Greetings, Friends. I'm starting to look at layouts for my new amp and have encountered a hum goblin. Amp circuit is nice simple Single-Ended EL84 output, discussed earlier in This Post. Power supply is lifted from the 6V6 Marblewood Amp from CascadeTubes.com (@Suncalc ) but since I don't need as much heater current I'm using the EDCOR XPWR013 and a pair of EDCOR GXSE15-5K OPTs. My initial layout had the iron lined up across the back of the chassis:
But when I wired a pair of headphones into the secondary of one of the OPTs and energized the PT, there was hum. Rotating the PT 90° helped, but did not eliminate the hum. I imagine a Horizontal/Laydown/Z PT would be required to cancel the hum completely, but I'm committed to this hardware. So the best I've come up with is this layout, with the iron at opposite ends of the chassis:
There's still a tiny bit of hum. not much at all. I'm tempted to simply order a larger chassis, Hammond has a model 4" longer than this one that might be a better choice. Or maybe I'm overthinking again. The layout for Matt's amp build shows a 2 3/16" gap between PT and OPT, and I've got a wider space than that with this layout. Rectifier tube goes in the bottom left corner and audio tubes occupy the area in the middle. As long as I keep the input wiring clear of the b+ wiring feeding the OPTs I should be ok. Right?
thanks for taking a look!
will
But when I wired a pair of headphones into the secondary of one of the OPTs and energized the PT, there was hum. Rotating the PT 90° helped, but did not eliminate the hum. I imagine a Horizontal/Laydown/Z PT would be required to cancel the hum completely, but I'm committed to this hardware. So the best I've come up with is this layout, with the iron at opposite ends of the chassis:
There's still a tiny bit of hum. not much at all. I'm tempted to simply order a larger chassis, Hammond has a model 4" longer than this one that might be a better choice. Or maybe I'm overthinking again. The layout for Matt's amp build shows a 2 3/16" gap between PT and OPT, and I've got a wider space than that with this layout. Rectifier tube goes in the bottom left corner and audio tubes occupy the area in the middle. As long as I keep the input wiring clear of the b+ wiring feeding the OPTs I should be ok. Right?
thanks for taking a look!
will
I think you're more likely to pick up more hum from running the input wiring and having the driver tube that close to the AC.
Good point. Even if I put the driver tube down front it's still the closest thing to all that flux.
You might try loading the OPT secondary with an appropriate resistor, say 8 ohms on the correct tap. Remember that it will be loaded down when operating. You have found a better layout with the ear buds, but now try with the load attached. It may not be as bad as you suspect. The ear buds are likely a fairly hisg resistance.
Earbuds measure 16ohm. I put a 8ohm 1w resistor across the 0-8 taps on the OPT and soldered the earbuds to that. Still had hum within about 3" of the PT. Spun the PT 90° same level of hum. It drops to inaudible if the OPT is along the far edge of the chassis. I think I need a wider chassis.
does an unloaded power transformer generate more flux/RF/noise?
is this why steel is a bad choice for chassis? It's so easy to use, strong, pre-finished. Plenty of rim shops in the Bronx could powder coat an aluminum box after I drill the holes, I suppose...
I'm worried about mounting the driver tube about 3" from this noisy power transformer. Wouldn't a tube shield help out here?
thanks!
does an unloaded power transformer generate more flux/RF/noise?
is this why steel is a bad choice for chassis? It's so easy to use, strong, pre-finished. Plenty of rim shops in the Bronx could powder coat an aluminum box after I drill the holes, I suppose...
I'm worried about mounting the driver tube about 3" from this noisy power transformer. Wouldn't a tube shield help out here?
thanks!
Last edited:
Yeah I was going to ask something similar. Can you measure the voltage across your earbuds? It might be inaudible with a speaker.
Normally, for headphone use, you terminate the output transformer 8 Ohm tap with an Ohm power resistor; and then use a 100 Ohm resistor (or even more Ohms) from the output tap to the 16 Ohm headphones.
Possible causes of hum . . .
Magnetic steel chassis.
Input circuit ground loop.
B+ first filter cap ground loop.
B+ filter choke orientation and spacing relative to the output transformers, if there is a choke.
Un-bypassed cathodes; especially for some tubes, not just the tube type, but certain manufacturers, and/or certain manufacturing production runs -(when using AC power to the filaments; but do not modify to use DC; instead select your tubes for low hum).
Un-grounded output secondary.
Wiring dress and position.
All but one of my amplifiers have less than 100uV into 8 Ohms; (it is Not easy to get it that low), and even that is Not good enough for headphones - unless you terminate the output with 8 Ohms, and use a series resistor to the headphones (like 100, 200, or whatever Ohms to cut back the hum). Then turn the volume up (all in the interest of a better signal to 'hum plus noise' ratio).
Possible causes of hum . . .
Magnetic steel chassis.
Input circuit ground loop.
B+ first filter cap ground loop.
B+ filter choke orientation and spacing relative to the output transformers, if there is a choke.
Un-bypassed cathodes; especially for some tubes, not just the tube type, but certain manufacturers, and/or certain manufacturing production runs -(when using AC power to the filaments; but do not modify to use DC; instead select your tubes for low hum).
Un-grounded output secondary.
Wiring dress and position.
All but one of my amplifiers have less than 100uV into 8 Ohms; (it is Not easy to get it that low), and even that is Not good enough for headphones - unless you terminate the output with 8 Ohms, and use a series resistor to the headphones (like 100, 200, or whatever Ohms to cut back the hum). Then turn the volume up (all in the interest of a better signal to 'hum plus noise' ratio).
Last edited:
Thanks for taking a look. I'm well ahead of any of the causes you mentioned, save the fist one. I'm just moving transformers around on the top of a box to see what makes the most/least hum. All leads of the PT are isolated, no current thru the secondary, and a 15w bulb in the current limiter. Primary leads on the OPT are isolated, secondary is connected to a headphone jack across an 8-ohm resistor, and I'm listening to a 16-ohm earbud.
I hooked up a DMM across the 8-ohm resistor and moved the OPT around the top of the chassis. I started hearing the hum on my earbud almost as soon as I started moving the OPT towards the PT, about 4" and the hum continued to get louder, but the DMM did not register until it was within half an inch of the PT. I got 0.007 volts when they were within 1/8".
It was at this point I started thinkin'
If I can measure an AC signal, I can scope it. So, I got a 60Hz wave that measures 7mVrms when the sides (stacks?) are touching, or within 1/8" or so.
Moving the OPT to the left, it drops below 1mVrms at 1" and drops below 1mVp-p at 2" but I can hear the hum until the OPT is 4" from the PT.
Rotating the PT 90° so the bell faces the OPT yielded significant improvement. While hum was audible, the scope was unable to detect a wave or any voltage when the iron was so oriented, at any distance.
So I i guess that's empirical evidence that the layout in the 2nd pic of Post #1 is ideal. For the iron.
w
I hooked up a DMM across the 8-ohm resistor and moved the OPT around the top of the chassis. I started hearing the hum on my earbud almost as soon as I started moving the OPT towards the PT, about 4" and the hum continued to get louder, but the DMM did not register until it was within half an inch of the PT. I got 0.007 volts when they were within 1/8".
It was at this point I started thinkin'
If I can measure an AC signal, I can scope it. So, I got a 60Hz wave that measures 7mVrms when the sides (stacks?) are touching, or within 1/8" or so.
Moving the OPT to the left, it drops below 1mVrms at 1" and drops below 1mVp-p at 2" but I can hear the hum until the OPT is 4" from the PT.
Rotating the PT 90° so the bell faces the OPT yielded significant improvement. While hum was audible, the scope was unable to detect a wave or any voltage when the iron was so oriented, at any distance.
So I i guess that's empirical evidence that the layout in the 2nd pic of Post #1 is ideal. For the iron.
w
Before you jump to conclusions, ground the end-bells of all the transformers. Leaving them floating will exacerbate the problem.
I knew there was a missing tidbit of wisdom that would fill out the picture. Thank you.
I'm gonna pause this build. Nice that the Holidays are here to take away all my time anyway. Right now I'm thinking I'll save the XPWR01 for a build with a wider chassis and maybe a pair of CSXE25s. As this is meant to be a budget build I'm switching to the cheaper, horizontal Hammond 270DAZ ($84 shipped from Hawk) that makes 520v@104mA and 6.3v@4A, plenty of heater power to use the 6CA4 or just a SS rectifier. You could also use the Champ's 290AX with the inner windings, that's 550v@100mA, 6.3v@2.25A and 5v@3A, barely enough heater power to use a 6N1P driver, prob better to use 6DJ8 or 12AU7, and a 5Y3 should yield enough voltage drop to keep the EL84s happy. Maybe bump the cathode resistors to 200 ohm.
I'm gonna pause this build. Nice that the Holidays are here to take away all my time anyway. Right now I'm thinking I'll save the XPWR01 for a build with a wider chassis and maybe a pair of CSXE25s. As this is meant to be a budget build I'm switching to the cheaper, horizontal Hammond 270DAZ ($84 shipped from Hawk) that makes 520v@104mA and 6.3v@4A, plenty of heater power to use the 6CA4 or just a SS rectifier. You could also use the Champ's 290AX with the inner windings, that's 550v@100mA, 6.3v@2.25A and 5v@3A, barely enough heater power to use a 6N1P driver, prob better to use 6DJ8 or 12AU7, and a 5Y3 should yield enough voltage drop to keep the EL84s happy. Maybe bump the cathode resistors to 200 ohm.
This is already good :
Below, one example with my U-KT120 : the amp is dead silent with "the head in a 12" woofer"...
T
Below, one example with my U-KT120 : the amp is dead silent with "the head in a 12" woofer"...
T
Last edited:
Thank you! You can see where the audio tubes will be placed on the top in that picture...will they be too close to the power transformer?
Well @dubadub , I don't clearly see where those tubes are 😕 , but you have my pictured example above and below again to give you an idea - aluminium chassis is 275x165x65mm to give the scale 😉 :
T
The * lines on the blue tape indicate the center of each tube socket. To put both OPTs on the far end, as in the 2nd picture of Post #1, would require all the audio tubes to be between the irons, with the driver tube 75-100mm from the power transformer. Seems a little close, but maybe a tube shield would be enough.
Are you planning to listen to this amp with high sensitive earbuds? Honestly you are gonna end up with more noise in the final amp putting the driver tube/input signal near the power transformer than what you are doing moving these transformers so far apart. I've built a bunch of amps using your initial layout, I have never heard any noise.
A good test for magnetically coupled hum is to remove all the tubes, then apply power.
If there is magnetic coupling between the PT & OPT(s) it will show up on the OPT output terminals.
A small sheet of scrap steel or a steel chassis bottom is a great tool to move between the magnetic devices on the chassis
& will indicate where the problems are. Aluminum sheet is OK at RF but not very useful at power frequencies.
Try this simple test, you will like it.👍
If there is magnetic coupling between the PT & OPT(s) it will show up on the OPT output terminals.
A small sheet of scrap steel or a steel chassis bottom is a great tool to move between the magnetic devices on the chassis
& will indicate where the problems are. Aluminum sheet is OK at RF but not very useful at power frequencies.
Try this simple test, you will like it.👍
Also note that if your amp is going to use global negative feedback, then that feedback should attenuate hum that is being injected into the forward loop (ie. the OPT) and is not part of the input signal.
PT to OPT Magnetic Coupling Tests.
The Amplifier I used for these tests is something I built in 1956. Basically it is a simple PP 6AQ5 pair driven by a 12AX7. There is a built in 6J6 Baxandall Bass & Treble tone control set, the front end a 12AX7 preamp corrected for use with the GE Variable Reluctance phono pickup. All the iron is Hammond including the box. The OPT is a 1618 while the PT is a 270BZ, the Rectifier a 6X4.
The PT & OPT are separated by only ¼ inch. The laminations are at right angles & that probably helps a bit.
The Amp has a handle & for the time was reasonably portable. I took it & a deck of 45s & LPs in a box, a Webcor turntable & a 10 inch Celestron speaker in a ported box to many house parties of our friends. Large quantities of alcoholic beverages were consumed. And a thick cloud of cigarette smoke also developed. At that time most people smoked, I didn’t. The smoke didn’t bother me at the party but as I drove home my eyes watered a lot. The next morning my clothes smelled very bad, they might need to be fumigated.👎
Smoking was so ubiquitous that nearly all the girls in the HP offices smoked at their desks. The tobacco lobby had US politicians tied up, there were many quietly on the Tobacco paylist. There was a joke often heard, ‘After trying Camels 9 out of 10 Doctors went back to their Wives’.
The test is very simple, remove all the tubes, hookup the test equipment to the OPT secondary to measures what comes out when the power supply is connected. The time & frequency domain measurements were taken by a Picoscope 3224. The voltage measured by a MetraHit 29S Precision DVM & Wattmeter.
I was surprised that the Aluminum piece when stuffed in gave the same attenuation as the piece of steel chassis bottom. But the Aluminum was 1/8 inch thick, the chassis bottom is 20 gauge, much thinner.
I did not apply ‘Cookes Variable Constant’ to fix the results some of us might like to see. What we seeing is what we are getting!😀
The Amplifier I used for these tests is something I built in 1956. Basically it is a simple PP 6AQ5 pair driven by a 12AX7. There is a built in 6J6 Baxandall Bass & Treble tone control set, the front end a 12AX7 preamp corrected for use with the GE Variable Reluctance phono pickup. All the iron is Hammond including the box. The OPT is a 1618 while the PT is a 270BZ, the Rectifier a 6X4.
The PT & OPT are separated by only ¼ inch. The laminations are at right angles & that probably helps a bit.
The Amp has a handle & for the time was reasonably portable. I took it & a deck of 45s & LPs in a box, a Webcor turntable & a 10 inch Celestron speaker in a ported box to many house parties of our friends. Large quantities of alcoholic beverages were consumed. And a thick cloud of cigarette smoke also developed. At that time most people smoked, I didn’t. The smoke didn’t bother me at the party but as I drove home my eyes watered a lot. The next morning my clothes smelled very bad, they might need to be fumigated.👎
Smoking was so ubiquitous that nearly all the girls in the HP offices smoked at their desks. The tobacco lobby had US politicians tied up, there were many quietly on the Tobacco paylist. There was a joke often heard, ‘After trying Camels 9 out of 10 Doctors went back to their Wives’.
The test is very simple, remove all the tubes, hookup the test equipment to the OPT secondary to measures what comes out when the power supply is connected. The time & frequency domain measurements were taken by a Picoscope 3224. The voltage measured by a MetraHit 29S Precision DVM & Wattmeter.
I was surprised that the Aluminum piece when stuffed in gave the same attenuation as the piece of steel chassis bottom. But the Aluminum was 1/8 inch thick, the chassis bottom is 20 gauge, much thinner.
I did not apply ‘Cookes Variable Constant’ to fix the results some of us might like to see. What we seeing is what we are getting!😀
