I figure at some point soon when the amp is running you will want to minimize / eliminate the hum in the amp. First place to start is by making sure that the AC wiring is done using best practices to minimize pick-up. I wouldn't change it yet however. One change at a time and only if we can measure it.
Our next step will be to do a similar check on the phase spliter, and then tackle the bias values for the output tubes. Once we know that all of that is correct we will give it a whirl and see if we can bias the output tubes without it going nuts.
Our next step will be to do a similar check on the phase spliter, and then tackle the bias values for the output tubes. Once we know that all of that is correct we will give it a whirl and see if we can bias the output tubes without it going nuts.
47K. (yellow, purple, orange, gold)
C= 197VDC
PIN3=1VDC
PIN6=91VDC
Havn't done that yet, will get around to it thou
My multi tip probes doesn't fit my meter so I'll just have to be careful. There is however plenty of space now, so no worries there.
Looking forward to it. With a little luck this could become a textbook example of how to debug an amp
Okay that is somewhat reasonable with the curve of the EF86 in triode. You are passing 2.3mA bias current. (C- Va)/Ra (197 - 91) / 47,000 = 0.0023 Amps.
Since we know that we are passing 2.3mA using V=IR again across the cathode bias circuit we find that V cathode (pin3) should be 0.0023 * 490 = 1.1V. 490R is the value of R6+R7. That is close to your measured value of 1V.
Now to your phase splitter. We need to know the voltage at Pin 1, Pin 6, Pin 8 and Pin 3. Pin 8 and Pin 3 should be the same as they are tied together, but it is worth checking. Also verify that R10 is really 82k.
As far as grounding the audio input just use an aligator clip or similar for right now. It will be important in a little while when we bring up the output tubes if my hypothesis is right.
Since we know that we are passing 2.3mA using V=IR again across the cathode bias circuit we find that V cathode (pin3) should be 0.0023 * 490 = 1.1V. 490R is the value of R6+R7. That is close to your measured value of 1V.
Now to your phase splitter. We need to know the voltage at Pin 1, Pin 6, Pin 8 and Pin 3. Pin 8 and Pin 3 should be the same as they are tied together, but it is worth checking. Also verify that R10 is really 82k.
As far as grounding the audio input just use an aligator clip or similar for right now. It will be important in a little while when we bring up the output tubes if my hypothesis is right.
Last edited:
Ok. Pin 1 and 6 are your anodes of the twin triodes. For right now we want them to be reasonably close. Makes calculations easier. Later we will tweek the balance pot to minimize distortion and you will see that the voltages won't be matched as close.
Okay we have 94,7 volts across the cathode resistor. ( I will use the EU notation for decimal place holder). V=IR we see that the total current through BOTH triodes is 1,2mA. So about 0,6mA per side. The Vgrid is about 4V negative. Remember the anode of the EF86 is directly coupled to the grid1 of the ecc83, so 91 - 94,7 = -3,7volts. That seems a little off. Please measure on the ecc83 the voltage difference between pin 3 and pin 7 and also the voltage between pin 7 and ground. With the voltage across the tube of 212 volts and a current of approx 0,6mA I was expecting to see around 2ish volts.
Okay we have 94,7 volts across the cathode resistor. ( I will use the EU notation for decimal place holder). V=IR we see that the total current through BOTH triodes is 1,2mA. So about 0,6mA per side. The Vgrid is about 4V negative. Remember the anode of the EF86 is directly coupled to the grid1 of the ecc83, so 91 - 94,7 = -3,7volts. That seems a little off. Please measure on the ecc83 the voltage difference between pin 3 and pin 7 and also the voltage between pin 7 and ground. With the voltage across the tube of 212 volts and a current of approx 0,6mA I was expecting to see around 2ish volts.
2,4 volt is reasonable and can be accounted for measurement error or tube variability but is close enough for the curves.
Now onto the output tube bias. What is the actual resistance of the Rvg resistor combination. We also need to know what the most negative voltage you can obtain using P1 (bias pot) at pin 5 of the output tube sockets. (no output tubes yet).
Please measure at both sockets and report both voltages.
How many DVM's do you have available? It would be nice to have three for the next step but will try to accomodate fewer.
Now onto the output tube bias. What is the actual resistance of the Rvg resistor combination. We also need to know what the most negative voltage you can obtain using P1 (bias pot) at pin 5 of the output tube sockets. (no output tubes yet).
Please measure at both sockets and report both voltages.
How many DVM's do you have available? It would be nice to have three for the next step but will try to accomodate fewer.
Ok. The range looks good for a test.
Okay, we are going to make another check on wiring. Please double check that the plate or anode wire goes to pin 3 on the output tube socket, and the corresponding UL wire goes to pin 4. I don't know how you OPT is labled but it should be clear which color goes to each tube and position. We want to make sure that the tubes are not crossed or the UL/Plate isn't backwards.
After you recheck, and you are absolutley sure that it is correct, install the output tubes and hook up your dummy load or CHEAP speaker". This is a time where three DVM's is nice as well as the incadescent light bulb in series with the power cord. We will want to watch B+, CT1, and CT2 voltage. Focus on the CT1 if you only have one. While watching the CT1 voltage power on. Again, the feedback wire should be disconnected and isolated. As the filaments heat up listen for howls, hums, and weird noises. You should see <0,1V on the meter and it should be stable. If it is check CT2 and it should be similar. Keep an eye on the plates, as they should not be glowing, (The filaments should however). If the voltages still look good, check B+. Do not adjust the bias yet.
Be ready to pull the power at a moments notice if something isn't right.
Okay, we are going to make another check on wiring. Please double check that the plate or anode wire goes to pin 3 on the output tube socket, and the corresponding UL wire goes to pin 4. I don't know how you OPT is labled but it should be clear which color goes to each tube and position. We want to make sure that the tubes are not crossed or the UL/Plate isn't backwards.
After you recheck, and you are absolutley sure that it is correct, install the output tubes and hook up your dummy load or CHEAP speaker". This is a time where three DVM's is nice as well as the incadescent light bulb in series with the power cord. We will want to watch B+, CT1, and CT2 voltage. Focus on the CT1 if you only have one. While watching the CT1 voltage power on. Again, the feedback wire should be disconnected and isolated. As the filaments heat up listen for howls, hums, and weird noises. You should see <0,1V on the meter and it should be stable. If it is check CT2 and it should be similar. Keep an eye on the plates, as they should not be glowing, (The filaments should however). If the voltages still look good, check B+. Do not adjust the bias yet.
Be ready to pull the power at a moments notice if something isn't right.
Pin3, the plate, goes to the upper most solder tag on the primary side of the opt (on the same side as the 8ohm tag on the secondary side) and pin4 goes to the second from the top. There are five so that makes the second one the UL-tag. There are no wires, just tags and the only labels are the 0,4,8 on the secondary side. And vice versa on the other tube, anode on the lower most tag and the UL one step above.
Since I managed to forget about the lightbulb, I'll have to get back tomorrow. Good evening to you all.
Since I managed to forget about the lightbulb, I'll have to get back tomorrow. Good evening to you all.
That it good. We will bias it it steps. First take both tubes to 200mV. Watch, listen. Let it sit there about a minute or so and then take them to 300mV. Again, watch listen. Repeat the step ramping to 400mV, 500mV, and finally 600mV.
Once at 600 mV we will let it sit for a while. Measure your B+ voltages and check the bias every couple of minutes and make adjustments as required. Once you are sure it is stable and not running away, and most importantly not letting any magic smoke out, you can remove the lights and rest your bias again.
Did you short the input? If not that is a good sign. It isn't too bad if you had too either.
Let us know how it goes.
P.S. The voltages are higher because we have not loaded the circuit yet. They will come down some when the output tubes are biased.
Once at 600 mV we will let it sit for a while. Measure your B+ voltages and check the bias every couple of minutes and make adjustments as required. Once you are sure it is stable and not running away, and most importantly not letting any magic smoke out, you can remove the lights and rest your bias again.
Did you short the input? If not that is a good sign. It isn't too bad if you had too either.
Let us know how it goes.
P.S. The voltages are higher because we have not loaded the circuit yet. They will come down some when the output tubes are biased.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.