Peavey TKO-65 hum

[cowasaki]

I've taken up fixing amplifiers for fun!  Anyway I've managed to fix 24 of 25 so far and I think that we could probably all guess that the one I haven't yet managed to fix is the only one where "my mate has already had a look and says it's knackered"....

Anyway I got it partially working before slipping up and shorting out of the transistors in the output section.  I've fixed that now with two new power transistors. The issues were many and included dirty pots, 2 knackered poly caps (I've replaced all of them now), a transistor removed by his mate to test and re-inserted the wrong way! plus lots of other rubbish......

It now works BUT there is a loud 100hz hum.  It is as loud as the input signal!  If you switch it off at the isolator the hum goes but the input signal continues till the caps loose their current. The actual signal is nice and clean otherwise. I’ve pushed music through it as well as test tones

I've attached a schematic and I'm sure it's coming from the mains being twice the UK 50Hz.

I've replaced the 10nF cap across the bridge rectifier, both 2200uF caps (C52/54) and C53/55.  None appeared to be faulty according to my Peak cap tester but grasping at straws now....  All other caps measure well within tolerance and low ESR.  All polys have been changed as I mentioned earlier.  All the transistors have been tested and having discovered that U5 isn't required to get a sound I've tried it in place of the other op-amps of the same type.  U3 is an op-amp I don't have and I'm not sure what I could substitute it with (even if just for testing).

All pots have been removed dis-assembled cleaned, de-oxitted, re-assembled and tested and are like new.

Input jacks are new high quality ones

Any help would be appreciated even if just pointing me at other things I may not have thought of.  I am quite new to amps having come from computers so I might have missed something obvious!

[don.r]

Routing of mains and/or input wiring? Could be from capacitive or inductive pickup.

[cowasaki]

Routing of mains and/or input wiring? Could be from capacitive or inductive pickup.

The mains is on the other end of the board so it's not that unfortunately.

[Andy Watson]

Input jacks are new high quality ones

Did you replace like with like? Is there a chance that you have created a ground loop? And/or not grounded something that should be grounded.

[cowasaki]

Input jacks are new high quality ones

Did you replace like with like? Is there a chance that you have created a ground loop? And/or not grounded something that should be grounded.

I photographed the original jacks - unless they had been messed up by the previous owner's "friend"

This is the original photograph:



Basically BLACK is ground whilst red and white are the two inputs.  The black goes through the eyelet and on to the one next to it.  This can't be seen too well in the image.

It is shown in the circuit diagram but I can't tell if it's right or wrong.  If you look at the input with the black and white wires going to it you can see that the two pieces of metal going to the hot connection don't make contact so the inputs were bent and corroded. 

[Andy Watson]

The wiring of the black wire does not look like something the factory would have done. However, the provision of three contacts in that white connecctor suggests that the jacks were wired to ground (rather than relying on the chassis.)
 What do you know of the history of this amp? You say it has been "worked on" before - do you know why? What was the original fault - i.e. the fault that caused it to be worked-on? How much of the original wiring has been changed?

[cowasaki]

The wiring of the black wire does not look like something the factory would have done. However, the provision of three contacts in that white connecctor suggests that the jacks were wired to ground (rather than relying on the chassis.)
 What do you know of the history of this amp? You say it has been "worked on" before - do you know why? What was the original fault - i.e. the fault that caused it to be worked-on? How much of the original wiring has been changed?

Unfortunately the person who tried to repair it was an idiot who didn’t know what he was doing. He managed to solder a transistor in backwards and his general level of soldering technique was akin to mine using a 6” nail in a pair of mole grips over a camp fire! The wiring to the inputs looks wrong but I can’t quite work out what Peavey are showing in the circuit diagram. The wiring is shown there but I’m not sure what it’s trying to say!

[Andy Watson]

The wiring to the inputs looks wrong but I can’t quite work out what Peavey are showing in the circuit diagram. The wiring is shown there but I’m not sure what it’s trying to say!

So there's a chance that the amp is not wired correctly. Do you have an actual wiring diagram - i.e. one than shows the wiring between modules? The schematic alone does not show how the manufacturer implemented the connections.
Obviously the jacks should be grounded - somewhere! And it looks as if they were intended to be grounded to chassis - so (I'm guessing) either there is an extra ground elswehere that shouldn't be there (creating a ground loop), or there's a ground missing such that the amplifer is floating with respect to the chassis. Is there any evidence of extra or missing grounds around the main electrolytics ? (i.e. where work has previously been done).

You say that the amplifier works (?) apart from the hum? - This suggests that the semiconducctors in the signal path are ok.

[cowasaki]

The wiring to the inputs looks wrong but I can’t quite work out what Peavey are showing in the circuit diagram. The wiring is shown there but I’m not sure what it’s trying to say!

So there's a chance that the amp is not wired correctly. Do you have an actual wiring diagram - i.e. one than shows the wiring between modules? The schematic alone does not show how the manufacturer implemented the connections.
Obviously the jacks should be grounded - somewhere! And it looks as if they were intended to be grounded to chassis - so (I'm guessing) either there is an extra ground elswehere that shouldn't be there (creating a ground loop), or there's a ground missing such that the amplifer is floating with respect to the chassis. Is there any evidence of extra or missing grounds around the main electrolytics ? (i.e. where work has previously been done).

You say that the amplifier works (?) apart from the hum? - This suggests that the semiconducctors in the signal path are ok.

I’ve followed the schematic and rewired the input jacks. It work perfectly for five to ten seconds then hum again.  I’m sure the jacks are now correct. Just looking at the zenor diodes and C2.

[cowasaki]

I’ve replaced the zenors as I had the right value here and it was a quick job. C2 cap is fine.

[cowasaki]

What doesn’t help is peavey using their own parts numbers. The main power transistors were SJ6392 which is couldn’t find at a reasonable price. I was then told that they are. 2N3055 I’ve got replacement op amps coming tomorrow as they were less then £3 and will go into my stock but there are two transistors 5332 and 5331 that I cannot find.  They test as good using my Peak meter but may be out of spec.

At this rate I will have replaced everything but the PCB. :-)

[Kwakerman]

Have you actually scoped the power supply rails to make sure they are free from ripple?  Could be something as simple as a dead diode in the bridge rectifier

And yes I'm the same kwakerman from Landyzone where we discussed speedo calibration!

[cowasaki]

Have you actually scoped the power supply rails to make sure they are free from ripple?  Could be something as simple as a dead diode in the bridge rectifier

And yes I'm the same kwakerman from Landyzone where we discussed speedo calibration!

Hello.

I’ve measured the ripple at 1.2v 100hz on the + line only. Basically there is a 100hz sine wave sat on top of the 30v rail. I checked the diodes and they show as ok. They have a Peavey part number but I’m sure that I can replace them with an alternative as they are only acting as a rectifier. 

[Kwakerman]

MR502s aren't anything too special and seem fairly common.  https://www.web-bcs.com/diode/dc/ma/MR502.html  https://www.ebay.co.uk/itm/MR502-Diode-200V-3A-DO-27-Motorola-lot-de-10-/172238616154

[cowasaki]

MR502s aren't anything too special and seem fairly common.  https://www.web-bcs.com/diode/dc/ma/MR502.html  https://www.ebay.co.uk/itm/MR502-Diode-200V-3A-DO-27-Motorola-lot-de-10-/172238616154

I ordered some diodes last night as I didn’t have anything up to that spec. Again if it’s not them they are good to have in stock. Not much else it could be though now!

[cowasaki]

Well the diodes turned up and made no difference....



So I hooked up my bench power supply at maximum voltage of 31/0/-31 which is less than the correct voltage but gave it a go and got a nice clean sound with the occasional crackle.  I then tapped round the circuit with my trusty chop stick finding no issues then found the two zenor diodes that I replaced are hot to the touch.  I think that the ones I fitted are not quite up to the job.  I will refit the originals tomorrow or new ones of the correct rating.

I am learning a lot with this amp which isn't worth what I have spent on bits other than the learning experience.  I have created quite a number of test leads to safely connect a speaker cab and signal generator etc etc and am quite happy with the progress.


So on to my next question......


I have two "Tenma 72-10495" PSUs which are EACH effectively two 31v current and/or voltage limited supplies in the same box.  You can run them independently or alternatively lock them together to give 61V between the positive of one side and the negative of the other side.  This also means that the negative of the first side is effectively a centre tap giving 31v GND -31V

NOW.....  Can I run BOTH PSUs in linked mode and connect the negative of the first to the positive of the second to create a fixed centre point and give me the ability to supply 61v GND -61V Is that reasonable or a stupid idea?

Most of the solid state amps I am repairing and looking at expect 40v+ on both positive and negative rails.  Or should I just run these amps at a lower voltage?

[sean0118]

NOW.....  Can I run BOTH PSUs in linked mode and connect the negative of the first to the positive of the second to create a fixed centre point and give me the ability to supply 61v GND -61V Is that reasonable or a stupid idea?

You'd have to check the maximum isolation voltage rating of the Tenma supplies.

But probably best to avoid if possible anyway, you'd have 120V present then.  It's better to use low voltages when troubleshooting if possible.


edit:
There's a Youtube video that shows the wiring of the jacks. Looks like the wiring of the black wire etc is stock:


Does the hum go away if you unplug the jacks? Also, what capacitors did you replace the 2200uF's with?

oh, have you checked the resistance between the different connections to chassis gnd on the PCB? Maybe there is a broken track / solder pad. 

[duak]

For a ripple voltage of 1.2 Vp-p on +43V with 2200 uF caps, I calculate a current draw of 264 mA.  This is a little higher than I would expect.  I don't see any adjustment for idle current though.

I wonder if the output stage isn't oscillating.  I understand that modern 2N3055s use a different construction (epitaxial?) than the ones from the 60s & 70s that gives them a higher fT, ie. current gain bandwidth, that could make them more prone to oscillation.  See https://en.wikipedia.org/wiki/2N3055   When I was working with power amps, an oscillation often manifested as a hum because of increased current draw as well as a modulation of the oscillation frequency due to the ripple on the power supply.  Are the output transistors hot with no signal?

This oscillation is a possibility when using emitter followers with capacitive loads and is more likely with faster transistors.  There's usually a small inductor in series with the speaker to isolate it from the transistor at high frequencies.  Most of these inductors are a number of turns of magnet wire wrapped around 10 ohm, 1 or 2 watt resistors.  The wire is soldered to the resistor leads and the resistor acts as a parallel damper.

Also, I think the 2N3055 is marginal here.  I've considered  them to be safely useable up to 60 V whereas this circuit can easily subject it to at least 86 V and probably more.  In an audio amp, the inductance of the speaker can cause high voltages to be generated when the load current is reversed.  I suggest something like an MJ15015G.

Best o' luck,

[cowasaki]

Ok VERY strange. I’m powering the amp using my bench supply and it is still breaking up and noisy. But when it starts up it sounds ok for a short while sometimes. 

But if I short out C48 it works !!! Not just for a while either, it sounds fine with no breakup pops bangs or anything. I replaced the cap and it’s back to not working then short it and it works again. However Q10 gets hot and Q11 stays stone cold.

[Renate]

Yow, don't short C48!
You are unbalancing the whole output stage.
You are putting a large amount of DC through your speaker.
Which is why Q10 is hot.

Did you check the DC voltage at the speaker?

Did you check the ripple on the +/- 15V?

Have you tried just putting your test signal in the "PWR AMP IN" socket?

Have you tried removing U3 (is it socketed?) or grounding that "T.P.1"?
(That defeats the whole overdrive protection.)

Odd things (or errors) in the schematic:
The bleeder 2k is shown as across only the +43, for under a Watt dissipation.
I would guess it is across the + and - in reality, for 4W or so, more logical for a 5W R.
The pilot LED is really off the secondary? It really gets hit with a reverse potential of 43V? A hearty LED.

[cowasaki]

Yow, don't short C48!
You are unbalancing the whole output stage.
You are putting a large amount of DC through your speaker.
Which is why Q10 is hot.

Did you check the DC voltage at the speaker?

Did you check the ripple on the +/- 15V?

Have you tried just putting your test signal in the "PWR AMP IN" socket?

Have you tried removing U3 (is it socketed?) or grounding that "T.P.1"?
(That defeats the whole overdrive protection.)

Odd things (or errors) in the schematic:
The bleeder 2k is shown as across only the +43, for under a Watt dissipation.
I would guess it is across the + and - in reality, for 4W or so, more logical for a 5W R.
The pilot LED is really off the secondary? It really gets hit with a reverse potential of 43V? A hearty LED.

Thanks. This is the only amp in 25 I haven’t been able to fix. As it seemed to work for a second when switched on from cold and then a short while when switches off I just tried to discharge all the electrolytics and suddenly it sounded right. I’m not running it like that for long it was just a few seconds. I have my suspicions about Q9 and Q8 but they test ok out of circuit. Q9 was the transistor I accidentally shorted in the first place. Splitting pre and power amp is one of the first things I do. With the bench power supply the pre amp section works perfectly. The issue is post R45.

I’ve replaced all the 4558 op amps, Q3 and 4, all the electrolytics. The input jacks are now wired exactly as they should be. All the large resistors and most other ones have been tested. I’ve done a visual check with my amscope. All the pots have been removed, disassembled, cleaned, lubed, tested and reinserted.

[Renate]

So, what's the DC output of the amp when it's clean and when it's not?
Is the "Comp" or "Clip" or whatever LED ever on?
Did you verify that 0.54V & 0.51V on R75, R78, R82?

[duak]

Here's a link to a Peavy parts cross reference: http://matthieu.benoit.free.fr/cross/competitive/Peavey_Transistor_Cross.pdf

According to the above, the 6392 devices cross to 2N3055A, an improved version of the 2N3055.  I still think it's marginal because of its low Vce rating.

When you have a moment, give us some circuit voltages, in particular the ones on Q3 & Q4, those called out near Q5, Q8 & Q9 and across R78 & R79.

Is anything getting hot?

[cowasaki]

Here's a link to a Peavy parts cross reference: http://matthieu.benoit.free.fr/cross/competitive/Peavey_Transistor_Cross.pdf

According to the above, the 6392 devices cross to 2N3055A, an improved version of the 2N3055.  I still think it's marginal because of its low Vce rating.

When you have a moment, give us some circuit voltages, in particular the ones on Q3 & Q4, those called out near Q5, Q8 & Q9 and across R78 & R79.

Is anything getting hot?

My 3055s are ones I've had for years but not used and likely not the A model.  I've been servicing my Peavey valve amp this evening so not looked at the TKO.  If I re-instate the capacitor I mentioned above and use the bench supply I get very poor quality and noisy output.  I will measure these locations though.  I may get chance tonight but it may be tomorrow.

[cowasaki]

Here's a link to a Peavy parts cross reference: http://matthieu.benoit.free.fr/cross/competitive/Peavey_Transistor_Cross.pdf

According to the above, the 6392 devices cross to 2N3055A, an improved version of the 2N3055.  I still think it's marginal because of its low Vce rating.

When you have a moment, give us some circuit voltages, in particular the ones on Q3 & Q4, those called out near Q5, Q8 & Q9 and across R78 & R79.

Is anything getting hot?

Q10 is mildly warm
Q11 is stone cold but eventually becomes warm
Q8 is very mildy warm eventually getting warm
Q9 is at room temp
Q5 is at toom temp


Across R78 = 0.26V
Across R79 = 12-13mV

Q3 EBC = 1.41 / 1.03 / 27V
Q4 EBC = 1.37 / 1.95 / 30.1V
Q5 EBC = 30.6 / 6 to 6.5 av 6.3 / 30
Q8 EBC = 30.6 / 30.9 / 7.3 (if physical pin out is same as Q5 ie left to right)
Q9 EBC = 3.6 / -30 / 3 to 3.2
Q10 EBC = -30.9 / -30.6 / 2.5
Q11 EBC = 2.2 to 2.7 / 3.2 to 3.6 / 30.9


-V rail = 31
+V rail = 30.96

15V rail = 15.37
-15V rail = -15.24