Different voltage at D at the channels is suspicious. The D voltage is the bias voltage at the gate of the P channel mosfet. -4.3V is about about right but -2.5V is off. So that channel needs investigating.
So connect a speaker to the -4.3V channel. Do you hear the tapping pulsing sound?
Connect the speaker to the -2.5V channel. Do you hear the tapping pulsing sound?
So connect a speaker to the -4.3V channel. Do you hear the tapping pulsing sound?
Connect the speaker to the -2.5V channel. Do you hear the tapping pulsing sound?
I'm a little confused, -DC is OK on D but +DC is not? Since P2 sets DC offset and P2 is between -V and D, wouldn't zeroing out the DC via P2 mean I am balancing the +/- DC on both the output and the input of the OS?
Yes, you are confused. 🙂
The voltage at D is the gate bias voltage of the P channel mosfet, which sets the current through the P channel mosfet. That voltage is always negative at the voltage that the amplifier operates at. As it drops lower and lower, at some point the current ceases to flow.
The bias pot adjusts the bias voltage of the N channel mosfets and the offset pot adjusts the bias voltage of the P channel mosfets. If the current flow through the P and N channel mosfets match, then the offset is zero.
The voltage at D is the gate bias voltage of the P channel mosfet, which sets the current through the P channel mosfet. That voltage is always negative at the voltage that the amplifier operates at. As it drops lower and lower, at some point the current ceases to flow.
The bias pot adjusts the bias voltage of the N channel mosfets and the offset pot adjusts the bias voltage of the P channel mosfets. If the current flow through the P and N channel mosfets match, then the offset is zero.
Trouble shooting suggestions:
1. Isolate and identify the channel or channels that sound bad. Is the channel with -2.5V at D the bad channel?
2. For the channel with -2.5V at D, measure the voltage between the anode and reference pins of the TL431. A proper functioning TL431 should measure 2.5V.
3. For the bad channel, measure the voltage drop across every mosfet source resistor. The mosfets should be matched so the voltages should not vary much.
edit:TL431 - measure between "anode" and reference pins.
1. Isolate and identify the channel or channels that sound bad. Is the channel with -2.5V at D the bad channel?
2. For the channel with -2.5V at D, measure the voltage between the anode and reference pins of the TL431. A proper functioning TL431 should measure 2.5V.
3. For the bad channel, measure the voltage drop across every mosfet source resistor. The mosfets should be matched so the voltages should not vary much.
edit:TL431 - measure between "anode" and reference pins.
@soundwavesteve Please verify that the tapping noise occurs with the front-end disconnected from the output circuit. (The output stage will be fine without it)
Last night updates:
On the -2.5v (left) channel, I am getting 1.5v DC on the output that I cannot zero out because the negative side of the left power supply is producing -37v while the positive side is 34v. When I pulled the transformer leads from the left supply, both secondaries were making about 29.5v due to my wall voltage. Seems there is an issue with that part of the power supply.
I disconnected the left channel completely and I also removed the left power supply from chassis ground, and the tapping noise remained on the working right channel.
The tapping noise does not occur when the FE is disconnected from the OS. The noise is still present in both channels when I attach the left channel FE to the right channel's power supply and OS. I did visually verify that the mosfets on the FE are installed in the correct spots. On a recent Mouser order, I did order some extras just in case I had to replace them. I already replaced the output caps, and the JFETs with a second set of matched units from the DIY store. I also checked each resistor is the correct value in the correct spot.
I did order a 30v lab supply to verify for sure the power supply is not the problem - should be here tomorrow. I did also borrow an oscilloscope but I have never used one before.
On the -2.5v (left) channel, I am getting 1.5v DC on the output that I cannot zero out because the negative side of the left power supply is producing -37v while the positive side is 34v. When I pulled the transformer leads from the left supply, both secondaries were making about 29.5v due to my wall voltage. Seems there is an issue with that part of the power supply.
I disconnected the left channel completely and I also removed the left power supply from chassis ground, and the tapping noise remained on the working right channel.
The tapping noise does not occur when the FE is disconnected from the OS. The noise is still present in both channels when I attach the left channel FE to the right channel's power supply and OS. I did visually verify that the mosfets on the FE are installed in the correct spots. On a recent Mouser order, I did order some extras just in case I had to replace them. I already replaced the output caps, and the JFETs with a second set of matched units from the DIY store. I also checked each resistor is the correct value in the correct spot.
I did order a 30v lab supply to verify for sure the power supply is not the problem - should be here tomorrow. I did also borrow an oscilloscope but I have never used one before.
On L channel Replace P2 with a 5K pot, if that doesn‘t give you enough throw to center the DC, you may need to drop the value of R25 a bit as well.
Or while keeping the 1K pot, if you have some resistors laying around (theres a better chance of that than a random 5K pot…) drop the value or R25 to about 12K and see if that moves the range of offset adjust in the proper direction. If it’s a lot worse, increase to 17K
No tapping with FE disconnected… this is good information.
Grab a small speaker and the amplifier and take it away from the area you are currently testing it. Heck, go to the room furthest in the house from where you are currently testing it and determine if the tapping remains with the FE attached. This may be environmental and that’s a straightforward and simple test.
Or while keeping the 1K pot, if you have some resistors laying around (theres a better chance of that than a random 5K pot…) drop the value or R25 to about 12K and see if that moves the range of offset adjust in the proper direction. If it’s a lot worse, increase to 17K
No tapping with FE disconnected… this is good information.
Grab a small speaker and the amplifier and take it away from the area you are currently testing it. Heck, go to the room furthest in the house from where you are currently testing it and determine if the tapping remains with the FE attached. This may be environmental and that’s a straightforward and simple test.
There may be nothing wrong with the power supply. The P channel mosfets are probably not conducting at -2.5V bias so they are not drawing current from the V- part of the power supply, resulting in no voltage drop in the power supply. The P channel mosfets not conducting is also the reason for the high offset voltage at the output.
The resistor adjustment as described by 6L6 will hopefully fix that.
The resistor adjustment as described by 6L6 will hopefully fix that.
Interesting observation, I can see how that would lead to the issue described. I will investigate all of these avenues. I appreciate the support! 🙂
You can easily measure the output mosfets' currents. The voltage across each mosfet's source resistor is proportional to the current through it.
Ohm's Law: V = IxR, so I = V/R
You can measure the voltage across the P channel and N channel mosfets' source resistors and compare.
Ohm's Law: V = IxR, so I = V/R
You can measure the voltage across the P channel and N channel mosfets' source resistors and compare.
I will test each resistor next chance I get and report back on that, as well as what's going on with R25.
Last night, I moved the amp to a circuit on the other leg of the panel and the tapping noise was still present. I also tried running the output wires outside of the case with no improvement. Then I tried hooking 'D' up directly to the speaker and that resulted in no tapping noise. I assumed it would have enough power to at least make some sound, I didn't try playing music though it at the time.
Last night, I moved the amp to a circuit on the other leg of the panel and the tapping noise was still present. I also tried running the output wires outside of the case with no improvement. Then I tried hooking 'D' up directly to the speaker and that resulted in no tapping noise. I assumed it would have enough power to at least make some sound, I didn't try playing music though it at the time.
Thank you for indulging the request. 🙂
What do you mean you hooked D up directly? No input stage attached and just the output stage hooked to speakers?
What do you mean you hooked D up directly? No input stage attached and just the output stage hooked to speakers?
No problem, I would try anything to experiment and learn. If it did make a difference I would looking into why next. 😉
I attached the front end D to the speaker, no output stage.
I attached the front end D to the speaker, no output stage.
Aha. Cool.
If you want to try something else, remove the FE from it’s mount (can stay on the riser panel) and get it physically as far away from the transformers as possible and see if the noise is still there.
But at this point it’s basically throwing spaghetti at the wall and seeing if anything sticks…
If you want to try something else, remove the FE from it’s mount (can stay on the riser panel) and get it physically as far away from the transformers as possible and see if the noise is still there.
But at this point it’s basically throwing spaghetti at the wall and seeing if anything sticks…
This weekend I hooked up the D output from the FE directly to the speaker and played music though it and it did play with no tapping noise. In fact it was completely silent with nothing playing, though of course pretty quiet without the output stage. Given this info, I think it's pretty certain the tapping noise is coming from the output stage. I wonder what could cause this 'sound' to occur there. Whatever it is, it happened to both channels...or I made the same mistake twice lol.
This weekend I took apart the amp again, took the output stage off of the heatsinks and verified every component was correct value and installed in the correct spot. I also reflowed every solder joint on both sides of the boards. I checked for continuity on every MOSFET at each pin. Everything looked good, I get voltage drop at every resistor yet one channel still has 2.5v difference between the positive and negative side rails. I switched heatsinks/OS between the left and right channels and the problem followed the output stage. On the other channel, with P2 set at 0, I have a positive .2v DC offset and when I increase P2, the DC offset increases, so it'll never be 0. This is from a matched set of output transistors bought from the DIY Store. Upon looking at the back of the MOSFETs where the VGS is listed, some of them are extremely close, like within .003 volts. All of them are better than .1v matched.
Before I bought the matched set, I had bought enough 240s and 9240s from Mouser to make an amp, so I could try using those or even starting with a fresh set of boards and all new parts. I see the store now uses 9141s so there is that benefit if I start fresh.
Before I bought the matched set, I had bought enough 240s and 9240s from Mouser to make an amp, so I could try using those or even starting with a fresh set of boards and all new parts. I see the store now uses 9141s so there is that benefit if I start fresh.
What were the voltage drops at each resistor? Knowing the values would be helpful in trouble shooting. Please measure and post.
Did you check the TL431 to see if it was working properly? Measure the voltage between anode and reference pins. A working TL431 should measure 2.5V.
What is the voltage at the mosfet gates?
The output stage has very few parts so there are not too many things that can go wrong.
Did you check the TL431 to see if it was working properly? Measure the voltage between anode and reference pins. A working TL431 should measure 2.5V.
What is the voltage at the mosfet gates?
The output stage has very few parts so there are not too many things that can go wrong.
What are the voltage drops across your source resistors?
Did you check or change your OS zeners, like proposed on the prev page?
Did you replace your TL431’s / verify function yet?
Sorry for messing up your mojo, Ben, but seems he isn’t able to bias up half his amp, and like you say: very few culprits.
And I still didn’t see no pics. No porn, no glory, and certainly no relief.
Not a problem, Andy. We all just want the amp working properly and have a happy amp builder . 🙂