Naim NAP 140 offset voltage

[jberry01]

I have just reset the bias current to 4.5 mv across each 0.22 ohm power resistors as per acousta.org advice. Before starting the offset voltage was 7mv each channel. The initial bias was 3mv on channel 1 and practically zero on 2. On completion the offset voltage was rechecked and it had gone mad nearly 0.5 v on ch 1 and 0.2V on the other. All I did in between was set bias and disconnect power supply to measure apprx 35ma. BUT I did not turn off the amp, just pulled out the power tag to the circuit board and inserted the ammeter leads. Then put the tag back. Could this have damged anything, if so what should be checked.
Any guidance appreciated

[bob91343]

Don't nitpick. The offset is tiny and will drift. It's only necessary to make sure it's well below the output signal to avoid loss of headroom.

An offset of as much as one Volt would probably cause no problems. It will drift with temperature so set it after the amplifier has been running for a few minutes. It's usually determined by the input stage of the power amplifier, and there is not always an adjustment for it.

[jberry01]

Maybe, but it started life at 7mV until I pulled out the power tag. I guess I could have caused damage by removing the 35V plus side of the power while leaving the negative 35V connected. The high offsets are both negative volts. What components are likely to be hit by this sort of exposure to the negative side of the power rail only.
cheers

[Mooly]

Historically anything under 100mv was "considered" as being acceptable... anything over and it's too high.

I'm pretty certain you won't have damaged anything. A faulty component (semiconductor) would normally give a huge offset. Disconnecting with it on doesn't sound like the smartest thing to do, but again it's semiconductors that would be damaged and again these would normally give a large offset toward one of the rails.

Bias... make sure adjust with no load connected.

Offset... check when the case is on the amp and it's been on a while. A temperature difference of only a few degrees between the transistors in a differential amp will give a large offset... so box it up and leave it on to check.
If it's still high then see if the bias setting is in fact altering it. It should not in any "normal" amp.

[jberry01]

Thanks for all the helpful tips. But the bottom line is that something has happened and I dont know where to look.

[Mooly]

Originally Posted by jberry01

Thanks for all the helpful tips. But the bottom line is that something has happened and I dont know where to look.

We don't know if anything has happened

You must measure and test as described perhaps resetting back to where it was.

Can you post a circuit ?

[jberry01]

I have followed everything you suggested and the end result is still high offsets, they stay the same even when I adjust the bias current.
So I am still unclesr where to look, you can download a circuit diagram from Modifying Naim Audio power amplifiers
cheers

[jberry01]

link was weird. Try Modifying Naim Audio power amplifiers

[Mooly]

OK I'm looking at the circuit, the one that says original circuit fig 1

Offset... I can not believe you have done any damage to this at all tbh as any failure would give a massive offset.

Having seen pictures and a circuit what you do is this,

1. Disconnect speakers and input leads. Make sure the amp is unplugged from the mains Having seen the circuit I would recommend you adjust the bias correctly first, at least approximately. This also must be done with no speakers attached.

2. You now need to apply a short across the inputs. Does it have RCA inputs ? If so get two phono plugs and just short centre to outer and plug them in. Or lightly solder a link across the 100k resistor at the input. This eliminates any noise due to stray pickup from the floating inputs. It's essential to ground in the way I describe. If you ground the input in any other way it will add noise which may distort the reading.

3. Now connect your meter to the speaker terminals of one channel. Do not read from chassis ground to the positive output. The meter must be across both terminals.

4. Switch on and read the offset. Does it vary with time ? If you put your finger on either TR1 or TR2 the offset will alter dramatically I imagine due to temperature imbalance... try it. Leave the amp on and with it's cover in place and see what the reading is after say 10 minutes. Repeat for the other channel.

[jberry01]

Changed all the tant caps and everything now fine offset around 4mV. I can only suspect a leaking tant although all measured ok when checked on meter after removal from circuit. Probably a very small out of tolerance leakage was the cause. Main suspect is 47 uF (C2) but I dont know for sure but it looks like the most likely to recieve some back voltage which Tants dont like.