I went ahead and replaced the two jumper wires with a short piece of wire.
![url], on Flickr\ }](/community/proxy.php?image=http%3A%2F%2F%7B%5Crtf1%5Cansi%5Cansicpg1252+%7B%5Cfonttbl%5Cf0%5Cfnil%5Cfcharset0+.HelveticaNeueInterface-Regular%3B%7D+%7B%5Ccolortbl%3B%5Cred255%5Cgreen255%5Cblue255%3B%5Cred0%5Cgreen0%5Cblue0%3B%5Cred255%5Cgreen255%5Cblue255%3B%7D+%5Cdeftab720+%5Cpard%5Cpardeftab720%5Csl480%5Cqc%5Cpartightenfactor0++%5Cf0%5Cfs40+%5Ccf2+%5Ccb3+%5Cexpnd0%5Cexpndtw0%5Ckerning0+%5Coutl0%5Cstrokewidth0+%5Cstrokec2+%5Burl%3Dhttps%3A%2F%2Fflic.kr%2Fp%2FDdm4Jf%5D%5BIMGDEAD%5Dhttps%3A%2F%2Ffarm2.staticflickr.com%2F1640%2F24424913150_5b0478bf64_b.jpg%5B%2FIMGDEAD%5D%5B%2Furl%5D%5Burl%3Dhttps%3A%2F%2Fflic.kr%2Fp%2FDdm4Jf%5Dimage%5B%2Furl%5D+by+%5Burl%3Dhttps%3A%2F%2Fwww.flickr.com%2Fphotos%2F97949719%40N03%2F%5DDennis+Kelley%5B%2Furl%5D%2C+on+Flickr%5C+%7D&hash=21f6dd9bbd4a9b4cfb65abcd179d2be6)
It seems to have lowered the hum level. It's not entirely gone, but it seems better. However now when I try to turn up the volume it gets really distorted sounding, so something isn't right. Tomorrow I will put the jumpers back and see if that is the cause of this new problem.
I just got disgusted and turned the lights out.
What I don't understand is why the filament voltage waveform has the flat top, and the rectifier turn off spike at the end? Unless that's what my AC looks like in my house? The filament should just be a pure AC waveform shouldn't it?
It seems to have lowered the hum level. It's not entirely gone, but it seems better. However now when I try to turn up the volume it gets really distorted sounding, so something isn't right. Tomorrow I will put the jumpers back and see if that is the cause of this new problem.
I just got disgusted and turned the lights out.
What I don't understand is why the filament voltage waveform has the flat top, and the rectifier turn off spike at the end? Unless that's what my AC looks like in my house? The filament should just be a pure AC waveform shouldn't it?
Last edited:
So go ahead and look at the Ac in your house. Also keep in mind all the windings in your PT are linked together, so load impulses on one winding can affect others.
Phew, idiot here!
OK, it's official. I am an idiot. I was checking the amp. It was acting really weird. it sounded good until I tried to turn it up, then it would get this rattling going on and heavy distortion? The clean channel wasn't working right at all, and was intermittent. Just really confused?
So I was probing around tracing the signal back through the preamp section, and I found a couple of resistors that some idiot put a shorting wire jumper between it and the resistor next to it. First thought was, "what idiot put a short on those two resistors?" Second thought was, "Doh!"
So I removed the , um temporary troubleshooting aid, and guess what the amp works just fine. The hum is still there, but greatly diminished. To the point that I don't think it will bother me. I made a recording of it before I started, so I will record it again to compare the two. It will be funny if it's the same. Kind of?
So those faster diodes should be here tomorrow, and what the heck it won't hurt anything to put them in. It may not make a difference, but then again it might. I need to do a little better job with my heater wire jumper. But I think I will go ahead and replace the reverb tank and glue down the caps, and call it a success.
Time to enjoy the amp, and quit working on it. Although that's fun too.
OK, it's official. I am an idiot. I was checking the amp. It was acting really weird. it sounded good until I tried to turn it up, then it would get this rattling going on and heavy distortion? The clean channel wasn't working right at all, and was intermittent. Just really confused?
So I was probing around tracing the signal back through the preamp section, and I found a couple of resistors that some idiot put a shorting wire jumper between it and the resistor next to it. First thought was, "what idiot put a short on those two resistors?" Second thought was, "Doh!"
So I removed the , um temporary troubleshooting aid, and guess what the amp works just fine. The hum is still there, but greatly diminished. To the point that I don't think it will bother me. I made a recording of it before I started, so I will record it again to compare the two. It will be funny if it's the same. Kind of?
So those faster diodes should be here tomorrow, and what the heck it won't hurt anything to put them in. It may not make a difference, but then again it might. I need to do a little better job with my heater wire jumper. But I think I will go ahead and replace the reverb tank and glue down the caps, and call it a success.
Time to enjoy the amp, and quit working on it. Although that's fun too.
Sound samples
I made a recording of the amp to compare with the before sound. The before was before I recapped it, but they didn't seem to make any difference, so the change was almost all due to the wire jumper for the filament that I moved.
Before
https://soundcloud.com/capndenny1/reverberockect-noise-before-recap-gain-at-300-mic-at-a-under-ampeg-11db
After
https://soundcloud.com/capndenny1/reverberockect-noise-after-recap-and-wire-jumper-w-gain-at-300-mic-at-a-under-ampeg-11db
Both signals were recorded with the mic in the same spot in front of the amp. All controls set to zero. Nothing plugged in. I truned the mic gain way up louder than I would normally use, an dafter I captured the sound I boosted the signal in reaper by +11dB. Repear read it as about a 10dB difference in level, but more importantly is that spikey noise is gone. The 60 Hz hum is a lot easier to listen to. I may go after that next. I might as well get every last bit of noise out so I can hear all of my mistakes in their naked glory!
And just to show the level of the hum relative the music level, this is with the amp at 9:00, or on 3 for the clean volume.
https://soundcloud.com/capndenny1/ampeg-with-fix
So it sounds pretty nice now. I put the new reverb in also. It a stiffer spring so not as long a delay, so it doesn't sound as lush? I think it's Ok though.
I made a recording of the amp to compare with the before sound. The before was before I recapped it, but they didn't seem to make any difference, so the change was almost all due to the wire jumper for the filament that I moved.
Before
https://soundcloud.com/capndenny1/reverberockect-noise-before-recap-gain-at-300-mic-at-a-under-ampeg-11db
After
https://soundcloud.com/capndenny1/reverberockect-noise-after-recap-and-wire-jumper-w-gain-at-300-mic-at-a-under-ampeg-11db
Both signals were recorded with the mic in the same spot in front of the amp. All controls set to zero. Nothing plugged in. I truned the mic gain way up louder than I would normally use, an dafter I captured the sound I boosted the signal in reaper by +11dB. Repear read it as about a 10dB difference in level, but more importantly is that spikey noise is gone. The 60 Hz hum is a lot easier to listen to. I may go after that next. I might as well get every last bit of noise out so I can hear all of my mistakes in their naked glory!
And just to show the level of the hum relative the music level, this is with the amp at 9:00, or on 3 for the clean volume.
https://soundcloud.com/capndenny1/ampeg-with-fix
So it sounds pretty nice now. I put the new reverb in also. It a stiffer spring so not as long a delay, so it doesn't sound as lush? I think it's Ok though.
Last edited:
Interesting article
I just found this today. i wish I had found this about a week ago?
http://dalmura.com.au/projects/Hum%20article.pdf
I just found this today. i wish I had found this about a week ago?
http://dalmura.com.au/projects/Hum%20article.pdf
A little more investigation
OK, I finally found my 500 ohm pot to try a humdinger on the filament circuit. I had to lift the center tap which was connected to circuit ground. It did lower the hum level, and was a little off center, so it is not quite where the center tap would put the ground. It was interesting if I went farther one way I got more 60 Hz hum, if I went farther the other way I got more buzz from the amp. The pot I had was like a 5W unit and was just too big to put in there, so I ordered a 1/2W pot.
I also got my 6V Little Tikes toy car battery all charged up and tried that. I used that to completely replace the filament supply with the battery voltage. To my surprise it didn't really lower the hum level at all? That was with the humdinger still in there, although that shouldn't matter with DC being supplied?
So I think the problem with the remaining hum is with the solid state rectifiers. That is the source of the high frequency buzz that I am still hearing.
I have 0.01uF caps at 1000V to put across each diode in the HV bridge. The same thing is done in a couple modern amps with solid state rectifiers. I also have a 220uF cap to put in place of the 150uF cap that is first, but that won't help with the high frequency.
I will also do a little investigation to see if the remaining hum is in the power amp output stage, or the signal inverter stage, or earlier. I determined the previous horrible hum was in the V2, but this may be somewhere else, and was just being hidden by the louder hum I fixed by moving the heater circuit wire.
I played my 1973 Fender Princeton Reverb for a while yesterday. That amp sounds so nice, and is almost totally silent. It has a lot of negative feedback too, and doesn't sound as live or dynamic as these modern amps, but man is it quiet! It is an old style turret board layout. Amazing what a good layout can do!
OK, I finally found my 500 ohm pot to try a humdinger on the filament circuit. I had to lift the center tap which was connected to circuit ground. It did lower the hum level, and was a little off center, so it is not quite where the center tap would put the ground. It was interesting if I went farther one way I got more 60 Hz hum, if I went farther the other way I got more buzz from the amp. The pot I had was like a 5W unit and was just too big to put in there, so I ordered a 1/2W pot.
I also got my 6V Little Tikes toy car battery all charged up and tried that. I used that to completely replace the filament supply with the battery voltage. To my surprise it didn't really lower the hum level at all? That was with the humdinger still in there, although that shouldn't matter with DC being supplied?
So I think the problem with the remaining hum is with the solid state rectifiers. That is the source of the high frequency buzz that I am still hearing.
I have 0.01uF caps at 1000V to put across each diode in the HV bridge. The same thing is done in a couple modern amps with solid state rectifiers. I also have a 220uF cap to put in place of the 150uF cap that is first, but that won't help with the high frequency.
I will also do a little investigation to see if the remaining hum is in the power amp output stage, or the signal inverter stage, or earlier. I determined the previous horrible hum was in the V2, but this may be somewhere else, and was just being hidden by the louder hum I fixed by moving the heater circuit wire.
I played my 1973 Fender Princeton Reverb for a while yesterday. That amp sounds so nice, and is almost totally silent. It has a lot of negative feedback too, and doesn't sound as live or dynamic as these modern amps, but man is it quiet! It is an old style turret board layout. Amazing what a good layout can do!
I've got the same model amp. One thing I noticed on the schematic copy that I have is that the signal voltage returned from the tone controls is very low, so I could see there would be a chance to pick up hum from the path on the upper circuit board.
Another thing, have you tried changing preamp tubes? Some are more susceptible to hum than others.
My amp has the infamous tube rattle problem (one of the output tubes is very close to the speaker.) It was an early version when they fitted EL34 tubes, I think the later ones came stock with 6L6GCs. On later versions they added a frame to hold the output tubes down tighter, does yours have that? I'd be interested in seeing a picture of that, if so.
Thanks,
jgf
Another thing, have you tried changing preamp tubes? Some are more susceptible to hum than others.
My amp has the infamous tube rattle problem (one of the output tubes is very close to the speaker.) It was an early version when they fitted EL34 tubes, I think the later ones came stock with 6L6GCs. On later versions they added a frame to hold the output tubes down tighter, does yours have that? I'd be interested in seeing a picture of that, if so.
Thanks,
jgf
Like Enzo wrote in post #19, the problem wil not be in the pcb board and neither in the schematic.
You tried with dc on the filaments; no change.
So we have to suspect the tubes.
But wich one?
As jgf wrote in post #27, swap a few tubes.
Maybe it can be a bad power tube, or even both.
Do you have the hum when the amp is out of the cabinet?
I once had a very stuborn problem in a Laney amp.
There the power tubes where to close to the speaker.
The magnetic field caused the power tubes to get red plated with plenty of hum.
Out of the box the amp played fine.
The solution: enlarge the cabinet with 2 cm and voila; no hum or redplating.
Try it out of the box and come back.
You tried with dc on the filaments; no change.
So we have to suspect the tubes.
But wich one?
As jgf wrote in post #27, swap a few tubes.
Maybe it can be a bad power tube, or even both.
Do you have the hum when the amp is out of the cabinet?
I once had a very stuborn problem in a Laney amp.
There the power tubes where to close to the speaker.
The magnetic field caused the power tubes to get red plated with plenty of hum.
Out of the box the amp played fine.
The solution: enlarge the cabinet with 2 cm and voila; no hum or redplating.
Try it out of the box and come back.
Picture of the frame?
Yes, I also have the EL34 output tubes. All tubes are new, and high quality. I have tried several tubes in different places. It is the amp. If you look up in the previous posts, I think there are two problems. The SS rectifiers are noisy as heck, and the layout is pretty bad in terms of putting the signal close to the noisy parts of the power supply.
You want to see the tube hold down, or hold up bracket for the power amp tubes? I can do that. It is just a metal strap with hooks in the ends that hook into a sheet metal part that is screwed down by the tube socket screw for the power amp tubes.
I had the head version before, the R50H and it had a metal cage around the power tubes with a foam pad on top. This is the same except, the back panel protects most of the tube, so the strap just needs to hold it in, not protect it.
I will try to get a picture tonight of the bracket.
So how would you compare the level of hum in your amp, with say other tube amps? I'm not sure if this is a design issue or if perhaps I am just fighting a bad part somewhere? You can hear my sound samples posted above. In particular the one with the guitar in it. That is at a reasonable room volume, similar to what an acoustic guitar would do with no amp.
Yes, I also have the EL34 output tubes. All tubes are new, and high quality. I have tried several tubes in different places. It is the amp. If you look up in the previous posts, I think there are two problems. The SS rectifiers are noisy as heck, and the layout is pretty bad in terms of putting the signal close to the noisy parts of the power supply.
You want to see the tube hold down, or hold up bracket for the power amp tubes? I can do that. It is just a metal strap with hooks in the ends that hook into a sheet metal part that is screwed down by the tube socket screw for the power amp tubes.
I had the head version before, the R50H and it had a metal cage around the power tubes with a foam pad on top. This is the same except, the back panel protects most of the tube, so the strap just needs to hold it in, not protect it.
I will try to get a picture tonight of the bracket.
So how would you compare the level of hum in your amp, with say other tube amps? I'm not sure if this is a design issue or if perhaps I am just fighting a bad part somewhere? You can hear my sound samples posted above. In particular the one with the guitar in it. That is at a reasonable room volume, similar to what an acoustic guitar would do with no amp.
Hey Tarzan,
Thanks for the reply. The worst problem was in the pcb layout, and was caused by the noise in the filament circuit induced by the diode rectifiers and coupled through the transformer. That was the worst offender, and was totally unacceptable. What I have now is a much lower level of hum, and I could probably just live with it. I am just trying a few things to see if I can get a little bit better.
I have tried all new tubes in all positions. And have also changed the speaker.
I have not tried it outside of the cabinet. Nobody is redplating. I have checked and adjusted the bias currents. They are fine.
It is the rectifier noise. The buzz is the problem. For two reasons.
First it is really irritating. When you hear it you have to get rid of it. 60 Hz hum is more forgivable, no as offensive, at least at low levels. This high frequency buzz is fingernails on chalkboard stuff!
Second, because the spikes from the diode noise contains high frequency components, it makes it easier for it to couple from one circuit to another. That is why the layout seems to be poorly done. For just 60 hz hum it is probably OK. But the high frequency buzz couple through capacitively more easily between circuits, or possibly even by RF radiation?
Thank you for the suggestions and tips. I will get this thing to where it is acceptable. It is almost there now. All I have really done is move a jumper wire a few inches. Perhaps I have another jumper wire to move, or something.
I am glad I don't need to do the DC filament supply.
Thanks for the reply. The worst problem was in the pcb layout, and was caused by the noise in the filament circuit induced by the diode rectifiers and coupled through the transformer. That was the worst offender, and was totally unacceptable. What I have now is a much lower level of hum, and I could probably just live with it. I am just trying a few things to see if I can get a little bit better.
I have tried all new tubes in all positions. And have also changed the speaker.
I have not tried it outside of the cabinet. Nobody is redplating. I have checked and adjusted the bias currents. They are fine.
It is the rectifier noise. The buzz is the problem. For two reasons.
First it is really irritating. When you hear it you have to get rid of it. 60 Hz hum is more forgivable, no as offensive, at least at low levels. This high frequency buzz is fingernails on chalkboard stuff!
Second, because the spikes from the diode noise contains high frequency components, it makes it easier for it to couple from one circuit to another. That is why the layout seems to be poorly done. For just 60 hz hum it is probably OK. But the high frequency buzz couple through capacitively more easily between circuits, or possibly even by RF radiation?
Thank you for the suggestions and tips. I will get this thing to where it is acceptable. It is almost there now. All I have really done is move a jumper wire a few inches. Perhaps I have another jumper wire to move, or something.
I am glad I don't need to do the DC filament supply.
I didn't notice a significant hum problem in mine, but maybe I didn't pay attention to it because of the other issues (the rattle and the noise/radio pickup). I think a significant layout issue is the lack of a grid stopper at pin 2 of v2, it's a long run from the previous stage through the relay and the volume control. I wouldn't rule out radiated rectifier noise getting picked up there.
Yes, please if you can post a picture of the output tube bracket.
Yes, please if you can post a picture of the output tube bracket.
Ok, curiosity got the better of me, and I pulled my amp out to fire it up. (haven't used it in a while) Some hum, but at about the same level as the background noise, I suppose it would depend on what you find objectionable. I do have a badly microphonic 12ax7 in there, I could record a cover of metal machine music, without even plugging in a guitar.
Hmmmm?
Kind of makes me wonder about my amp?
Here is the bracket. There are two small clips that get bolted down to the chassis. Then this bracket just clips in and forces the tubes up. There are two layers of spongy padding against the tubes. I suspect some kind of silicon rubber foam?
[/url]IMG_0814 by Dennis Kelley, on Flickr[/IMG]
Kind of makes me wonder about my amp?
Here is the bracket. There are two small clips that get bolted down to the chassis. Then this bracket just clips in and forces the tubes up. There are two layers of spongy padding against the tubes. I suspect some kind of silicon rubber foam?
I had that amp w/el34s. The electrolytic caps were shot after 5 years. They were a Korean brand I think...Samwa or something. I replaced them and all was good. I also put greenbacks in the cab. It was a nice amp but a little modern sounding and stiff compared to my vox and 5e3.
@CapnDenny: Thanks for the pic, I might try to rig up something similar. I think it's a good sounding amp (I like the 'clean' channel, anyway), I'm not trying to cause despair by my tales of woe! The line out / line in jacks are very close to the rectifier diodes, so that could be where noise is getting in. It looks like there's long enough leads on the diodes to tack small ceramic caps on top off them to reduce noise. I haven't had to work on the upper circuit board yet, that looks difficult to remove, I suppose all the nuts on the top panel will have to be removed. Have you or Hikari removed that board? Any tips?
Last edited:
Aha, sort of.
I have been struggling with the question of whether this is a bad part or a poor design. I can't find a bad part, which makes me think poor design. But I had a version of the same amp in a head version the R50H, and two of you were nice enough to come forward and give me feedback that your Ampeg Reverberocket with the EL34 is not so noisy, and neither was my R50H?
I was strolling through my iphone for pictures and I discovered I had some pictures of the old amp I sold, the R50H. So I zeroed in on the area of R13 and guess what I found? There was a weird circuit on V1 for the filament circuit where they ran both sides of the filament circuit together to the back of the tube for pin 9. This was in spite of the fact that no current was going through one of those traces. They clearly ran pins 4&5 from the filament buss in front of the tube. But they still ran both wires to the back? This made no sense to me.
On the R50H they also did that on the V2 tube. But on my R12R they only had one filament trace running back to pin 9. Here are the pics.
First the R12R that was loud as heck with the raspy hum.
[url=https://flic.kr/p/Dwjj8m]
[IMG]https://farm2.staticflickr.com/1598/24628268290_19dbde8847_b.jpg[/url]r12r_detail_r13 by Dennis Kelley, on Flickr[/IMG]
Next the R50H which I remember as being pretty quiet, not silent, but not offensive.
[/url]R50H_R13_detail by Dennis Kelley, on Flickr[/IMG]
Now for the theory part of our program.
The noise is being generated by the HV power supply. When the HV power supply rectifier diodes turn off they no longer draw current. This causes the voltage on the transformer side of the diodes to suddenly become disconnected from anything. This puts a glitch into the magnetic field of the transformer. i simulated this in pspice using coupled inductors to make a multiple winding transformer. The filament supply has the same characteristic discontinuity on the falling part of the AC waveform right where the diodes stop conducting. This is on the filament winding which isn't really connected to the HV winding directly. Just through the magnetic flux of the transformer.
This waveform is then getting coupled into the circuit of R13, the grid input of the second half of V2. I made a big improvement in this by using the flying jumper for the filament trace, but it is still not too quiet. By running the other half of the filament winding along, even though it carries no current, the idea is that it's voltage will also couple into the same places as the other filament trace. Since the filament winding is center tapped and that center tap is connect to circuit ground, the two signals will be more or less equal and opposite in polarity. So they should effectively cancel each other out at the grid input at R13.
So my flying lead helped, but I left the traces on the board and just added the new jumper. What I need to do is add a second wire to that jumper and twist them together, and just terminate the one that is needed at pin 9. The other wire is just a signal cancelling shield, sort of an active noise cancellation circuit.
When I turn the clean volume up all the way I hear hum coming from V1 as well, so even with the double jumpers like they did on V1 there is hum still being introduced. So the thing to do here is to remove the jumpers from both V1 and V2 completely. Cut the traces to pin 9 on both V1 and V2. Then I will add a twisted pair from the filament trace near where pin 4-5 is attached (red wire), and to pin 4-5 itself (black wire). This twisted pair will then travel over the top of the tube socket, and I will terminate the red wire at pin 9 of both the V1 and V2 sockets.
I have been struggling with how to get rid of this discontinuity in the HV waveform going through the transformer. I may still need to do that, since running the filament on a battery I still had hum? But at this point, I have a worse problem at R13 still.
PS Hey, JGF, which board are you asking about? I've had them both out completely. You have to take the nuts off the pots and the I/O jacks, but not the switches or pilot light. The hardest parts is keeping track of what wires go where. You need a schematic and a parts location diagram to get them back where they belong. i usually take a bunch of pictures too, and use that to help. But sometimes I still can't tell one black wire from another?
I think it's pretty interesting to find they changed the pcb in the same area where I thought it needed to change.
I have been struggling with the question of whether this is a bad part or a poor design. I can't find a bad part, which makes me think poor design. But I had a version of the same amp in a head version the R50H, and two of you were nice enough to come forward and give me feedback that your Ampeg Reverberocket with the EL34 is not so noisy, and neither was my R50H?
I was strolling through my iphone for pictures and I discovered I had some pictures of the old amp I sold, the R50H. So I zeroed in on the area of R13 and guess what I found? There was a weird circuit on V1 for the filament circuit where they ran both sides of the filament circuit together to the back of the tube for pin 9. This was in spite of the fact that no current was going through one of those traces. They clearly ran pins 4&5 from the filament buss in front of the tube. But they still ran both wires to the back? This made no sense to me.
On the R50H they also did that on the V2 tube. But on my R12R they only had one filament trace running back to pin 9. Here are the pics.
First the R12R that was loud as heck with the raspy hum.
[url=https://flic.kr/p/Dwjj8m]
[IMG]https://farm2.staticflickr.com/1598/24628268290_19dbde8847_b.jpg[/url]r12r_detail_r13 by Dennis Kelley, on Flickr[/IMG]
Next the R50H which I remember as being pretty quiet, not silent, but not offensive.
Now for the theory part of our program.
The noise is being generated by the HV power supply. When the HV power supply rectifier diodes turn off they no longer draw current. This causes the voltage on the transformer side of the diodes to suddenly become disconnected from anything. This puts a glitch into the magnetic field of the transformer. i simulated this in pspice using coupled inductors to make a multiple winding transformer. The filament supply has the same characteristic discontinuity on the falling part of the AC waveform right where the diodes stop conducting. This is on the filament winding which isn't really connected to the HV winding directly. Just through the magnetic flux of the transformer.
This waveform is then getting coupled into the circuit of R13, the grid input of the second half of V2. I made a big improvement in this by using the flying jumper for the filament trace, but it is still not too quiet. By running the other half of the filament winding along, even though it carries no current, the idea is that it's voltage will also couple into the same places as the other filament trace. Since the filament winding is center tapped and that center tap is connect to circuit ground, the two signals will be more or less equal and opposite in polarity. So they should effectively cancel each other out at the grid input at R13.
So my flying lead helped, but I left the traces on the board and just added the new jumper. What I need to do is add a second wire to that jumper and twist them together, and just terminate the one that is needed at pin 9. The other wire is just a signal cancelling shield, sort of an active noise cancellation circuit.
When I turn the clean volume up all the way I hear hum coming from V1 as well, so even with the double jumpers like they did on V1 there is hum still being introduced. So the thing to do here is to remove the jumpers from both V1 and V2 completely. Cut the traces to pin 9 on both V1 and V2. Then I will add a twisted pair from the filament trace near where pin 4-5 is attached (red wire), and to pin 4-5 itself (black wire). This twisted pair will then travel over the top of the tube socket, and I will terminate the red wire at pin 9 of both the V1 and V2 sockets.
I have been struggling with how to get rid of this discontinuity in the HV waveform going through the transformer. I may still need to do that, since running the filament on a battery I still had hum? But at this point, I have a worse problem at R13 still.
PS Hey, JGF, which board are you asking about? I've had them both out completely. You have to take the nuts off the pots and the I/O jacks, but not the switches or pilot light. The hardest parts is keeping track of what wires go where. You need a schematic and a parts location diagram to get them back where they belong. i usually take a bunch of pictures too, and use that to help. But sometimes I still can't tell one black wire from another?
I think it's pretty interesting to find they changed the pcb in the same area where I thought it needed to change.
Hi CapnDenny - I saw your post over on The Gear Page about this & got curious & came over here. Maybe you have already solved this w/your jumpers - but just in case:
Wouldn't you be a lot better off putting a signal tracer plus spectrogram generator on these nodes, not trying to look at waveforms at various frequencies with a scope? An FFT spectrogram will show you a consistent, easy to read "signature" across all audio frequencies, whereas with the scope you are still trying to assemble a jigsaw puzzle - quite difficult.
Assuming you have a suitable probe - anything with a low-enough value cap that you can get decent read on 60Hz without too much rolloff, e.g. 0.022uF is good enough - all you need, if you don't have a "real" spectrogram tool for your scope, is something like a smart tablet (iPad) or computer (Mac or Windows is fine). Quite a few spectrogram apps are available for very little $$ that will be good enough. You might want to put an attenuator in line after the probe to protect whatever sound card you use - rather than the computer sound card, which would be harder to replace, you can use a USB soundcard as an intermediary. Yes it's a little fuss to rig up but it would make things much easier than w/ the scope.
Also, if you are thinking that it's diode noise in a rectifier, trobbins knows that stuff cold so rather than have to reinvent 3 or 4 wheels you could just see which existing wheels he suggests you check out. E.g. you mention getting "faster" diodes earlier in the thread but no mention of what happened when you put those in? Plus, if you do put in fast diodes, e.g. UF4007, he would tell you to remove the snubber caps you've got in there. A strategy like this might work much better for diode noise than jumper wires, if you haven't tried it already. IF you are even dead sure it's diode noise - a PT and its secondaries can cause a buzz that might seem to be diode noise, but isn't. That's where a spectrogram would be so much better than a scope for separating out different noise profiles and tracking them to their source.
Anyway trobbins is a member here - I don't what the etiquette is in terms of getting his attention, but he was very helpful to me over on another forum. As you know he is the author of the hum article you mention in comment #25. He was also the one who first clued me in about the advantages of a signal tracer + spectrogram for chasing noise.
Wouldn't you be a lot better off putting a signal tracer plus spectrogram generator on these nodes, not trying to look at waveforms at various frequencies with a scope? An FFT spectrogram will show you a consistent, easy to read "signature" across all audio frequencies, whereas with the scope you are still trying to assemble a jigsaw puzzle - quite difficult.
Assuming you have a suitable probe - anything with a low-enough value cap that you can get decent read on 60Hz without too much rolloff, e.g. 0.022uF is good enough - all you need, if you don't have a "real" spectrogram tool for your scope, is something like a smart tablet (iPad) or computer (Mac or Windows is fine). Quite a few spectrogram apps are available for very little $$ that will be good enough. You might want to put an attenuator in line after the probe to protect whatever sound card you use - rather than the computer sound card, which would be harder to replace, you can use a USB soundcard as an intermediary. Yes it's a little fuss to rig up but it would make things much easier than w/ the scope.
Also, if you are thinking that it's diode noise in a rectifier, trobbins knows that stuff cold so rather than have to reinvent 3 or 4 wheels you could just see which existing wheels he suggests you check out. E.g. you mention getting "faster" diodes earlier in the thread but no mention of what happened when you put those in? Plus, if you do put in fast diodes, e.g. UF4007, he would tell you to remove the snubber caps you've got in there. A strategy like this might work much better for diode noise than jumper wires, if you haven't tried it already. IF you are even dead sure it's diode noise - a PT and its secondaries can cause a buzz that might seem to be diode noise, but isn't. That's where a spectrogram would be so much better than a scope for separating out different noise profiles and tracking them to their source.
Anyway trobbins is a member here - I don't what the etiquette is in terms of getting his attention, but he was very helpful to me over on another forum. As you know he is the author of the hum article you mention in comment #25. He was also the one who first clued me in about the advantages of a signal tracer + spectrogram for chasing noise.
Last edited:
Thanks for your input. I sent Mr. Trobbins a pm and asked for help.
I have been using Pspice to simulate the transformer and the flattopped waveform that I see in the heater voltage is not due to the charging current of the filter caps. If that were the case it would still be sinusoidal, just flatter. That thing is flat, like a clipped SS amp? I believe that is from the PT being designed to operate into saturation at nominal voltage levels. I have read where this acts liek a voltage regulator and keeps the peak voltages at a more or less constant level regardless of the actual line voltage variations. That way your incoming AC line can vary by +/- 20%, but the HV and heater supply will stay more or less constant. That keeps the power tube's idle current constant, and also prevents the heater from being too hot, or too cold.
I think the sharp edge is where the problem is, but it's not coming through the transformer. I believe it is being coupled from the HV into the heater circuit.
I replaced the 6.3 Vac with a Little Tike's 6V car battery, and I still had hum.
It was interesting when I had the 500 ohm pot across the 6.3 Vac heater circuit. I could null out more of the hum by using that and get it even lower. But i found a spot where if I turned it cw more I got more 60Hz hum, but if I tried to go more ccw I got less 60Hz hum but then I got more high frequency buzz. But it was buzz at 120Hz.
I have the caps, and wires for the new heater connections, and the humdinger. I will probably make those changes, and call it a day. By then it should be very quiet. If it still has a little hum, no big deal. I am pretty sure I found the design flaw in the amp, which is supported by the fact that Ampeg also made a change to later models.
I have been using Pspice to simulate the transformer and the flattopped waveform that I see in the heater voltage is not due to the charging current of the filter caps. If that were the case it would still be sinusoidal, just flatter. That thing is flat, like a clipped SS amp? I believe that is from the PT being designed to operate into saturation at nominal voltage levels. I have read where this acts liek a voltage regulator and keeps the peak voltages at a more or less constant level regardless of the actual line voltage variations. That way your incoming AC line can vary by +/- 20%, but the HV and heater supply will stay more or less constant. That keeps the power tube's idle current constant, and also prevents the heater from being too hot, or too cold.
I think the sharp edge is where the problem is, but it's not coming through the transformer. I believe it is being coupled from the HV into the heater circuit.
I replaced the 6.3 Vac with a Little Tike's 6V car battery, and I still had hum.
It was interesting when I had the 500 ohm pot across the 6.3 Vac heater circuit. I could null out more of the hum by using that and get it even lower. But i found a spot where if I turned it cw more I got more 60Hz hum, but if I tried to go more ccw I got less 60Hz hum but then I got more high frequency buzz. But it was buzz at 120Hz.
I have the caps, and wires for the new heater connections, and the humdinger. I will probably make those changes, and call it a day. By then it should be very quiet. If it still has a little hum, no big deal. I am pretty sure I found the design flaw in the amp, which is supported by the fact that Ampeg also made a change to later models.
I don't know where you're getting the idea that the PT would be designed to operate into saturation at all times, but it doesn't sound right. I am far from an expert but from what I have read, the notion of sag (for a PT) or saturation (for an OT) is only when a note is struck - and not all the time to somehow act like a rectifier.
Most amp designers and DIY hobbyists who've built a lot of amps would tell you that if you push a transformer too hard, you will destroy it. And the scenario you are describing would push the transformer too hard all the time - hardly what Ampeg would have designed for. E.g. here is a comment from a saved post by Randall Aiken, who knows amp design, over on blueguitar.org - he is talking about output transformers here, but just imagine that this is the power transformer in your Ampeg as you think Ampeg designed it - to be constantly saturating:
And here is a similar comment from Rod Elliott over on his Eilliott Sound web site:
Capn - just a final comment - I know it's tough to put in so much work & not get the result you want, but a systematic approach is still your best bet. Spice models are fun but not the same as what is actually happening in the amp. A spectrogram tool can be very useful. For example you could check pin 2 of V2, as rfg suggested, to see what the FFT noise spectrum looks like there; and you could compare it to a node on the output of the HT rectifier to see what that looks like. If they match, that could be a start. If not then you've eliminated that, at least. The scope just does not compare in this regard.
But I can also understand that you might just want to let it go awhile & play the amp for what it is. Good luck!
But I can also understand that you might just want to let it go awhile & play the amp for what it is. Good luck!