How much output offset voltage is too much?

[tubetvr]

Hi

I am experimenting with different amplifier circuits and have experienced that output offset voltage can be quite different in different circuits, also some circuits drift with temperature so even if offset is very low initially it can increase quite a lot when the amplifier gets warm

Now my question, how much offset is too much? It is obvious that if offset is very high, i.e. several volts most electrodynamic speakers will react and the speaker membrane will be in a offset position either out or in from the resting position. In this case either the voice coil will go to the end position or the speaker will give excessive distorsion as the voice coil is not centered in the magnet gap.

So how much offset is tolerable? 50mV, 100mV, 200mV, 500mV or 1V or even more?

Regards Hans

 

[Tube_Dude]

Hi Hans...

For me , my target is less than 50 mV at the output...with the normal +- 5 Ohms speaker DC resistence ...10 mA is the of set current circulating in the voice coil...

But some people can be less stringent...

 

[Circlotron]

My class A has a constant 325mV at the output. No problemo. Zero.

 

[Richard C]

I'd go for less than 50mV if possible. Moving coil loudspseakers are only reasonably linear over a few milimeters of cone travel so it pays to operate with as little offset as possible.

 

[fmak]

My class A has a constant 325mV at the output. No problemo. Zero.

--------------------------------------------------
This is way too high and biases your speaker cone popsition.

----------------------------------------------------
I'd go for less than 50mV if possible. Moving coil loudspseakers are only reasonably linear over a few milimeters of cone travel so it pays to operate with as little offset as possible. [/QUOTE]
----------------------------------------------------

I agree; I'd go for <25mV and my systems have less than 10mV.

 

[peranders]

I think this question has a rather floating answer. Take a small 10 W speaker and a huge PA speaker. I'll guess that over 1 volt isn't good if we talk distortion but under 100 mV is good I think and normal also.

 

[EchoWars]

I think this question has a rather floating answer.

Probably sums it up. It's a personal thing. On amps I'm repairing, anything over 50mV warrrants investigation. On my personal amps, anything over 15mV needs attention. I like to have them at <2mV.

 

[markp]

My amps are within 5mv without a servo. If you have lots of offset then try to match your transistors better or build a servo to control it.

 

[EchoWars]

I'm with you Mark. But the problem with matching is that they aren't going to stay matched. Check your amp in a few years.

I have only recently been introduced to the whole servo thing...I like. Long-term stability, no muss, no fuss. I do believe I am a convert.

 

[fmak]

Probably sums it up. It's a personal thing. On amps I'm repairing, anything over 50mV warrrants investigation. On my personal amps, anything over 15mV needs attention. I like to have them at <2mV.

--------------------------------------------------
I agree with your numbers. However, this is more than a personal thing and doesn't dpend on the size of the speakers but on sensitivity and cone displacement.

 

[Jennice]

Odd...

Hi folks...

I have the odd feeling that this topic has been brought up not long ago in this forum...!?
Deja Vu, anyone...???

Anyways... I'd say less than 100 mV is acceptable. Let's bring in some math and kill the :

Using the P = U * U / R
formula, even a DC resistance of as low as 1 ohm would still mean less than 10mW. Compare that with the power levels of a headphone amplifier to get an idea of the magnitude.
Also, I would say that 10 mW can't cause any heat damage.

Just my 2 cents -
Jennice

 

[EchoWars]

Well, if the amp has a diff-pair input, there are other considerations here than simply the power lost at the speaker.

From Mr. Self

Exact DC balance of the input differential pair is essential for minimum distortion. It seems almost unknown that even minor deviations from equality of collector current (Ic) in the input devices seriously upset the 2nd-harmonic cancellation

Here.

No dude, just math..

 

[fmak]

Re: Odd...

[Anyways... I'd say less than 100 mV is acceptable. Let's bring in some math and kill the :

Using the P = U * U / R
formula, even a DC resistance of as low as 1 ohm would still mean less than 10mW. Compare that with the power levels of a headphone amplifier to get an idea of the magnitude.
Also, I would say that 10 mW can't cause any heat damage.

-------------------------------------------------------
Wattage and heat issues are not the reason. Cone displacement is, as is dc balance of amp. If you allow 100 or 325 mV steady state, dynamic excursions can be very much higher and you are doing your system no favours.

 

[Jennice]

fmak,

I totally agree that zero DC offset is desireable, but I understood the original question as asafety / "serious concern" issue, and not as much as a perfectionist issue.

Jennice

 

[cunningham]

The circuit I designed from scratch only has an offset of less than 10mV rated 100W RMS into 4Ohms