the pcb with washer, I rinse it with deionized water.

[velik_kazakov]

All done, but still have no any ideas what to do with all caps that need to be exchanged - most of it is tanatlum caps and when clean the pcb find some cracked. Will check tonight the caps and maybe tomorrow will exchange some if I find at home the same. Now power on only for a minute to test all and will check all when exchange the caps. Please, if someone try to exchange this caps with electrolytic and all is OK to write some words . Thank you in advance!

[David Hess]

After reading some pages in the service manual I find that the PS is the same as tek 485 (I service one before 2-3 month and all is still in my head).

I was going to correct you but I checked my notes and you are right.  The 485 was released in 1972 and the first 7000 mainframes  to use the same switching regulator controller were the 7704A and 7904 released in 1973.  The 7704 however was released in 1969 and has a discrete version of the same power supply which could be considered the prototype for this design.  The 7704 service manual is worth studying to understand how this power supply works.

This was the problem in the tec 485 that I service and I exchange power transistors with 2 new BUX48A.

The only modern replacement in a TO-3 package that I know of is the 2N6545 and I think the only reason it is still produced is because it was the best of the high Vcbo fast switching transistors and should be able to replace all of them.

I have not finished researching it but in later 7000 mainframes, Tektronix was selecting *slower* 2N6308 transistors unless "Toff>=3us" was a typo and they were suppose to be faster than 3us.  Why would they want them to be slower?

After all is is done think to recap the scope and to keep it in my lab for using and maybe will put in the old age pension my 7604.

I have thought much the same comparing my 7904 to my 7603.  The 7603 has a much larger screen but the 7904 is so much brighter and sharper.

Want to ask when recap the scope, because it is difficult for me to find good tantalum caps here can I replace all with electrolytic caps (maybe low esr caps) in the power distribution of the scope. Or to put electrolytic caps in parallel with small tantalum cap (do this in other of my old equipment and all work well).

Unless the tantalum capacitors have failed or leaked, then I would leave them alone.  Short of failure and unless they are dipped solid tantalum capacitors, they do not wear out and are very reliable.

If you do want to replace them, then low ESR high frequency aluminum electrolytic capacitors suitable for switching power supply use with from 2 to 4 times the capacitance are suitable.  The higher capacitance is needed to get an equivalent ESR at higher frequencies.

The input capacitors can be normal aluminum electrolytics because there is LC decoupling between them and the switching transistors.

[alm]

I would double check if the tantalum caps are used for power supply decoupling. Then replacing them (if necessary, like David I am not a fan of shotgun part replacements without a good reason) with higher value low ESR aluminum electrolytics sounds good. But make sure they did not use one of the caps for something like timing. An RC oscillator with an aluminum electrolyic would make for a pretty terrible oscillator . I have not checked the schematics if any tantalum caps are used outside of power supply decoupling.

[capt bullshot]

I second the opinions. Don't fix it if it isn't broken.
Of course there are exceptions from this rule, like SMT capacitor plague, drying electrolytic caps, mains input filter caps of certain age and brand - I don't think these exceptions apply here.
I'd check for ripple on the power supply rails and if suspect, check or replace the electrolytic caps.

[WastelandTek]

my personal experience... the tantalums fail short, the PSU crowbars out, no real damage...replace tantalum...use scope


just my experience

[capt bullshot]

my personal experience... the tantalums fail short, the PSU crowbars out, no real damage...replace tantalum...use scope

That's right. I've had a similar case with a FG503 module. A tantalum failed short, the power supply just went into current limit. Replacing the tantalum fixed it, no further damage. Tek did a great job back then in protecting the circuitry from internal fails to cascade.

[velik_kazakov]

The only modern replacement in a TO-3 package that I know of is the 2N6545 and I think the only reason it is still produced is because it was the best of the high Vcbo fast switching transistors and should be able to replace all of them.

I have not finished researching it but in later 7000 mainframes, Tektronix was selecting *slower* 2N6308 transistors unless "Toff>=3us" was a typo and they were suppose to be faster than 3us.  Why would they want them to be slower?

I check to find 2N6545, but I can't. Need to get it from Germany and maybe will order 4 pcs for testing. I check the BUX48A and look that have the similar specifications like 2N6545: Ton - 1uS; Storage time - max 3uS;  Fail time -max 0.8uS. Maybe can be used here. But will keep for now original one that I install from 485.

I would double check if the tantalum caps are used for power supply decoupling. Then replacing them (if necessary, like David I am not a fan of shotgun part replacements without a good reason) with higher value low ESR aluminum electrolytics sounds good. But make sure they did not use one of the caps for something like timing. An RC oscillator with an aluminum electrolyic would make for a pretty terrible oscillator . I have not checked the schematics if any tantalum caps are used outside of power supply decoupling.

Yes I talk for the power decoupling caps. I have a service manual, but find that have some differences inside with the scop hare. Today try to find some caps on generator board (yesterday I look some with cracks on it) but it is different.

my personal experience... the tantalums fail short, the PSU crowbars out, no real damage...replace tantalum...use scope


just my experience

Yes, I have some experience with tantalum caps on other scopes that was repaired, but now start to write to ask for better way to replace it (aluminium or tantalum or in parallel of both).

Here have some pictures from generator board that I repair today - there have some bad caps, but have no schematic and need to spend a time to check the traces to the caps and all 5 pcs are in the power distribution. Now replace all with 47uF/35v and generator is back to life. Old was 4,7uF tantalum 16v.
Other that I find in the scope is that when work some time the focus is going bad. Maybe need to start with Power supply and to recap it before to keep wolking ahead or to looking to find other bad caps in the power distribution. just don't know when will find a time to do this .

[WastelandTek]

OK...lol

those ones probably need replacing, sorry to doubt you

Personally I replace with modern tantalum, double voltage rating, others have different views.

[David Hess]

I check to find 2N6545, but I can't. Need to get it from Germany and maybe will order 4 pcs for testing. I check the BUX48A and look that have the similar specifications like 2N6545: Ton - 1uS; Storage time - max 3uS;  Fail time -max 0.8uS. Maybe can be used here. But will keep for now original one that I install from 485.

The 2N6545 is just what is available where I am and I have a parts drawer full of them.  They were commonly used in off-line switching power supplies before good MOSFETs became available.

The BUX48 or BUX48A look like another good alternative but nobody around me carries it.

[ivaylo]

Good job, Velik! Ping me if you need parts from the states.

[velik_kazakov]

The 2N6545 is just what is available where I am and I have a parts drawer full of them.  They were commonly used in off-line switching power supplies before good MOSFETs became available.

The BUX48 or BUX48A look like another good alternative but nobody around me carries it.

I can get BUX48A on good price here in Bulgaria and can't find 2N6545 (can to buy it from Germany for more money - maybe 2-3 times). When have a time will power my 485 for a long time to show how will work with BUX. When make it it work maybe 1-2 hours w/o any problems, but have other components for exchanging inside.
I look that you have lots of info about Tek and want to ask you something. I buy lot of tek scopes (48 pcs) and have 10+ scopes 463. I try to find any information for this model, but no success. I know only that it is 75MHz and when open to look it inside I find this is the same as 465 with very small differences. Will be hapy to get any information for this model. I think that this scope are designed from Tektronix and maybe is produced here or in East Europe in the past.

Good job, Velik! Ping me if you need parts from the states.

Hi and thank you for the support. Your name sounds like you are from Bulgaria .
Always have very interesting items for me from USA, but the cost of shipment and customs clearance is not cheap and need some time to do all here. But thank you in advance, maybe will need to find anything in the future.

Hope to find some time in next 2-3 days to continue to check this scope and will write details. Now have lots of other jobs that need to be finished.

[David Hess]

I buy lot of tek scopes (48 pcs) and have 10+ scopes 463. I try to find any information for this model, but no success. I know only that it is 75MHz and when open to look it inside I find this is the same as 465 with very small differences. Will be hapy to get any information for this model. I think that this scope are designed from Tektronix and maybe is produced here or in East Europe in the past.

The only thing I have in my notes about the 463 is the model number and it is not in any of my Tektronix catalogs.  Tekwiki has a page about it but no documentation.  Since the 463 supports mixed sweep, it is obviously a derivative of the 465 and I suspect the 465 service documentation will describe it pretty well.

It is likely that someone over at TekScopes@yahoogroups.com or TekScopes2@groups.io would know more about the 463.

An internet search turns up the 463 schematics here but not the rest of the service manual.

[N2IXK]

There are a few wet slug tantalum bypass caps on the "main interface" board (the board that the plugins seat into) that are notorious for failing shorted. They can take out other components in the SMPS when they do.

They are a pain to replace, as you have to separate a stack of 3 PCBs connected together with long pin connectors and LOTS of those delicate "Peltola" coax connectors, but I would definitely suggest swapping them out since you already have the whole thing torn down.

[velik_kazakov]

The only thing I have in my notes about the 463 is the model number and it is not in any of my Tektronix catalogs.  Tekwiki has a page about it but no documentation.  Since the 463 supports mixed sweep, it is obviously a derivative of the 465 and I suspect the 465 service documentation will describe it pretty well.

It is likely that someone over at TekScopes@yahoogroups.com or TekScopes2@groups.io would know more about the 463.

An internet search turns up the 463 schematics here but not the rest of the service manual.

This schematic maybe is new in the net. When looking for it before 5-6 months I can't find anything. Will check maybe in the next week this with the schematic froim 465 - there jave some small differences but now I don't remember it. But will be interesting go find it. BTW have a very interesting repair with some photos and video of tek 463 and maybe can post it here in the future. Maybe will be interesting for someone.

There are a few wet slug tantalum bypass caps on the "main interface" board (the board that the plugins seat into) that are notorious for failing shorted. They can take out other components in the SMPS when they do.

They are a pain to replace, as you have to separate a stack of 3 PCBs connected together with long pin connectors and LOTS of those delicate "Peltola" coax connectors, but I would definitely suggest swapping them out since you already have the whole thing torn down.

Make all today - 3-4 hours to do all.
First start with this caps in the main board. Disassembly only the vertical interface and exchange the caps (some was good some bad, but exchange all because the time for disassembly is not cheap like 3-4 caps ). But after look at bottom, find some wire that is soldered on the cap and one was desoldered and need to disassembly all 3 pcb to do all. After exchange some caps in main horizontal amplifier and Zaxis board. All have some problems. Now all is ok, and it work from 4 hours on the desk. Just to putt all power voltages in the limits need to adjust 50v to 50.2v - _15 and +15v are lower the limits if adjust 50v to 50.0v. But thing that this will be no problem for now.
When the scope was disassembled I exchange some wire because all blue connectors are bad (maybe if we use blue plastic bags will save the planet ) with other colors have no problems. I get some wires with connectors from one 7603 that I have for parts.
2236 show the ripple of one of the power rail - all is ok for me.

[WastelandTek]

BTW have a very interesting repair with some photos and video of tek 463 and maybe can post it here in the future. Maybe will be interesting for someone.

please do

[David Hess]

There are a few wet slug tantalum bypass caps on the "main interface" board (the board that the plugins seat into) that are notorious for failing shorted. They can take out other components in the SMPS when they do.

I cannot say that I have ever run into a shorted wet tantalum capacitor.  I have seen a couple that leaked though.

I have a datasheet for the T11 series wet tantalum capacitors but it lacks enough detail to make a good evaluation of potential replacements or to know why Tektronix used a wet tantalum part instead of an axial solid tantalum part.  One or the other was required to get a 75 volt or higher voltage rating without using a film or aluminum electrolytic capacitor.

Besides their low leakage, reliably, and high temperature range which are all irrelevant in this application, they have a high ripple current rating.  What would be a suitable replacement type?

[velik_kazakov]

There are a few wet slug tantalum bypass caps on the "main interface" board (the board that the plugins seat into) that are notorious for failing shorted. They can take out other components in the SMPS when they do.

I cannot say that I have ever run into a shorted wet tantalum capacitor.  I have seen a couple that leaked though.

I have a datasheet for the T11 series wet tantalum capacitors but it lacks enough detail to make a good evaluation of potential replacements or to know why Tektronix used a wet tantalum part instead of an axial solid tantalum part.  One or the other was required to get a 75 volt or higher voltage rating without using a film or aluminum electrolytic capacitor.

Besides their low leakage, reliably, and high temperature range which are all irrelevant in this application, they have a high ripple current rating.  What would be a suitable replacement type?

I read some info about this caps and the interesting for me is that Tektronix exchange t11 (8.2uf/75v)caps with BMIcaps aluminium (500d series 10uf/100v) and after this with Nichicon aluminium (22uf/100v) TLB2A220TCAANA. This is in the power rail of +/-50v. In 5V rail, they start to put inside tantalum (100uf/20v - tdc107m020wld) and in the end of production put Panasonic electrolytic ECEA1EV221 (220uf/25v).
Maybe I make mistake with some caps (it is difficult to find all info for this caps, if somebody can to share datasheet for TLB2A220TCAANA will be thankful for this). But there have installed tantalum wet and aluminium electrolytic caps in the different serial numbers of 7904. I'm in two mind what to install here . Now put low esr aluminium caps from Panasonic and Fujicon (this I can find easy here) and think that work well but maybe need to check more when put more plugins and to check how they will work on big frequency of the input signals. Maybe when have a time will look for ripples in this power rail when the scope work with more plugins installed.
Something else for this scope. I have no calibration plugin for this series. Can be possible to calibrate this scope with cg5001 (I have it on my lab, but for now always use it with 4xx, 22xx and 24xx series scopes). In the manual I read that need to do this with calibration plugin.

[David Hess]

I read some info about this caps and the interesting for me is that Tektronix exchange t11 (8.2uf/75v)caps with BMIcaps aluminium (500d series 10uf/100v) and after this with Nichicon aluminium (22uf/100v) TLB2A220TCAANA. This is in the power rail of +/-50v. In 5V rail, they start to put inside tantalum (100uf/20v - tdc107m020wld) and in the end of production put Panasonic electrolytic ECEA1EV221 (220uf/25v).

I have been compiling notes on the 7000 series switching power supplies to track how they evolved over time because I have several which need to be refurbished but that has not included the bulk decoupling capacitors on the interface board.

Your observations are consistent with the ESR and frequency performance differences between tantalum and aluminum electrolytic capacitors.  I suspect Tektronix found that the original 10uF replacements had too high of an ESR at high frequencies and switched to 22uF to solve that which is inline with my rule about replacing tantalum capacitors with aluminum electrolytic capacitors.  Replacing the 100uF/20V tantalum with a 220uF/25V aluminum electrolytic is consistent with that as well.

So it was the ESR at high frequencies which matters as usual for a bulk decoupling capacitor.

I do not know why Tektronix used wet instead of solid axial tantalum capacitors for the 50 volt supplies early on.  They used solid axial tantalum capacitors elsewhere for the same application before replacing them with aluminum electrolytic capacitors.

Now put low esr aluminium caps from Panasonic and Fujicon (this I can find easy here) and think that work well but maybe need to check more when put more plugins and to check how they will work on big frequency of the input signals. Maybe when have a time will look for ripples in this power rail when the scope work with more plugins installed.

That is what I would do.  You could even quadruple the capacitance from the original tantalum value just to be sure.  I would use the Nichicon PW series (105C, long life, inexpensive, miniature, low ESR, high frequency) but there are plenty of good options and it would be difficult to go wrong.

Something else for this scope. I have no calibration plugin for this series. Can be possible to calibrate this scope with cg5001 (I have it on my lab, but for now always use it with 4xx, 22xx and 24xx series scopes). In the manual I read that need to do this with calibration plugin.

Except for transient response, the calibration can be done with a vertical amplifier by connecting test points for an AC/DC multimeter to the differential signals at the interface.  I use a 7A16A (1) but any of the single trace high input impedance vertical amplifiers are suitable.  The dual trace amplifiers will work also but access is more difficult.

Transient response calibration is more difficult and would require a 7A19 or 7A29 and an external adapter to sample the differential signal between the plug-in and interface using the procedure Tektronix used for calibrating the 7104.  (2) I just settled for verifying the transient response at the BNC with my fastest vertical amplifier plug-in which is a 7A24 and fastest reference level pulse generator; it looked good enough to me so I did not mess with it.

(1) The 20 MHz bandwidth limit of the 7A16A helps a lot by producing a low noise trace so the 7A15A with its 10 MHz bandwidth limit may be even better.

(2) Tektronix made a special plug-in extender for 7104 calibration with coaxial connections for the differential vertical and trigger signals that could be disconnected to be swapped or sampled.

[WastelandTek]

(2) Tektronix made a special plug-in extender for 7104 calibration with coaxial connections for the differential vertical and trigger signals that could be disconnected to be swapped or sampled.

067-0655-00?

this thread may be of interest

someone built a homemade version

[David Hess]

(2) Tektronix made a special plug-in extender for 7104 calibration with coaxial connections for the differential vertical and trigger signals that could be disconnected to be swapped or sampled.

067-0655-00?

That is the official one but I was thinking of a different one and an alternative way to do it.

The extender I was thinking of (1) has larger diameter coaxial cable for the differential vertical and trigger signals with connectors in the middle so they could be swapped.  Place a 50 ohm resistive splitter (2) between the connectors and use the additional output to drive the reference oscilloscope which would normally be a sampling oscilloscope.  Now the transient response of the 7A19 or 7A29 plug-in can be calibrated and then the transient response of the mainframe can be calibrated.

The problem here is that the connector on the extender which plugs into the vertical plug-in needs to be one of the constant impedance connectors used on the 500 MHz or 1 GHz mainframes so I do not know if the connector on the lower frequency plug-in extender will work.

(1) Rigid Plug-In Extension (067-0589-00)

(2) So the outputs are 6dB (x2.5) down but have good pulse fidelity.

[velik_kazakov]

Except for transient response, the calibration can be done with a vertical amplifier by connecting test points for an AC/DC multimeter to the differential signals at the interface.  I use a 7A16A (1) but any of the single trace high input impedance vertical amplifiers are suitable.  The dual trace amplifiers will work also but access is more difficult.

Transient response calibration is more difficult and would require a 7A19 or 7A29 and an external adapter to sample the differential signal between the plug-in and interface using the procedure Tektronix used for calibrating the 7104.  (2) I just settled for verifying the transient response at the BNC with my fastest vertical amplifier plug-in which is a 7A24 and fastest reference level pulse generator; it looked good enough to me so I did not mess with it.

(1) The 20 MHz bandwidth limit of the 7A16A helps a lot by producing a low noise trace so the 7A15A with its 10 MHz bandwidth limit may be even better.

(2) Tektronix made a special plug-in extender for 7104 calibration with coaxial connections for the differential vertical and trigger signals that could be disconnected to be swapped or sampled.

I understand how to do this, but can't find some info. What is the voltage/division for vertical and horizontal deflection of the scope. Find some info for 50mv/div, but not sure for this (after some reading of the calibration procedure for 067-0587-02 — Calibration Fixture I'm sure now - it is 50mv/div).
I can try to make this extender - have some mainframe and plugins for spares and think that can be use connectors from it and to make a pcb to put all together with bnc connectors inside the extender.

[David Hess]

I understand how to do this, but can't find some info. What is the voltage/division for vertical and horizontal deflection of the scope. Find some info for 50mv/div, but not sure for this (after some reading of the calibration procedure for 067-0587-02 — Calibration Fixture I'm sure now - it is 50mv/div).

I think the easiest place to find it is in the standardizer service manuals and as I recall, it is 25mV/div per side or 50mV/div differential.  (1)

I can try to make this extender - have some mainframe and plugins for spares and think that can be use connectors from it and to make a pcb to put all together with bnc connectors inside the extender.

I just left the cover off of the mainframe and off of the 7A16A vertical plug-in so my multimeter leads could reach the test points that I added.

DC volts on my multimeter allowed zeroing the output from the vertical plug-in to do the zero calibration of the mainframe and AC volts allowed adjusting the function generator and plug-in to produce exactly 1 div, 2 div, ect. peak-to-peak low frequency square waves for vertical calibration.  An average responding AC meter reads 1/2 the peak-to-peak value 11% high and an RMS responding AC meter reads 1/2 the peak-to-peak value.

Then I moved the plug-in to a horizontal slot to do the same for the mainframe's horizontal circuits.

The transient response calibration needs a standardizer or the extender with the signal splitters to drive another faster oscilloscope.

(1) Confusingly, Tektronix sometimes calls this push-pull but that is fair since people now confuse differential with difference.

[velik_kazakov]

I think the easiest place to find it is in the standardizer service manuals and as I recall, it is 25mV/div per side or 50mV/div differential.  (1)

Yes I find this in 068-0587-02 SM - page 5.20 and 5.21. About difference and differential - maybe because of my bad English, but I know what is and how to measure differential  voltage . This will be no problem to check the scope and to calibrate it.

The transient response calibration needs a standardizer or the extender with the signal splitters to drive another faster oscilloscope.

Think to make this extender and to use CG5001 with pulse head in fast edge mode and other oscilloscope to check this. Just need to find a time to make it. Think that is the way to trigger the scope to show good edge w/o 7s11 and 7t11.

[David Hess]

The transient response calibration needs a standardizer or the extender with the signal splitters to drive another faster oscilloscope.

Think to make this extender and to use CG5001 with pulse head in fast edge mode and other oscilloscope to check this. Just need to find a time to make it. Think that is the way to trigger the scope to show good edge w/o 7s11 and 7t11.

Just to be clear, I was writing about using your CG5001 to drive the fastest 7000 vertical amplifier plug-in that you have, calibrating the transient response of the vertical amplifier plug-in with the extender, and then using the combined CG5001 and vertical amplifier plug-in to calibrate the mainframe.  The flaw with this method is the requirement for a faster reference oscilloscope which is why Tektronix used a 7S11/7T11/S4 sampling oscilloscope.

I am not sure if using the extender to directly drive the mainframe from the CG5001 fast edge would work well.  The standardizer plug-ins go to considerable effort to create an accurate differential pulse and the CG5001 only produces an accurate singled ended pulse.  The PG506 creates an differential pulse but it may not be time aligned accurately and it is too slow for calibration at 500 MHz.

Note that if you only have a limited set of vertical amplifier plug-ins, then it is more important that they all agree with each other in producing a clean transient response on your 7904 mainframe.  If they do, and this is exactly what your CG5001 was intended to test, then messing with the transient response of the 7904 will not be necessary and may just lead to trouble.

[velik_kazakov]

The transient response calibration needs a standardizer or the extender with the signal splitters to drive another faster oscilloscope.

Think to make this extender and to use CG5001 with pulse head in fast edge mode and other oscilloscope to check this. Just need to find a time to make it. Think that is the way to trigger the scope to show good edge w/o 7s11 and 7t11.

Just to be clear, I was writing about using your CG5001 to drive the fastest 7000 vertical amplifier plug-in that you have, calibrating the transient response of the vertical amplifier plug-in with the extender, and then using the combined CG5001 and vertical amplifier plug-in to calibrate the mainframe.  The flaw with this method is the requirement for a faster reference oscilloscope which is why Tektronix used a 7S11/7T11/S4 sampling oscilloscope.

I am not sure if using the extender to directly drive the mainframe from the CG5001 fast edge would work well.  The standardizer plug-ins go to considerable effort to create an accurate differential pulse and the CG5001 only produces an accurate singled ended pulse.  The PG506 creates an differential pulse but it may not be time aligned accurately and it is too slow for calibration at 500 MHz.

Note that if you only have a limited set of vertical amplifier plug-ins, then it is more important that they all agree with each other in producing a clean transient response on your 7904 mainframe.  If they do, and this is exactly what your CG5001 was intended to test, then messing with the transient response of the 7904 will not be necessary and may just lead to trouble.

yes, I think to do this with 7a19 and 7b92a instaled on 7904. I can't use CG w/o amplifier (or maybe I dont know how to do this). Just think to use CG because  I have some other generators (pg501, pg508, hp 8012 and one philips 5770), but CG with pulshead installed have smallest Rise/fail time (~200ps).  Something else - I can see positive and negative edge with GC and pulse head. It trigger well to show this - I use this to check other scopes before (4xx, 2236, 2445, 2455), but never with 7000. Will check when find  a time for this and will write the results.

Last night make some photos of what I mean to use for checking 7904. Make it with a scope that was on the desk for repairing - it is not calibrated but show good response. First two photo is in fast edge mode positive and negative edge and 50uS/div. Second two is in voltage mode 100uS/div. This is the best that I have at home, try some other generators but they have lower rise/fail time than the pulse head of the CG.
On 7904 maybe will get better responce, because the rise/fail time in the pulse head specifications is <= 200ps. Tonight just for info will check what will show 2455A or 485. Maybe will be better than 2236.