Hi
I have an Audiolab 1997 8000S and 1986 8000A, both of which are working fine.
However, the 8000S resistors 721, 722, (724, 725, 821, 822, 824 and 825) are visibly burnt. They seem to be 4 sets of parallel configured resistors to give higher power ratings. They have been obviously getting hot for a while as the circuit board is discoloured. My schematics are dated 2005 so are for the Chinese version amp. The diagram says these resistors should be 3K0. which in parallel, will be 1K5. Actually, with my amp built in 1997, they are 2K0, so have a total R of 1K0.
The 8000A circuit is similar but not identical. It has 3 x 4K7 resistors in parallel
(R447, 449 and 451 according to my circuit) to give 1k56, which is similar to the Chinese version of 8000S.
Using a DVM, there is upto 32V dropped across the resistors in the 8000S, giving a power requirement of about 1Watt which might explain why they are burning.
Although I have made measurements etc, Im no Designer.
So what do these resistors actually do?
Will changing them to 3K0 on the 8000S be a good idea, so as to reduce the current and therfore lower the wattage needed?
What effect will this have on the sound of the amp, if any?
I have an Audiolab 1997 8000S and 1986 8000A, both of which are working fine.
However, the 8000S resistors 721, 722, (724, 725, 821, 822, 824 and 825) are visibly burnt. They seem to be 4 sets of parallel configured resistors to give higher power ratings. They have been obviously getting hot for a while as the circuit board is discoloured. My schematics are dated 2005 so are for the Chinese version amp. The diagram says these resistors should be 3K0. which in parallel, will be 1K5. Actually, with my amp built in 1997, they are 2K0, so have a total R of 1K0.
The 8000A circuit is similar but not identical. It has 3 x 4K7 resistors in parallel
(R447, 449 and 451 according to my circuit) to give 1k56, which is similar to the Chinese version of 8000S.
Using a DVM, there is upto 32V dropped across the resistors in the 8000S, giving a power requirement of about 1Watt which might explain why they are burning.
Although I have made measurements etc, Im no Designer.
So what do these resistors actually do?
Will changing them to 3K0 on the 8000S be a good idea, so as to reduce the current and therfore lower the wattage needed?
What effect will this have on the sound of the amp, if any?
With Audiolabs recommendation, I changed the resistors to 3K, 0.6 or 0.75W, 1% metal film. According to Audiolab, the resistors in question have been modded to 3K resistors in newer amps because of this prob. Its a design fault. They are also 3K in my old 8000A. The 2 resistors are in parallel so old total = 1K. Chnaging them to 3K means total = 1.5K. This will reduce the I through them, resulting in them not getting so hot. The previous owner admitted to leaving the amp on all the time and had done for over 10 years!! Its still not within x5 of the power rating but its better than it was. Audiolab also said reset the bias, after changing them. I found that when I did this change, it def ran cooler and after rebiasing, it sounded less harsh.
I found it difficult to adjust the pots when resetting the bias. They just wouldnt stay fixed. I was told its best to be slightly high with the bias V, if it couldnt be exactly set, to stop a harsh sound. It should be 22.5mV.
I have heard people say their Audiolabs sound a bit harsh sometimes....perhaps a slightly higher bias may help??
I found it difficult to adjust the pots when resetting the bias. They just wouldnt stay fixed. I was told its best to be slightly high with the bias V, if it couldnt be exactly set, to stop a harsh sound. It should be 22.5mV.
I have heard people say their Audiolabs sound a bit harsh sometimes....perhaps a slightly higher bias may help??
The trick to adjusting the bias in the 8000 amps is to solder the DMM test wires to the test point on the board and then rest the amp chassis on the base plate with the test wires running out between a gap in the plate and the chassis, and to replace the top cover after every adjustment and wait for the amp to fully stabilise before checking the bias. You don’t need to screw the base plate and covers on – you just need to make sure the electronics are as close to their normally ‘enclosed’ condition as possible, so the amp under test can build to its normal running temperature.
As I understand it, not being an EE, the amps are not temperature compensated in anyway (at least the old UK models aren’t), so apparently the bias drops as the amp warms up. You can watch this happening on a meter when you take the top cover on and off – it’s quite noticeable.
So unless you follow the above test arrangements, and you’ve adjusted the bias with the covers removed – you can guarantee the bias will be well below the required tolerance when the covers are replaced and the amp is up to normal operating temp.
I’ve got two very early 90’s black fronted MKIII 8000Ps - this pair are so early in the production run they still have the older models grey top covers. Actually, it’s a shame they didn’t keep the grey covers – they look really nice with the black front.😀
Anyway, the inherent design of the P is much the same as the power section in the 8000A – essentially, just more caps and power transistors. I replaced the original cermet bias pots with some encapsulated 25-turn variety a year back, and then spent quite a few hours tweaking the bias on both amps.
I came up with a somewhat overkill procedure. Where, after setting up an amp for testing, I’d put the top cover on and let the amp warm up for half an hour, check the bias on each channel, remove the cover, give them a small tweak, switch the amp off and let it cool down for another half hour, and then start the procedure again. The result being, I managed to get both channels in both amps stable at +/-1mV of 44mV (two paralleled sets of power transistors in the P) – and they remain stable within these values when the amps are repeatedly allowed to cool off and warm up again.
Whether all this mucking about was actually necessary, and admittedly it was a lot easier with two matched DMMs to hand – it seems a bit of patients is what is required when dealing with the ‘Slabs’. They still have that characteristic glary tinge on occasions, but are never overly harsh.
As to the resistors. The Ps have a single 2W 2K2 Noble made metal oxide in place of the three parallel resistors in the early A, and these run incredibly hot. So hot, they’ve discoloured the area around the board where their leads pass though it. But they’ve been in there 22 years now so there’s no doubting their reliability, but given how hot they get they must be getting pretty noisy by now. I’m in the process of overhauling both amps at the moment, so I’ve got a set of 5W Mills replacing the Nobles in the near future – so hopefully they should cool things down a bit.
Oh fun, a few more hours bias setting again. 🙄
To reduce interaction by the input stage even by full power, the dropped voltages by the resistors are very high.
Best solution is an extra independend input stage power supply for +/-12V
More easy is the use of TO-220 resistors for Chassis mounting to avoid burned PCB aeras.
Tim
I cannot read the resistor labels in the 8000S schematic.
The 8000A sch is quite different.
Those three resistors are feeding the Zener regulated 12V line and the same in the upper rail.
P= V^/R = [42-12]^2/4700 = 0.191W. If they used 500mW, or 600mW, resistors then they should run warm, not hot.
I cannot read the resistor labels in the 8000S schematic.
The 8000A sch is quite different.
Those three resistors are feeding the Zener regulated 12V line and the same in the upper rail.
P= V^/R = [42-12]^2/4700 = 0.191W. If they used 500mW, or 600mW, resistors then they should run warm, not hot.
The two circled parallel to the rail are 3K0, and the third, in series above the 3K0 on the left, is 47R.
I still can't read the sch, but I can guide your calculations
Power in one dropper resistor = {Vrail - [Zener voltage + Vbe] } ^2 / 3k
If Vrail = 51V and Zener = 12V then Pdiss = 492mW
If mains voltage is high, often 245Vac at my various UK homes, then the Pdiss will be even higher.
Power in one dropper resistor = {Vrail - [Zener voltage + Vbe] } ^2 / 3k
If Vrail = 51V and Zener = 12V then Pdiss = 492mW
If mains voltage is high, often 245Vac at my various UK homes, then the Pdiss will be even higher.
Old thread I know ... but someone may find this useful in the future:
I had the same problem with setting the bias on a recent repair and the problem turned out to be instability (high frequency oscillation) which was evident on the 'scope. The reason was that all the bypass electrolytics in the power amp section had gone low capacity/high ESR (dried up, probably due to the heat). Replacing all of them (C704, C713, C714, C715, C716 and C804, C813, C814, C815, C816) settled the oscillations and allowed for the bias to be set stably. I think the overheating resistors (I used 2W MF ones) and the caps should be replaced as routine. Also check for dry joints as there were very many of them. The other problem, and what it initially came to the workshop for, was the plastic disintegrating on the input sockets which were all replaced as well.
