Yorkville 100watt Mixer/Amp - Replaced Transistors

[Jason G]

Hey all, new here and hope for some direction and assistance. Here is the backstory.

This amp has a bass/treble pot (50k linear) and the shaft was broken. I wanted to replace for cosmetics, amp functioned just fine.

Bought 2 pots from digikey and replaced. They were 50k linear.

Fired amp up and it didn't quite sound the same, played with bass/treble and turned amp up and then POP. Blew the 5amp fuse.

I replaced the fuse and every time i powered on the fuse pops again. Took it to a buddy and we decided I probably fried the 2 output transistors (MJH11018 and MJH11017). Now when i was testing/playing I never put the transistors back on the metal frame so we assumed they overheated and fried.

For added measure we replaced the BC550C transistor as well.

Order 2 orig 50k pots from yorkville and replaced them as well.

The leads on the BC550C were too long and was shorting out, we figured this out as BD139 was getting really hot. Fixed the leads and now the amp runs and sounds good at low volume. No fuse issues.

As soon as we push the amp more than 30% volume the sound is distorted and sounds terrible. No matter what input/channel.

The headphone jack works just fine and i can drive the amp loud and the headphones don't sound distorted.

However as normal the amp is not right, i don't know what i'm missing and what to do next? We're thinking of replacing BD139 as maybe the heat ruined the chip.

Here are the 2 transistors i replaced from DigiKey, slightly different spec but we were told they should be just fine.

https://www.digikey.ca/product-detail/en/on-semiconductor/MJH11022G/MJH11022GOS-ND/919545
https://www.digikey.ca/product-detail/en/on-semiconductor/MJH11021G/MJH11021GOS-ND/919544

I've also attached a schematic.

Could we somehow have switched up the NPN PNP transistors? if so would the amp work at all?

any tips would be super helpful! thank you!!

 

[kellys_eye]

That's not a schematic.

 

[Jason G]

Sorry, is this better?

 

[kellys_eye]

Check the main supply voltages first.

The +/- 38V to the amplifier and the +/- 16V to the pre-amp circuitry.

Report your findings.

 

[Jason G]

Ok I will do that tonight. Would you mind giving me a tip on best points on the circuit to place my multi-meter and test those voltage rails?

thank you for your time.

 

[kellys_eye]

At the power supply itself or at the output transistors.

Next check would be to measure the DC voltage at the junction of the two emitter resistors - should be zero volts with no signal applied.

 

[Jason G]

Sorry haven't had time yet to check. I'm hoping tonight I can pull it apart and confirm those voltage readings.

 

[Jason G]

Ok 16 volt rail i got 15.9+ and 15.9-
38 volt rail i got 36.9 and 37.1

I tested voltage to the 2 output transistors

Ground to JA10 = -37.2 vdc
Ground to JA13 = 37.2 vdc

What would be the next best step in testing or figuring out why my signal distorts quickly.

 

[Jason G]

Another thing i noticed.

R76, R75 - really hot to the touch, can only handle a finger on them for 4 seconds.
R88, R89, R90 - same. really hot

I don't know if this is normal behaviour?

 

[kellys_eye]

What about the voltage at the junction of the output transistors? Measure at TP1 and TP2.

'Hot' is a relative term but given you have the correct voltages on the 16V rail there is nothing to worry about.

 

[kellys_eye]

As an aside - double check the pinout for the BC550. These packages often have different lead outs from what's posted on the schematic.

 

[Jason G]

Ok ground to TP1. I tested JA11 and its -0.66 dcv same point on RA18 -0.65 and other end of resister is +0.20

ground to TP2. I tested JA14 +0.114 and RA19 -0.12. other end of RA19 is +0.024

From what i see the BC550 pin layout is the same.

https://www.fairchildsemi.com/datasheets/BC/BC546.pdf

thank you again for your time!

 

[kellys_eye]

What is JA11 (or JA14 for that matter?)

Connect your meter negative lead to ground (the ground at the main smoothing capacitors would be best) and then use the positive probe to measure TP1 and TP2.

They should both read the same (albeit one will read positive, one negative) and at the junction of both those resistors it 'should' read 0V.

 

[Jason G]

Sorry those points were right on the pin of the output transistors. see below.

Could you highlight where you would prefer me measure ground from? i've been using the main chassis ground

 

[kellys_eye]

Speaker negative leads or anywhere on the schematic that is labelled 'AG'.

 

[Jason G]

Ok speaker ground (-) to JA11 (TP1)= -0.06
Speaker Ground (-) to JA14 (TP2) = 0.11

What do you think?

 

[kellys_eye]

The voltage sounds ok - ideally they are equal but to achieve that would require messing with the biasing resistors but the fault isn't being caused by such a small difference.

As soon as we push the amp more than 30% volume the sound is distorted and sounds terrible. No matter what input/channel.

The headphone jack works just fine and i can drive the amp loud and the headphones don't sound distorted.

Since the headphone jack is at the amp output it points to the SPEAKER being damaged - probably by the original fault causing a spike in the output. If you can find a different speaker to use temporarily this will prove the cause.

 

[Jason G]

Hey there, already tried another speaker and get the exact same result. Clean signal until i hit a certain volume point, then distortion.

At low output i can push the bass and it sounds great, then i push the volume just a bit more and its distorted.

Question:

Do you see an issue with the replacement output transistor specs? I listed them in the first message. My hunch has been that maybe the bias registers are now not right? I know other types of amps have pots for bias control so not sure?

Where do you suggest I go from here? any specific components you suggest I just replace to see if that helps?

The headphone jack output was clean but i never pushed the volume that high as there isnt a way to listen to them as they would be blaring loud and un-listenable.

Thanks again for the replies.

 

[73's de Edd]

Sir Jason . . . . . . .Gee

R76, R75 - really hot to the touch, can only handle a finger on them for 4 seconds.
R88, R89, R90 - same. really hot

Observation . . . .
R77, R37, R39 . . .should also be saying . . . . me too . . . . ME TOO !

In initially looking at the main power amp section . . . . using your preliminary analysis info . . .
It seems that the output transistors that you are PRESENTLY using, seem to be balancing out
evenly and they are NOT producing more than a 500 millivolt imbalance / difference at their emitter resistors.

PLUS . . . . you do seem to be getting proper power amplification on up to a common " TV listening" volume level.
Let's temporarily give them credit for operating properly . . and direct our attention over to the . . . . .

POWER SUPPLY SECTION . . . .

Initial inspection reveals the use of a comom 60Hz power transformer, so no complex power supply circuitry will need to follow.
Just notice that the first Full wave rectifier circuitry is moving up to the top via D6-7 which is to supply a - DC bias to a FET mute circuit to preclude that the units power amp doesn't PLOP on turn on nor give a decaying and frequency shifting "fart" on turn off.

The Full Wave rectifiers of D1-2 provides your PLUS 38V power for the POWER OUTPUT stage, with about a 2 amp capacity with those 2200 ufd caps values.
The companion Full Wave rectifiers of D3-4 provides your MINUS 38V power for the POWER OUTPUT stage, also being at a 2 amp capacity.

Back at prototype design engineering time the need of dropping the +38 down to a sub +16 DC supply was approached by finding the worst current consuming condition of the +16V supply line, in order to come up with that current value being needed.
Then the 1N4745 would be seen to approach its 1 watt power capacity at a 62ma current consumption level.
To derate the diode load 50% . . . ~ 30 ma would be the new preferred power drain.
Then they go back to the current pull of the +15 supply and correlate in the extra 30 ma drain.
THEN they compute the needed dropping resistor between the +38 V main supply and the +15 reduced supply.
In your situation that would be your four series-parallel 1/2 watt dropping resistors which are sharing the load.
Then . .. the 1 watt zener only gets involved in corrective regulation, if the circuitry drain upon the +15VDC supply is minimized. ( However, those 15 V supply's are not being UBER critical. )

NOW . . .. where there is being a possibility of your problem source.

The C25 and C26 E-caps are the BIG DADDYS of the power source to your audio ouputs.

" What if " C25 has declined in capacitance and is NOW only being 1000 ufd ? ( 470-? 220?)
At low volume, voltage/current demands are minimal, but, as the volume is increased current demands ramp up and if there is not enough stored voltage in the cap, that voltage is pulled down after its newly lowered 1 amp capacity of a 1000 ufd cap. or (500ma ?-250 ma ?)

Incrementally, and progressively, a quasi AC ripple starts appearing in the supply instead of the prior EVEN and pure DC that was being supplied at lower power levels.
Then you will have a complementary- symmetry amp output stage starting to TRY to operate with a normal -38 supply on half of it, but only having an imbalanced +35 ? +30? +25? +20 ? ? ? ? as its companion ? supply.

Now can't you visualize the encroaching deviance, from there being a normal SYMMETRY of the output waveform ?
At some high power point, the overworked POWER transistor says " HELL . . . I can't take it anymore !!!!! . . and goes Pffffffffft collector to emitter.
Then, according to which is faster, either the fuse(s) or the other POWER transistor also goes Pfffffft.
REPEAT . . . the same situation, if it is being the C26 Negative supply capacitor that has the decline in capacitance.

TESTING . . . . .

Using meter common ground between C25 and C26 for your DC metering ground and moving the + meter probe (hoping that you have an autoranging and auto polarity meter) between the Pos and Neg of those caps . . . or F2 and F3 fuses if being nearby.
Power up the unit and get music playing at low volume and start increasing volume and remeasuring the + and - supplies to see if one is pulling down under that musical load level. It should be pronounced when your music starts limiting / distorting.

You could also measure the AC ripple on the Positive and Negative 38 supplies BUT some DMMs act real squirrelly in their AC readings, when having a combined AC and DC presence .
I avoid this by having a 1 ufd @ (max voltage expected) poly or paper capacitor . . .NO electrolytics . . . in series with my meter probe for DC / blocking isolation, while metering in AC mode.

73's de Edd
.....

 

[Jason G]

Hey 73's de Edd, thank you so much for your reply!

I started with the test of speaker ground to F3, at low volume it reads 37.00 and as i boost volume to distortion it drops to 36.8/36.7

speaker ground to F2 starts at -37.2/1 to -36.8 when the volume is hitting distortion.

Now for testing the C25 and C26, should i still do this test as it will require me remove the PCB board from the frame. My fear is the transistors as they are mounted to the metal frame and i don't want them to overheat again. Here is what i just tested.