Hello,
While I realize this question is from a green level standpoint, I'd be very interested in what you have to say or think about this.
As a guitarist having done a few successful tube amp builds, the current availability of very nice stereo effects pedals (primarily those with time-based reverb/chorus/delay) has got me seriously thinking about arranging for a single cabinet combo that contains two full and adjustable 20 watt 6V6 push pull amps, providing as much compactness as can reasonably be had. (To mention, this amp would likely have two independent power transformers that I concurred may be best located at opposite ends of the chassis rear with input section at the front middle.)
While experimenting with layouts and doing my best to follow prudent and time-tested layout rules, I kept arriving at a place where the two output transformers were in somewhat closer proximity than I might have optimally preferred. Now, while I have in the past set PT to OT proximity on a chassis using the "heaphone trick" and dancing the OT around for minimal hum/buzz/noise with PT powered, I am not able to establish any intuition at all on what the modes of OT-to-OT interferences might be, at least to the point where they would be creating a practical compromise to each other's performance. It was very eye opening to actually hear in real time how a PT will induce audible noise in an OT but, I admittedly have no way yet to determine that in two output transformers and how proximity will affect each other. What I can say is that this is just a guitar amp that is by its own nature likely very forgiving of interference elements that would not be tolerated in a decent hi-fi stereo amp. A touch of minor crosstalk would not call for the end of civilization as we know it for my application although, this would be a clean pedal platform that is out there in the local venues helping Dad pay the bills.
Can anyone relay their experiences in what the concerns and effects are regarding output transformers in proximity to each other and possibly any guidlines in establishing safe proximity?
Thank you for your time and efforts in my behalf.!!
Best,
Phil D
While I realize this question is from a green level standpoint, I'd be very interested in what you have to say or think about this.
As a guitarist having done a few successful tube amp builds, the current availability of very nice stereo effects pedals (primarily those with time-based reverb/chorus/delay) has got me seriously thinking about arranging for a single cabinet combo that contains two full and adjustable 20 watt 6V6 push pull amps, providing as much compactness as can reasonably be had. (To mention, this amp would likely have two independent power transformers that I concurred may be best located at opposite ends of the chassis rear with input section at the front middle.)
While experimenting with layouts and doing my best to follow prudent and time-tested layout rules, I kept arriving at a place where the two output transformers were in somewhat closer proximity than I might have optimally preferred. Now, while I have in the past set PT to OT proximity on a chassis using the "heaphone trick" and dancing the OT around for minimal hum/buzz/noise with PT powered, I am not able to establish any intuition at all on what the modes of OT-to-OT interferences might be, at least to the point where they would be creating a practical compromise to each other's performance. It was very eye opening to actually hear in real time how a PT will induce audible noise in an OT but, I admittedly have no way yet to determine that in two output transformers and how proximity will affect each other. What I can say is that this is just a guitar amp that is by its own nature likely very forgiving of interference elements that would not be tolerated in a decent hi-fi stereo amp. A touch of minor crosstalk would not call for the end of civilization as we know it for my application although, this would be a clean pedal platform that is out there in the local venues helping Dad pay the bills.
Can anyone relay their experiences in what the concerns and effects are regarding output transformers in proximity to each other and possibly any guidlines in establishing safe proximity?
Thank you for your time and efforts in my behalf.!!
Best,
Phil D
If your stereo guitar amp is anything like what I am working on, you will essentially take the same guitar signal through different preamp and FX chains, then apply those two signals to two similar power amps which feed two different speakers. My version will have one 15 inch speaker and one pair of 6 inch speakers in the same cabinet. My cabinet is nearly finished, but the amp is just beginning to take shape. I will however use identical OPT's (I have a bunch of them) and they will sit right next to each other. The single power transformer will be somewhere else and rotated at a 90 degree angle to the OPT's. There could be some slight crosstalk between the OPT's, but a few milliwatts bleeding through from one OPT to the other will not be noticed when several watts or more of a very similar signal is already present. If interested, that amp design will unfold here, but progress may be slow and stall from time to time:
https://www.diyaudio.com/community/threads/tubelab-wants-a-new-guitar-amp.408097/
https://www.diyaudio.com/community/threads/tubelab-wants-a-new-guitar-amp.408097/
Actually, I take a tiny bit of this back - there is a layout I just created that enables the two OT's to be moved away from each other more if I moved the power tubes toward the middle of the amp, which may be just fine. As long as the decreased proximity between the PT's on the end are still OK with the OT"s having been brought closer to them. That I can determine with the "Headphone trick" and even rotating the OT's would be OK. It might not looks as nice but, I did find by experimenting on another amp that rotating the PT's and OTs can have even a more effect on creating quiet OT secondary taps than proximity even has to offer - but, it did end up being a compromise of some proximity coupled with some rotational angle. We have the right to do that!
thank you,
Best,
Phil
thank you,
Best,
Phil
Thanks for responding tubelab, When I get out of work, I will definitely take a closer look at what you are doing.
My idea was for two discrete Fender Deluxe reverb type amps to be built into a Twin reverb chassis, each outputting to its own ten-inch speaker. Pedal stereo effects would be routed to the inputs of each amp. Generally, the amp would be clean enough to not malform the quality of the effects, too too much. Then, live, each speaker would be miced up separately and panned full left and right on the PA board for a nice wide stereo effect to the listening audience. that way, rushing to a gig from work, I would have an easy, uncomplicated and quick set up that would both give me a satisfying and somewhat realistic monitoring of my guitar rig, while also giving the audience and band a quality wide stereo rendition of my guitar and pedals. Thats what I have so far but, being early in the stereo conquest, there is plenty of time for change or improvement to the concept.
Thank you,
Phil
My idea was for two discrete Fender Deluxe reverb type amps to be built into a Twin reverb chassis, each outputting to its own ten-inch speaker. Pedal stereo effects would be routed to the inputs of each amp. Generally, the amp would be clean enough to not malform the quality of the effects, too too much. Then, live, each speaker would be miced up separately and panned full left and right on the PA board for a nice wide stereo effect to the listening audience. that way, rushing to a gig from work, I would have an easy, uncomplicated and quick set up that would both give me a satisfying and somewhat realistic monitoring of my guitar rig, while also giving the audience and band a quality wide stereo rendition of my guitar and pedals. Thats what I have so far but, being early in the stereo conquest, there is plenty of time for change or improvement to the concept.
Thank you,
Phil
lavane,
Good looking!
Some of the factors that might have contributed to your "dead Quiet" amplifier:
A non-steel chassis?
Proper reduction of all the types of ground loops, such as input circuit wiring, B+ wiring; all with short wire-length well planned controlled ground loops?
Perhaps DC for the 300B filaments?
Good looking!
Some of the factors that might have contributed to your "dead Quiet" amplifier:
A non-steel chassis?
Proper reduction of all the types of ground loops, such as input circuit wiring, B+ wiring; all with short wire-length well planned controlled ground loops?
Perhaps DC for the 300B filaments?
Last edited:
pjd3,
An unloaded power transformer's external magnetic field tends to be less sonically offensive; with few upper harmonics, and those are low amplitude.
A loaded power transformer's external magnetic field tends to be more sonically offensive; especially if the B+ uses solid state rectifiers driving a capacitor input filter.
That will have more upper harmonics, and more amplitude of those upper harmonics, all the way to your ear's most sensitive frequency range.
An unloaded power transformer's external magnetic field tends to be less sonically offensive; with few upper harmonics, and those are low amplitude.
A loaded power transformer's external magnetic field tends to be more sonically offensive; especially if the B+ uses solid state rectifiers driving a capacitor input filter.
That will have more upper harmonics, and more amplitude of those upper harmonics, all the way to your ear's most sensitive frequency range.
Last edited:
I wonder what the details are, as the headphone typically has higher impedance than a speaker - easier to drive - and I wonder if the same level would actually show up across 4 / 8 / 16 Ohms or whatever.
The cheap vintage amps (Bogen, Thomas) I'm currently working on have steel chassis and use that steel as "ground". A technique I've seen in an Eico amp is the ground is an elevated "Bus" wire running throughout the amp, connecting to the - I assume it was - steel chassis at only one point.
I assume with magnetics floating about in close proximity, that steel will have an induced voltage in it. "Good" choice for ground -
Typical vintage push pull amplifiers, use magnetic steel chassis, and use global negative feedback.
Global Negative feedback does at least 4 things:
Reduces Distortion (+)
Increases Damping Factor (+)
Reduces Hum (+)
And . . . Increases Saturation, once saturation has begun, no matter what is the reason that started the saturation (-) Oh!
Push Pull output transformers (with interleaved E-s and I-s, and No Intentional spaced Air Gap) tend to have Less hum pickup from magnetic fields.
Single ended output transformers (with all the E-s on one side, all the I-s on the other side, and an Air Gap between them) tend to have More hum pickup from magnetic fields.
Please do not ask me about my Dyna Stereo 70 Magnetic Steel Chassis . . . and all the hum I got when I built various single ended, and various push pull amplifiers, all those circuits did Not have negative feedback. Salt in the Wound!
Instead, think about all the single ended and push pull amplifiers I designed and built, that did Not have negative feedback, but Did have Aluminum Chassis.
Hum < 100uV.
Yes, less than 100 microvolts hum.
As always, Your Mileage May Vary.
Factoid Central
(Just my opinions)
Have Fun!
Global Negative feedback does at least 4 things:
Reduces Distortion (+)
Increases Damping Factor (+)
Reduces Hum (+)
And . . . Increases Saturation, once saturation has begun, no matter what is the reason that started the saturation (-) Oh!
Push Pull output transformers (with interleaved E-s and I-s, and No Intentional spaced Air Gap) tend to have Less hum pickup from magnetic fields.
Single ended output transformers (with all the E-s on one side, all the I-s on the other side, and an Air Gap between them) tend to have More hum pickup from magnetic fields.
Please do not ask me about my Dyna Stereo 70 Magnetic Steel Chassis . . . and all the hum I got when I built various single ended, and various push pull amplifiers, all those circuits did Not have negative feedback. Salt in the Wound!
Instead, think about all the single ended and push pull amplifiers I designed and built, that did Not have negative feedback, but Did have Aluminum Chassis.
Hum < 100uV.
Yes, less than 100 microvolts hum.
As always, Your Mileage May Vary.
Factoid Central
(Just my opinions)
Have Fun!
Hello, just a quick "leaving work" response about "The headphone trick".
I understood that is merely a way to determine induced proximity of 60 cycle hum, and moving the OT around with headphones across the full secondary coil gave you enough to hear when the 60 cycle hum has been minimized due to some kind of transformer magnetic poles orientation coil cancellation. I can imagine once the PT is underload and supplying AC to all sorts of rectifiers, coils, caps etc....., that there would be a whole new list of potential issues to contend with. I've heard reports that when the 60 cycle hum is transduced badly enough, it can be heard even potently from way out in a room. That sounded like a good thing to avoid, and the purported solution was easy enough to follow through with. And as I recall, the angle of the OT to the PT has just as much bearing on minimizing the hum as the actual distance apart. I also recall that it was a fairly small window of angular rotation that brought the greatest minimization of buzz/hum.
Thank you
Phil D .
I understood that is merely a way to determine induced proximity of 60 cycle hum, and moving the OT around with headphones across the full secondary coil gave you enough to hear when the 60 cycle hum has been minimized due to some kind of transformer magnetic poles orientation coil cancellation. I can imagine once the PT is underload and supplying AC to all sorts of rectifiers, coils, caps etc....., that there would be a whole new list of potential issues to contend with. I've heard reports that when the 60 cycle hum is transduced badly enough, it can be heard even potently from way out in a room. That sounded like a good thing to avoid, and the purported solution was easy enough to follow through with. And as I recall, the angle of the OT to the PT has just as much bearing on minimizing the hum as the actual distance apart. I also recall that it was a fairly small window of angular rotation that brought the greatest minimization of buzz/hum.
Thank you
Phil D .
60Hz power mains.
Power transformers
Solid state rectifiers, cap input filters.
Magnetic fields.
60Hz, 120Hz, 180Hz, 240Hz, . . . 900Hz, 960Hz, 1020 Hz, . . .
Does that sound good? I have heard such things from tube amplifiers.
Did you ever connect that output transformer's magnetically coupled hum signal to an oscilloscope?
The waveform is as ugly as a Rat in the pumpkin pie.
The difference is, generally with an unloaded power transformer, there is no rat in the pumpkin pie.
Yes, there is line noise, but if you do not know how to use a B+ filter of choke, cap, resistor, cap, then you need to try it . . . Clean B+
But there still may be line noise magnetically coupled to the output transformer.
Power transformers
Solid state rectifiers, cap input filters.
Magnetic fields.
60Hz, 120Hz, 180Hz, 240Hz, . . . 900Hz, 960Hz, 1020 Hz, . . .
Does that sound good? I have heard such things from tube amplifiers.
Did you ever connect that output transformer's magnetically coupled hum signal to an oscilloscope?
The waveform is as ugly as a Rat in the pumpkin pie.
The difference is, generally with an unloaded power transformer, there is no rat in the pumpkin pie.
Yes, there is line noise, but if you do not know how to use a B+ filter of choke, cap, resistor, cap, then you need to try it . . . Clean B+
But there still may be line noise magnetically coupled to the output transformer.
This picture shows the very first Tubelab TSE boards being tested over 20 years ago. Note the three transformers all lined up in a row looking nice. I had not yet made a case for any of them. There were three boards, one with 45 tubes, one with 2A3's and one with 300B's. All worked well and sounded great with my 87 dB Yamaha NS-10M Studio monitors. I put one board in a DIY clear Lexan case. Since the three transformers in a row looked nice, that's how I mounted them. I took it to a friend's house who had some Lowther based 102 dB horn speakers with active powered subwoofers built in and it was ugly, lots of HUMMMMMMM!
I brought the amp home and scoped the output and there was indeed a good bit of 60 Hz voltage on the output of the transformer next to the power transformer, and some measurable hum on the other OPT. Oddly the hum was loudest with no output tubes in the amp. Apparently, the DC current through the OPT will reduce the susceptibility to hum pickup. Rotating the power transformer 90 degrees made the hum vanish and it was indeed quiet on its next meeting with the 102 dB speakers. We drove one channel to full output (maybe 2 watts) and unplugged the input cable from the other channel. It was hard to tell if there was any sound from the quiet channel with the loud channel going, but no vibration could be felt on the subwoofer cone. If there was any coupling from one OPT to the other, it was not an issue.
After listening to the Lexan TSE on $10,000 worth of speakers, I fired up the first production quality SSE board for some OPT auditioning. Knowing that the OPT's mounted on the particle board were the worse sounding pair I had, I didn't bother to conceal them. All others were in unmarked cardboard boxes and connected into the SSE with clip leads. With 5 people present and 5 different sets of OPT's there were no clear cut winners, but the big Hammonds were unanimously the worst sounding through real sensitive speakers. I still have them and now, 20 years later I know how to make them sound good. They need to be driven by a tube with a very low plate resistance. A big TV sweep tube in UNSET mode with lots of local feedback feeding my old Yamaha NS-10s will do it.
The TSE was heard here with some NX-483 tubes from 1929. they sounded great then and are STILL IN THE AMP TODAY. Will they make it to 100 years old? I was rummaging through a flea market in Mount Dora Florida when I spotted an old Sparton radio chassis in the dirt seemingly unwanted. The owner wanted $20, I gave him $10. I kept the tubes and sold the chassis to a collector for $20. That's where the NX-483's came from. They are like a 45 with a 5 volt filament. Will any of the tubes being made today live for 90+ years?
I brought the amp home and scoped the output and there was indeed a good bit of 60 Hz voltage on the output of the transformer next to the power transformer, and some measurable hum on the other OPT. Oddly the hum was loudest with no output tubes in the amp. Apparently, the DC current through the OPT will reduce the susceptibility to hum pickup. Rotating the power transformer 90 degrees made the hum vanish and it was indeed quiet on its next meeting with the 102 dB speakers. We drove one channel to full output (maybe 2 watts) and unplugged the input cable from the other channel. It was hard to tell if there was any sound from the quiet channel with the loud channel going, but no vibration could be felt on the subwoofer cone. If there was any coupling from one OPT to the other, it was not an issue.
After listening to the Lexan TSE on $10,000 worth of speakers, I fired up the first production quality SSE board for some OPT auditioning. Knowing that the OPT's mounted on the particle board were the worse sounding pair I had, I didn't bother to conceal them. All others were in unmarked cardboard boxes and connected into the SSE with clip leads. With 5 people present and 5 different sets of OPT's there were no clear cut winners, but the big Hammonds were unanimously the worst sounding through real sensitive speakers. I still have them and now, 20 years later I know how to make them sound good. They need to be driven by a tube with a very low plate resistance. A big TV sweep tube in UNSET mode with lots of local feedback feeding my old Yamaha NS-10s will do it.
The TSE was heard here with some NX-483 tubes from 1929. they sounded great then and are STILL IN THE AMP TODAY. Will they make it to 100 years old? I was rummaging through a flea market in Mount Dora Florida when I spotted an old Sparton radio chassis in the dirt seemingly unwanted. The owner wanted $20, I gave him $10. I kept the tubes and sold the chassis to a collector for $20. That's where the NX-483's came from. They are like a 45 with a 5 volt filament. Will any of the tubes being made today live for 90+ years?
Attachments
Thanks.
Yes, aluminum plate, short paths, and DC regulated heaters.
I also.used a choke instead of the resistor. The Trancendar OPT's I used sound really good. I wish he was still winding transformers.
It's not obvious in the pictures but in my hum reduction quest I also ditched the resistor and stuffed a small choke inside the plastic case up under the power transformer and added a poly propylene capacitor from the B+ terminal of the OPT's to ground. It's the white thing seen hiding behind the 45 tubes to the right of the green filament regulator heat sink in the picture. The OPT's are Electra Prints. The power transformer is an Allied 6K56VG which is a Hammond in disguise. The PCB eliminates ground loops with a single point ground at the negative terminal of the big fat electrolytic in the foreground to the left of the volume pot. There should be only one connection from the PCB to the chassis, usually at the input connector.
What was the biggest difference between your ver.1 and ver.2 boards?
My ver.1 has been worry free all these years. I did mount a big TO-220 heat sink on the dual 5v regulator and mounted flat against the bottom of the aluminum plate.
Heck, the amp was dead quiet with everything just wired together with jumper leads sprawled all over my banch before I mounted everything to the plate.
My ver.1 has been worry free all these years. I did mount a big TO-220 heat sink on the dual 5v regulator and mounted flat against the bottom of the aluminum plate.
Heck, the amp was dead quiet with everything just wired together with jumper leads sprawled all over my banch before I mounted everything to the plate.
Attachments
The Lexan box has near zero airflow so I used a heat sink from an old Pentium 1 CPU chip. Somewhere between the 386 chips and the Pentium, CPU's needed heat sinks, but no fans. I saved all the heat sinks from all the dead computers I parted out over the years that I build PC's.
The Sharp filament regulator chip that's used in the original TSE went extinct and there is no drop in replacement. That meant a new board layout, so I asked the forums what else they would want in a new TSE. A larger board size was the most popular request. Heat mitigation in the mosfets was also a common request. The transformation from TSE to TSE-II took some time, and most of the valid requests were accommodated. The audio path did not change though. It all unfolded and was pretty well documented here:
https://www.diyaudio.com/community/threads/after-a-14-year-run-the-tse-must-die.331038/
The Sharp filament regulator chip that's used in the original TSE went extinct and there is no drop in replacement. That meant a new board layout, so I asked the forums what else they would want in a new TSE. A larger board size was the most popular request. Heat mitigation in the mosfets was also a common request. The transformation from TSE to TSE-II took some time, and most of the valid requests were accommodated. The audio path did not change though. It all unfolded and was pretty well documented here:
https://www.diyaudio.com/community/threads/after-a-14-year-run-the-tse-must-die.331038/
Tubelab_com
I am always glad to read your posts.
But darn, then I start thinking about 2GHz amplifiers, and EVM. I am trying to forget.
You said:
"Oddly the hum was loudest with no output tubes in the amp. Apparently, the DC current through the OPT will reduce the susceptibility to hum pickup."
I think the real cause is the difference between a primary that has the tube plate impedance, rp, across it, versus no damping reistance across the primary.
Exmple:
With a 2A3, 800 Ohms plate impedance, rp, is across the 2400 Ohm primary.
The hum transmitted magnetically from the power transformer to the output transformer is loaded by the 800 Ohms rp that is across the output transformer primary.
Damping Factor 3, but with an 8 Ohm resistor across the secondary, the effective damping is about 4 (to the magnetically transmitted hum.)
With a Missing 2A3, a very high impedance is across the 2400 Ohm primary.
The hum transmitted magnetically from power transformer to output transformer is un-loaded, except by the 8 Ohm resistor across the secondary;
The effective damping of the magnetically transmitted hum signal is 1 (unity).
That is a magnetically transmitted hum amplitude difference of 1 to 4, (1, with the 2A3 installed; versus 4, 2A3 not installed, Respectively).
I do not think the DC current through the primary has any effect, because the DC current is not high enough to cause saturation.
The only thing DC current does, is it causes the 2A3 to have an rp = 800 Ohms.
Just my theory, my experience, and my opinion.
I am always glad to read your posts.
But darn, then I start thinking about 2GHz amplifiers, and EVM. I am trying to forget.
You said:
"Oddly the hum was loudest with no output tubes in the amp. Apparently, the DC current through the OPT will reduce the susceptibility to hum pickup."
I think the real cause is the difference between a primary that has the tube plate impedance, rp, across it, versus no damping reistance across the primary.
Exmple:
With a 2A3, 800 Ohms plate impedance, rp, is across the 2400 Ohm primary.
The hum transmitted magnetically from the power transformer to the output transformer is loaded by the 800 Ohms rp that is across the output transformer primary.
Damping Factor 3, but with an 8 Ohm resistor across the secondary, the effective damping is about 4 (to the magnetically transmitted hum.)
With a Missing 2A3, a very high impedance is across the 2400 Ohm primary.
The hum transmitted magnetically from power transformer to output transformer is un-loaded, except by the 8 Ohm resistor across the secondary;
The effective damping of the magnetically transmitted hum signal is 1 (unity).
That is a magnetically transmitted hum amplitude difference of 1 to 4, (1, with the 2A3 installed; versus 4, 2A3 not installed, Respectively).
I do not think the DC current through the primary has any effect, because the DC current is not high enough to cause saturation.
The only thing DC current does, is it causes the 2A3 to have an rp = 800 Ohms.
Just my theory, my experience, and my opinion.
Last edited:
I'm glad I ordered afew extra regulators when I sourced all the parts back then. I bought a few extra of other sand parts as well.
That makes perfect sense. My initial observation was that the speaker hum was loud when the amp was first turned on and got quieter as the rectifier warmed up, and was affected by varying the bias current through the output tube. I assumed that the current "did it" and never gave it much more thought.
It seems that that old Sharp regulator was pretty hard to kill. It will simply turn itself off if it gets too hot and limit its output current to 5 amps when a pair of 2A3's are cold but feeding it too much input voltage will kill them. I think the only dead ones I ever saw was to stupid mistakes with input voltage.
Interesting, this would explain why a few of my amps hums for 20 seconds or so until the tubes starts conducting.