Hi Everybody,
just finished a new amp project.
This time an EL34 amp.
I did build quite a few 6L6 amps.
The new one has basically the same layout, just different transformers and power tubes.
The hum level though is a lot higher.
Basically I have a JMP and a Blackface (without Reverb) channel, the long tail resistor is 15k.
If I remove the input from the phase inverter, the amp is dead quiet.
If I attach each channel single, I get different sounding hum which get´s louder as I turn the volume up.
I checked with an cell phone app called Spectroid which frequencies occur,
and the biggest peak is at 150Hz (my power is at 50Hz here in Germany).
There is a smaller peak at 100Hz and several at 200, 250, 300 etc Hz (these ones are higher than the 100Hz one, but all are lower than 150Hz).
Nothing to see at 50Hz though....
I changed several tubes, especially the phase inverter (also from 12AX7 to 12AT7),
I checked all voltages, everything as it should be.
And the amp sounds great, just the hum is bad....
Any hints where to dig in further?
Thank you in advance,
Stephan
too much hum
Moderators: pompeiisneaks, Colossal
Re: too much hum
Ground reference issue?
Tube junkie that aspires to become a tri-state bidirectional buss driver.
-
dorrisant
- Posts: 2689
- Joined: Tue Sep 21, 2010 1:27 pm
- Location: Somewhere between a river and a cornfield
- Contact:
Re: too much hum
Did you use a filament CT to ground? If not, did you use virtual CT resistors?
Is your input jack referenced to ground?
Is your input jack referenced to ground?
"Education is what you're left with after you have forgotten what you have learned" - Enzo
Re: too much hum
Does the build implementation comply with the principles shown in figs 15.13 and 15.14 of http://www.valvewizard.co.uk/Grounding.pdf
My band:- http://www.youtube.com/user/RedwingBand
-
Stevem
- Posts: 4750
- Joined: Fri Jan 24, 2014 3:01 pm
- Location: 1/3rd the way out one of the arms of the Milkyway.
Re: too much hum
The first simple fast check I would make would be looking at what power supply ac ripple voltage is at the input to the plate load resistor on V1.
You do not want to see much more then about 250 mv.
You do not want to see much more then about 250 mv.
When I die, I want to go like my Grandfather did, peacefully in his sleep.
Not screaming like the passengers in his car!
Cutting out a man's tongue does not mean he’s a liar, but it does show that you fear the truth he might speak about you!
Not screaming like the passengers in his car!
Cutting out a man's tongue does not mean he’s a liar, but it does show that you fear the truth he might speak about you!
Re: too much hum
Good thinking, looking for which harmonics are in the overall hum.sbirkenstock wrote: ↑Sat Apr 01, 2023 5:34 pm and the biggest peak is at 150Hz (my power is at 50Hz here in Germany).
There is a smaller peak at 100Hz and several at 200, 250, 300 etc Hz (these ones are higher than the 100Hz one, but all are lower than 150Hz).
Nothing to see at 50Hz though....
It is very strange that you have a high peak at 150, and even stranger that you have 200, 250, 300 ... at higher levels than 100hz. The 100hz component should be very much more than the others, given full wave rectification. Generating 150Hz from 50 hz requires symmetrical clipping of a 50hz wave, or intermodulation of 100Hz and 50Hz (and a few other intermods...). Here are the things I can think of without a lot more study.
- Something clipping the incoming 50Hz mains to make 50Hz+150Hz. Do you have a MOV or other transient/surge protector ahead of the power transformer?
- Something funny about your power line; Maybe there's a DC offset on your AC power and this amp is more sensitive to it. Have you tried it in a different building?
- Really different rectifiers on each half of the PT secondary; Maybe one side is only partly working.
- PT running at near saturation on the primary, causing clipping of the incoming effective AC mains in the transformer; could be from using a 60Hz transformer at 50Hz
- PT radiating third harmonic of the AC mains and that being picked up by the preamps
That's a few places to look. Let us know what you find.
I don't "believe" in science. I trust science. Science works, whether I believe in it or not.
Re: too much hum
Are you using a toroidal power transformer in this build? They are very sensitive to any D.C.on the mains. The core saturates easily and could cause the problem.
Tube junkie that aspires to become a tri-state bidirectional buss driver.
-
sbirkenstock
- Posts: 78
- Joined: Fri Oct 17, 2014 10:50 pm
Re: too much hum
Hi Everybody!
First thank you so very much for your help!
I use this power transformer:
https://www.tubeampdoctor.com/en/mains- ... t-plexi-45
It´s a Marshall remake.
It does have a filament center tap and I use it. Actually I measured 6.305 Volt, so PERFECT
I actually did find one bad solder joint (the 1 Meg at V1), but fixing it did not change the hum.
I do use a PicoScope 4262 USB Oscilloscope, it has 16 bit and at about 1.000 US$/Euro it is not a toy, but also not perfect. If you ground one probe with it´s own ground, and "magnify" / zoom that in, it will show some ripple if there should be none.
So I am not sure how accurate it is to measure 250mV at around 400 volts.
I checked many places, I did not find any concerning ripple. Especially, this transformer does not have a bias tap, so you have to get it from the full voltage, like Marshall does.
Also I did not find any ripple there.
What I did find:
I do use a ground bus. That is in my case a 2 mm pure copper wire/rod (it is not coming from a spool, I buy it as a 10 feet "rod").
That runs around the amp in a C-shape, along the front, on the right side where the power transformer is it goes to the back and goes back left at the back of the amp (to ground my tube driven effect loop (which is not there jet in this humming amp)).
Every grounding element is soldered to this ground bus. Including all potentiometers, they are wired to the ground bus where grounding is required and I do not bend the solder log back and ground it to the casing of the pot.
And then this ground bus is soldered to the ground of the power jack, where the ground bus is closest to the power jack (so kind of in the middle of it).
I hope you can follow me!
Also somewhere I have to ground the chassis (I use 3mm Aluminium chassis, lasered to my plans). So I have this solder lug, that I fix with the screw holding the power jack.
Then I did wire it to the ground bus. We are talking about 2 inch here.
So there is a wire of 2 inch from the power plug ground to the ground bus and a 2 inch wire from the side of the power plug to ground the chassis, both wires meeting on the ground bus. Like a V shape.
I did remove the chassis ground wire from the ground bus and soldered it directly to the ground of the power jack (and shortening the wire to a bit over 1 inch).
Result: 3dB less hum!!!!!
Still not good enough, but what a difference for 1 inch of wire....
Ok, one more.
My perfect ground bus is of course not perfect... (yet)
I use Switchcraft jacks (11 and 12a)
On the inputs it is the 12a and it grounds itself if nothing is plugged in.
All the humming is happening with nothing plugged in, so the input is grounded.
These Switchcraft jacks of course are connected = grounded to the chassis.
But I also wired them to the ground bus.
So the ground bus and the chassis are also connected here (additionally at multiple points). Which is definitely not perfect.
I took off the wires from both inputs to the ground bus, so the inputs are only grounded direct to the chassis.
Result: brutal hum (very brutal like in ear damaging)
My Switchcraft speaker jacks I did not ground the the ground bus (so chassis only).
Well I hoped, if I additionally connect them to the ground bus, it would improve the situation like with the inputs.
Result: nothing
I got isolating washers for the switchcraft (they are for that purpose, but I am not convinced. The tread of the Switchcraft could still touch the chassis in the hole I think)
Will add some shrink tubing I guess.
That would result in a 100% ground bus only connected the the chassis in one point, not multiple.
Any further hints on isolating Switchcraft? Or any other high quality isolating jacks?
Any question, hint, feedback are very welcome!
Thank you and happy Easter!
Stephan
p.s.
I don´t use any surge protector.
And will of course intensivly read this:
http://www.valvewizard.co.uk/Grounding.pdf
First thank you so very much for your help!
I use this power transformer:
https://www.tubeampdoctor.com/en/mains- ... t-plexi-45
It´s a Marshall remake.
It does have a filament center tap and I use it. Actually I measured 6.305 Volt, so PERFECT
I actually did find one bad solder joint (the 1 Meg at V1), but fixing it did not change the hum.
I do use a PicoScope 4262 USB Oscilloscope, it has 16 bit and at about 1.000 US$/Euro it is not a toy, but also not perfect. If you ground one probe with it´s own ground, and "magnify" / zoom that in, it will show some ripple if there should be none.
So I am not sure how accurate it is to measure 250mV at around 400 volts.
I checked many places, I did not find any concerning ripple. Especially, this transformer does not have a bias tap, so you have to get it from the full voltage, like Marshall does.
Also I did not find any ripple there.
What I did find:
I do use a ground bus. That is in my case a 2 mm pure copper wire/rod (it is not coming from a spool, I buy it as a 10 feet "rod").
That runs around the amp in a C-shape, along the front, on the right side where the power transformer is it goes to the back and goes back left at the back of the amp (to ground my tube driven effect loop (which is not there jet in this humming amp)).
Every grounding element is soldered to this ground bus. Including all potentiometers, they are wired to the ground bus where grounding is required and I do not bend the solder log back and ground it to the casing of the pot.
And then this ground bus is soldered to the ground of the power jack, where the ground bus is closest to the power jack (so kind of in the middle of it).
I hope you can follow me!
Also somewhere I have to ground the chassis (I use 3mm Aluminium chassis, lasered to my plans). So I have this solder lug, that I fix with the screw holding the power jack.
Then I did wire it to the ground bus. We are talking about 2 inch here.
So there is a wire of 2 inch from the power plug ground to the ground bus and a 2 inch wire from the side of the power plug to ground the chassis, both wires meeting on the ground bus. Like a V shape.
I did remove the chassis ground wire from the ground bus and soldered it directly to the ground of the power jack (and shortening the wire to a bit over 1 inch).
Result: 3dB less hum!!!!!
Still not good enough, but what a difference for 1 inch of wire....
Ok, one more.
My perfect ground bus is of course not perfect... (yet)
I use Switchcraft jacks (11 and 12a)
On the inputs it is the 12a and it grounds itself if nothing is plugged in.
All the humming is happening with nothing plugged in, so the input is grounded.
These Switchcraft jacks of course are connected = grounded to the chassis.
But I also wired them to the ground bus.
So the ground bus and the chassis are also connected here (additionally at multiple points). Which is definitely not perfect.
I took off the wires from both inputs to the ground bus, so the inputs are only grounded direct to the chassis.
Result: brutal hum (very brutal like in ear damaging)
My Switchcraft speaker jacks I did not ground the the ground bus (so chassis only).
Well I hoped, if I additionally connect them to the ground bus, it would improve the situation like with the inputs.
Result: nothing
I got isolating washers for the switchcraft (they are for that purpose, but I am not convinced. The tread of the Switchcraft could still touch the chassis in the hole I think)
Will add some shrink tubing I guess.
That would result in a 100% ground bus only connected the the chassis in one point, not multiple.
Any further hints on isolating Switchcraft? Or any other high quality isolating jacks?
Any question, hint, feedback are very welcome!
Thank you and happy Easter!
Stephan
p.s.
I don´t use any surge protector.
And will of course intensivly read this:
http://www.valvewizard.co.uk/Grounding.pdf
Re: too much hum
Grounding is a little mysterious to most people because you have to think in terms of what currents flow and where they are flowing instead of what voltages happen.
Ground busses are based on the idea that if you make the ground bus thick enough, the currents have so little resistance that currents flowing through them will cause zero voltage offset in the bus wire resistance. Clearly, this is not true for anything except a superconductor.
The best ground bus practice is to (1) figure out exactly what currents have to flow in the "ground" to/from each amplifying stage and (2) connect these ground currents to different places on the ground bus so they either don't matter or even better cancel out. This is HARD to do, and time consuming. As you can imagine, getting a good result this way is very dependent on which part and tube stage is connected to exactly where on the ground bus. And the best place for low ground noise on the ground bus may not be the best place on the chassis for other reasons.
I personally classify "ground" into different categories.
> Safety ground; the AC mains safety ground must be connected to all of the metal that the user can touch. In many cases, this is required by law.
> Shield ground; the chassis itself shields the internal wires and parts from RF and magnetic fields. For lowest noise, this is connected to signal ground at exactly one point.
> Signal reference ground; this is the voltage point that tells an amplifying stage what exactly zero volts is. In a single ended triode stage, this is the voltage on the grid leak resistor end and the end of the cathode resistor.
> Power return ground, which I call "sewer ground"; this is the path that the used-up current from running the amplifying stage has to follow to get back to the power supply common point - that is, the negative terminal of the first filter capacitor.
> Signal return ground; this is the path that signal currents have to follow to return to the place where the signal current was generated. For instance, a triode stage may generate a plate voltage that causes signal current to flow to tone and volume controls. For lowest noise, this current must flow out of the tone and volume controls back to the ground point in the stage that generated it, that being the cathode ground point of the triode.
Ground busses make getting low ground noise harder by requiring you to mix up all the ground currents in one conductor. While is it possible to do this with enough work on exactly how to connect which part to the bus, it can be very difficult. You have uncovered one indication of this - that one inch of wire you changed made a big difference in hum. That can only happen if the currents flowing in the wire are unexpectedly big, or if the one inch of wire was picking up a large magnetic field and converting that into a current. It's possible that the inch of wire was part of an unexpected loop of the chassis plus the wire plus the ground bus forming a closed loop of conductor.
Your idea about the input jacks connecting to the chassis is a good line of thought. Ideally there would be exactly one connection from the signal ground voltage to the chassis. If you make this the input jacks, this also serves for RF shield connection of the input jack to the chassis/shield. But ideally there would be no other connections of signal ground to the chassis anywhere else. Insulating jacks remove the possibility of ground hum loops from the inputs being connected to the chassis too, but usually require a small-value ceramic capacitor from the signal ground contact on the insulated jack to the chassis metal right at the input jack itself. Switchcraft and other companies make 1/4" phone jacks which have insulating bushings but still fit in the standard sized hole so you don't have to drill out bigger holes for insulating washers.
Ground busses are based on the idea that if you make the ground bus thick enough, the currents have so little resistance that currents flowing through them will cause zero voltage offset in the bus wire resistance. Clearly, this is not true for anything except a superconductor.
The best ground bus practice is to (1) figure out exactly what currents have to flow in the "ground" to/from each amplifying stage and (2) connect these ground currents to different places on the ground bus so they either don't matter or even better cancel out. This is HARD to do, and time consuming. As you can imagine, getting a good result this way is very dependent on which part and tube stage is connected to exactly where on the ground bus. And the best place for low ground noise on the ground bus may not be the best place on the chassis for other reasons.
I personally classify "ground" into different categories.
> Safety ground; the AC mains safety ground must be connected to all of the metal that the user can touch. In many cases, this is required by law.
> Shield ground; the chassis itself shields the internal wires and parts from RF and magnetic fields. For lowest noise, this is connected to signal ground at exactly one point.
> Signal reference ground; this is the voltage point that tells an amplifying stage what exactly zero volts is. In a single ended triode stage, this is the voltage on the grid leak resistor end and the end of the cathode resistor.
> Power return ground, which I call "sewer ground"; this is the path that the used-up current from running the amplifying stage has to follow to get back to the power supply common point - that is, the negative terminal of the first filter capacitor.
> Signal return ground; this is the path that signal currents have to follow to return to the place where the signal current was generated. For instance, a triode stage may generate a plate voltage that causes signal current to flow to tone and volume controls. For lowest noise, this current must flow out of the tone and volume controls back to the ground point in the stage that generated it, that being the cathode ground point of the triode.
Ground busses make getting low ground noise harder by requiring you to mix up all the ground currents in one conductor. While is it possible to do this with enough work on exactly how to connect which part to the bus, it can be very difficult. You have uncovered one indication of this - that one inch of wire you changed made a big difference in hum. That can only happen if the currents flowing in the wire are unexpectedly big, or if the one inch of wire was picking up a large magnetic field and converting that into a current. It's possible that the inch of wire was part of an unexpected loop of the chassis plus the wire plus the ground bus forming a closed loop of conductor.
Your idea about the input jacks connecting to the chassis is a good line of thought. Ideally there would be exactly one connection from the signal ground voltage to the chassis. If you make this the input jacks, this also serves for RF shield connection of the input jack to the chassis/shield. But ideally there would be no other connections of signal ground to the chassis anywhere else. Insulating jacks remove the possibility of ground hum loops from the inputs being connected to the chassis too, but usually require a small-value ceramic capacitor from the signal ground contact on the insulated jack to the chassis metal right at the input jack itself. Switchcraft and other companies make 1/4" phone jacks which have insulating bushings but still fit in the standard sized hole so you don't have to drill out bigger holes for insulating washers.
I don't "believe" in science. I trust science. Science works, whether I believe in it or not.
Re: too much hum
A schematic might prove useful.
Tube junkie that aspires to become a tri-state bidirectional buss driver.
-
SoulFetish
- Posts: 191
- Joined: Wed Apr 03, 2013 1:50 pm
- Location: Norwood, MA
Re: too much hum
Without touching on all the potential causes of hum issues, maybe I can help trying to eliminate 2 of them. The first: Easy one, a bad tube can cause noise problems which can sound like a number of different problems. You can start at V1, pulling each tube and listening to whether the tube you pulled caused the noise to go away. If when you pull a certain tube kills the noise, try and replace the tube and see if that fixes it. Sometimes, it can be that simple.
The second pertains to something you mentioned seeing spikes at harmonic frequency of fundamental mains (assuming you live in a 50Hz region).
This makes me suspicious that you might be dealing with something I was dealing with on my most recent design and build...
Sometimes, diode switching spikes in the HT winding can cause distortion in the peaks of the heater winding AC waveform. This manifested itself in a kind of noise that had a very "buzzy" kind of quality In my amp (even with the heater wiring and lead dress being being solid). The power transformer I'm using has no center tap on the heater winding, so I used 2 high quality metal film 100 ohm resistors tied to ground as the zero voltage reference. However, I tried a number of different things to find the source of the problem, and ultimately ran out of ideas. Out of a bit of desperation, I removed the fixed resistors and installed a 200 ohm pot to try and see if I could null the noise. I connected the each outside legs of the pot across the heater winding, and the wiper to the top of the output tubes cathode resistor to elevate the heater supply to around 12.3V. Turned the amp on, and to my surprise, was able to make the adjustment to almost zero noise. Like with the volume turned all the way up, you can barely tell it's on.
I've since installed it permanently, however I use a potential divider off the HT line and elevate the reference voltage to around 25V, with a 10uF capacitor across the lower resistor to stabilize the voltage reference. Works like a charm.
It's interesting to me that the pot made such a profound difference than fixed ballast resistors.
The second pertains to something you mentioned seeing spikes at harmonic frequency of fundamental mains (assuming you live in a 50Hz region).
This makes me suspicious that you might be dealing with something I was dealing with on my most recent design and build...
Sometimes, diode switching spikes in the HT winding can cause distortion in the peaks of the heater winding AC waveform. This manifested itself in a kind of noise that had a very "buzzy" kind of quality In my amp (even with the heater wiring and lead dress being being solid). The power transformer I'm using has no center tap on the heater winding, so I used 2 high quality metal film 100 ohm resistors tied to ground as the zero voltage reference. However, I tried a number of different things to find the source of the problem, and ultimately ran out of ideas. Out of a bit of desperation, I removed the fixed resistors and installed a 200 ohm pot to try and see if I could null the noise. I connected the each outside legs of the pot across the heater winding, and the wiper to the top of the output tubes cathode resistor to elevate the heater supply to around 12.3V. Turned the amp on, and to my surprise, was able to make the adjustment to almost zero noise. Like with the volume turned all the way up, you can barely tell it's on.
I've since installed it permanently, however I use a potential divider off the HT line and elevate the reference voltage to around 25V, with a 10uF capacitor across the lower resistor to stabilize the voltage reference. Works like a charm.
It's interesting to me that the pot made such a profound difference than fixed ballast resistors.
-
WhopperPlate
- Posts: 1082
- Joined: Fri Apr 16, 2010 9:04 am
- Location: Santa Cruz, CA
Re: too much hum
Did you get this buzz with master on zero?SoulFetish wrote: ↑Tue Apr 11, 2023 1:29 am This manifested itself in a kind of noise that had a very "buzzy" kind of quality In my amp (even with the heater wiring and lead dress being being solid).
Charlie
-
SoulFetish
- Posts: 191
- Joined: Wed Apr 03, 2013 1:50 pm
- Location: Norwood, MA
Re: too much hum
Well, I don't have a traditional master volume on my amp. There is just a single volume, however, with the volume on zero the noise was not really a problem. I don't know if it killed the buzz altogether, but it would have passed as a tolerable background noise anyway. But, as soon as you started to dial in any volume, it would steadily get worse. So, because the volume controls the gain coming from the 1st 2 stages, these seemed to be the most sensitive to the buzz. (which makes sense).WhopperPlate wrote: ↑Tue Apr 11, 2023 5:55 pmDid you get this buzz with master on zero?SoulFetish wrote: ↑Tue Apr 11, 2023 1:29 am This manifested itself in a kind of noise that had a very "buzzy" kind of quality In my amp (even with the heater wiring and lead dress being being solid).
-
WhopperPlate
- Posts: 1082
- Joined: Fri Apr 16, 2010 9:04 am
- Location: Santa Cruz, CA
Re: too much hum
amp did this without a guitar input?SoulFetish wrote: ↑Tue Apr 11, 2023 9:28 pmWell, I don't have a traditional master volume on my amp. There is just a single volume, however, with the volume on zero the noise was not really a problem. I don't know if it killed the buzz altogether, but it would have passed as a tolerable background noise anyway. But, as soon as you started to dial in any volume, it would steadily get worse. So, because the volume controls the gain coming from the 1st 2 stages, these seemed to be the most sensitive to the buzz. (which makes sense).WhopperPlate wrote: ↑Tue Apr 11, 2023 5:55 pmDid you get this buzz with master on zero?SoulFetish wrote: ↑Tue Apr 11, 2023 1:29 am This manifested itself in a kind of noise that had a very "buzzy" kind of quality In my amp (even with the heater wiring and lead dress being being solid).
Charlie
-
SoulFetish
- Posts: 191
- Joined: Wed Apr 03, 2013 1:50 pm
- Location: Norwood, MA
Re: too much hum
Oh yeah. I shorted the input