BJT replacement - pulling my hair out

[jaunty]

I'm trying to replace a complementary pair of BJTs in a power supply  - they are 2SD234 and 2SA490 ... i happen to have on hand some TIP31B and TIP32B BJTs which LOOK like they may fit the bill but I'm not 100% sure about it ... as this is something I've never dealt with before.

https://datasheet.octopart.com/2SA490-Inchange-Semiconductor-datasheet-15978228.pdf

http://server-die.alc.upv.es/electroweb/datasheet/bjt/tip32.pdf

it seems the breakdown voltage is compatible mostly and other current limits - or maybe there's something i am missing? they are not a perfect match spec wise but I am not sure how much it matters? In the circuit I'm working on they are being used as regulators it seems.

here is the basic schematic .... in case that matters - let me know if it's not readable
they are TR2 (2SA490) and TR5 (2SD234) in the diagram

https://www.gearslutz.com/board/attachments/geekslutz-forum/480464d1435615098-help-troubleshooting-yamaha-cs50-sensing-psu-cs50_15.jpg


HELP! thanks much in advance

[drussell]

They should work fine in that circuit.  The hfe might be a bit different depending on the version of the 2SD234 and 2SA490 that are in there but that shouldn't be an issue in this case.

I find it a bit strange, though, for the transistors in positions tr2 and tr5 to die, I would have expected tr1 and tr6 to take the brunt of any abuse.

What is the issue with the supply?

[madires]

Are you sure about Tr2 and Tr5?  My guess would be Tr1 and Tr6. The TIP32 has a lower hFE (max. 50) than the 2SA490 (three ranges up to 240) which can cause some issues. Please check the hFE classification of the 2SD234 and 2SA490.

[jaunty]

They should work fine in that circuit.  The hfe might be a bit different depending on the version of the 2SD234 and 2SA490 that are in there but that shouldn't be an issue in this case.

I find it a bit strange, though, for the transistors in positions tr2 and tr5 to die, I would have expected tr1 and tr6 to take the brunt of any abuse.

What is the issue with the supply?

As they are pretty old BJTs they seem to be the most basic form of those transistors - i can't find any other versions of them. Actually - they didn't 'die' as far as i can tell - but the power supply started blowing fuses - so I started rebuilding the entire thing, not having found any obvious culprits. In the process of trying to resolder everything back together i managed to break off one of the legs of one of the 2SA490s (there are actually two of each) - so I figured maybe rather than risk circuit instability it might be better to simply replace them all anyhow (esp. considering the thermal compound has now fossilized etc) ... so here i am. I hope that explains it well enough.

[jaunty]

Are you sure about Tr2 and Tr5?  My guess would be Tr1 and Tr6. The TIP32 has a lower hFE (max. 50) than the 2SA490 (three ranges up to 240) which can cause some issues. Please check the hFE classification of the 2SD234 and 2SA490.

yes it APPEARS the hFE max is lower on the TIP packages ... though i thought it was breakdown voltage that was more critical in this application - though (seems more than adequate at -80?) ... does it make a difference for the way the transistor is being used?

from the schematic

TR2: 2SA490
TR 4,8,10: 2SA561
TR 3, 9, 11: @SC828
TR5: 2SD234

it IS a very unconventional circuit however - each of the supply feeds going downstream have a sense line (you might have noticed the 15 vs 15S outputs on the right above) ... so I guess there are two 'halves to the circuit - of which the depicted is only one. I just figured what i showed might was probably more than enough to suss out the function of this transistor pair ?

[David Hess]

They will work fine as replacement transistors.  If the circuit is that finicky, then something is wrong with it.

Note that the circuit does not include base-emitter shunt resistors across TR1, TR2, TR5, or TR6 which limits performance and is generally not a good idea.  This indicates that the designer was either being cheap or did not know any better.

The 330K resistor across TR1 makes sure that the circuit starts up.

[jaunty]

i don't imagine that the circuit is designed in such a way to make use of the full limits of the transistors? is there any way to tell by the overall current draw (rated by manufacturer) or perhaps some other spec? It only uses a 1.5 ampere fuse ... at 110v

i can't imagine there's that much wrong with the circuit - the unusual topology helps with it's protection of everything downstream ... i guess (?) the SCRs help protect everything else on the PS ... but you almost never hear of any 'routine' PS failures with these yamaha keyboards except what comes of the ravages of time ...

but like i said i'm pretty new to this power supply stuff

[David Hess]

The base-emitter shunt resistors are very important for good line and load transient response.  They also protect against failure because of a leaky transistor.

[jaunty]

The base-emitter shunt resistors are very important for good line and load transient response.  They also protect against failure because of a leaky transistor.

are you trying to imply they should be replaced? or ...?

[David Hess]

The base-emitter shunt resistors are very important for good line and load transient response.  They also protect against failure because of a leaky transistor.

are you trying to imply they should be replaced? or ...?

Obviously the design worked without them.  But I have seen similar discrete linear regulators in consumer equipment which would have been more reliable if the base-emitter shunt resistors had been added.

These days if I have time, I reverse engineer the design, test it, and then add the resistors and improve it which entails adjusting the frequency compensation as well although this is not strictly necessary.

[dacman]

h_FE is low on  the TIP32.  Here are some crosses I found at nteinc.com:

https://www.newark.com/nte-electronics/nte152/bipolar-transistor-npn-90v-to/dp/29C4445?st=NTE152

https://www.newark.com/nte-electronics/nte153/bipolar-transistor-pnp-90v-to/dp/29C4446?st=NTE153

https://www.newark.com/nte-electronics/nte153mcp/match-complementary-pair-of-nte152nte153/dp/47W7140?st=NTE153

[alsetalokin4017]

FWIW, UTSource has both original Toshiba transistors listed, for about a dollar apiece.

[jaunty]

thanks - i actually received a 2SA490 today that i purchased on ebay a few days ago... so i guess i'll play it safe and try that to replaced the damaged one ... here goes nothing!

[David Hess]

h_FE is low on  the TIP32.  Here are some crosses I found at nteinc.com:

Hfe varies over a wide range and circuit operation does not depend on controlled hfe.  Current gain-bandwidth product for both parts is 3MHz minimum at 0.5 amps.  Hfe is given under different test conditions for each part but is consistent with similar construction as is everything else.  It would be nice to know the output capacitance of the TIP32 but it is not specified.  In short, the different parts are about as close as they can be.

Based on saturation voltage, the NTE parts actually differ more but I am sure they would also work fine.  Any modern 3MHz triple diffused transistor of sufficient current, voltage, and power rating will work.

[jaunty]

UGH! well i finished all the work ... recapacitoring and a few of the BJTs but now failing the dim bulb test and won't even light up the power switch bulb ... I guess the next step would be what ...? Investigating the power transistors? (2SA745 and 2SD203) ... I guess i WOULD try to investigate while hooked up but it CAN'T be healthy for most of the components to keep it fired up under dim bulb even while looking for the fault ...?

[jaunty]

quick question... if i'm measuring open circuit between the hot and neutral of the power cable - would that not indicate an issue with the transformer?

[drussell]

quick question... if i'm measuring open circuit between the hot and neutral of the power cable - would that not indicate an issue with the transformer?

Uhhh...  Check every component, wire and connection between one lug of the power cord and the other for continuity.  Check the cord itself, any fuses and holders, power switch, transformer, every wire, every connection, connector, etc....

Something in that path is open, it is not difficult to check each and find out where it is.

[jaunty]

yes well i thought i'd been through it with a fine toothed comb ... but apparently i misssed something ... i actually thought i saw something on one of dave's videos that mentioned checking continuity between power cable prongs as a kind of general health check (?)

[drussell]

i actually thought i saw something on one of dave's videos that mentioned checking continuity between power cable prongs as a kind of general health check (?)

Well, it is...  If you don't have any continuity, then it isn't going to do anything. 

In that case, you then start going over it with the fine tooth comb, checking everything, one at a time, from one end to the other and see where you LOSE continuity.

[jaunty]

starting to wonder if the problem isn't happening at the SCRs ... would make sense i guess ... i just don't have experience dealing with troubleshooting them ... i guess one has to take them all out first?

[dacman]

If it has a 115/230 switch, be sure to check it.

[jaunty]

If it has a 115/230 switch, be sure to check it.

thanks - no such switch ... just extra capped off primary winding wiring  for japanese and european markets ... it hit me over the head today that I NEVER even THOUGHT to check the primary winding wires (an open there would certainly explain the 'open' in the cable prong measurement i figured ... but apparently no ... at least all the wires going into the primary side seem to have continuity between them. But the neutral and hot pins on the power cable do not ... i guess the X type capacitor could explain that??