When doing some searching for hum on a PC19 board, power arced to the frame (I think from the B- side of the board, to the heat sink, then to the chassis, but it happend too fast to know if it was + or - for sure). There was a noticeable spark, but no smoke or obvious burned wires or components. The rail fuses on the channel I was working on blew, but both channels were affected. It seems that current went through the frame to the other channel. I think I smelled burned plastic, but it was very, very faint if at all.
After the spark, I could not adjust the bias current at all on either channel. It is fixed at 20.3mA, with the amp switch on or off. Rail voltage is 69. This is a bit high, but I put in a higher capacity toroidal capacitor, and I think the voltage is what it was before the spark.
I replaced the PC19 boards with a spare set, but this made no difference. Voltage is zero at R21 and R30.
It seems like the problem would be behind the PC19 board, so
I checked the wires behind the PC19 board for continuity (okay), and the resistors (R38 to R41) check out also. Although they tested okay, I also replaced C24 and C30. This raised the bias current to about 54mA, but I still could not adjust it, and after turning the amp off and on, it is back to steady 20mA. Perhaps these caps drew a little current for a while.
The toroidal tranformer seems fine, but I put the original transformer and bridge rectifier back in anyway, but this did not help.
My meter does not have enough voltage (1.5) to test the MOSFET's, which I think need 5. I could build a tester (like Pass's on his web site. Frank Van Alstine told me that he's never heard of all MOSFETS breaking all at once, and they rarely fail anyway.
Even so, as a last resort, I bought matched NOS MOSFET's (2SK134, 2SJ49), and put them in. The steady 20mA bias current remains. Now I doubt that any of the FETS are broken.
The star ground is okay, and is continuous with the ground on the heat sink.
I am stumped. I may build a MOSFET tester anyway, to be sure, but I think I am missing something obvious. I looked at another thread that dealt with low bias current (20mA) on a Hafler 200, but that problem was solved by replacing FET's on the board. As mentioned above, I replaced the boards with ones I know worked before.
Could there be a problem with the MOSFET sockets? These look pretty simple and rugged to me, with not much to go wrong.
After the spark, I could not adjust the bias current at all on either channel. It is fixed at 20.3mA, with the amp switch on or off. Rail voltage is 69. This is a bit high, but I put in a higher capacity toroidal capacitor, and I think the voltage is what it was before the spark.
I replaced the PC19 boards with a spare set, but this made no difference. Voltage is zero at R21 and R30.
It seems like the problem would be behind the PC19 board, so
I checked the wires behind the PC19 board for continuity (okay), and the resistors (R38 to R41) check out also. Although they tested okay, I also replaced C24 and C30. This raised the bias current to about 54mA, but I still could not adjust it, and after turning the amp off and on, it is back to steady 20mA. Perhaps these caps drew a little current for a while.
The toroidal tranformer seems fine, but I put the original transformer and bridge rectifier back in anyway, but this did not help.
My meter does not have enough voltage (1.5) to test the MOSFET's, which I think need 5. I could build a tester (like Pass's on his web site. Frank Van Alstine told me that he's never heard of all MOSFETS breaking all at once, and they rarely fail anyway.
Even so, as a last resort, I bought matched NOS MOSFET's (2SK134, 2SJ49), and put them in. The steady 20mA bias current remains. Now I doubt that any of the FETS are broken.
The star ground is okay, and is continuous with the ground on the heat sink.
I am stumped. I may build a MOSFET tester anyway, to be sure, but I think I am missing something obvious. I looked at another thread that dealt with low bias current (20mA) on a Hafler 200, but that problem was solved by replacing FET's on the board. As mentioned above, I replaced the boards with ones I know worked before.
Could there be a problem with the MOSFET sockets? These look pretty simple and rugged to me, with not much to go wrong.
Get a Variac.
Pull the pair of fuses from both cards. Dial up the voltage on your power supply and make sure it is working OK.
Then, when the PS caps have discharged through a 6000 ohm resistor, try the fuses in one card at a time, slowly dialing up the voltage to them. Check the bottom of page 12 of the manual for the voltage table and compare the readings you get with them. These voltages can vary somewhat but your readings should be close else something is awry.
Or (shameless plug ) check my message at the Trading Post for a pair of DH-200 cards to sell. They could get you sound cheaply.
Pull the pair of fuses from both cards. Dial up the voltage on your power supply and make sure it is working OK.
Then, when the PS caps have discharged through a 6000 ohm resistor, try the fuses in one card at a time, slowly dialing up the voltage to them. Check the bottom of page 12 of the manual for the voltage table and compare the readings you get with them. These voltages can vary somewhat but your readings should be close else something is awry.
Or (shameless plug ) check my message at the Trading Post for a pair of DH-200 cards to sell. They could get you sound cheaply.
Thanks again, Dick.
I don't have a variac. But with about 120 VAC line voltage, I could make some approximate checks using the voltage chart on page 12.
I am still not 100% sure the transformer is okay, and will re-hook up the old one again along with a spare rectifier to make sure. But if I have 69 volts at the rails, wouldn't this indicate an okay transformer and rectifier? There just does not qppear to be a load on the curcuit to draw much current.
I don't have a variac. But with about 120 VAC line voltage, I could make some approximate checks using the voltage chart on page 12.
I am still not 100% sure the transformer is okay, and will re-hook up the old one again along with a spare rectifier to make sure. But if I have 69 volts at the rails, wouldn't this indicate an okay transformer and rectifier? There just does not qppear to be a load on the curcuit to draw much current.
A main purpose of a variac is to help making repairs without letting the magic smoke out of your parts. I got a variac for $60 at an eBay auction and its use has saved me many times over the cost by avoiding the damage of parts.
Your voltage is high but that depends on what the incoming AC voltage is. Nominally you should get 65 VDC with 120 AC incoming. Perhaps your AC is running around 128.
Throwing parts at a non-working amp before its problem is defined is not wise.
Your voltage is high but that depends on what the incoming AC voltage is. Nominally you should get 65 VDC with 120 AC incoming. Perhaps your AC is running around 128.
Throwing parts at a non-working amp before its problem is defined is not wise.
Dick, your advice got me closer to solving the problem, thank you.
I put the original transformer back in to do the voltage tests (page 12 in the manual), and realized that I could not measure bias current above 20mA because one of the fuses in my meter had blown. It took while to discover this, as it functioned normally for everything but current from about 20mA to 500 mA
I can provide more details if anyone likes, but my problem is not low bias current, but current that's way too high. Voltage at the fuse is about 75 (maybe 78, I cannot remember exactly), but bias current was over 10A! Voltage at the speaker teminal is about 68 DC. I did not bother to adjust it
The output MOSFETS are okay, and with spare set of PC19s, I can set the bias and DC offset to specs, and the sound is fine in one channel. The other channel functions okay, but it still has the muffled sound (see muffled channel thread).
Both PC19's that were in during the spark are bad, so the current probably did get to the opposite side through the case. I think rail current went to the frame.
In pulled boards, I checked all diodes and transitors in-circuit, using the spare board as a reference. They all compared well, but this is probably not definitive. I may pull Q12 and 13, based on their location and what I think they do, but if anyone has an idea where to start, this would be helpful.
I was expecting a short somewhere, but have not found it yet. I checked C8, C12, and C13 for shorts. I would expect failed resistors to be open, but I checked R32 to R37.
The toroidal transformer that was in during the spark is okay. I put the original back in for the voltage check.
I am learning more about amps, but it is definately the hard way! I am reading more about them as well.
I put the original transformer back in to do the voltage tests (page 12 in the manual), and realized that I could not measure bias current above 20mA because one of the fuses in my meter had blown. It took while to discover this, as it functioned normally for everything but current from about 20mA to 500 mA
I can provide more details if anyone likes, but my problem is not low bias current, but current that's way too high. Voltage at the fuse is about 75 (maybe 78, I cannot remember exactly), but bias current was over 10A! Voltage at the speaker teminal is about 68 DC. I did not bother to adjust it
The output MOSFETS are okay, and with spare set of PC19s, I can set the bias and DC offset to specs, and the sound is fine in one channel. The other channel functions okay, but it still has the muffled sound (see muffled channel thread).
Both PC19's that were in during the spark are bad, so the current probably did get to the opposite side through the case. I think rail current went to the frame.
In pulled boards, I checked all diodes and transitors in-circuit, using the spare board as a reference. They all compared well, but this is probably not definitive. I may pull Q12 and 13, based on their location and what I think they do, but if anyone has an idea where to start, this would be helpful.
I was expecting a short somewhere, but have not found it yet. I checked C8, C12, and C13 for shorts. I would expect failed resistors to be open, but I checked R32 to R37.
The toroidal transformer that was in during the spark is okay. I put the original back in for the voltage check.
I am learning more about amps, but it is definately the hard way! I am reading more about them as well.