UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Specific Vintage Equipment > Vintage Test Gear and Workshop Equipment

Notices

Vintage Test Gear and Workshop Equipment For discussions about vintage test gear and workshop equipment such as coil winders.

Reply
 
Thread Tools
Old 31st Jul 2018, 9:56 pm   #21
Noopy2014
Triode
 
Join Date: May 2014
Location: Eichstätt, Germany
Posts: 11
Default Re: Gould 4072 bad PSU

You are right! The power supply is a little bit tricky.

The main problem of the time was the low voltage rating of the power FETs at the time of the development. The solution for Gould was the construction of a switching power supply working with something around 70V (the upper supply).

To get the 70V-Supply the lower supply was developed. This supply is a step down regulator which has lower voltage requirements.
In detail the ground of the upper supply is connected to the inductor L1 and the lower supply transistor switches in this way that the upper supply get“s 70V.

To make it a little bit more confusing the regulation of the output voltages at the secondary side is realized over the lower supply. It monitors the auxiliary winding 9-5-10. This voltage is distributed to the 3524 over D19 and D20.
(D21 and D22 are for synchronisation of the to supplys. The upper supply works with 19kHz while the lower supply works with 38kHz.)
Noopy2014 is offline   Reply With Quote
Old 2nd Aug 2018, 7:25 am   #22
mole42uk
Octode
 
mole42uk's Avatar
 
Join Date: Aug 2010
Location: Bristol, UK
Posts: 1,534
Default Re: Gould 4072 bad PSU

There are several things about which I'm puzzled. Referring the the enlarged circuit diagram I posted earlier (post 11):

The -VE rail of the pre-regulator is connected to the -VE of the input bridge rectifier at PLD1,2. There is a note on the drawing that "These points have high potentials relative to each other and to mains earth" with asterisks to indicate the pre-regulator -VE and the main regulator -VE rails. However, at PLD9,10 (upper part of the circuit diagram) there is a connection that goes between the main regulator -VE and the connection to L1, where there is also a connection to the pre-regulator -VE via PLD11. Is this a drawing error or have I missed something?
__________________
Richard

Index:
recursive loop: see recursive loop
mole42uk is offline   Reply With Quote
Old 2nd Aug 2018, 5:56 pm   #23
Noopy2014
Triode
 
Join Date: May 2014
Location: Eichstätt, Germany
Posts: 11
Default Re: Gould 4072 bad PSU

In my opinion the ground sign between L1 and PLD9,10 is a mistake. It makes absolutely no sense and would short the the pre-regulator.
Noopy2014 is offline   Reply With Quote
Old 2nd Aug 2018, 8:29 pm   #24
mole42uk
Octode
 
mole42uk's Avatar
 
Join Date: Aug 2010
Location: Bristol, UK
Posts: 1,534
Default Re: Gould 4072 bad PSU

Thanks Richard, that sets my mind at rest a little.
I will be testing the PCB to make sure that the ground connection on the circuit diagram is, in fact, an error.
I have been working in entirely the wrong direction due to that error and hope that my efforts to recover the power supply might be now a little more successful
__________________
Richard

Index:
recursive loop: see recursive loop
mole42uk is offline   Reply With Quote
Old 3rd Aug 2018, 11:24 am   #25
mole42uk
Octode
 
mole42uk's Avatar
 
Join Date: Aug 2010
Location: Bristol, UK
Posts: 1,534
Default Re: Gould 4072 bad PSU

I checked the main power supply PCB this morning before work, and there is no evidence that I could measure or see that the main regulator ground is connected to the pre-regulator ground, so the drawing is in error.

I've a plan now to set up test cells for each half of the high-voltage primary side of the power supply so that I can verify that each does what it should before committing the whole to operation. As I have two complete supplies I think that makes sense so that I end up with two working supplies rather than two dead ones as at present. After that the secondary side should be straight-forward.
__________________
Richard

Index:
recursive loop: see recursive loop
mole42uk is offline   Reply With Quote
Old 3rd Aug 2018, 2:04 pm   #26
dave cox
Octode
 
dave cox's Avatar
 
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 1,190
Default Re: Gould 4072 bad PSU

Quote:
After that the secondary side should be straight-forward.
That's surely tempting fate!

This is my understanding of this circuit, without the complexity of the switch controller …

dc
Attached Thumbnails
Click image for larger version

Name:	switcher.jpg
Views:	39
Size:	43.1 KB
ID:	166985  
dave cox is offline   Reply With Quote
Old 3rd Aug 2018, 4:06 pm   #27
mole42uk
Octode
 
mole42uk's Avatar
 
Join Date: Aug 2010
Location: Bristol, UK
Posts: 1,534
Default Re: Gould 4072 bad PSU

Quote:
Originally Posted by dave cox View Post
Quote:
After that the secondary side should be straight-forward.
That's surely tempting fate!
After finding a pretty fundamental drawing error, I might be feeling like it's downhill all the way now!!

A couple of weeks should tell......
__________________
Richard

Index:
recursive loop: see recursive loop
mole42uk is offline   Reply With Quote
Old 5th Aug 2018, 12:56 pm   #28
mole42uk
Octode
 
mole42uk's Avatar
 
Join Date: Aug 2010
Location: Bristol, UK
Posts: 1,534
Default Re: Gould 4072 bad PSU

A much improved Sunday morning. After some testing of the main regulator by supplying 70V across C5 and measuring 5V and 16V at the correct places, I decided to assemble the pre-regulator and test the whole thing.

Winding the Variac up from 0 towards 100V I was rewarded with acceptable DC levels at 5V and 16V, a little higher than I'd like but that's what you get running an SMPS without load.

Shortly after I'd begun to congratulate myself there was a huge 'crack' and the 1µF tantalum at C55 keeled over. After settling my nerves and replacing that with a proper electrolytic, the output looks good, I even wound up to 160V while measuring the 70V across C5 which remained within limits.

I'm having a break from the bench now. I will be testing the other power supply next, and replacing all the tantalums as a matter of course. Is it permitted to hate tantalum bead capacitors?

There's also the EHT supply to test. I have an EHT probe so we'll see if that works.

The other photo is of parts that had also failed - a couple of diodes, a resistor and some tantalums that I outed just because they were there.
Attached Thumbnails
Click image for larger version

Name:	Dead C55.jpg
Views:	43
Size:	97.4 KB
ID:	167052   Click image for larger version

Name:	outed bits.jpg
Views:	33
Size:	32.5 KB
ID:	167053  
__________________
Richard

Index:
recursive loop: see recursive loop
mole42uk is offline   Reply With Quote
Old 5th Aug 2018, 3:17 pm   #29
mole42uk
Octode
 
mole42uk's Avatar
 
Join Date: Aug 2010
Location: Bristol, UK
Posts: 1,534
Default Re: Gould 4072 bad PSU

Back to it, I couldn't leave it alone. The front panel indicates it's doing a ROM-TEST, then AUTO-SETUP but sticks there. No display, but the AC heater voltage is present and correct.

Further checking revealed no 260V supply, the 1µF tantalum that I thought exploded was actually okay, it was the 150R resistor next to it. Replacing that revealed a high current demand on the 260V rail, which is connected to the pre-amp board at PLNV14. From there is a mess of surface mount components, so that's an investigation for another day.

On the positive side, I did test the other pre-regulator daughter board which works okay.
__________________
Richard

Index:
recursive loop: see recursive loop
mole42uk is offline   Reply With Quote
Old 5th Aug 2018, 8:32 pm   #30
Noopy2014
Triode
 
Join Date: May 2014
Location: Eichstätt, Germany
Posts: 11
Default Re: Gould 4072 bad PSU

Do not lose patience!
Noopy2014 is offline   Reply With Quote
Old 5th Aug 2018, 8:41 pm   #31
karesz*
Hexode
 
Join Date: Jun 2010
Location: Zurich, Switzerland
Posts: 478
Default Re: Gould 4072 bad PSU

Hi Richard,
look the double electrolythics please over C2--C55: you have to change minimum the 10uF/350V than its strong "blowed" up...
In my opinion are tantalums not so bad, but the ~30 years old versions are usualy to chhange for better ones as Rifas epoxy potted capacitors and some mains filters too
Wish to end much fan with your repaired PSU!
Karl
karesz* is offline   Reply With Quote
Old 6th Aug 2018, 7:22 am   #32
mole42uk
Octode
 
mole42uk's Avatar
 
Join Date: Aug 2010
Location: Bristol, UK
Posts: 1,534
Default Re: Gould 4072 bad PSU

Hello Karl,

That capacitor is not bulging, it's a photographic artifact, but I will change all the capacitors now that I know that the power supply and a significant amount of the oscilloscope works. I intend to repair both supplies completely so that I have a spare. I would like to put the 4-channel version of this instrument into use and not have to repair the power supply again!

All of the RIFA capacitors show evidence of cracking and several of the wet electrolytics show signs of running at elevated temperature. See post #1 for pictures. I know about the mains filters - I have a Gould 1602 that exploded the mains filter and the smell remained for several weeks.
__________________
Richard

Index:
recursive loop: see recursive loop

Last edited by mole42uk; 6th Aug 2018 at 7:30 am.
mole42uk is offline   Reply With Quote
Old 6th Aug 2018, 11:27 am   #33
karesz*
Hexode
 
Join Date: Jun 2010
Location: Zurich, Switzerland
Posts: 478
Default Re: Gould 4072 bad PSU

Hi Richard,
Im sorry, than I forget the first photo with some Rifas, than in my opinion are the Schaffner mains filters the most wors-now adays; after 2-3 decades...
Otherweise; you have physically in your hands these condensers; I have only your photo and on it is hes neighborro IO, undsitorted-why?...
Rgds, Karl
Attached Thumbnails
Click image for larger version

Name:	Dead C55.jpg
Views:	30
Size:	27.5 KB
ID:	167114  
karesz* is offline   Reply With Quote
Old 6th Aug 2018, 11:40 am   #34
mole42uk
Octode
 
mole42uk's Avatar
 
Join Date: Aug 2010
Location: Bristol, UK
Posts: 1,534
Default Re: Gould 4072 bad PSU

Hello Karl,
It is no problem, the photos are often mis-leading!
The capacitors will be changed anyway because several others exhibit signs of heat degradation and I want this supply to operate correctly for a long time. One other electrolytic (in the picture of post 1) had leaked electrolyte and corroded one of the connection wires completely away.
Schaffner mains filters do seem to have RIFA capacitors in them and when they break they are not good at all.
__________________
Richard

Index:
recursive loop: see recursive loop
mole42uk is offline   Reply With Quote
Old 10th Aug 2018, 7:15 pm   #35
mole42uk
Octode
 
mole42uk's Avatar
 
Join Date: Aug 2010
Location: Bristol, UK
Posts: 1,534
Default Re: Gould 4072 bad PSU

Having decided to replace all the tantalum capacitors in the power supply with high temperature electrolytic, I’ve found another exploded tantalum on one of the data boards.

I wonder if I should plan on replacing them all?
__________________
Richard

Index:
recursive loop: see recursive loop
mole42uk is offline   Reply With Quote
Old 12th Aug 2018, 3:10 pm   #36
mole42uk
Octode
 
mole42uk's Avatar
 
Join Date: Aug 2010
Location: Bristol, UK
Posts: 1,534
Default Re: Gould 4072 bad PSU

More work on the 4074 power supplies today. What Karl thought was a bad electrolytic was a bad electrolytic although the top remains flat, removing it revealed the cause of the high current demand on the 260V supply - tracking on the PCB between the two terminals of the capacitor, pictured below.

I have scraped the carbon away and reduced some of the copper ground track, hoping that this will stop the problem. I'm waiting for new electrolytics to arrive before proceeding.

The other supply is slowly coming back to life. One of the BUZ31 MosFETs was dead and changing that got the main regulator working after a fashion with 30V on the supply rails. I put the pre-regulator back on and introduced 100V AC into the board, only to release smoke from a fast schottky diode in the secondary side. I have some fast diodes which I may try now.
Attached Thumbnails
Click image for larger version

Name:	260V tracking.jpg
Views:	29
Size:	73.4 KB
ID:	167428  
__________________
Richard

Index:
recursive loop: see recursive loop
mole42uk is offline   Reply With Quote
Old 12th Aug 2018, 3:52 pm   #37
mole42uk
Octode
 
mole42uk's Avatar
 
Join Date: Aug 2010
Location: Bristol, UK
Posts: 1,534
Default Re: Gould 4072 bad PSU

After fitting the fast diodes the power supply seemed to be fairly happy. I connected it to the oscilloscope and was rewarded with a display on the screen. After some fifteen minutes, the heatsinks felt a bit warmer than I'd like to I've switched it off.

There are a lot of new capacitors arriving in a day or two so I'll leave it alone now until I get to replace those.
Attached Thumbnails
Click image for larger version

Name:	Progress!!.jpg
Views:	36
Size:	36.2 KB
ID:	167435  
__________________
Richard

Index:
recursive loop: see recursive loop
mole42uk is offline   Reply With Quote
Old 12th Aug 2018, 5:09 pm   #38
mole42uk
Octode
 
mole42uk's Avatar
 
Join Date: Aug 2010
Location: Bristol, UK
Posts: 1,534
Default Re: Gould 4072 bad PSU

Now I have been running the 'scope again, increasing the AC input to 160V and the PSU heatsinks remain slightly warm with the fan running. I did smell something getting hot though, and switched off after about 20 minutes. The heatsinks on the CCD board were very hot.

I did get a signal display on the screen and upon pressing the Auto Setup button the 'scope tried to present a useful display. I was attempting to provide an input when I decided to switch off to investigate the hot smell.

I'm not sure what the devices are on the CCD board and why they might be getting hot. I just read in the service manual that "the aquisition clock cable from the Digital Timebase must be connected while the instrument is switched on. Operation for extended periods of time without this connected could lead to excessive heat dissipation..." I must check!
__________________
Richard

Index:
recursive loop: see recursive loop
mole42uk is offline   Reply With Quote
Old 14th Aug 2018, 4:20 am   #39
Noopy2014
Triode
 
Join Date: May 2014
Location: Eichstätt, Germany
Posts: 11
Default Re: Gould 4072 bad PSU

Without proper ventilation (CCD-board on extender cards) the CCD-heatsinks go up to 55°C. There should be no problem. The eight resistor arrays arround these heatsinks can go up to 100°C. But that“s ok.


Quote:
"the aquisition clock cable from the Digital Timebase must be connected while the instrument is switched on. Operation for extended periods of time without this connected could lead to excessive heat dissipation..."
I haven“t had any problem with this one. And I did a lot of testing.
Noopy2014 is offline   Reply With Quote
Old 19th Aug 2018, 5:16 pm   #40
mole42uk
Octode
 
mole42uk's Avatar
 
Join Date: Aug 2010
Location: Bristol, UK
Posts: 1,534
Default Re: Gould 4072 bad PSU

Today should have been one of celebration. Having replaced all the electrolytics, polys and tantalums in the power supply I connected it up and all was well. The 'scope worked as well as I could expect given that I haven't set up the signal inputs yet. The CCD board was running cooler as well.

Then the power supply was replaced in the metal housing and the lot re-installed in the case. The 'scope switched on again, same display as before. I set about trying the get an image of a suitable signal on the screen - simples maybe, just apply a signal and press the auto setup button. That didn't quite work, and I noticed a smell from the rear of the casing. Sigh.

So I took the power supply out of the case again and ran it exposed on the bench. This time I heard cracking from the EHT board and noticed that the screen was a little dimmer than previously. Okay, thinks I, then it may be that I need to replace the 4kV capacitors as well. Having another EHT board to hand, I tried that but the screen was way too bright and the front panel controls would not reduce the brightness.

I've left it alone for now, it seems I may need to buy even more capacitors before I can transfer this thread to the Success Stories section.
__________________
Richard

Index:
recursive loop: see recursive loop
mole42uk is offline   Reply With Quote
Reply

Thread Tools



All times are GMT +1. The time now is 3:49 am.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2018, vBulletin Solutions, Inc.
Copyright ©2002 - 2018, Paul Stenning.