Several questions about bias

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Forgive me this :) What is the right procedure for setting a bias for a car amp??? Will too-low bias contribute to distortion, loss of gain, and how much, if it is?

I always used either current draw increase measurement method, or voltage measuring across FET's resistor... Until I got one very nice amp, with a very tricky channel - it's bias, setted to normal value (amp is at operating temp, 12v at bench) tends to be very dependant of external conditions, being installed in car - temperature, voltage. All ends up in blown FET's at random times. When I set a bias significantly lower then normal from the beginning, it works reliable, but sounds not the way it should. All parts in bias circutry for this channel seems to be good.

:confused:
 
Hi,
judging by the sound quality of all the car systems I have heard, I am surprised the designer of many of them have heard about bias never mind making it adjustable.
Maybe I have a biased view?

If the bias moves around with temperature it sounds like a bad design.
If it blows up from bias instability it does not deserve to be on the market.
If it blows up because it has been asked to do too much at the running temperatures found in a car, then choose more carefully how you locate and ventilate it.
 
valterdaw said:
Forgive me this :) What is the right procedure for setting a bias for a car amp??? Will too-low bias contribute to distortion, loss of gain, and how much, if it is?

I always used either current draw increase measurement method, or voltage measuring across FET's resistor... Until I got one very nice amp, with a very tricky channel - it's bias, setted to normal value (amp is at operating temp, 12v at bench) tends to be very dependant of external conditions, being installed in car - temperature, voltage. All ends up in blown FET's at random times. When I set a bias significantly lower then normal from the beginning, it works reliable, but sounds not the way it should. All parts in bias circutry for this channel seems to be good.

:confused:

Try setting your bias with a scope. just do a simple sine wave setup and look for notch distortion at the zero crossover point. It will be there when the bias is turn off all the way. Then adjust your bias up to the point that the notch disappears. This will be you minimum set point for your bias, any more than this is a matter of listening tastes and SQ.
Remember that these car amps have a single common issue, a general lack of cooling capacity when driven hard in a hot trunk of a car.
I always like to let the amp run on a bench with a purely resistive dummy load under a 30% drive condition. This will heat up even the most expensive amp made by anybody. (class AB only)
I then recheck my bias setpoints for thermal related drift, and see how far the little monster has driven itself to the brink of no return.

Your dealing with a fairly common issue with most car amps. The other issue is the moron that goes inside his amp and turns the bias set up all the way thinking hes gonna get more power out of the amp as he sets the assend of his car on fire doing so:whazzat: :D :bawling:

Anyway give this method a gee-whiz and tell me what you think. Oh for Honesty purpose I borrowed this from Steve Mantz of ZED.
He should know being hes one of the biggest names in car audio, and a well respected one at that. Thank you Steve :)
 
If only one channel is not working properly, it's likely that there is a defect in that channel and not necessarily a design flaw. It could be a defective electronic component or it could be that the bias compensating transistor is not thermally mated to the sink.

There are many amplifiers that have the bias transistors too far from the output transistors. For these amps, if you set the biasing at low temps and drive the amp really hard, the outputs will heat up but the biasing transistor doesn't compensate because it's part of the heatsink is still cool. For these amps, the bias compensation works well to compensate for changes in ambient temps but if the bias is set just slightly too high at low temps, the amp will go into thermal avalance when driven really hard.

You can test to see how quickly the bias compensates by setting the bias where the outputs are at their threshold (at room temperature). Then drive the amp to full power into a dummy load equal to the lowest rated ohm load for the amp. After a few seconds at full power, remove the drive signal (set the signal source volume to zero) and watch the amplifier's current draw. Class AB amps with good bias compensation will almost immediately return to whatever the idle current was when you set the bias current. Amps with poor compensation (generally poorly located bias transistors) wil draw several more amps than normal when the signal is initially removed. The current will eventually return to the initial level but for some amplifiers, it takes a few seconds.

Zapco and a few alpine amps were the only amps that I've seen that really had the bias transistors mounted correctly. They were in direct contact with the output transistor.
 
Thanks a lot for all help!!!
I'll try to excercise several options of bias setup/stabilization tonight :smash:

Bias transistors are mounted poorly, no question - they are not even touching heatsink. And it looks like I am over-biasing a little. It's also very possible, that something on a repaired driver board is out of spec. I would compare voltages side-by-side with good channel as function of temperature.

Reminds me valves adjustment on SBC :D
 
Re-checked everything yesterday, set a conservative bias, looked at it closely for a half-hour at medium load. At first, while heating up, on this channel it was increasing quicker, then on others, and was harder to adjust (way too sensitive). But it did stabilize, and worked well with no bias float for 30 min.
Today, on my way to work (I am always breaking amps this way ;) ), after 40 minutes of beautiful performance, FET's are blown again :hot:
It worked for a week before under heavy load with almost zero bias in this channel (but sounded not right). Damn :cannotbe:
It is JL 500/5 amp, which suffered blown FET's before and blown driver transistor on driver board. Nothing more intelligent comes to my head, other then idea to rebuild this small driver board completely with all new transistors :confused:
 
Using a lighted magnifying glass, inspect all of the surface mount resistors on the driver board. If there are any that have the slightest surface defect, it should be double-checked.

On those driver boards, there are two diodes marked 6.8y (it could be different on your amp). Sometimes one shorts. I'm sure that you would have found this but I felt it was worth mentioning.

If you're able to get the driver boards out without doing damage to the main board, you may want to swap driver boards from one channel to another. If the problem follows the driver board, you know the problem is on the driver board. If the same channel has problems, you know to look on the main board.

If you are going to swap boards, while you have them out, you may want to compare resistance measurements on the transistors of the two boards. This may help you to find a leaking transistor.
 
Yeah I have to agree with Perry, you have a bad driver board. With the rising results you had during the test I would have failed that channel .

You definitly have a issue with the driver board. I have a few spare for reasonable price, if you want or need one PM me. This should solve your issue quckly.

Then you can look at and tinker on the bad one while your music plays on in your vehicle:)

Hope this helps :)
 
Thanks for reply, justonemoreamp. I am willing to dedicate a part of Columbus Day celebration to this driver board repair, and I'll be damned if it won't work after it :D
But, from any point of view, it would be very nice to have a spare board, or two. Please PM/email me detailes, I really appreciate it. :)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.