Generally do you prefer to ground a shielded cable connecting the grid to a jack or volume pot at the grid end or at the Jack/pot end? On my breadboard it seems better to ground the grid side and leave the pot/Jack floating. But one can make the theoretical argument of keeping the shield intact all the way across your external interconnect right up to the grid.
Just to be “âhnal”, why not both ends? I know the signal levels of the input of an amplifier are modest. And for “preamp output” on the order of volts.
So, really “volts” are much greater than the millivolts associated with magnetic cartridge pickups for records. They clearly need shielding to protect against sub-microvolt AC induction, or pickup of errant AM radio signals, and so forth.
But amp-inputs with 0.775 V sensitivity (i.e. 'dBu', which is to say, 1 milliwatt of power into a 600 Ω load! …
Anyway, good shields are good to have. Grounding them at either end to a good star or big-fat-copper ground bus ought to be perfectly acceptable. Especially for nominal amplifier input levels. Not magnetic pickup levels.
GoatGuy ✓
So, really “volts” are much greater than the millivolts associated with magnetic cartridge pickups for records. They clearly need shielding to protect against sub-microvolt AC induction, or pickup of errant AM radio signals, and so forth.
But amp-inputs with 0.775 V sensitivity (i.e. 'dBu', which is to say, 1 milliwatt of power into a 600 Ω load! …
P = E²/R …
PR = E² …
E = √(PR) …
E = √( 0.001 W × 600 Ω ) …
E = 0.775 VRMS
… bet many people don't know that!!!) PR = E² …
E = √(PR) …
E = √( 0.001 W × 600 Ω ) …
E = 0.775 VRMS
Anyway, good shields are good to have. Grounding them at either end to a good star or big-fat-copper ground bus ought to be perfectly acceptable. Especially for nominal amplifier input levels. Not magnetic pickup levels.
GoatGuy ✓
For my low wattage power amps, I do the following:
The RCA input jack ground 1, goes to the 'zero' end of the volume control 2, and to the bottom of the input tubes self bias resistor 3, and bypass cap 4. Those 4 ground points connect together before connecting them with a separate wire that goes to the amplifier's central ground point. Those first 4 points make a purposeful short local ground loop.
I do not need to use a shielded wires there, I keep this part of the circuitry short and away from the other amplifier wiring.
I also make another short purposeful local ground loop:
The center tap of the B+ secondary 1, negative of the first filter cap 2, and the second filter cap 3. Only after that local loop do I connect another wire from there to the amplifier's central ground point.
Of course I also have to pay attention to the spacing and orientation of the output transformer, power transformer, and B+ filter choke.
I use aluminum chassis (steel has the problem of coupling those electro-magnetic parts).
Then there is the requirement to watch where you allow the other wiring to be placed.
Those practices give my amplifier's output hum to be equal to, or less than 100uV.
The RCA input jack ground 1, goes to the 'zero' end of the volume control 2, and to the bottom of the input tubes self bias resistor 3, and bypass cap 4. Those 4 ground points connect together before connecting them with a separate wire that goes to the amplifier's central ground point. Those first 4 points make a purposeful short local ground loop.
I do not need to use a shielded wires there, I keep this part of the circuitry short and away from the other amplifier wiring.
I also make another short purposeful local ground loop:
The center tap of the B+ secondary 1, negative of the first filter cap 2, and the second filter cap 3. Only after that local loop do I connect another wire from there to the amplifier's central ground point.
Of course I also have to pay attention to the spacing and orientation of the output transformer, power transformer, and B+ filter choke.
I use aluminum chassis (steel has the problem of coupling those electro-magnetic parts).
Then there is the requirement to watch where you allow the other wiring to be placed.
Those practices give my amplifier's output hum to be equal to, or less than 100uV.
Alert / Apology… While it is rather amazing what “the mind will come up with at 3 AM”, this very early I woke up in a sweat: my opinion of “ground at both ends” is wrong. It creates a kind of low impedance ground loop which can and will pick up everything from low RF all the way up to UHF.
Now, the “Faraday Shield” effect keeps the oscillating electric fields on the outside of the shield for the most part. So the “ground loop problem” isn't in reality a very large one.
But the best advice is just to attach shield/braid at either end, and let it do its work. There might be a subtle advantage in grounding nearest to the wiper of the front-end attenuator pot. But if so, it is subtle.
Again, my apologies for misrepresenting the ground-connection advice. Tho' he didn't claim it, this discussion and my incorrect advice was corrected by none other than DF96 a few years back. See? Even at 3 AM, I remember!!!
GoatGuy ✓
Now, the “Faraday Shield” effect keeps the oscillating electric fields on the outside of the shield for the most part. So the “ground loop problem” isn't in reality a very large one.
But the best advice is just to attach shield/braid at either end, and let it do its work. There might be a subtle advantage in grounding nearest to the wiper of the front-end attenuator pot. But if so, it is subtle.
Again, my apologies for misrepresenting the ground-connection advice. Tho' he didn't claim it, this discussion and my incorrect advice was corrected by none other than DF96 a few years back. See? Even at 3 AM, I remember!!!
GoatGuy ✓
There is a measurement of coax cables, it is called Reverse Transfer Impedance.
The simplified explanation:
If I remember correctly, the Reverse Transfer Impedance measurement determines how much signal will be induced into the center conductor, when an external field impinges on the coax.
That field is non contacting, and is away from the body of the coax.
"Grounds are Commonly Misunderstood" - Me
The simplified explanation:
If I remember correctly, the Reverse Transfer Impedance measurement determines how much signal will be induced into the center conductor, when an external field impinges on the coax.
That field is non contacting, and is away from the body of the coax.
"Grounds are Commonly Misunderstood" - Me
That's what I was trying to say, without sounding smug (when I'm often very wrong)
In my amateur experience of 2 amps, it almost always creates a ground loop, if we are building stereo amplifiers.
This could just be because I'm stupid though.
If the L and R ground are isolated throughout then it's moot
Last edited:
I always like jacks at the back and volume at the front of the chassis so interconnects can be hidden. Seeing coax and power cords and spealers entering anywhere other than the rear is weird looking to me. So a small diameter coax is necessary. Breaking whatever ground is coming in on the interconnect, I believe is good, who knows where that cable is coming from or if it has any ac potential different than my amps ground.. So I run the jack ground to the star and the volume ground to the same star which is established as close to the grid as I can get it. It seems to break any hum loops caused by the source device. Basically I stop the interconnects ground at the front door and hand it all over to grounds I can control. Does this sound like a good approach for a wide variety of unknown sources?
My house power outlets are 3-wire with ground.
My signal sources, CD players and Tuners have 2-wire power cords.
My Turntable and Outboard Phono Preamp also have 2-wire power cords, but the turntable has a ground wire that goes to the outboard phono preamp.
I do not use any other preamp.
All my power amplifiers have 3-wire power cords. I use the ground, connect it to the chassis at the IEC socket in the amp.
That plus my info in post # 4 works for me.
Take a look at many of Tektronix early tube oscilloscopes, plug-ins, and other products. They often have a knob on the front panel, and then a coupling, a rod, and another coupling that goes to a Potentiometer on the rear panel.
That technique will allow you to use a knob on the front panel of your amp, but at the same time have inputs and volume controls on the back of your amp.
My signal sources, CD players and Tuners have 2-wire power cords.
My Turntable and Outboard Phono Preamp also have 2-wire power cords, but the turntable has a ground wire that goes to the outboard phono preamp.
I do not use any other preamp.
All my power amplifiers have 3-wire power cords. I use the ground, connect it to the chassis at the IEC socket in the amp.
That plus my info in post # 4 works for me.
Take a look at many of Tektronix early tube oscilloscopes, plug-ins, and other products. They often have a knob on the front panel, and then a coupling, a rod, and another coupling that goes to a Potentiometer on the rear panel.
That technique will allow you to use a knob on the front panel of your amp, but at the same time have inputs and volume controls on the back of your amp.
Last edited:
No, on the contrary it looks like a good way to encourage hum. You need to be aware that the incoming signal is not on the signal wire; it is the potential difference between the signal wire and the return wire. Your voltage reference point for source X is whatever X uses as a ground, which it has conveniently sent to you on the cable shield. Far from 'breaking' this ground you should use this ground.Windcrest77 said:
I thought of putting the pot in the rear as well as one of those Chinese source switching boards for 10 bucks. However I'd still want the input tube at the front so I'd run shielded up to the grid once from all that stuff in the back I guess.
Shaft Coupler - 1/4" to 6mm - ROB-12211 - SparkFun Electronics
AIYIMA Audio Stereo Two Channel Relay 4 Way Sound Source Selection Audio Switch Input Selection Board-in Amplifier from Consumer Electronics on Aliexpress.com | Alibaba Group
Thanks, that's correct I don't know what I was thinking. All this is still on the breadboard where I cant really simulate a metal chassis.
I do know that on my Schitt DAC I have to plug the DAC unit into a ground lift adapter before it will not hum when plugged into a tube amp. Its dead quiet with no AC ground, Schitt even tells you to do that in the manual if its humming. Same goes if I want to take a signal out of my AV receiver pre-out into a tube amp, I have to un-ground one or the other. I'm just thinking of ways and where to break these potentials coming in on an interconnect.
Thanks, thats what I'm finding, any source signal equipment I have works fine if it has 2 wire power or if it does have 3 wire power I just use a ground lift adapter and its fine. My AV receiver is big and has a 3 prong, I'm hesitant to unground that because it has 11 power amps in it, so I can rarely use the pre-out on the AV receiver into a tube amp, or any other power amp, it always hums. But you're right 2 prong sources work fine into just about anything.
The success of running a shielded wire from a pot at the back, then all the way to the input tube that is up front, depends on ground loops, hum, and potential oscillation from the reactances.
It also depends on the capacitance of the shielded wire versus the impedance of the pot with the wiper at mid resistance. Also, with the wiper between mid and full, it also depends on the the driving impedance of the signal source. High frequency roll off is the potential issue here.
I remember some very low capacitance coax from years ago. It was used for AM whip antennas on cars. It was very stiff, with quite a large diameter. It had an insulating outer jacket, shield, center insulator, and center conductor, similar to other coax. But the usual solid center insulator was not employed. Instead, the center conductor actually had lots of distance to the insulator, because the center insulator was a thin-walled tube, and so most of the dielectric from the shield to the center conductor was air (with a much lower dielectric constant). If I had to use a shielded wire from back to front, I might try to find some of this coax.
Years ago, I learned to test the high frequency response of my amplifiers with the volume pot at the -6dB point. That is the Mid-Resistance point. For a log or audio pot, it is not mid-rotation. Then, I would also test the amp high frequency response with the pot all the way up. I would compare the two results. I changed out some 100k pots, and put in 50k pots, when the particular amp's high frequency response needed that.
It also depends on the capacitance of the shielded wire versus the impedance of the pot with the wiper at mid resistance. Also, with the wiper between mid and full, it also depends on the the driving impedance of the signal source. High frequency roll off is the potential issue here.
I remember some very low capacitance coax from years ago. It was used for AM whip antennas on cars. It was very stiff, with quite a large diameter. It had an insulating outer jacket, shield, center insulator, and center conductor, similar to other coax. But the usual solid center insulator was not employed. Instead, the center conductor actually had lots of distance to the insulator, because the center insulator was a thin-walled tube, and so most of the dielectric from the shield to the center conductor was air (with a much lower dielectric constant). If I had to use a shielded wire from back to front, I might try to find some of this coax.
Years ago, I learned to test the high frequency response of my amplifiers with the volume pot at the -6dB point. That is the Mid-Resistance point. For a log or audio pot, it is not mid-rotation. Then, I would also test the amp high frequency response with the pot all the way up. I would compare the two results. I changed out some 100k pots, and put in 50k pots, when the particular amp's high frequency response needed that.
Last edited:
Thanks for all the inputs about cable capacitance, grounding, HF roll off, etc. associated with jacks on back volume on front chassis and the long cable length in-between... Especially if those cables have to pass basically across everything else in the chassis. I have decided to move the input tube to the back of the chassis! Radical I know, considering every amp out there puts that cute little input tube front and center. The powr xformer torroid will be front right. OPT's somewhere in the middle. Power tubes in front. I ordered a source selector relay board and a bunch of pot extension rods and bushings. All my input wiring should now be under two inches and the amp will have jacks on back, controls on front.
Aluminum 300mm Volume Potentiometer Extension Shaft Audio Connecting Long Rod | eBay
6 Channel Unbalanced Stereo or Balanced Mono Audio Input Selector Relay Module 6921407430846 | eBay
I have a 5 amp LM338 regulated DC filament supply board which will also power the selector relays.
5 Amps 1.5 to 32V Adjustable Voltage Regulator Module. 6921407452145 | eBay
Bringing the input tube to the rear will also open up some aesthetic possibilities of putting the output tubes up front.
The powr xformer torroid will be front right. OPT's somewhere in the middle. Power tubes in front. I ordered a source selector relay board and a bunch of pot extension rods and bushings. All my input wiring should now be under two inches and the amp will have jacks on back, controls on front.Aluminum 300mm Volume Potentiometer Extension Shaft Audio Connecting Long Rod | eBay
6 Channel Unbalanced Stereo or Balanced Mono Audio Input Selector Relay Module 6921407430846 | eBay
I have a 5 amp LM338 regulated DC filament supply board which will also power the selector relays.
5 Amps 1.5 to 32V Adjustable Voltage Regulator Module. 6921407452145 | eBay
Bringing the input tube to the rear will also open up some aesthetic possibilities of putting the output tubes up front.
For sure I will, this would be my first build since around 1975! Available time, knowledge (lack of), waiting on parts, my old employer calling me back to consult, wife activities, but mostly constantly revising and experimenting with things... all take time. Now I know why I'll see people posting back here a year later about an amp. This first one will just be a finished headphone amp. Subsequently I'll get more complex.
My latest revision is to possibly build the input tube stage all in its own little aluminum mini box, have the pot extension shaft couple to the pot on the minibox, have the input wire coming out to the relay source switcher through another grommet, bring the power in through another grommeted hole. Then mount that mini box as a shielded sub chassis with nylon spacers so it is isolated from the safety ground but is star grounded to my audio circuit ground. The tube will sit in a "well" out the amp top plate but the socket is on the minibox a half inch lower or so than the top plate. This gives me a way to hold down the remote pot, which I was originally just going to mount on an L bracket, So the whole minibox idea came about as a substitute for an L bracket with an Alps pot on it, I thought why not just build the whole stage in a minibox so its totally shielded? I'd like to try to get away with no grid stopper, I have a bag of ferrite beads that are welded to wire (they look like resistors but are just a ferrite fused on a wire). Also like to get away with a high value grid leak resistor. With the minibox I figure I can keep the wiring real close and tight in the first stage. I may even use a nylon shaft coupler to isolate the pot shaft extension rod!
Project "scope creep" for sure.
My latest revision is to possibly build the input tube stage all in its own little aluminum mini box, have the pot extension shaft couple to the pot on the minibox, have the input wire coming out to the relay source switcher through another grommet, bring the power in through another grommeted hole. Then mount that mini box as a shielded sub chassis with nylon spacers so it is isolated from the safety ground but is star grounded to my audio circuit ground. The tube will sit in a "well" out the amp top plate but the socket is on the minibox a half inch lower or so than the top plate. This gives me a way to hold down the remote pot, which I was originally just going to mount on an L bracket, So the whole minibox idea came about as a substitute for an L bracket with an Alps pot on it, I thought why not just build the whole stage in a minibox so its totally shielded? I'd like to try to get away with no grid stopper, I have a bag of ferrite beads that are welded to wire (they look like resistors but are just a ferrite fused on a wire). Also like to get away with a high value grid leak resistor. With the minibox I figure I can keep the wiring real close and tight in the first stage. I may even use a nylon shaft coupler to isolate the pot shaft extension rod!
Project "scope creep" for sure.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.