Traynor TS-50B - Negative half of sine wave clipped - low bias?

[garytoosweet]

I have a Traynor TS50. I replaced output and driver transistors. MJ15015G for outputs, BC139 and BC140 for drivers.

As the amp volume is increased the negative half of a sine wave is severely clipped. I am thinking bias is too low.

Voltage drop across 0.27R emitter resistors is -14mVDC on one and 19mVDC on the other.

It is my understanding the voltage drop across emitter resistor can be used to calculate the idle current of that transistor, so if I=V/R, idle current is -52mA for one transistor and 70mA for the other.

How do I determine an appropriate idle current for this amp? Should I bias for symmetrical clipping and make sure idle current is within a reasonable range?

What resistor do I change to adjust bias? I'm thinking I would adjust the 39k coming off the negative rail but not sure.

Thanks in advance for any help!

[Enzo]

Why do you think it is a bias problem?

[garytoosweet]

Why do you think it is a bias problem?

I guess in my limited knowledge I do know (have read) that incorrect bias can cause the same symptom that I am seeing. So my thought was to check and adjust bias and see if it resolves the issue. Something that should be done in any case with the Output transistors swapped.

My limited experience biasing solid state amps have always have some notes from manufacturers, usually to adjust some trimpot for no crossover notch or a certain mV value across a specific resistor.

The sine at output looks identical to the wave on the right of the attached graphic.

[Jon Snell]

Ah, you maybe confusing Bias with DC offset.
With no signal, is the junction of the 0R27s at zero volts? It should be.
When you drive the output to clip on the lower half, does the dc voltage across the -ve voltage smoothing capacitor, reduce? If it does then the common from the transformer may be badly connected.

Edit; more questions asked.

[garytoosweet]

Ah, you maybe confusing Bias with DC offset.
With no signal, is the junction of the 0R27s at zero volts? It should be.
When you drive the output to clip on the lower half, does the dc voltage across the -ve voltage smoothing capacitor, reduce? If it does then the common from the transformer may be badly connected.

Edit; more questions asked.

Thank you will do some more research on the two and their differences.

130mV at junction of 0R27's. Is that close enough to zero? I have seen a wide range of numbers for acceptable dc offset voltage, normally up to 100mV is ok?

Voltage across -ve 2200uf cap is super stable.

[Helmholtz]

130mV with or without load?
Idle currents should always be measured without load.
130mV will cause 16mADC to 20mADC through the load and would explain the difference between emitter resistor drops.

[garytoosweet]

130mV with or without load?
Idle currents should always be measured without load.
130mV will cause 16mADC to 20mADC through the load and would explain the difference between emitter resistor drops.

I had an 8 ohm load plugged in! My mistake.

I removed load and turned amp on. After 8min the emitter resistor voltage drop stabilized to 8.9mV and -8.9mV So about 33mA idle current per transistor.

The DC offset measures 79mV.

[g1]

When you have it clipping like that, is the signal out of the master pot unclipped?

[garytoosweet]

When you have it clipping like that, is the signal out of the master pot unclipped?

The signal at the volume pot is unclipped, although showing something odd that is probably indicative of a separate issue in the pre that I will also have to investigate.

Attached photos of the scope at both points.

Gain control on 1/10, Volume on 3/10.

Wave @ volume pot wiper 50mV/div.
Wave @ speaker out 500mV/div.

[The Dude]

Are both of your rails still at 35V and clean when you drive the amp to clip?

[Enzo]

I bet your scope is set for AC coupling. Set it to DC coupling, and center the trace. Now scope the output. You need to differentiate between the bottom clipping and simply missing the whole bottom half. Scope on AC, and the whole trace will drift down to center on the average. On DC, the scope will show exactly what is ther, and you may find the top half works and the bottom not.

I'd wager that was the case, and in that case you could have an open resistor or broken trace.

[vintagekiki]

What resistor do I change to adjust bias?

- When checking that the amplifier clipping symmetrically, bring the signal from the generator directly to the power amp input (via 4k7 resistor)

- Bias adjust by changing the value resistor 18 Ohm to MJ15015 IC = 25mA.
- Check (replace) 2 x 1N4148 in bias circuit.

- Feel free to temporarily remove diodes 1N4148 in protection circuit (collectors MPSA18/MPS8599) and check that the amplifier clipping symmetrically. If the amplifier without the protection transistors (MPSA18/MPS8599) clipping symmetrically, replace the defective components in the protection circuit.

- If change the drivers (2N6553/6556) or output (MJ15015) transistors, they should be relatively matched pair.
- If change output transistors check (replace) 2 x OR27/10W.

[J M Fahey]

Yes.

You are SEVERELY clipping the bottom half, or straight missing all of it.
It is NOT a bias problem, so for now leave that aside, until amp is working.

1) scope amp with NO load to begin with and drive it to clipping.

a) Do top and bottom peaks clip (almost) symmetrically now?

b) And applying load we get the severe bottom clipping?

If so, amp is "almost" working but can not feed **current** into a load, power transistors are not doing their job or not driven , so drivers or even the predribers are trying to drive speakers.
They can supply *voltage* into a "no load" but no current.

Do the A and B tests and post results.

2) or failed short protection clips current even without a short circuit.

Just for testing, lift 1 leg of 1N4148 diodes on collectors of MPSA18/MPS8599 protector transistors.

Repeat tests A and B.

[garytoosweet]

Yes you are all right I was missing the bottom of the sine wave. I had misinterpreted what was happening on the scope as I was seeing a very small complete sine wave at first, then the bottom quickly flattening out while the top half was normal. Setting to DC coupling as you suggested Enzo, was much clearer. Now I understand the full small sine I was seeing was likely from the predriver.

So I traced the signal back from the output and found it was not passing through the driver transistor, swapped it out and now amp is working, passing a nice signal and making right around 50W.

Voltage drop across emitter resistor is 1.4mVDC, so idle current is 5mA per transistor, seems low. However, there is no crossover notch at all. The amp sound good at any volume. Should the bias still be adjusted to ~25ma per transistor?

DC offset at idle is 120mV.

Thank you all for a ton of great help and info. Very much appreciated.

[g1]

No one at the factory or it's authorized service centers ever adjusted bias on those. They would put a trim pot for bias adj. in there if they thought it was needed. If there is no more fault, the bias is good.