Suitable bias for 4 ohm subwoofer over 200W?

[EWorkshop1708]

I'm finishing up my discrete mono amp I built for driving my 4 ohm 12" sub in my HT system.

Using a small 200W 36-0-36 trans to run the amp until my 350VA 55-0-55 toroid arrives. (if I find 350 is not enough, I'll get another 350VA toroid and parallel )

anyway...

I built the amp with 5 pairs of output transistors, and a strong VAS, and I'm wanting good control of the subwoofer, while keeping good SQ, and I'm wondering what's the optimal bias to use.

I've set it to 200mA so far, and it sounds clear, but because of fans, it does not get warm at all. Does this mean I should crank up the bias more, or should I leave it low to not waste power? Does it make much difference with a subwoofer to have high bias?




*also side question, because of my intake air fan location, it cools my NPN heatsink more than my PNP heatsink slightly. Is that a problem for them to be uneven in temp? I will also be adding 80mm fans to the heatsinks directly when they come with the transformer I ordered.

[Nigel Goodwin]

The actual bias value won't make any appeciable audible difference, as long as it's high enough to overcome cross-over distortion. 200mA sounds pretty high bias though, personally I'd probably run it lower?.

From the sound of things, with a massive amplifier and fan cooling, you are planning running LOT'S of volume?. In which case running with zero bias would probably make no audible difference either - cross-over distortion is only a problem at very low volume levels.

[lineup]

Originally posted by Nigel Goodwin
The actual bias value won't make any appeciable audible difference, as long as it's high enough to overcome cross-over distortion.
200mA sounds pretty high bias though, personally I'd probably run it lower?

Somewhere like 30-100 mA per transistor is a value often seen.
50mA would be most usual, what I have seen

See also Douglas Self paper, with thoughts on bias:

Distorton In Power Amplifiers
It is not generally appreciated that moving into Class-AB, by increasing the quiescent current, does NOT simply trade efficiency for linearity.
If the output power is above the level at which Class-A operation can be sustained, THD increases as the bias advances into AB operation.
This is due to so-called "gm-doubling" (ie the voltage-gain increase caused by both devices conducting simultaneously in the centre of the output-voltage range, in the Class-A region) putting edges into the distortion residual that generate high-order harmonics much as under-biasing does.
This vital fact is little known, presumably because gm-doubling distortion is at a relatively low level and is obscured in most amplifiers by other distortions.

[sreten]

Originally posted by EWorkshop1708
Does this mean I should crank up the bias more, or should I leave it low to not waste power? Does it make much difference with a subwoofer to have high bias?

Originally posted by Nigel Goodwin
The actual bias value won't make any appeciable audible difference, as long as it's high enough to overcome cross-over distortion. 200mA sounds pretty high bias though, personally I'd probably run it lower?.

Hi,

crossover distortion doesn't really show up at bass frequencies so
there is little point in overbiasing the stage into Class AB. Class aB
is when bias is set for minimum distortion at higher powers but
there is no attempt to extend class A operation at low power.

The optimum bias current depends on the type of output stage and
the value of the emitter resistors, for five pairs its the five emitter
resistors in parallel.

The actual important parameter is Vbias, not the current, see :
(the voltage across the emitter resistors sets the current)

http://www.dself.dsl.pipex.com/ampins/dipa/dipa.htm#5

section 5.3 .

/sreten.

[EWorkshop1708]

Thanks for all the info folks.

So for low frequencies, crossover isn't very audible, cool.

I also tested the amp fullrange, and I can't hear any distortion, so I'll just leave the bias set about 200mA, seems to work good without making much heat.

[AndrewT]

Hi,
200VA and 2*35Vac for a 4 ohm speaker is seriously under powered. You are proposing an even worse situation when changing to 350VA and 2*55Vac.

Your philosophy of using a low power transformer will result in PSU voltage droop under heavy loading. Similar to many others, so you are not alone. It does help prevent output stage overloading as pointed out in another thread recently.
However I believe that this leads to loose uncontrolled bass that in general terms could sound a lot better by balancing the design more evenly.

You can retrieve some of the situation by increasing the quantity of the smoothing caps from the usual 2mF per Apk output to 3mF or 4mF/Apk. for 80V PSU = +-56000uF to +-75000uF.

I recommend a bigger transformer about 1000VA to 1500VA 2*55Vac.

To answer your bias question, running the amp at minimum bias for optimal sound quality into your bass speaker will keep your output stage cooler. This increases the ability of the output stage to survive misuse and may be a consideration if using only 5pairs.

[EWorkshop1708]

If the 350VA 55V CT trans isn't good for a 4 ohm load, then should I unwind a few turns to get 45 or 50V?

Or should I just stick with the 55V and get two transformers? (700VA)

I'm guessing I'll have to see how it performs when it arrives. It weighs more, and is larger than the AVEL transfomers of similar size, so I hope it should be good enough.

BTW, I'm using 5 x 4700uf 80V Caps per rail for the output (23,000 uf per rail). The PCB with VAS and LTP stages also has two 4700's for the rails that feed it.

I'm looking for approx 300W RMS @ 4 ohm. I already have the speaker, so I have to stick with 4 ohm.

As far as biasing, I'm keeping it low to keep it cool.

Thanks folks, I'll keep you posted.

[Eva]

I would try with a single 350VA transformer first. Then I would watch rail sagging and ripple with oscilloscope and decide whether to buy or not another 350VA unit.

Ripple and sagging with music signals is probably not going to be as severe as some people suggest, altough it depends a lot on actual loudspeaker impedance, being sealed boxes the easier to drive since they show a big impedance bump at LF.

[AndrewT]

Hi,
to get 300W into 4ohm you will need about +-60Vdc for a bjt amp and 5pairs (130W or above) should easily cope. 150W or 200W devices will eat it.

2 * 40Vac should give about +-58Vdc. This may turn out to produce a little less than 300W (-0.5db) but if you optimise all the component values you may just achieve it. About 400VA or 500VA should be about right.

49Vpk into 4r =12.2Apk and requires about +-24mF for smoothing.
So your 23mF is spot on.

If you want you can try to find a 41Vac or 42Vac but more likely adding a dozen turns of same diameter enamelled wire to achieve a slightly higher voltage will hit your target. Another thread suggested that adding turns gives a slight increase in VA.

If you remove turns from a toroid you will REDUCE the VA rating of the transformer. You can only draw the SAME current at a lower voltage. The core will now have some spare capacity to allow you to add other windings for ancilliares.

DO NOT use 80V caps on a 55Vac transformer. You may overload them when incoming voltage is above nominal.

[EWorkshop1708]

The caps would be at 77V, which is near the limits, but I don't think it's something to worry about.

I expect some rail sag, so say if the rails sag from 77V to 60V, I'd still have enough power. I'm not concerned with stressing the transformer, enough airflow in the case would keep it cool.

Could I use an MOV across the 120V for surge protection?

Also, I have an additional circuit that shuts the amp off for overheating, power on/off, fan speed, etc. What if I made a resistor divider for the amp rails, and used a spare section of the LM339 chip to sense if the amp voltage goes too high to make sure the supply V does not exceed the caps voltage?