All,
I have a challenging and rather unique bias issue I cannot solve. The issue resides only in the right channel. And here is the problem.
When you first turn the amp on, the bias is stable. I keep my voltmeter connected to the right channel only and bias is rock steady 1.25 volts. If I take my finger and rotate the tube around the socket without disconnecting the tube the bias voltage begins to rise and will keep going past 2 volts. (I never let it go beyond 2 volts) If I take the voltmeter and check the negative grid voltage on the near side .1 uf capacitor it shows something like -20 volts then increases to -31 volts. Then when I recheck the bias voltage it is back to the original value of 1.25 volts and remains at a steady state. Also, if I turn the amplifier off and wait two minutes or so and turn it on the bias is perfect once again. The left channel does not exhibit this behavior. I built the new card from Dynakit and it has been a flawless board up until this problem arose. The board is well constructed and I would buy another. I have done the following for trouble shooting.
Replaced tube sockets.
Replaced near side .1 uf capacitor
Installed new tubes drivers and outputs tubes and even installed the original 7199's (same problem with the 7199's in the right channel)
Polished tube pins
Deoxit tube sockets
Re-tensioned sockets
Fixed and re-soldered many joints in right channel signal path and more
Checked continuity on every trace directly to the tube sockets
Check voltages and all are in specification.
Checked the bias supply and it is healthy
Checked for cracked traces, none found and checked continuity hot and cold. There are no issues.
New power transformer in 2017 during rebuild.
So, does anyone know why one channel would be affected? More like only one side of the right channel, but since the bias is tied together it is hard to say. It is almost as if there is some sort of gremlin that caused the capacitor to build up some type of opposing charge and the DVM discharges the negative side through the meter.
I am open to anyone's expertise so that i don't have to work about bumping the right driver tube and have a runaway bias situation. Right now I have my voltmeter almost always connected.
Thanks to anyone in DIY forum that has experience with this issue short of replacing the driver board.
I have a challenging and rather unique bias issue I cannot solve. The issue resides only in the right channel. And here is the problem.
When you first turn the amp on, the bias is stable. I keep my voltmeter connected to the right channel only and bias is rock steady 1.25 volts. If I take my finger and rotate the tube around the socket without disconnecting the tube the bias voltage begins to rise and will keep going past 2 volts. (I never let it go beyond 2 volts) If I take the voltmeter and check the negative grid voltage on the near side .1 uf capacitor it shows something like -20 volts then increases to -31 volts. Then when I recheck the bias voltage it is back to the original value of 1.25 volts and remains at a steady state. Also, if I turn the amplifier off and wait two minutes or so and turn it on the bias is perfect once again. The left channel does not exhibit this behavior. I built the new card from Dynakit and it has been a flawless board up until this problem arose. The board is well constructed and I would buy another. I have done the following for trouble shooting.
Replaced tube sockets.
Replaced near side .1 uf capacitor
Installed new tubes drivers and outputs tubes and even installed the original 7199's (same problem with the 7199's in the right channel)
Polished tube pins
Deoxit tube sockets
Re-tensioned sockets
Fixed and re-soldered many joints in right channel signal path and more
Checked continuity on every trace directly to the tube sockets
Check voltages and all are in specification.
Checked the bias supply and it is healthy
Checked for cracked traces, none found and checked continuity hot and cold. There are no issues.
New power transformer in 2017 during rebuild.
So, does anyone know why one channel would be affected? More like only one side of the right channel, but since the bias is tied together it is hard to say. It is almost as if there is some sort of gremlin that caused the capacitor to build up some type of opposing charge and the DVM discharges the negative side through the meter.
I am open to anyone's expertise so that i don't have to work about bumping the right driver tube and have a runaway bias situation. Right now I have my voltmeter almost always connected.
Thanks to anyone in DIY forum that has experience with this issue short of replacing the driver board.
Lift one end of (or remove) the coupling capacitors to the output tube grids.
Test again. If the problem is gone, the coupling capacitor(s) leak.
Avoid measuring the grid bias voltage directly at the output tube grid.
Instead, measure at the other end of the 270k grid resistor, where loading
effects are much less.
Test again. If the problem is gone, the coupling capacitor(s) leak.
Avoid measuring the grid bias voltage directly at the output tube grid.
Instead, measure at the other end of the 270k grid resistor, where loading
effects are much less.
Last edited:
Rayma, The caps are Cornell 940c6p1k-f. They came with the new board and have been running great since 2017. I have some new old stock .1 uf 400WVDC brown dips marked EM. I also have the same uf rating, same EM markings at 600WVDC.
Mark, I do not have any instrumentation to check for oscillations or even where to begin. I am going to check and replace the coupling caps.
Mark, I do not have any instrumentation to check for oscillations or even where to begin. I am going to check and replace the coupling caps.
The Stereo 70 cannot oscillate at RF, there is not enough bandwidth there.
It's just DC leakage, developing a positive DC voltage across the 270k grid resistor.
If the bias is off by +10VDC, the capacitor leakage would be 10V/270k = 37uA.
Also check for residue on the pcb, if the bias is still off with the coupling caps removed.
Last edited:
First remove both output tube coupling capacitors in the bad channel, and recheck the bias over an hour or so.
It should be stable and at an appropriate value for the tubes, somewhere between -30VDC and -45VDC.
If it is ok, go ahead and install 4 new capacitors. The 600V parts would be fine if they fit properly.
The 940c6p1k-f should be a good part, but perhaps one was damaged during handling or soldering.
Last edited:
You are not connecting a preamp to the ST70 Octal sockets on the front panel, Right?
Remove one of the wires from that socket (the B+ wire)
Disconnect the bad channel's EL34 cathodes from each other.
That allows you to use individual 3.1 Ohm current sense resistors, one for each cathode.
(1.56 x 2 = 3.1 Ohm). Connect the new 2nd resistor, 3.1 Ohm to the old B+ terminal on the octal socket on the front panel.
Now, you can monitor the individual currents of each EL34 in that channel.
That is good for troubleshooting to see if the current of each EL34 is behaving, or which one is not.
And, keep the amp wired that way, because it is good to allow you to see how well the two EL34 currents are matched (the output transformer will love you for that).
Just use common 3 Ohm resistors, or 3.3 Ohm resistors (you do not need to use a 3.1 Ohm resistor).
50mA and 3 Ohms = 1.5V
50mA and 3.3 Ohms = 1.65V
After that, you might see the value of doing that to the other channel too.
Remove one of the wires from that socket (the B+ wire)
Disconnect the bad channel's EL34 cathodes from each other.
That allows you to use individual 3.1 Ohm current sense resistors, one for each cathode.
(1.56 x 2 = 3.1 Ohm). Connect the new 2nd resistor, 3.1 Ohm to the old B+ terminal on the octal socket on the front panel.
Now, you can monitor the individual currents of each EL34 in that channel.
That is good for troubleshooting to see if the current of each EL34 is behaving, or which one is not.
And, keep the amp wired that way, because it is good to allow you to see how well the two EL34 currents are matched (the output transformer will love you for that).
Just use common 3 Ohm resistors, or 3.3 Ohm resistors (you do not need to use a 3.1 Ohm resistor).
50mA and 3 Ohms = 1.5V
50mA and 3.3 Ohms = 1.65V
After that, you might see the value of doing that to the other channel too.
Last edited:
Rayma, I ordered more of the same caps. I should be able to check this tomorrow. I have a newer heat adjustable soldering iron and it is set to 535 degrees and I use an old transistor heat sink clamp on the lead into the capacitors when I originally soldered them in place. I will connect my DVM to the center of the two 270k ohm resistors. I have cleaned and inspected the PCB for residue and use a 30 power loop to check all wire connections. Thanks to you and everyone for advice.
It's almost certain that the capacitor is leaking. Of course you can swap the tubes with those in the other channel,
and verify the tubes are not the problem. Maybe someone hit the capacitor with a soldering iron during assembly,
or used too much heat and damaged it.
The junction of the two 270k resistors should not show much (if any) change in DC voltage, at the same time
that the grid does. The resistance to ground at that point is much lower than at the grid.
Of course a good path must exist from the tube pin to the socket, 1k resistor, the wire, and the pcb.
and verify the tubes are not the problem. Maybe someone hit the capacitor with a soldering iron during assembly,
or used too much heat and damaged it.
The junction of the two 270k resistors should not show much (if any) change in DC voltage, at the same time
that the grid does. The resistance to ground at that point is much lower than at the grid.
Of course a good path must exist from the tube pin to the socket, 1k resistor, the wire, and the pcb.
Last edited:
I decided not to do the bias resistor change. I did receive four new .1uf coupling caps and replaced them all. I reflowed all the tube sockets and fixed about a dozen cold solder joints and one wire that was loose on the pcb. Hum and noise is much lower than previously. Bias is rock steady with minimal work minimal work. Thanks to everyone. Now if I could get a blown hafler se 120 operational? Finding Hitachi mosfets is next to impossible. I would send it out for repair in case you know of someone.