dc offset revisited

This is my first post at Carveraudio. I am a relatively new Carver TFM35 owner. I am also quite new to high fidelity in general. I always had cheap gear, but accidentally came across a TFM35 for sale, and decided to make the plunge into the audiophile world.

In any event, I am reading like crazy trying to learn more about audio equipment, and decided to measure the dc offset of my TFM35. I am getting -8mv on the left and 16mv on the right.

I would like to adjust this for slightly better agreement. Are there pots inside the TFM35 for me to do this? If so what is the easiest way to I.D them, and make the adjustment?

I apologize if this has been specifically addressed before. I did do a search and found some relevent info. but not the specifics I need.

I should note that I am otherwise very happy with my Carver. Right now it is being driven by a Philips PA1000B preamp.

First off, I will apologize for not reading extensively enough, and I realize it must be exasperating to constantly face a barrage of newbie questions. There was significant information on bias of other Carver units, and I learned enough to forge ahead.

The best I could obtain was a warm reading of 2.0mv on the left across TP201/203 and 2.1mv across TP202/204. I saw this value for some Carver amps, but not specifically for TFM 35. I tried to get them exact, but the trimpots are really sensitive and perfect agreement seems near impossible.

Are these acceptable values?

In addition, I get about -14mv on the left at the speaker terminals, and about 14mv on the right at startup, but this drifts over several minutes to about -8 mv on the left and 20mv on the right. Is this acceptable, or does it indicate something? I figured out that mv readings at the speaker terminals are not affected by changing VR201. I thought I read (at some other forum) that some people check bias at the speaker terminals, but it is all new to me.

I am also curious, what actual differences would exist in a unit biased at say 1mv (both sides) vs. one biased at say 5mv 9 (both sides)?

Would they sound different, run hotter or cooler, or some other manifestation?

RichP is our resident expert on this stuff, but he's been busy with him own forum lately....perhaps BillD can offer some assistance?

Thanks for the welcome in the other thread.

I should note that I have since managed to achieve 2.3mv on both sides.

I also made a fascinating discovery (although this might be old news to some). In trying to get the right side to dead match the left, I noticed a fluctuation from blowing on the heat sink. The control is so sensitive and so fast, I could move the reading from 2.3 to 1.9 in about one second from a gentle breeze.

It seems that one could affect this setting by almost half a mv just by gentle air currents from body movement, or working close to a source of heat with one side of the amp closer to the source than the other. This is something that might be factored in when trying to get exact readings.

There is a procedure for adjusting them, but I'm not the one to ask. I've seen it discussed here before but don't recall the details....

Well, as far as adjusting bias, there is what is recommended (which may or may not be optimal) and what sounds best. RichP714 did some tests to see what bias settings produced the lowest distortion on his Mk.II amps. I would generally go, however, with the factory specs for your amp, and they always match left to right. Does the bias drift — sure. The procedure usually says to let your amp warm up for, say, 5 minutes before performing the bias set. I don't know what the factory setting is for a TFM-35, but I don't think 2.3mv is too high. My M-500 (250 WPC int 8 ohms) is 6mv.

Thanks. Which reportedly sounds better, lower or higher values? Is the better sound due to less distortion or some other factor?

What you're trying to do is a compromise between having the output transistors turn off for low signals (not wanted), and high voltage swings in each output device, wanted. The ideal is to have the bias voltage as low as possible without the output devices turning off at all. For instance, a Class A amplifier is fully biased. In other words all the voltage is there to meet the needs of the largest signal all the time. When less than the maximum signal is there, the device dissipates the voltage as heat. Class B amplifiers turn off the output devices when no signal is present. Class A/B, which is what the vast majority of high-fi amps are, keep the output device with just enough voltage so that they don't fully turn off when a signal crosses zero to minimize crossover distortion. So less is better, but not too low. Also, some output devices aren't all that linear at really low bias, so getting above their non-linearity can improve the sound.

Thanks for the comprehensive and interesting reply. I have one more question related to your remarks. You said "Also, some output devices aren't all that linear at really low bias, so getting above their non-linearity can improve the sound". Presumably, there is some trade off or negative aspect if one tries to improve sound by getting above the non-linearity. What is it?

I think that the major point is that each model has its own specs concerning bias settings which should be the real target and not getting the bias as low as possible. For those values, you should always refer to the manufacturers sepcifications in the service manual for the particular unit being adjusted.

Hi, I entirely agree, however, I am still curious about how such settings affect sonics even though I would be inclined to go strictly by manufacturers recommendations (which I do not know at present, but will try to secure on Monday).

Knowing something about cause and effect simply gives one greater insight into how things work. Good hi fi is a new thing for me, and only started with my acquisition of the TFM 35 and some Pioneer S-T500 speakers. I am enjoying the learning curve, and my inquisitive nature compels me to ask such questions like above.