Topics

Tek 7904 In Pulse Mode Related to C1511 on LV Regulator


rgpierce1@...
 

Hey Guys,

I'm in the process of (mostly done I think) repairing a 7904 power supply that goes into pulse mode. I replaced a number of the bad tantalums on the cap rectifier board which brought some supplies closer to where they needed to be but but did not keep the supply from going into pulse mode. Continuing to diagnose things I found that pulling C1511 on the LV regulator board solved the problem, however the cap I pulled isn't reading as a short and replacing with a known good causes the pulse mode to return. C1511 seems to be working as a speed up capacitor for the regulator (allowing high frequency transients to bypass R1514 in the ladder) but seems awful high at 0.1 uF for that purpose. I'm considering replacing with a much smaller value ceramic (like 100 pF). Does anyone have any experience/thoughts/suggestions with this? Could my -50V supply have a lot of noise on it from somewhere else that it shouldn't which is causing this cap to be a problem? I'm currently running the supply into a bank of resistive loads so it would have to come from a connection to one of the other supplies if this was the case.

Thanks for the help.
Ricky


Tom Lee
 

I haven't looked at the schematic, so I can only speak in general terms. If that cap is indeed part of a loop compensator, I think it's ill advised to go about modifying it in any ad hoc way. Whether 0.1uF is "high" or "low" isn't related just to its value, it's a function of the impedance it sees, and of the gain of the loop. Speeding up the loop by dropping the value of the cap could endanger loop stability. A wildly oscillating switcher is not what you want!

-- Tom

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 1/3/2021 16:05, rgpierce1@gmail.com wrote:
Hey Guys,

I'm in the process of (mostly done I think) repairing a 7904 power supply that goes into pulse mode. I replaced a number of the bad tantalums on the cap rectifier board which brought some supplies closer to where they needed to be but but did not keep the supply from going into pulse mode. Continuing to diagnose things I found that pulling C1511 on the LV regulator board solved the problem, however the cap I pulled isn't reading as a short and replacing with a known good causes the pulse mode to return. C1511 seems to be working as a speed up capacitor for the regulator (allowing high frequency transients to bypass R1514 in the ladder) but seems awful high at 0.1 uF for that purpose. I'm considering replacing with a much smaller value ceramic (like 100 pF). Does anyone have any experience/thoughts/suggestions with this? Could my -50V supply have a lot of noise on it from somewhere else that it shouldn't which is causing this cap to be a problem? I'm currently running the supply into a bank of resistive loads so it would have to come from a connection to one of the other supplies if this was the case.

Thanks for the help.
Ricky




rgpierce1@...
 

Thanks for the reply Tom. I thought better of the change as well since Tek made a boat load of these with absolutely no change in component value over the lifetime of the instrument. My load board was a purely resistive implementation initially and I decided to add the shunt capacitance the main interface board applies to this rail and that solved the problem. The supply now correctly drives the artificial load perfectly. On to the rest of the mainframe.


Göran Krusell
 

The purpose of C1511 is to compensate for the stray capacitance that you have from Q1508B base to ground. This is why a typical value would be for example 10 to 47pF and not 0.1uF.
Göran


David Campbell <k_8_b_y_p@...>
 

This scope is 40 ( +) years old.* It is not a repair. It is a restoration.*

Replace ALL the caps in the PS. Try again.


Roy Thistle
 

On Wed, Jan 6, 2021 at 04:58 PM, David Campbell wrote:


Replace ALL the caps
Didn't we have that conversation?


Roy Thistle
 

On Tue, Jan 5, 2021 at 01:25 PM, <rgpierce1@gmail.com> wrote:


I'm currently running the supply into a bank of resistive loads
I'm interested in that. Why the load bank? Is the scope a relative "basket case?"
Did you follow the "hi eff ps" guide posted on Tekwiki?


 

For sure don't replace the metal cased tantalums in the PSU. It's highly unlikely they have failed.

If the power supply is in burst mode (ticking), then my experience of fixing 485s strongly suggests looking for a shorted dipped bead tantalum hanging off one of the power rails as the first probable culprit. Sure there are other possible causes like the 120V Zener going low but that's not very common fault.

Is the PSU ticking?

-----Original Message-----
From: TekScopes@groups.io <TekScopes@groups.io> On Behalf Of David Campbell
Sent: 07 January 2021 00:59
To: TekScopes@groups.io
Subject: Re: [TekScopes] Tek 7904 In Pulse Mode Related to C1511 on LV Regulator

This scope is 40 ( +) years old.* It is not a repair. It is a restoration.*

Replace ALL the caps in the PS. Try again.


 

Oops I mentally cross wired 485 and 7904 in my reply below, but as the PSU topologies and implementations are almost identical my comment still stands.

The most common tick mode culprits in the (early) 7904 PSU are C1314, C1318, C1353, and C1354 (all on the Cap. Rectifier board at the top of the board).

If the bulk filter caps go OC, then they can also cause tick mode, easy to check by measuring the voltage across them - they should both read about the same voltage (within 10-20%). Beware that they run at line (mains) voltage (in USA). Problem is that it's almost impossible to find replacements for these ☹, so you're reduced bodging in a modern snap-lock to cap to replace (unless you know of a 1000uF 200V 3/4" pitch part).

In the inverter board, VR1246 (110V in the early 7904, and IRC 120V in later ones) can go low voltage, and C1249 can go leaky or short which also causes problems, but neither of these are common problems.

On very rare occasions the Controller IC (U1275) can also go bad, only way to be sure on this is to swap parts.

Its also possible for faults in the VR circuits on the LV regulator board to cause tick mode because all the LV supplies contribute via a resistor array to a "summing junction" that should sit at 0V when the supplies are at their nominal values - these can be a pain to track down.

HtH
David

-----Original Message-----
From: TekScopes@groups.io <TekScopes@groups.io> On Behalf Of David C. Partridge
Sent: 07 January 2021 10:06
To: TekScopes@groups.io
Subject: Re: [TekScopes] Tek 7904 In Pulse Mode Related to C1511 on LV Regulator

For sure don't replace the metal cased tantalums in the PSU. It's highly unlikely they have failed.

If the power supply is in burst mode (ticking), then my experience of fixing 485s strongly suggests looking for a shorted dipped bead tantalum hanging off one of the power rails as the first probable culprit. Sure there are other possible causes like the 120V Zener going low but that's not very common fault.

Is the PSU ticking?

-----Original Message-----
From: TekScopes@groups.io <TekScopes@groups.io> On Behalf Of David Campbell
Sent: 07 January 2021 00:59
To: TekScopes@groups.io
Subject: Re: [TekScopes] Tek 7904 In Pulse Mode Related to C1511 on LV Regulator

This scope is 40 ( +) years old.* It is not a repair. It is a restoration.*

Replace ALL the caps in the PS. Try again.