Forte 1A Help Needed

[Panelhead]

I have recently picked up a nice Forte 1A to play with. The unit works fine, the prevous owner said he never touched after buying new. It sounds great.
There is a small issue. On one side the heatsink is about 20 degrees hotter than the other. One side is barely warm, the other is what I expect a 50 watt class A amp to run.
My system only requires 10 watts or so, I would like to idle down the hot side. And maybe idle up the cool side. Right now, running less than a watt the sound is the same.
There is a single turn pot on each circuit board. On the heatsinks there is a NTC and something labeled 26G +085A and the other 49G + 085A. Assume these are to control runaway.
Both channels are only running a mv or so dc offset, and ac is 0.4 mv unloaded for each channel. Outstanding!
Can anyone help me with details on adjusting bias, or where to look if there is a problem causing one channel to run so cool?

TIA,
George

 

[Joe Berry]

You want to measure the bias on each channel and adjust it. To do this, first locate the 0.47 ohm power resistors on each circuit board. There are ten per board, one for each output transistor, and they will be among the larger resistors present.

The factory adjustment (IIRC) was 120mV across each resistor with the amplifier idling hot. There will be some variation across the resistors, so you may want to measure all of them to find one that represents the group average, and use it as the reference.

The single-turn pot is the bias adjustment. The pot on the cold channel may have an open wiper due to dirt or oxidation, but this problem may clear up when you turn the wiper. If it does, be prepared for a sudden increase in bias when you touch it.

If you set the bias a little below factory, say at 100mV hot, the heat sinks may feel better and the amp will still run class A up to half power or so.

The devices marked +085A are thermal cutouts with an 85 degree C opening temperature. They don't provide thermal feeback for the bias; that job is handled by a thermistor soldered on the backside of each board and in physical contact with the heat sink.

 

[Panelhead]

Many thanks!

You want to measure the bias on each channel and adjust it. To do this, first locate the 0.47 ohm power resistors on each circuit board. There are ten per board, one for each output transistor, and they will be among the larger resistors present.

The factory adjustment (IIRC) was 120mV across each resistor with the amplifier idling hot. There will be some variation across the resistors, so you may want to measure all of them to find one that represents the group average, and use it as the reference.

The single-turn pot is the bias adjustment. The pot on the cold channel may have an open wiper due to dirt or oxidation, but this problem may clear up when you turn the wiper. If it does, be prepared for a sudden increase in bias when you touch it.

If you set the bias a little below factory, say at 100mV hot, the heat sinks may feel better and the amp will still run class A up to half power or so.

The devices marked +085A are thermal cutouts with an 85 degree C opening temperature. They don't provide thermal feeback for the bias; that job is handled by a thermistor soldered on the backside of each board and in physical contact with the heat sink.

Joe,
Thanks, I willl look into biasing it properly. Looking the single turn pot is in about the same position on both channels so it most likely has lost the wiper connection.
I like the sound of the unit. It is warm sounding, need to let it cook longer to evaluate better.

George

 

[GRollins]

The position of the pot is not a good indicator of where the bias is set. Pots are not exactly what you'd call precision devices, so the resistance can vary, even at the same number of degrees of rotation. By the same token, even if the resistors and such are 2%, 1% or better, the output devices will likely only be matched for that channel, not for the whole amp, so the bias can easily vary widely between the two channels.
I don't know about Forte, but in the Threshold product, Nelson set final bias by the temperature of the heatsinks. That neatly sidesteps the variations from device to device, since the thermal mass of the heatsinks serves to average the heat into one neatly measurable quantity. You could do worse than to use a digital kitchen thermometer to set the two channels to the same temperature.

Grey

 

[Panelhead]

The position of the pot is not a good indicator of where the bias is set. Pots are not exactly what you'd call precision devices, so the resistance can vary, even at the same number of degrees of rotation. By the same token, even if the resistors and such are 2%, 1% or better, the output devices will likely only be matched for that channel, not for the whole amp, so the bias can easily vary widely between the two channels.
I don't know about Forte, but in the Threshold product, Nelson set final bias by the temperature of the heatsinks. That neatly sidesteps the variations from device to device, since the thermal mass of the heatsinks serves to average the heat into one neatly measurable quantity. You could do worse than to use a digital kitchen thermometer to set the two channels to the same temperature.

Grey

That is a very good idea. The value of the resistor used to measure the voltage drop may have not been close, and each of the five pair of transistors may not be matched any more.
I have a long way to go though, the one channel most likely has a problem of sort type, not just the pot setting. The heatsink is barely above room temp. Hopefully turning the pot will get it in the ballpark.
In my system, the two channels sounded about the same. This surprised me. But with 100 dB efficient speakers playing at mid 80's level the power demand is almost nothing. The bass was full and lush on the cold channel. Makes me think the bias could be dropped off the page on the normal channel.

George

 

[Panelhead]

You want to measure the bias on each channel and adjust it. To do this, first locate the 0.47 ohm power resistors on each circuit board. There are ten per board, one for each output transistor, and they will be among the larger resistors present.

The factory adjustment (IIRC) was 120mV across each resistor with the amplifier idling hot. There will be some variation across the resistors, so you may want to measure all of them to find one that represents the group average, and use it as the reference.

The single-turn pot is the bias adjustment. The pot on the cold channel may have an open wiper due to dirt or oxidation, but this problem may clear up when you turn the wiper. If it does, be prepared for a sudden increase in bias when you touch it.

If you set the bias a little below factory, say at 100mV hot, the heat sinks may feel better and the amp will still run class A up to half power or so.




It did just as you thoughgt. As soon as the pot was turned the bias came up. Set it a 80 mv. The other channel was 150 mv, cranked it down.

I don't know about Forte, but in the Threshold product, Nelson set final bias by the temperature of the heatsinks. That neatly sidesteps the variations from device to device, since the thermal mass of the heatsinks serves to average the heat into one neatly measurable quantity. You could do worse than to use a digital kitchen thermometer to set the two channels to the same temperature.

Grey

Learn from the experts. After dialing in the bias, it was checked and is now heating both sides up 15 degrees C. This is about half of what the hot channel was running when the voltage across the source resistor was 150 mv.


Thanks for the help. This was a very easy problem to fix, once I knew where to start.




George