• TalkBass has been independent since 1998. Add your voice.
    Create a free account to reply to discussions, view embedded media, and browse with fewer display ads.
    Join free Log in
    Want zero display ads or expanded classifieds tools? Compare plans.

SVT3Pro Bias Setting

#121
Ummm, there are no paper resistors in your amp, you can't possibly measure bias current (or voltage due to current) between the bridged amp halves, it's just not even a possibility.

Can you at least identify the .47 ohm source resistors?

Side note, there are some unusual features about this amp that make it different than a traditional topology. Of course if you don't understand the traditional topologies, it won't matter but there is no differential amp (global feedback is handled differently) and DC offset is part of this network fed by an integrator that is part of an unusual current source which provides for the bias circuit. Current limiting is handled differently as well, in addition to gate voltage limiting there is an additional dual slope VI limiter.

I would expect there to be additional damage other than just the FETs from your accident...
 
#122
agedhorse said:
Ummm, there are no paper resistors in your amp, you can't possibly measure bias current (or voltage due to current) between the bridged amp halves, it's just not even a possibility.

Can you at least identify the .47 ohm source resistors?

Side note, there are some unusual features about this amp that make it different than a traditional topology. Of course if you don't understand the traditional topologies, it won't matter but there is no differential amp (global feedback is handled differently) and DC offset is part of this network fed by an integrator that is part of an unusual current source which provides for the bias circuit. Current limiting is handled differently as well, in addition to gate voltage limiting there is an additional dual slope VI limiter.

I would expect there to be additional damage other than just the FETs from your accident...
Click to expand...

I'll double check but I believe the .47 ohm resistors are the green ones in pairs on each side, so yes we're on the same page there. On the panel that experienced the burn, it was only the MOSFETs that went up. Is there another place on the main board that I should be concerned in checking? I just don't want to have to pull many more components if I can help it. Will a simple continuity test suffice to check the main board?
 
#124
The schematic is available in the schematics section of the Loud service center: Link Removed. Unfortunately, the bias procedure is not included through that portal. The SVT-3Pro power stage is similar, the 6Pro is a scaled up version with some twists. The 3Pro documents are on the same server so you can compare them.

Static electricity can zap the components. You need to be properly grounded to avoid imparting a shock. The electrolytic caps should be drained to prevent you getting a hazardous shock.

It usually isn't as simple as replacing the Mosfets. Some resistors can act as fuses and blow. Carefully inspect the printed circuit board for damage. There could be damaged traces. Any carbon traces should be cleaned (Q-tip with isopropyl alcohol) as they can be conductive. There are reference voltages on the schematic. You can take readings as per the schematic notes and compare them as a means of diagnosing the circuit. The 3Pro docs indicate how to set the pots and switches. A signal generator and an accurate voltmeter are required to fully check out the amp.
 
  • Like
Reactions: agedhorse
#126
BassmanPaul said:
Did you match the Gate - source voltage of the new mosfets to each other? This is normally done so the the work is evenly spread across the bank.
Click to expand...
Always a good idea, but not as essential (within reason) on this amp due to the high source degeneration resistors.

Again, there are design features on this amp that make it different than most others...
 
#127
BassmanPaul said:
Did you match the Gate - source voltage of the new mosfets to each other? This is normally done so the the work is evenly spread across the bank.
Click to expand...

++ on that...I got a SVT4PRO on the bench this summer with only one power amp working. 7 out of 10 MOSFETs were toast + 15 resistors in the power amp + other stuff in the bias servo. I ordered a set of 10 + 10 matched MOSFETS (PNP & NPN) and replaced everything in PowerAmp "B". Setting bias showed that they were indeed matched closely. For fun , I checked the bias in PowerAmp "A" (which was working but had signs of some earlier repair )...it was all over the place...at least one of the FET's were on the high side. The customer also reported that PowerAmp "B" now sounded better and that PowerAmp "A" has some distortion....so I will replace all FET's in PowerAmp "A" too even if it's working. So, ordering matched FET's is recommended, and change all while you're at it...
 
#129
To give an update on my SVT-6 Pro, I found some pretty thorough damage on the bottom side and first SS power section biased by AP1. I'm currently almost done with the 2nd SS power section and just about done checking EVERYTHING on the lower side. I'm going to need all FETS replaced and a good number of diodes and resistors. The price won't break my back, but the labor is extensive. I've learned so much about diagnoses, soldering, and how amplifiers work.
 
  • Like
Reactions: beans-on-toast
#131
beans-on-toast said:
Absolutely, it seems that with some amps all you have to do is touch the adjustment pot and the bias goes way off. The problem with this is when the amp is getting bounced around, the bias can change. With a multi turn pot (depending in the pot, it can take 10 or 20 full turns to go from one end of travel to the other) life is easier because the adjustment is fine. Makes it less touchy. They are more expensive so manufacturers shy away from them.
Click to expand...

When I went to bias my 3Pro this thread was 4 pages long. I had the classic constant static symptom. My .02 after adjusting the bias on both of the SVT-3Pros in my family. (1 is mine, the other my son's.) I set the first 3Pro to ~20mV average with a high of 37.5 and a low of 9.5, so not well matched MOSFETs. I played that amp for a couple of months. But the fan started to sound too loud and it was running a little hot. I rechecked bias and the high value was just over 50 mV. As soon as I readjusted it to ~19mV average the fan slowed down and the amp ran cooler. I played it daily for another week just to be sure it was OK, then gave it back to my son. Time will tell if we suffered any damage.

I bought the 2nd 3Pro for myself last month. It has a minor crackling issue, which from this thread I believe will turn out to be a tube, or cold solder, or something else. While re-seating the tubes and inspecting for shipping damage (found a bent transformer bolt and the shaft of the DI pot was broken off), I adjusted the bias trim on this one too. (I have a hard time leaving well enough alone.) Better matched MOSFETs this time and I set them to ~24 mV avg. This one also started running hot with the fan going full speed. It has not left the studio yet so it hasn't even been jostled and the bias still drifted into the 50's. I readjusted to ~19mV.

Take home for me: The bias can drift well out of range for no apparent reason.
 
  • Like
Reactions: beans-on-toast and JimmyM
#132
Bias can change with time. That's why it is good to check it once a year as part of a regular maintenance program if the amp is being used regularly. The bias is not regulated. This means that the bias will change with the power line level. Always measure the AC line voltage and write it down as a reference. You'll not that the reference values on the schematic are relative to a state line voltage, probably 120VAC (I am not at my computer to check).

If you set the bias at 120 VAC and go somewhere else where the line voltage is higher, the bias will change. Often the voltage is lower in dives with bad power. This can affect the performance of the amp. Too low and it can cause problems.

You don't have to be nuts about getting the bias perfect. It was meant to operate within a range. Set it close, bias it to sound and perform the best, then make sure it is within a safe operating range. This takes some experience and is a bit of a problem with amps that have a sensitive bias pot.
 
#133
Bias changes with temperature, this is normal. The thermal tracking device must be in contact with the middle of the devices mounting surface in order to track temperature closely. The bias must set at a specified temperature so that the Vgs generated will start in roughly the middle of the thermal curve and any deviations due to normal mis-tracking will remain within the predefined range (specified by the designer). The bias voltage circuit is a servo, it has gain and offset. The offset is responsible for the operating point and the gain is responsible for the slope of the Vgs compensation curve.

I suspect that many of the problems encountered may be due to not setting the bias at the specified heatsink temperature, and/or poor thermal contact of the thermal tracking device with the heatsink. The contact of the tracking device with the heatsink is the servo's feedback loop, and an open loop bias generator will never track properly.
 
#134
Thanks you two for the expertise. :)
beans, this is a relief to me:
beans-on-toast said:
. . . You don't have to be nuts about getting the bias perfect. It was meant to operate within a range. Set it close, bias it to sound and perform the best, then make sure it is within a safe operating range. This takes some experience and is a bit of a problem with amps that have a sensitive bias pot.
Click to expand...

and horse too:
agedhorse said:
. . . I suspect that many of the problems encountered may be due to not setting the bias at the specified heatsink temperature, and/or poor thermal contact of the thermal tracking device with the heatsink. The contact of the tracking device with the heatsink is the servo's feedback loop, and an open loop bias generator will never track properly.
Click to expand...

I might then ask you both, would you renew the thermal paste - if that's what Ampeg uses - between the heatsink and thermal sensor while doing a sort of in depth maintenance routine on these amps? And does that hold true for the transistors also? I guess I'm just wondering what the lifespan of thermal paste is, and whether it's the same stuff used on computer CPUs for example.

cheers,
theo

ps - I should be working out the bass lines from Deja Vu and mixing down the demo tunes for my new band instead of jawing with you guys . . .:)

pps - let it be known I'm just a geezer with a soldering iron, not an electronics guy per se.
 
#135
Generally if undisturbed, modern thermal compound is good for a lifetime. However, there are some old compounds that did not age well but they are pretty obvious.

Usually, things are best by leaving well enough alone.
 
  • Like
Reactions: Crazy John
#136
If you change a transistor, change the heatsink compound. Some of those not so good compounds can dry out with time. Otherwise, I would leave it.

You would do better by ensuring that there isn't a buildup of dust and gunk inside the amp. This can affect cooling and make the fan run faster.

If you have a thermometer, you can measure the effectiveness of the cooling system by taking readings on the transistor cases, heatsink, even the vents.
 
Last edited:
#138
In order to normalize heating conditions to prevent heat fluctuation, it's best to run the amp for a half an hour before adjusting bias. Noting the AC voltage was a great suggestion earlier as well. A followup question would be how should you adjust the bias differently if the incoming AC is high or low? If you have 115 coming inline do you set the bias higher and by how much? Does it really affect the adjustment at all after AC has run through the transformer and has been converted to DC?

Also, I know you're meant to run the amp without load or signal, but where should you set the preamp dials? Would all gain, volume, and eq dials be set at 12:00 with the compressor turned fully off? What about the tube control setting? Obviously the amount of work you're placing on the tubes is going to create a change in current somewhere, although that may very well be beyond the bias setting and have no effect on it.
 
#139
As long as the amp's output topology is not compound feedback, the thermal sensing should be tight to the output device heatsink. For CF topologies, the thermal sensing is usually done at the driver.

(Static) Bias is always set with no signal and no load. I recommend master volume down, but it usually doesn't matter.

Line voltage (within reason) also shouldn't matter either because of how bias is usually generated.

There are other ways to set bias, but they are usually reserved for the design phase where specific behaviors are being investigated.

Disclaimer... It is very possible to destroy your amp if you do not know and understand what you are doing, or have an accidental slip of a test probe. Yes, I've seen it happen more than you would expect.
 
#140
agedhorse said:
Disclaimer... It is very possible to destroy your amp if you do not know and understand what you are doing, or have an accidental slip of a test probe. Yes, I've seen it happen more than you would expect.
Click to expand...

He's right, I did it, although now that everything has been checked, I'm almost glad, because I'm seeing a lot that needed rework anyway. Bought this amp used and quite a few components were running too close to the outlying recommended conditions.

So EQ flat? Gain and Volume at 0. Compressor off. Tube adjustment? or is that irrelevant?

AgedHorse might have said this earlier as well, but there was a recommendation made somewhere to turn the bias back a bit prior to turning it forward. If I had done that, I would have had a better feel for how fast the dial spins. The least bit of force sent it to far right. Is there any risk to underbias?
 

Latest posts

Top Bottom
Back