Tektronix 2247A power supply problems

[AMR Labs]

I was using the scope and it just turned off and will no longer power up. I check the fuse in the back and it is good.

Pulled off the case and removed the processor board and shield underneath to access the power supply board. I can see the fan spins for about 2 seconds when power is turned on then it just quits. I checked the primary power supply and it shows 163V, and the preregulator output is at 40V steady. It seems the problem is in the inverter stage or secondary power supplies. On the board there is nothing obviously damaged or burned that might give a clue on what happened.

When I got the scope about 10 years ago I did replace all the diodes on the power supply that had either ZM or ZS marked on them as they are well known to get leaky after some time and might fail, and used MUR160 diodes from Mouser. Back then I did not replace any capacitors as the ripple on all power supplies was well within normal limits.

I pulled out the board and removed all the capacitors which all tested good (between 0.12 to 0.14 ohms ESR and capacitance values are ok). Still figured this might be a good time to do that and went ahead and replace all the 1000uF 12V with 1000uF 16V, and the 470uF 25V with same values, finally both 39uF 150V where replaced with 47uF 250V for the 130V supply, and 47uF 100V for the 58V supply, which was what I had on hand.

Needless to say the capacitor replacement had no effect on the problem.

Does anyone have any suggestions on what to check next? I was wondering if the power supply can be tested outside of the scope and perhaps put some load with resistors on the required DC outputs for stability. I do have the service manual but it does not mention anything about this.

Any ideas or assistance appreciated.

[David Hess]

Did you miss 100 microfarad C2204 which is what actually powers the control circuits?  47 microfarad C2208 is for decoupling.

You may have reached the point where measurements need to be taken.  Note that voltages marked on the schematic, like "+56V", are relative to ground but the control circuits are floating at the +44 volt output so they actually see +12 volts.

The voltage across C2208 will reveal if the control circuits are even powered.

[AMR Labs]

I measured the 43VTP test point and was seeing 40V so although it seemed a bit low was assuming the pre-regulator looked like it was working. The black negative test lead was at the -DC Test point. Wanted to put a load on those 40V to see if they would hold up under load, but not sure how much current can I expect that rail to supply without the over current protection kicking in, or causing some damage. Would it be safe to assume I should at least be able to get about 1amp from there? So probably a 40 ohm resistor would be the right load. I  got some 33 ohm 5w resistors that might be also useful by themselves, otherwise I have 10 ohm 10w that I could put in series to bring the load value up to 43 ohms.

Will also check on the +56V to make sure its there. And yes, will check C2204 and C2208 and report back.

BTW the high voltage section seems to be at least starting up, as I accidentally touched one of the 0.01 HV caps and got a bit of a nip. I'll be more careful from now on with that area.

Thank you for the information so far.

EDIT:

BTW, is it safe to test the power supply board outside of the scope? Does any of the secondary power supplies need a load connected so it will have some stability, or can the whole board work as expected without any loads?

[David Hess]

BTW, is it safe to test the power supply board outside of the scope? Does any of the secondary power supplies need a load connected so it will have some stability, or can the whole board work as expected without any loads?

I do not know.  The service manual may say something about it.

[WattsThat]

It needs a load. At least the 2245A supply needed it and AFAIK, the 2247A supply is identical.

I have parts unit so I removed the supply connector from the main board and wired it up with 10 ohms across the +/-5 supplies and 20 ohms on the +15V. I placed high value bleeders on the other outputs as I quickly became tired of being bitten by the 130V supply cap holding charge.

My repair story was a low time 2245A from work that just failed to turn on one day. No pop, no bang, fuse fine. It was replaced with a digital unit and was going to be trashed so I grabbed it. It sat dead on a shelf for 15 years and I finally decided in 2019 it was time to fix it.

Everything was fine visually and all active devices checked okay. My +43V supply was at 40V. Q2209 and Q2210 simply refused to oscillate. Eventually, out of desperation, I started swapping parts with a good unit. The good unit continued working and the defective remained dead. It finally came down to T2204. The core wasn’t cracked, resistances and inductances matched the good unit but it just refused to work. It still puzzles me as to why but the problem was solved and I moved on.

[AMR Labs]

Well I must have done something wrong as I no longer was getting the 40V of the pre-regulator output and actually the fuse in my isolation transformer open up. With the light bulb in series with the AC supply trick it came up quite bright so definitively a short somewhere. Started checking and found Q2214 dead short. Also while I was at it also checked Q2209 and Q2210, both OK, and Q2201 mosfet also Ok.

Q2214 is labeled as IRF830 but the closest I have in my parts stock is IRF820, a slightly lower spec mosfet. Not sure I should try to use it as a temp substitute or just order 2-3 of these and wait.

Will continue to check other active parts but it makes me wonder if I might end up on a dead end as you did, and and perhaps instead just buy a cheap parts 2245-46 on ebay that is shown powering up fine, or order a pulled power supply board outright, which will no doubt cost more than a whole "for parts" scope.

As usual these switching power supplies are an immense pain in the posterior to diagnose and to ultimately find the problem. Last time I had to deal with one of these in a scope it was about 14 years ago while repairing my then 2213A only scope, which in the end it also turned out it was a bad main secondary supply transformer, T948 I believe the number was. I was lucky to get my hands on another partially populated 2213A main board and by swapping out said transformer the power supply finally started to work after weeks of diagnosing, understanding, blowing and replacing various parts.

[David Hess]

It finally came down to T2204. The core wasn’t cracked, resistances and inductances matched the good unit but it just refused to work. It still puzzles me as to why but the problem was solved and I moved on.

I have read about more failures in the buck transformer, T2203 in the 2247A series, with most being unexplained.  I have never gotten hold of any of these mysteriously failed transformers or oscilloscopes to test, so I do not have any good ideas about what is going on.  All of mine have continued to operate with no power supply failures other than electrolytic and line capacitors.

[WattsThat]

PM me before you buy anything on eBay, I have extra supplies that I bought that I’d be happy to be rid of at my cost, just not sure about shipping.

The IRF820 has higher charge numbers but the other basic specs are similar/same. It’s only running ~30KHz so it’s worth a try if you have one on hand. I’m thinking they might be the same device and the 830’s are just selected for the better specs.

[AMR Labs]

To replace the IRF830 is just going to get me back to the same point where I was yesterday. Could you ship one of those boards to my PO Box in Miami FL?
That would be great if possible, I'll PM you. Thanks.

PM sent.

[WattsThat]

Replied. We’ll keep it going via PM.

[AMR Labs]

@WattsThat

Tried to reply to you privately a couple of days ago to follow up, but not getting any further response from you.
Was hoping to finalize purchase of your power supply at your convenience.
Thank you.

[philwong5176]

Hello:

What is the replacement  part number for Q2209 and Q2210? It seemed to be a Motorola part made specifically for Tektronix.

[David Hess]

What is the replacement  part number for Q2209 and Q2210? It seemed to be a Motorola part made specifically for Tektronix.

The book says they are TIP31C but they were probably selected for high hfe or matched saturation voltage or something.

[AMR Labs]

Hello:

What is the replacement  part number for Q2209 and Q2210? It seemed to be a Motorola part made specifically for Tektronix.

I don't think there is anything special about the TIP31C transistors, I used off the shelf transistors to fix a 2213A scope power supply which also uses them in the inverter stage and there where no problems. I actually had to replace them twice as I blew the first replacement pair because of a mistake I made while taking some measurements, so basically no issues with 4 regular transistors. Just make sure not to use anything less than the "C" version rated at 100V.

[David Hess]

I don't think there is anything special about the TIP31C transistors, I used off the shelf transistors to fix a 2213A scope power supply which also uses them in the inverter stage and there where no problems. I actually had to replace them twice as I blew the first replacement pair because of a mistake I made while taking some measurements, so basically no issues with 4 regular transistors. Just make sure not to use anything less than the "C" version rated at 100V.

There is *something* special about them because Tektronix did not specific a standard TIP31C.  The part number is 151-0476-03 and not 151-0476-00.  We just do not know what selection criteria was used because Tektronix did not say.  Some information is available in older documentation for the same circuit which says the transistors are "gain selected", but is that for high or low gain, or even matching gain?  I do not think it was matching gain because they would have said that.