B & O Penta amplifier repair and Tr equivalents

[punkieys17]

Hi all, just taken on a project - a 150w B&O amplifier from the base of a Penta loudspeaker. The ouput uses 3 cascaded pairs of SC3281 and SA1302.

The unit was blowing the main power fuse - suggesting one or more of these transistors had gone short circuit. With nothing on the resistance measured from the rails (68v-0v -68V) across the emmitter/collector and a 0.47 ohm wire wound resistor was 0.9ohms. Visually theres nothing overheated on the boards - but 2 of the transistors have clearly been replaced in the past as they are a different colour case !

I just found the forum's comments about many replacements being fake. Unfortunately I've already ordered some....from Hong Kong. I'm worried I've just wasted my money.

I understand there are alternatives in the form of the 2SC5200 and the 2SA1943. Are these plug and play or will I need to play with the driver circuitry for the base currents of these.

Any help greatly appreciated. Also any help on how to identify the short circuit - I trained as an electronics engineer over 30 years ago but this is my first foray since then.

 

[dorke]

On paper, you can replace the 2SC3281/2SA1302 with the 2SC5200/2SA1943.

But you have to realize there is no such thing as plug and play regarding the replacement of output stage power transistors.
In any case the Idle current should be adjusted(even when replacing with the same number Tr.s exactly).

You should also check the drivers and bias transistors.
It is very common that when the powers short/blow they blow as well.
Shorts are identified with an ohm-meter.
When using a DMM better check in "diode" mode.

Do you have the schematics of the output stage?

 

[punkieys17]

Hi Dorke, many thanks for your reply. I'm uploading the B&O manual with the amplifier section on page 2-4. Right hand side you will see the tr's I was asking about.

Seperating the TR's seemed a good idea so I isolated the last 2 on both power rails by disconnecting the current limiting reistors R101, R102, R103, R104. I baulked at removing the first two of that series of resistors (R99, R100) as I was concerned what action might have on the feed back via R88 and R83.

 

[punkieys17]

B&O Amp is overview. Amp2 is tighter view on the o/p stage

 

[dorke]

Well,
The PA final Tr.s are not cascaded they are paralleled.

You should check all the finals(in blue) drivers(in pink).
Test them in "diode mode" B-C ,B-E and C-E that is a total of 6 measurements for each Tr.(3 and 3 in reverse polarity)
If you need help with that post the test results in a table form.
You can start in circuit if something looks suspicious take it out and re-measure.

The power resistors(green) should be check as well (in ohm test).
The pot.(in red) is for setting the idle current.

The pics should be at higher resolution they aren't clear enough.

 

[punkieys17]

Will do. Yes sorry about the shots. They were screen grabbed from a b&o pdf manual. Unfortunately the repair manual was too large to upload to the forum. Thanks for the reply will get back to you within 12 hours with the measurements.

 

[punkieys17]

Have uploaded a tigher view of the area of interest.

 

[punkieys17]

These are the values from the trs in the circuit board. The meter on diode setting shows O/L when over 200 ohms is detected.

 

[dorke]

If the measurements are correct than all the Tr.s are you measured are bad:

RED -shorted C-E .
Blue-open B-E.
Green- somewhere in the middle but also bad.

Have you measured out of circuit?

Need to check the power resistors and also the next bunch in red below:

 

[punkieys17]

Haven't measured out of circuit yet as bench space at a premium and the amp is a nightmare built into a hexagonal aluminiumheat sink.

Bad news is I checked the next series of Trs (15,16,17,18) - all >200 ohms in each of the six measurements.Similarly the diodes (D18, D19) and the IC1 at the bottom. I went on to check all the TRs so TR6,7,8,9,10,11 & 12. All were >200 except TR12 which reported 0 ohm resistance B-C and 0 ohm C-B - all others were >200.

On powering up before hand the pre-amp section was reporting O volts at the test points so I'm guessing that section survived.

I'm guessing my next action is to remove all the active components starting from Tr6 onwards. That means facing the dceonstruction from the heat sink.

 

[dorke]

Have you any good Tr.s to test for reference?
For instance,check on another channel.

Another thing ,
make sure there is no insulating coating on the pads you have checked(just to be on the safe side of the measurements)

That is a very long line of destruction!
I would start removing them from the finals inwards ,
and test them again outside the board as I go.

Very important !
Make sure you mark each Tr. on the board when you pull it out including direction.
A good idea is to take a photo before you start .

 

[punkieys17]

Will check the stores to see what I have in terms of good kit to check against. Will probably get back after the weekend - might as well tear it down now and keep going back till I find a good semi !!. Shows the problem of not knowing the background to it.

Next few days are getting busy - not least erecting a chicken coup for the mother in law....or rather her new chickens!

Have a great weekend and many thanks for the help.

 

[73's de Edd]

.

Sires . . . . .


Would a mo' bettah schematic of that little jewel , be of any further assist ?

I'll be, safely, way up on the very top row in the audience at the peanut gallery, when it comes time to "Crash and Burn"

73's de Edd

.

 

[punkieys17]

.

Sires . . . . .


Would a mo' bettah schematic of that little jewel , be of any further assist ?

I'll be, safely, way up on the very top row in the audience at the peanut gallery, when it comes time to "Crash and Burn"

73's de Edd

.

Many thanks for the schematic - it immediately solved one problem of a seemingly mismatched pair of driver transistors. Looking further back down the circuit it seems only the bf423 and bf858 are going to be hard to source. More on Monday hopefully. By the way did your schematic have a parts list attached - mine did but weirdly no mention of the final driver transistors - such as tr27.

 

[punkieys17]

So stripped out the semis back to TR3 (MPF4392). ALL THE SEMIS WERE BLOWN ! Would like to know how to test TR3 - not sure if the test meter on diode setting will work on this type (n-channel JFET). Once I know the fate of TR3 I will get the replacements ordered. Once D5 & D6 are replaced I'm tepted to energise the 15V lines to see if the first stages are working. I am sure all the supply rails (15V & 68V) are workings as they should.

 

[davenn]

mine did but weirdly no mention of the final driver transistors - such as tr27.

They are all on the above circuit diagram

TR 25, 26, 27 = 2SC3281

TR 30, 31, 32 = 2SA1302


Dave

 

[73's de Edd]

punkieys17 . . . . . .

Did you actually find fault with D5 and D6, with them being so remote from the typical problem area ?
Or . . . . . after having found so many bad parts in the power audio area, that you are now being fully into the
frenzied . . . . . "off with their heads ". . . . . mode.


Back when you were testing TR12 . . . . do note that it is being shorted B to C to be configured for a thermal compensating diode function for
the base of TR15 of the TR15-16 series pair .



I don't think that there will be problems with TR3 at all.
Its function is as a mute element, if its gate is activated, the Drain to Source short/conduction mutes the incoming audio.

When your care package with ALL of your semis comes in, hold off on installing ANY of your big power output devices, and just test up to the driver stages.

That then lets you see if all is in order and that the frontal circuitry feeding to TR24 and 29 is balancing them out and they will have have very close to zero volts at their common point of their R98 and R99 emitter resistors tieing together.
To be on the safe side, I would even initially be testing with a series 100 watt incandescent lamp in series with the AC line input

Did you find your mentioned :

" only the bf423 and bf858 are going to be hard to source"

I found:

I don't think that you would have to worry about these "past prime time" / or / "not considered for new design" / or / "NLA " devices being
counterfeit units.

http://www.ebay.com/itm/Siemens-BF858-Silicon-NPN-High-Voltage-Power-Transistor-TO-202-/271071597359

http://www.ebay.co.uk/itm/1-PC-BF423-PNP-Silicium-Low-Power-LF-Transistor-CS-TO92-/221706505211
( Limit, 6 to a customer from out of that pile)

Thasssit . . . . . .

Addenda after seeing just above:

The TR24 and 29 are the "final" DRIVER transistors while the

TR25 and 30
TR26 and 31
TR27 and 32
are being the complementary NPN / PNP power output transistors

73's de Edd

 

[dorke]

So stripped out the semis back to TR3 (MPF4392). ALL THE SEMIS WERE BLOWN ! Would like to know how to test TR3 - not sure if the test meter on diode setting will work on this type (n-channel JFET). Once I know the fate of TR3 I will get the replacements ordered. Once D5 & D6 are replaced I'm tepted to energise the 15V lines to see if the first stages are working. I am sure all the supply rails (15V & 68V) are workings as they should.

Testing a JFET:
Your T3 is an N-channel JFET.
Test it like you test an NPN-BJT in diode test mode.
i.e Base replaced by Gate .
you should get Froward diodes when gate is + lead (G-D,G-S).
and Backwards diodes when gate is - lead (G-D,G-S).

The only difference between JFET and BJT is :
same resistance(relatively low) between D-S and S-D
while with C-E or E-C resistance is "infinity".

One important test to perform:
Tets resistance between output (L1) to GND and ±68V rails

Another thing is testing all diodes in the P.A stage ,
including Zeners (D16,D21),
and all "power resistors"/low resistance resistors(R89-R106).

I would also replace ,or at least Deoxit the trimmers R59,R95.

Did you check "IC1" the Dralington transistor?

 

[punkieys17]

Hi guys, apologies for the radio silence, thought I would wait till I had new replacement semis to test against. At this point I have to hold my head in shame....finger trouble and not checking the multimeter....so I was a tad surprised when having removed the semis then tested them they came back with real measurements. The attached file is using the diode test so is showing the voltage drop reading. All diodes tested OK...also after having removed them ! I ended up ordering from 4 different mainland UK suppliers (so RS, Farnell, Cricklewood & Profusion) who can be trusted not to have fakes. You may notice TR8 is untested - dropped it and its disappeared...ah well. I am wondering if the collectors of the power transistors were correctly insulated from the heatsink which is common to all. - I only found one insulating tab under 1 of the 6 main power transformers. So would appreciate input on my take of the semi test results. I noticed it was the -68V line which was the primary short - it rose to close to zero before the main transformer fuse blew. No signs of shorts with the semis removed so feeling slightly more confident on going to reassembly.

 

[73's de Edd]

.

Sir punkieys17 . . . . .


Sooooooo it looks like if you place all new power transistors in this unit that you will have one matched set of old transistors to use on a ( lower power) amplifier after you get this (POWER) amplifier running again.
I think that the 1302 / 1382 devices had the bare metal for its heat sinking exposed , thereby requiring an insulator, whereas some of the later transistor spin offs had a high thermal conductivity epoxy case.
The only situation that I might suspect creating a variant, was if the heat sink itself has a HARD anodized surface itself, that might have relied upon that, instead of a separate insulator.
Through the years I have even seen . . . . and received engineering samples . . . .of high thermal conductivity hard anoidized aluminum insulators offered by AAVID engineering for TO-126, TO-220 and TO-3 case profiles of transistors.
http://www.aavid.com/product-group/interface/insulators/aluminum

73's de Edd

.