With all that you have to replace, is there a good reason for NOT replacing the original base resistors? You may as well do it right. 🙂
If you order parts from a distributor that stocks the 2sc945 (QA8), order a few of them. If you can't get the bias/idle current to hold, you may need to relocate QA8 to the heatsink near the output transistors. This should help stabilize it.
If you order parts from a distributor that stocks the 2sc945 (QA8), order a few of them. If you can't get the bias/idle current to hold, you may need to relocate QA8 to the heatsink near the output transistors. This should help stabilize it.
It's really my marginal soldering skills and choice of two 1/4 watt resistors in parallel (all I had at the time to make 2.2ohm) that's discouraging me from replacing all the base resistors (originals were 1/5 watt). Although time consuming, I'll replace these along with the emitters. Much easier than doing all this over again!
Does relocating QA8 have to do with heat dissipation? I'll let you how the idle current holds when I'm able to test more thoroughly tomorrow.
Other question is about the distortion-- Even with the amp crossover set to full-range or low-pass, there is a large amount of distortion when input voltage falls < 11v. I think this would be a problem as I've measured < 11v when both amps (I have another Audiobahn amp) are under heavy load when installed in the vehicle. Any way to get rid of the distortion?
Does relocating QA8 have to do with heat dissipation? I'll let you how the idle current holds when I'm able to test more thoroughly tomorrow.
Other question is about the distortion-- Even with the amp crossover set to full-range or low-pass, there is a large amount of distortion when input voltage falls < 11v. I think this would be a problem as I've measured < 11v when both amps (I have another Audiobahn amp) are under heavy load when installed in the vehicle. Any way to get rid of the distortion?
I wouldn't be concerned about the distortion at this point. It could be due to the 12v supply. I doubt that you'll have this problem when it's reinstalled in the vehicle.
If you order the correct size resistors, they will be much easier to replace. To remove the old ones, grab one leg of the resistor with a pair of hemostats or needle-nosed pliers and heat the solder for that leg while pulling on the resistor. It will pull right out (unless they're bent over on the bottom). It will take no more than 5 seconds to remove each one. If you use desoldering braid to clean up the pads and relatively small solder (0.03-0.04 diameter), soldering them back in will be very quick and easy.
If you have a poor quality iron, that will make things much more difficult. If you want a relatively inexpensive iron, a Weller WP35 is a very good choice. If you plan on doing this in the future, a Weller WES51 is a good choice. If you're patient, you can find them for ~$100 new.
In many higher quality amps, the bias transistor is on the sink VERY near the output transistors. As you've already seen, the bias current increases when the outputs heat up. The bias transistor (under normal operating conditions) will be at approximately the same temperature as the outputs (if the outputs don't heat up too quickly and the entire amp heats up at the same rate). When the bias transistor is cooler than the outputs, it doesn't compensate properly. That's why I told you not to use a fan. It would have cooled the bias transistor too much. Mounting it on the sink will provide better tracking for the bias at all temperatures (in most amps).
If you order the correct size resistors, they will be much easier to replace. To remove the old ones, grab one leg of the resistor with a pair of hemostats or needle-nosed pliers and heat the solder for that leg while pulling on the resistor. It will pull right out (unless they're bent over on the bottom). It will take no more than 5 seconds to remove each one. If you use desoldering braid to clean up the pads and relatively small solder (0.03-0.04 diameter), soldering them back in will be very quick and easy.
If you have a poor quality iron, that will make things much more difficult. If you want a relatively inexpensive iron, a Weller WP35 is a very good choice. If you plan on doing this in the future, a Weller WES51 is a good choice. If you're patient, you can find them for ~$100 new.
In many higher quality amps, the bias transistor is on the sink VERY near the output transistors. As you've already seen, the bias current increases when the outputs heat up. The bias transistor (under normal operating conditions) will be at approximately the same temperature as the outputs (if the outputs don't heat up too quickly and the entire amp heats up at the same rate). When the bias transistor is cooler than the outputs, it doesn't compensate properly. That's why I told you not to use a fan. It would have cooled the bias transistor too much. Mounting it on the sink will provide better tracking for the bias at all temperatures (in most amps).
Thanks for the tip on the soldering irons. It's a combo of not having the right tool and trying to use parts I've had on hand (or can bring home from work) vs buying new ones.
I may move the bias transistor over just for the sake of having a more reliable amp. Can I just cut the transistor legs, and jumper some wires over to the new location?
The amp in it's stock form has a fan very close to the bias transistor. Probably not the best design.
I may move the bias transistor over just for the sake of having a more reliable amp. Can I just cut the transistor legs, and jumper some wires over to the new location?
The amp in it's stock form has a fan very close to the bias transistor. Probably not the best design.
Yes, you can use jumpers for the bias transistor. I generally use ribbon cable from old 40 conductor IDE cables. Be absolutely sure that you don't allow the legs on the bias transistor to short together.
You're correct, not the best design.
If you need solder, Kester 44 (24-6040-0039 or 24-6040-0027).
You're correct, not the best design.
If you need solder, Kester 44 (24-6040-0039 or 24-6040-0027).
Trying to get this thing tested and I can't heat it up enough without connecting a much larger speaker to it. I Don't want to make the neighbors mad either.
Can I wire the two coils of a DVC driver out of phase with each other, or is this a bad idea?
Can I wire the two coils of a DVC driver out of phase with each other, or is this a bad idea?
Correct! Bad idea.
It will load the amp but the speaker won't be moving any air over the voice coil so it's very likely to burn the coil.
It will load the amp but the speaker won't be moving any air over the voice coil so it's very likely to burn the coil.
Any stores out there that sell capacitors locally?
Only ones I can find at radio shack are 50v max. I've got a few 100v caps here, but they're only 470uf.
Only ones I can find at radio shack are 50v max. I've got a few 100v caps here, but they're only 470uf.
Update-- Added a couple 100v 470uf caps in parallel with the single one just for safe measure. Closer I can get in capacitance to the other side the better I suppose until I can get a replacement.
Replaced all emitter resistors with 0.1ohm 5 watt variety. Replaced all base transistors with the same 2.2ohm parallel combo I'd be using. Also moved the bias transistor over. Used thermal epoxy to mount it on the heatsink near the outputs.
Installed amp in vehicle. Purrs along like new as far as I can tell. I have a 12" sub w/ 2 ohm load that will handle all the power-- Sound output and clarity were fine at max volume. Only played for a minute or two till amp began to get warm. At that point the voltage @ emitter resistors was still 0.000. I haven't gotten much farther with it, but it's safe to say it's definitely working-- I can't discern any difference in sound output vs when the amp had it's original components. I will keep testing the voltage @ emitter resistors as the amp heats up.
Only other thing worth mentioning is that I didn't have any MAXI type fuses (for the B+ cable nearest the vehicle battery), so used a 40A ATC (auto) fuse that I had. Even with the amp working at full load, it didn't blow the fuse. Reason I find this strange is that the amp itself is equipped with 80A worth of fuses. It doesn't seem right that it would draw less than half it's normal power.
Perry, you were correct-- There's no distortion at lower voltages. I had the other amp playing with this one, with B+ voltage dropping well under 11V (vehicle was off, just on battery power) and it sounded clean.
Replaced all emitter resistors with 0.1ohm 5 watt variety. Replaced all base transistors with the same 2.2ohm parallel combo I'd be using. Also moved the bias transistor over. Used thermal epoxy to mount it on the heatsink near the outputs.
Installed amp in vehicle. Purrs along like new as far as I can tell. I have a 12" sub w/ 2 ohm load that will handle all the power-- Sound output and clarity were fine at max volume. Only played for a minute or two till amp began to get warm. At that point the voltage @ emitter resistors was still 0.000. I haven't gotten much farther with it, but it's safe to say it's definitely working-- I can't discern any difference in sound output vs when the amp had it's original components. I will keep testing the voltage @ emitter resistors as the amp heats up.
Only other thing worth mentioning is that I didn't have any MAXI type fuses (for the B+ cable nearest the vehicle battery), so used a 40A ATC (auto) fuse that I had. Even with the amp working at full load, it didn't blow the fuse. Reason I find this strange is that the amp itself is equipped with 80A worth of fuses. It doesn't seem right that it would draw less than half it's normal power.
Perry, you were correct-- There's no distortion at lower voltages. I had the other amp playing with this one, with B+ voltage dropping well under 11V (vehicle was off, just on battery power) and it sounded clean.
Gave the amp a real hard workout just now. Gave me a real scare once, the amp just shut off while in the middle of running hard, and that darn red LED came on again. Luckily it was just the 40A fuse having blown-- Replaced it with the correct one and all is well.
Long story short-- Amp runs fine, heats up quick but I guess that's typical of these. I ran it hard while measuring voltage every 5 mins or so. It measured 0.000v @ emitter resistors. Once the amp was hot enough so that it was no longer comfortable to touch (I assume > 75% of it's thermal limit) I measured 0.001 volts across the emitter resistors. I believe I'm using a resistance value of ~2.1k on the bias circuit currently.
Other than that no issues encountered. Caps ran cool, I ditched the useless fan, may add more heatsink material to the bottom of the amp at a later time. I'm very very happy right now to have this thing operational.
Kudos to you Perry and the website for all the generous help and time spent. If there's anything further I can do to increase the reliability of this thing (I still need to restore the correct rail cap value), I'm all ears.
Long story short-- Amp runs fine, heats up quick but I guess that's typical of these. I ran it hard while measuring voltage every 5 mins or so. It measured 0.000v @ emitter resistors. Once the amp was hot enough so that it was no longer comfortable to touch (I assume > 75% of it's thermal limit) I measured 0.001 volts across the emitter resistors. I believe I'm using a resistance value of ~2.1k on the bias circuit currently.
Other than that no issues encountered. Caps ran cool, I ditched the useless fan, may add more heatsink material to the bottom of the amp at a later time. I'm very very happy right now to have this thing operational.
Kudos to you Perry and the website for all the generous help and time spent. If there's anything further I can do to increase the reliability of this thing (I still need to restore the correct rail cap value), I'm all ears.
The fan may be needed to keep air moving over the transformer so it doesn't overheat. I'd reinstall it. If you want to cool it better, get some air moving over the bottom of the heatsink.
If/when you reinstall the fan, you should check the biasing again.
If a 40 amp fuse was nearly enough to run the amp, I'd suggest going to a 50 or a 60 amp fuse. If you only need (and use) a 50 amp fuse, the amp's power supply has a much better chance of surviving if the outputs fail again.
If/when you reinstall the fan, you should check the biasing again.
If a 40 amp fuse was nearly enough to run the amp, I'd suggest going to a 50 or a 60 amp fuse. If you only need (and use) a 50 amp fuse, the amp's power supply has a much better chance of surviving if the outputs fail again.
Thanks for the advise. Will replace fuse with a 60A.
I had to ditch the fan because it had an obnoxious whine to it, the bearings or something were shot. I'll replace it with a CPU fan.
For cooling I was thinking of epoxying CPU heatsinks (about 25mm tall) along the two sides where the outputs set. In it's present configuration, it's just poorly designed flat aluminum. Very little surface area to dissipate heat, and in it's installed location sits flat on carpet. I'm thinking the heatsinks will increase surface area and raise the bottom of the amp up for air circulation.
Also another oddity I've found-- When the amp body touches the vehicle chassis (ground) I get a large arc. It's almost as if B+ is running through the amp chassis, but I don't see how this is possible. I'm still looking to see if I can find the cause.
I had to ditch the fan because it had an obnoxious whine to it, the bearings or something were shot. I'll replace it with a CPU fan.
For cooling I was thinking of epoxying CPU heatsinks (about 25mm tall) along the two sides where the outputs set. In it's present configuration, it's just poorly designed flat aluminum. Very little surface area to dissipate heat, and in it's installed location sits flat on carpet. I'm thinking the heatsinks will increase surface area and raise the bottom of the amp up for air circulation.
Also another oddity I've found-- When the amp body touches the vehicle chassis (ground) I get a large arc. It's almost as if B+ is running through the amp chassis, but I don't see how this is possible. I'm still looking to see if I can find the cause.
The voltage on the sink is generally due to a defective insulator. Use your meter to confirm that the tabs of all heatsink mounted components are isolated from the heatsink.
If it's still in the vehicle, you may be able to determine the most likely shorted component by measuring the voltage on the heatsink.
You can fix the fan. Pull the label and remove the retaining ring. Slide the impeller out of the bearing and clean the shaft an bearing with solvent to remove the old lubricant. Put a drop of motor oil on the shaft and/or bearing and reassemble it. It works virtually 100% of the time.
If it's still in the vehicle, you may be able to determine the most likely shorted component by measuring the voltage on the heatsink.
You can fix the fan. Pull the label and remove the retaining ring. Slide the impeller out of the bearing and clean the shaft an bearing with solvent to remove the old lubricant. Put a drop of motor oil on the shaft and/or bearing and reassemble it. It works virtually 100% of the time.
Well I made the mistake in thinking "off" is "off"... I believe in the course of testing resistance from amp chassis to center terminal of transistors, I touched two legs together.
Never seen a transistor on fire until now. It looks like two are shorted. Any other components likely to have failed?
Never seen a transistor on fire until now. It looks like two are shorted. Any other components likely to have failed?
When the amp is off, you still have 12v on the power supply FETs. If you shorted the gate to the drain (1-2), you turned one bank of FETs on and they tried to drag the 12v supplied by the battery to ground. The battery generally wins when that happens. 🙂
If the FETs shorted, it's possible that the driver transistors have failed. If you removed power quickly, they may have survived.
If the FETs shorted, it's possible that the driver transistors have failed. If you removed power quickly, they may have survived.
Wasn't much I could do about removing power quickly, the burning transistor resembled a 4th of July sparkler-- I wasn't going anywhere near it, I was more worried about the interior of my vehicle catching fire. Surely a fuse must have blown, correct? I took the fuses out but my garage is littered with fuses both good and bad so it's hard to tell which ones came out of the amp at the time.
I believe only two are failed. Is this possible. When you say bank of transistors, are we talking 3? 6?
I'm not familiar with which one is the driver transistor. Is there anyway to test this component? I found two groups of transistors (images below) on the power supply side.
The guy who sold me this amp on Ebay should have labeled it cursed.
In between power FETs
Near B+ fuse holder
I believe only two are failed. Is this possible. When you say bank of transistors, are we talking 3? 6?
I'm not familiar with which one is the driver transistor. Is there anyway to test this component? I found two groups of transistors (images below) on the power supply side.
The guy who sold me this amp on Ebay should have labeled it cursed.
In between power FETs
Near B+ fuse holder
There are 2 banks of FETs per transformer.
It's possible that only two failed but you should at least replace all in the bank with the ones that failed.
The driver transistors are going to have the emitter (leg 1 most likely) connected to the gate resistors. They will probably have a diode connected from the base to the emitter (1-3). The A1013 transistors are probably the drivers.
It's possible that only two failed but you should at least replace all in the bank with the ones that failed.
The driver transistors are going to have the emitter (leg 1 most likely) connected to the gate resistors. They will probably have a diode connected from the base to the emitter (1-3). The A1013 transistors are probably the drivers.
I could be crazy but is it possible that the banks are arranged diagonally? Meaning three on one side of the board, and three in the opposite corner making one bank?
Reason I ask is that there is little to no resistance measured from the emitter of one driver transistor to the gate resistors on diagonal sets of three. The other driver transistor shows this same trait. I hope what I'm saying makes sense, to simplify it 1 would denote a FET in the first bank, and 2 the second bank.
111 222
222 111
Reason I ask is that there is little to no resistance measured from the emitter of one driver transistor to the gate resistors on diagonal sets of three. The other driver transistor shows this same trait. I hope what I'm saying makes sense, to simplify it 1 would denote a FET in the first bank, and 2 the second bank.
111 222
222 111
111 222
trans A
trans B
444 333
FETs 1 and 2 drive transformer A. FETs 3 and 4 drive transformer B. Driver X probably drives FETs 1 and 3. Driver Y probably drives FETs 2 and 4.
trans A
trans B
444 333
FETs 1 and 2 drive transformer A. FETs 3 and 4 drive transformer B. Driver X probably drives FETs 1 and 3. Driver Y probably drives FETs 2 and 4.
- Status
- Not open for further replies.