How to bias a Plinius Odeon 6 channel amplifier

[AudioGuy88]

Hi everyone,

I am working on a Plinius Odeon 6 channel amplifier. Not much wrong with it, but I replaced some caps (20 years old) and decided to check the bias current.

The manual says

1.

[1] Check the right DC servo is set within its operating window.
Use VR2 to set TP5 to within –3.0 to +3.0 volts.

[2] Check the bias current is correct once the amplifier has warmed up, typically 15 minutes.
Use VR1 to adjust the rail current to 140mA or 52.5mV across any emitter resistor (35mA per
device).

I am assuming the green 5W resistors next to each transistor is the emitter resistor. When I measure that, every one has a slightly different value.
For example on 1 channel one of them is 54mV and the others are more like 40mV.


My question is

• Am I measuring it wrong? Or
• Is this normal? Or
• Is my multimeter just no good? Or
• They should all be the same and that it's a sign there is a fault somewhere?

Thanks

[redwire]

Re: How to bias an amplifier

You might have two problems. The loudspeaker/load must be disconnected for setting idle current. If you are getting different values, some of the bias current is going to the loudspeaker/load because there is DC offset.
I see an extra multi-turn trimpot which could be for adjusting offset, and should first be set after it's all warmed up, for as close to 0mV at the output- maybe, I don't know what it's for but check that. After DC offset then you'd set bias and hopefully the two controls don't interact much which can mean going back and forth.

If the amplifier has a DC offset say 50mV, that's 6.25mA flowing through the loudspeaker stealing bias from the output transistor to say 28.75mA and sometimes you can't achieve perfection and bias a little higher if the extra heat is OK.

[AudioGuy88]

Re: How to bias an amplifier

You might have two problems. The loudspeaker/load must be disconnected for setting idle current. If you are getting different values, some of the bias current is going to the loudspeaker/load because there is DC offset.
I see an extra multi-turn trimpot which could be for adjusting offset, and should first be set after it's all warmed up, for as close to 0mV at the output- maybe, I don't know what it's for but check that. After DC offset then you'd set bias and hopefully the two controls don't interact much which can mean going back and forth.

If the amplifier has a DC offset say 50mV, that's 6.25mA flowing through the loudspeaker stealing bias from the output transistor to say 28.75mA and sometimes you can't achieve perfection and bias a little higher if the extra heat is OK.

Thanks for the quick reply.
I should have been more clear in my question.
The first part is to set the DC servo. It says the operating window is -3V to +3V which I don't understand. However I have set it to as close as 0V as I could.
I have added that to the original post to prevent further confusion.

The manual also tells you to have no speakers connected and to let the amp warm up for 15 mins.

So these measurements are done with no speaker load, powered on for at least 15 mins, and with DC servo adjusted to as close to 0V as possible beforehand.

So am I measuring the right thing? Those green 5W resistors? Are they all supposed to have the same value? Because I get a big variation between the 8 different ones.

Thanks

[redwire]

Re: How to bias an amplifier

What are you measuring across the emitter resistors? Can you post them.
In a perfect world each pair would have the 35mA but variations are due to transistor mismatch and tolerances of the emitter resistors.
So I expect some compromise but if the values differ vastly I'd have to see the schematic for possible reasons.
Another issue can be AC hum and noise, so the input should be shorted as well so the amplifier is as quiet as possible when measuring mV.

[AudioGuy88]

Re: How to bias an amplifier

What are you measuring across the emitter resistors? Can you post them.

Not sure if I'm doing it right, but I find the values very hard to measure. They will change and sometimes by quite a bit. Also when I remove the mm then put it back the value may be different.
I have used the highest value and have got that as close to 52.5mV as I could.

In any case here are the values I got (mV).

• R5 - 50.5 - 51
• R6 - 51.2 - 52
• R8 - 46.9
• R9 - 46.3
• R15 - 52.3 - 52.6
• R16 - 49.5
• R22 - 48.3 - 48.5
• R24 - 46 - 46.3

In a perfect world each pair would have the 35mA but variations are due to transistor mismatch and tolerances of the emitter resistors.
So I expect some compromise but if the values differ vastly I'd have to see the schematic for possible reasons.

I can send you the schematics if you want.

Another issue can be AC hum and noise, so the input should be shorted as well so the amplifier is as quiet as possible when measuring mV.

I have not shorted the inputs when doing the measurements

[redwire]

Re: How to bias an amplifier

That looks reasonable for matching but the bias mV readings wandering around is a bit strange.
Mains voltage can move around, semiconductors have popcorn noise, the amp can pick up WiFi/cellphone/AM radio RF, multimeter leads with oxide on them and near the hum and noise from the power transformer etc. - all can make the bias current measurement move around.

You can listen to the amp's (idling) output with cheap headphones, of course they will blow up if any thumps or more happen (don't use a shorting jack). But you can instantly listen to the noise floor of the amp and hear the dirt - crackling, rumbling, cellphone bleeps, hiss, microphonics etc.

If one channel is way different than the other, then it might be worth investigating.

The servo window is just what it can accommodate for DC offset +/-3V is the max it will correct for. If it is noisy (op-amp) or the electrolytic capacitors are misbehaving that might also make it the culprit, a noisy servo.

[AudioGuy88]

Re: How to bias an amplifier

That looks reasonable for matching but the bias mV readings wandering around is a bit strange.
Mains voltage can move around, semiconductors have popcorn noise, the amp can pick up WiFi/cellphone/AM radio RF, multimeter leads with oxide on them and near the hum and noise from the power transformer etc. - all can make the bias current measurement move around.

You can listen to the amp's (idling) output with cheap headphones, of course they will blow up if any thumps or more happen (don't use a shorting jack). But you can instantly listen to the noise floor of the amp and hear the dirt - crackling, rumbling, cellphone bleeps, hiss, microphonics etc.

If one channel is way different than the other, then it might be worth investigating.

The servo window is just what it can accommodate for DC offset +/-3V is the max it will correct for. If it is noisy (op-amp) or the electrolytic capacitors are misbehaving that might also make it the culprit, a noisy servo.

Thanks again for the quick reply.

I take it from the response that variations from 46-52 mV is reasonable and not unexpected. This is good, so I will not worry about it.
I am taking the highest value and setting it to 52.5mV as the manual instructs.

I have connected some speakers to the output without any input and there is a little bit of hiss and static coming from the amp. Not much, but more than you'd expect from a very expensive amp. All the channels do it, and it's certainly not audible unless your ear is right next to the speaker.

I will bias all the channels as close as I can to match.

Thanks again for your help

[cdphobe]

Re: How to bias an amplifier

Apologies for dragging up an old(ish) thread but I'm just about to dive into a Plinius Odeon that's been sat unused for a couple of years and came across this while looking for some technical info. @AudioGuy88 you make reference to some instructions for bias setting and a manual (which is clearly not the user manual) - could you point me in a direction where I could find similar and/or any circuit diagrams?

Many thanks,

Al