Power supply with only negative voltages

[roadrash]

Re: Power supply with only negative voltages

Well, I didn't expect that package type. Commonly referred to as a TO-3 package, it's actually a TO-204 package. And the part I referred to earlier was in that package. So that is probably a TV horizontal flyback output transistor. The part I figured out was made by Fujitsu, I think, and had a 2SC#### type part number. If you had to replace it a 2N6545 would work. At the die level, the 2N654x and MJE1300x series were of the same family, just in different packages.

just found a bad 2sb561c transistor so I am going to order some more and replacements for all the others in case they fail. glad you think you have found what this TO-3 (Q2) transistor is so I can get one too if its faulty. Can i check this in situ or do i need to remove it and do i test it like a normal transistor ie. (2 diodes). Also can you tell me what a TL431CLP does and how to test it?

[stj]

Re: Power supply with only negative voltages

TL431 aka KA431 aka LM341 etc. is a voltage reference i.c.

[roadrash]

Re: Power supply with only negative voltages

TL431 aka KA431 aka LM341 etc. is a voltage reference i.c.

Thanks got that and found a explanation elsewhere on the web explaining how to test shunt regulators like this one. Bit of disappointment that 2sb561c transistor I thought was faulty turned out was a PNP and I was mixing up my ECB so getting the wrong readings. So swill have to keep testing the remaining components.

[PeteS in CA]

Re: Power supply with only negative voltages

The TL431 can used as a programmable reference or as an amplifier/comparator with a 2.495V reference at one input. Astec probably used it as a comparator with reference. If you experiment with a TL431, pay attention to the minimum shunt current (too little current, and it will oscillate) and that if used as a comparator, the "low" state will be around 2V.

[roadrash]

Re: Power supply with only negative voltages

The TL431 can used as a programmable reference or as an amplifier/comparator with a 2.495V reference at one input. Astec probably used it as a comparator with reference. If you experiment with a TL431, pay attention to the minimum shunt current (too little current, and it will oscillate) and that if used as a comparator, the "low" state will be around 2V.

I found reference to a TO-3 cased power transistor on a Apple iie computer power supply (also made by Astec) and very similar my one and this psu uses a 2SC1875 power transistor. The BBC computer also uses a Astec PSu that looks identical to mine and it uses a BU208A.
Could one of these be the mystery transistor that is used on mine?

[stj]

Re: Power supply with only negative voltages

probably.

[PeteS in CA]

Re: Power supply with only negative voltages

Very possibly. Precisely matching the original part isn't as important as having a part that works - max current & voltage, good beta at high current, speed and storage time. re that second factor, a good rule-of-thumb is that the collector current at which beta falls to 10 is the practical maximum current of the transistor. For example:

2SC1875 - rated max continuous current is 3.5A; collector current where beta = 10 is ~1.8A (Mospec);

BU208A - rated max continuous current is 5A; collector current where beta = 10 is 1.0-1.4A (Mospec);

2N6543 - rated max continuous current is 5A; collector current where beta = 10 is 3.7A (Mospec).

Of the 3 examples, only the 2N6543 was designed specifically for switchmode power supplies. The other 2 were designed for use in TV horizontal deflection circuits (= similar switching speeds).

Speaking personally (and hypothetically) if all 3 were equally available and of similar price, I'd go with the 2N6543. It's the most robust device among the 3.

[R_J]

Re: Power supply with only negative voltages

I checked the astec power supply I have, (the one pictured in post #10) and it uses a 2SC3459 different style but specs should be close.
I also found this which might help http://www.kev.pulo.com.au/osborne1/

[redwire]

Re: Power supply with only negative voltages

Do we have the 853-00400050 transistor original part number? It's an Astec house number.

Bipolar switching transistors- newer parts have better (faster) switching times, so look at that parameter too in substitutes. TV-use ones are slower.

The TO-3P package will fit in a TO-3 outline, so you could use that. If you bend the leads and snip the middle pin (goes to tab) they work and easy to rip one out of an old PC power supply.

Comparing numbers: {ton, tstorage, tfall}
2SC1875 (500V 3.5A 50W ?usec/10usec/1usec)
2SC3459 (800V 4.5A 90W 0.5usec/3usec/0.3usec)
BU208A (700V 5A 150W ?usec/7usec/0.55usec)
2SC2625 (400V 10A 80W 1usec/2usec/1usec) from old PC power supply
2N6543 (400V 5A 100W 0.7usec/4usec/0.8usec) I think hard to find?

The 2SC3459, 2SC2625 looks good to me for speed.

If Q2 shorted, I would check R11, D4, Q1 also- any parts around Q2 can get damaged too depending on its particular short (C-E, C-B etc.)

Awww an old friend the Z-80 CPU. I did a lot of assembly programming on that beast, tons of fun.

[roadrash]

Re: Power supply with only negative voltages

Do we have the 853-00400050 transistor original part number? It's an Astec house number.

Bipolar switching transistors- newer parts have better (faster) switching times, so look at that parameter too in substitutes. TV-use ones are slower.

The TO-3P package will fit in a TO-3 outline, so you could use that. If you bend the leads and snip the middle pin (goes to tab) they work and easy to rip one out of an old PC power supply.

Comparing numbers: {ton, tstorage, tfall}
2SC1875 (500V 3.5A 50W ?usec/10usec/1usec)
2SC3459 (800V 4.5A 90W 0.5usec/3usec/0.3usec)
BU208A (700V 5A 150W ?usec/7usec/0.55usec)
2SC2625 (400V 10A 80W 1usec/2usec/1usec) from old PC power supply
2N6543 (400V 5A 100W 0.7usec/4usec/0.8usec) I think hard to find?

The 2SC3459, 2SC2625 looks good to me for speed.

If Q2 shorted, I would check R11, D4, Q1 also- any parts around Q2 can get damaged too depending on its particular short (C-E, C-B etc.)

Awww an old friend the Z-80 CPU. I did a lot of assembly programming on that beast, tons of fun.

Thank you thats very informative. I dont understand why they would grind off the markings on the transistor in the first place. Well hopefully we have found what this transistor is now for future reference. I tested this in circuit with a multimeter and it tested as OK but decided to remove it from the PCB to do a proper test on my component tester and here is a picture of what the tester found. I assume this proves its working ok.

[roadrash]

Re: Power supply with only negative voltages

Do we have the 853-00400050 transistor original part number? It's an Astec house number.

Bipolar switching transistors- newer parts have better (faster) switching times, so look at that parameter too in substitutes. TV-use ones are slower.

The TO-3P package will fit in a TO-3 outline, so you could use that. If you bend the leads and snip the middle pin (goes to tab) they work and easy to rip one out of an old PC power supply.

Comparing numbers: {ton, tstorage, tfall}
2SC1875 (500V 3.5A 50W ?usec/10usec/1usec)
2SC3459 (800V 4.5A 90W 0.5usec/3usec/0.3usec)
BU208A (700V 5A 150W ?usec/7usec/0.55usec)
2SC2625 (400V 10A 80W 1usec/2usec/1usec) from old PC power supply
2N6543 (400V 5A 100W 0.7usec/4usec/0.8usec) I think hard to find?

The 2SC3459, 2SC2625 looks good to me for speed.

If Q2 shorted, I would check R11, D4, Q1 also- any parts around Q2 can get damaged too depending on its particular short (C-E, C-B etc.)

Awww an old friend the Z-80 CPU. I did a lot of assembly programming on that beast, tons of fun.

That's a really useful bit of info. Just been re-capping a Sinclair ZX81 & Spectrum 48k that were working last year but not this year. Got them both working again but the Spectrum has a small display fault but that's for a different thread.

Back to this ancient Osborne 01 Computer PSU, I managed to pick up a genuine Astec repair manual for this PSU and have been following through the test procedures which brought my attention to the thyristor which when tested my component tester points to being faulty. I have a problem though in trying to find a replacement or equivalent at a sensible price. Those I have found have been very far away from me in Cambridge UK and cost of shipping is just crazy. Does anyone know of a cheap readily available equivalent to the C122UX5 that’s currently fitted?

[stj]

Re: Power supply with only negative voltages

that part number looks bogus.

[PeteS in CA]

Re: Power supply with only negative voltages

that part number looks bogus.

Nope. The C122 is an 8A SCR originally made by GE or RCA. Boschert used the things like water back in the late 70s and early 80s for OVP crowbars. Since it was used on the +5V O/P, they probably used the C122F, the 50V part.

So an 8A, 50V TO-220 SCR of whatever type should suffice.

If the crowbar SCR is shorted, you should hear a, "Tick, tick, tick,...," as the P/S starts, current limits, restarts, current limits, etc.. It would also act the same if the +5V is not regulated and the crowbar SCR fires with every start-up attempt.

[roadrash]

Re: Power supply with only negative voltages

Nope. The C122 is an 8A SCR originally made by GE or RCA. Boschert used the things like water back in the late 70s and early 80s for OVP crowbars. Since it was used on the +5V O/P, they probably used the C122F, the 50V part.

So an 8A, 50V TO-220 SCR of whatever type should suffice.

If the crowbar SCR is shorted, you should hear a, "Tick, tick, tick,...," as the P/S starts, current limits, restarts, current limits, etc.. It would also act the same if the +5V is not regulated and the crowbar SCR fires with every start-up attempt.

Thanks for the reply PeteS. According to one component supplier a NTE5461 is an equivalent but that is just as hard to get hold of. No this one wasnt ticking just extremely low millivolts if any output.
This description and referring to the crowbar SCR may have a similarity to another problem I have. Have you read my other thread on here about another PSU I am trying to repair on a AMIGA 2000 because because what you describe is exactly the fault I have been try to cure on this psu without success.

[stj]

Re: Power supply with only negative voltages

stay away from NTE, they just buy surplus junk and re-print the numbers

[R_J]

Re: Power supply with only negative voltages

I checked the scr (scr1) on my board, it is a 2N6395

[PeteS in CA]

Re: Power supply with only negative voltages

Thanks for the reply PeteS. According to one component supplier a NTE5461 is an equivalent but that is just as hard to get hold of. No this one wasnt ticking just extremely low millivolts if any output.
This description and referring to the crowbar SCR may have a similarity to another problem I have. Have you read my other thread on here about another PSU I am trying to repair on a AMIGA 2000 because because what you describe is exactly the fault I have been try to cure on this psu without success.

Pretty much anything that trips the P/S current limit on a discontinuous flyback P/S will cause that audible symptom.

I'd suggest looking at the catalog for whatever distie sells in the UK and see what they offer in a >/= 8A SCR. Go with a brand name like On Semiconductor or TI (examples, not a comprehensive list). And don't worry whether it's 50V or 100V or ... it's in a 5V circuit.

[roadrash]

Re: Power supply with only negative voltages

I checked the scr (scr1) on my board, it is a 2N6395

That's interesting. I wonder what is the year & version of your osborne 01 As I believe there were about three versions starting with the one like mine (the first) which was white & black with Large Black OSBORNE name down the side. The last model was Blue & grey with Osborne name Recessed into the cover. They must have changed the component I will have to do comparison of the specs to see if I can use this one instead.

[roadrash]

Re: Power supply with only negative voltages

I had a 2N6395 arrive today and fitted and powered up the PSU. At first I was getting nothing. So got the repair manual out again and started following through the checklist again and when I got to checking for 310v between Diode block and the emitter of Oscilator transistor, it for some reason started working and was giving the chirping noise so powered it off for a bit before powering it again and this time it was a constant whine. Again powered off. Did this a few more times then "POP" went a blue tantallum cap on the main logic board. Seen this before old old computers that have sat for years so hoping when I replace it will be ok. I was trying to get some voltage readings when it went pop so dont know what the voltage outputs were. The cap was a 15uf 16v tantallum bead.. If i replace this will it be ok or will it just blow again?

[R_J]

Re: Power supply with only negative voltages

I don't own a Osborne, I only have the power supply, these things were used everywhere
why don't you make sure the power supply is working properly before connecting it to the logic board, It should operate by itself and you can make sure the voltages are correct before connecting it. It's hard to say why the cap went, It could have been bad or it could have received a voltage higher than it could handle, I would replace it and then connect a meter to it so you can check the voltage thats going to it but I suspect it was just the one cap. or there would have been more damage
That whine noise was from the power supply because there was a short on one of its outputs