Tek 2235 Power Supply Revisited

[piloteer]

AARRRGGGHHHH ! As you can see the supply is giving me heart burn. For the second time after replacement R926, C925, Q9070, U930, Q930 gets blown out. The supply never did get repaired. The crow bar worked fine indicating the voltage at TP 940 was too high. It should be about +43 volts. It seems for some reason the voltage acroos C925 a 100 mfd 25 vdc cap blows the cap shorting R926 then taking out most of the pre-regulator parts. Any one have some ideas?

From my calculation the voltage at the base of Q928 should rise to 6.9 volts to turn Q928 on.
The voltage at the top end of R926 is 175 vdc charges up C925 through R926 a 150K resistor.
Seems to me the voltage drop across R926 is 132 vdc. C925 is only a 25 volt cap. Something crazy here. When the cap blows it takes R926 with it. Yey theirs no short to TP150 or to chassis ground.

The serial number is high starting with a B70--- Option 0

25 years ago I repaired a number of these scopes with wierd power supply issues, but this one is the worse. Another strange thing with this scope (milltary version) is the markings on the board for Q928 is marked Q944. In the correct Q944 position it's not marked at all. Traced out every line to verify this. However on the circuit board layout page of the manual it's shown correctly. Scope appears to have been extensibly worked on before. Got this one at an auction.

 

[shrtrnd]

How's your C906 (75ufd cap), and that bridge configuration CR901 thru CR904?
Did you check that? It's supplying your circuit that keeps frying.
(And CR920 can screw you, if it's bad).
TP940 should give you a good indication of the health of that circuit, Check Q908, if it doesn't look good.

 

[piloteer]

Did all that. C906 tests good on a tester I built and measures about 87 mfd on a cap checker.
CR 920 forward and back checks fine. Q928, 930 and 908 are ok. Did find C940 somewhat leaky and measured 840 rather than 1000 or higher. The crowbar SCR and 51 volt Zener that limits the voltage at TP140 are also ok. In the process of building a tester for U930.

 

[shrtrnd]

I'll look at the schematic again. But I'm telling you right now, I'd replace C906, whether you figure it tests good or not.
I'll send another post after I recheck the schematic.

 

[shrtrnd]

I can see why you're frustrated, there's not much else left to check.
I'd check Q938, Q939, Q944, Q946 & Q947 (Back to back resistances, for a short).
If you can't find the problem, you're going to have to recheck those previous repairs you said somebody else already did.
If I was you, I'd disconnect the output of the power supply itself, and energize it. If the
supply doesn't blow with no circuit load, you'll know something outside of the power supply circuit itself, is what's causing the damage.
Sorry I don't have the magic bullet, you've done what I would have done.

 

[piloteer]

But I never give up but I found that the schematic of the invertor and circuit board layout diagram (not the schematic) are not in agreement. What I need is the correct schematic of the inverter. As an example there is no R944. It's an R941. The darlingtons 938 and 939 appear to be inverted that are connected to the collectors. As I stated in my first post this scope serial number is B070415 and option1 but now I find the circuit board layout does not agree with what I see for the invertor. Need to try and find the correct diagrams if at all possible. While all the front to backs look the way they should I afraid to turn it on lest it blows up again. Thinking the person who tried to repair it had the same problem. Gave up on it and it was turned over to the auction, This scope is the miitary version.

 

[shrtrnd]

I've only got the one manual on this scope.
I agree, you're going to need the manual for the military version you've got since it's different.
Did you check with Tektronix, to see if you can get one from them? They may have a way for you to just get the schematic, without having to pay for a full manual.

 

[piloteer]

Yeah, checked with Tek not too long ago. They have list of who and where might have one.
Found 1 who wants $80 bucks for one. The problem is I paid $50 for the scope and already paid $45f or the service manual I got now in good condition. But the military versions also have many revs. I won' tpay $80 for a maybe. Gonna fire it up once more later as I just replaced a 2K resistor that measured 2.3K If it blows up again I don't care. It goes back to the auction house I got it from. If I recover $25 I'll be happy.

 

[shrtrnd]

I would've been screwed too. I didn't even KNOW they had some different military version of a 2235 (I have one, it's my main hobby scope). If you couldn't find the problem in the power supply itself, I'd be betting whoever worked on the scope last, made one too many mistakes, elsewhere in the scope, and it's shorting out your power supply. Fixing a failure is one thing, finding somebody else's mistakes before you can even GET to the original problem, can be a real pain.

 

[piloteer]

Yeah, tell me about. I ran the metrology department at General Instrument and at Loral
electronics corp. Started at Emerson Radio and Phono back in 1959. Last fixes were HP Spectrum analyzers. Went into business started an FAA certified aviaonics repair station at L.I. MacArthur airport in NY. Then worked at Raytheon at Tampa airport as a weather radar and navigation engineer until I fell off the wing of a plane and busted my arm. Now started a computer repair business working from home. 73 now but I keep very active. No rocking chairs for me sitting down doing nothing except waiting for the meat wagon to come and get my carcass. Ha Ha.

 

[davenn]

hi there,

I thought I had commented earlier, scrolling through didnt see it so it much have been another Tek user topic.

as I said to him, have you tried the tektronix yahoo group ? I have been a member there for quite some time as I own a couple of bits of Tek gear. there's a huge knowledge base there... maybe some one can help you

I'm still trying to sort out some niggling faults in my Tek 465B scope.

cheers
Dave

 

[shrtrnd]

Yeah, 40 years doing this myself.
We're dinosaurs now. All the new gear is just swap boards, and throw away the old one. Ain't a whole heck of a lot of guys like us left, ....and when the old gear is gone, we'll be too.

 

[piloteer]

Hi Davenn and Shrtnd,
Didn't have a chance to look at the Yahoo group yet.

But I tried out the changes I did with new zeners and resistor replacement. It looks like it wants to start because the leds blink once every two seconds. Don't have a 45v volt supply to test the invertor with but I built a regulated supply around a 317t regulator and drive it with a 24 vac power pack. It gives me about 15 vdc. The leds light dimmly. The 5.2 volt supply puts out about 3.5 volts. The other supplies put out very weakly. I'm sure with 45 volts the scope would work. (maybe)

Couldn't find a source for C906. Sphere sold the last one they had. I had some old boards laying around and found a 100mfd 450vdc cap. Soldered that one in but it made no difference.

The thought about it and said if the U930 doesn't seem to want to start, it may be the resistor and timing cap pin 5 and 6 of the PWM. So that's next.

Now I need to find a cheap ESR tester. I can build one but by the time I order the parts pay for it then add shipping and tax, it would be more cost and time effective to buy 1 for $80. Seen one on ebay.

 

[piloteer]

Very interesting find. Have any noticed while repairing or trouble shooting the power supply this anomaly. Measuring R926 the 150k resistor that it does not measure a resistance but looks like a cap discarging? It may start out measuring 1.5 megohm and slowing comes down to the 150K. value. If I remove it from the circuit it measure immediately correct. I tried shorting out C906 with jumpers and also C925.

 

[piloteer]

FOUND the problem unfortunately can't find a supplier. The part is U975 the high voltage multiplier rectifier Is shorted.
Tried all the sources all have been sold out. Speres, Qwest, Ebay. All sold out. Any other ideas?

 

[shrtrnd]

Your inquiry to Sphere must have motivated them. They're now showing two used ones (they claim fully tested) for $69. Also one, looks like outside US for $49 on ebay. I'll look around some more, and post another note if I find something later.
Nice job, finding it.

 

[shrtrnd]

I went to the Sphere site. I see the one you were looking at is sold (Unitrode CMX647). But they also list the two Murata MSL8513's I mentioned. Several times I've emailed Sphere stuff I was looking for, that they didn't have listed. And they actually
went out and found some for me. You might try that, if you want a NEW one.
Everybody else that claims to have them, are places that don't sell to us regular people, only businesses. It might take you a while, but one will turn-up somewhere.

 

[piloteer]

Thank you shrtrnd. Went that route but I'll contact them again. The only issue I see is that there are TWO 152-0806-00. They are slightly different and might not bolt up. These mulitplier units are very similar to the types used in the many weather radar units I've repaired. The older ones use a series of diodes and "charge capacitors" in a lattice configuration and that is what these are. Can't find a spec sheet for it otherwise I make my own. I hope I can turn one up or else it goes back to the auction house I got it from.

Some of the Murata's might not bolt up either. But I'll talk to Spere and see what happens.

 

[piloteer]

Hey shrtrnd thanks.
Contacted Sphere and spoke to Susan. She checked and said the Murata part IS identical to the Unitrode but is a later manufacturer. $69 + $22 shipping so it's on the way. Should get it within 7days maybe she said. Murata no longer makes them I checked with them.

 

[shrtrnd]

Yeah, I didn;t see how Tektronix would put the same part number on something that wouldn't work. You probably know it's not unusual for several different manufacturers to make parts for Tek or HP. They might still put their own numbers on them, but they'd
have to meet the Tek specs, for Tek to bag them under their Tek number.
What I like about Sphere, if I can't find it in the U.S., is that if you ask, they'll usually
TRY to find what you're looking for. Most companies don't bother anymore.
Good luck with the 2235. Hope to keep mine going for quite a while.