Hi!
I have build a PP KT88. Works greate but there is one problem.
When i have the rca cabels connected there is a small hum in the speakers (ground loop?), but's not a big problem. The hum is not hearable from 10 feet from the speakers.
The problem is when there is no rca cabel connected to the rca jacks there is a loud ground hum. The input jacks are grounded and the cable are shielded..
Anyone know what to do?
Thanks // Daniel
I have build a PP KT88. Works greate but there is one problem.
When i have the rca cabels connected there is a small hum in the speakers (ground loop?), but's not a big problem. The hum is not hearable from 10 feet from the speakers.
The problem is when there is no rca cabel connected to the rca jacks there is a loud ground hum. The input jacks are grounded and the cable are shielded..
Anyone know what to do?
Thanks // Daniel
It depends on the kind of hum. There are basically three kinds.
1. Power supply noise, due to poor filtering. At double the power line frequency, a sort of slightly raucous sound.
2. Power line leakage, a smooth clean hum that is at the power line frequency.
3. Pickup, also known as open grid hum, which is really raucous and often includes local radio station modulation. It has high pitch overtones that dominate.
Knowing what kind you have will help track it down. First, if you still hear it with the cables disconnected, try shorting the input jacks. If you still hear it, the problem is in the amplifier. However, if this amplifier is new, I presume it always had the hum, so it may be the way it was assembled.
1. Power supply noise, due to poor filtering. At double the power line frequency, a sort of slightly raucous sound.
2. Power line leakage, a smooth clean hum that is at the power line frequency.
3. Pickup, also known as open grid hum, which is really raucous and often includes local radio station modulation. It has high pitch overtones that dominate.
Knowing what kind you have will help track it down. First, if you still hear it with the cables disconnected, try shorting the input jacks. If you still hear it, the problem is in the amplifier. However, if this amplifier is new, I presume it always had the hum, so it may be the way it was assembled.
Yeah, short the inputs and see if it gets quiet.
If it does, then you might have a look at how the front end wires and jacks are shielded.
Also, what is the input impedance? I lowered mine to 10 kΩ to help that.
If shorting the input does not help, then you probably have a ground loop in the amp.
Generally, ground loop hum gets louder when you hook up other equipment.
If it does, then you might have a look at how the front end wires and jacks are shielded.
Also, what is the input impedance? I lowered mine to 10 kΩ to help that.
If shorting the input does not help, then you probably have a ground loop in the amp.
Generally, ground loop hum gets louder when you hook up other equipment.
Bingo!
That's the problem. Think about what you have when the input jack is unplugged!
You have a center conductor running all the way to the isolated jack, but where is the shield connected?
It is floating. So, it does not act as a shield. You have a large AC antenna for the input circuit.
Ground the shield at the input side of the circuit and try it again. I will bet it will be quiet as a mouse.
Once you confirm that you can lift it at the jack and it should be good to go.
Can I offer an alternative?
Grounding the shields at the input sockets only is perfectly valid.
The input pins at the amplifier compare the voltage at the two pins and amplifies the difference.
If the ground pins are connected to the sheilds then there are two routes to the signal ground. Two routes must form a loop. This loop will pick up interference and that in turn generates a voltage around the loop.
The two hot input pins will read the voltage differences between the two shields and at least one of these input pins will get that interference signal amplified.
Why?
If the input jacks are isolated from the chassis and nothing is connected to the input jacks, then the shield is not grounded to anything!
Than means the shield is simply floating.
All that changes when you plug something into the jack and the shield becomes an extension of the cable from the preamp.
However, the original post stated that he got lots of hum when the input was unplugged from the source. His shield was then disconnected from everything and that is why all the hum.
It may be perfectly valid if the input jacks are grounded to the chassis (no isolation) and the shield only connects to the jack. However, he isolated jacks.
the two input socket signal grounds (=signal returns) must be connected to the Signal Ground if provided on the PCB, or if not provided on the PCB, then taken to the main Audio Ground. The route of this link should closely follow the signal route to minimise loop area. That's why I asked the questions earlier.
Good questions, but in this case he gets hum with no inputs connected, which can be explained by the fact that the shield is disconnected at both ends.
I have found that you can get a loop between channels. I tie the signal grounds of the left and right channel at the RCA inputs. If your shielded cable only has the shield connected at the input RCA, how are you connecting the signal ground to your circuit? Do you have a shielded cable with two central conductors with one carrying the signal and the other the signal ground? If you have a single core shielded cable, you would use the central conductor as the signal and connect the shield at both ends, one to the RCA ground and one to (typically) the ground side of the first stage grid leak resistor.
If your input wiring is a single conductor with shield, there's no point isolating the input jack and then grounding the shield at the chassis at the input.
The point of isolating the input jack is so you can terminate the shield, which is also the input signal "return", to the star ground along with all the other grounds in the amp. Connecting the grounds this way should be the quietest.
Wherever you terminate your single end signal shield/return to ground is where the music signal will have to go... If you don't terminate the shield/return inside the amp, then the music signal takes the long path through your house safety ground.
If it's a balanced input with +,- and shield, then the shield should terminate at one end only. With a balanced input the shield carries no signal (hopefully...)
Cheers,
Michael
The point of isolating the input jack is so you can terminate the shield, which is also the input signal "return", to the star ground along with all the other grounds in the amp. Connecting the grounds this way should be the quietest.
Wherever you terminate your single end signal shield/return to ground is where the music signal will have to go... If you don't terminate the shield/return inside the amp, then the music signal takes the long path through your house safety ground.
If it's a balanced input with +,- and shield, then the shield should terminate at one end only. With a balanced input the shield carries no signal (hopefully...)
Cheers,
Michael
Last edited:
Since all the expertise are here, I will thread jack a little.
My problem is kind of the opposite. When I have nothing plugged in the input RCA's jacks, there is no hum. But once I plug in the turntable (yes, this is a phono stage) or simply short the input with a shorting plug, I get a hum which is highly likely caused by a ground loop. This happens even I isolate to only one channel. My grounding look like:
1. RCA female jacks are isolated from the chassis.
2. The RCA grounds are tied together and wired to the star ground of the whole circuit.
3. One 12AX7 per channel. When I had each channel connected to one local star
ground which was connected to the circuit star ground, the hum was loud. I then
split the grounds of the 2 sections and wired them to the circuit star ground
individually, the hum was lower considerably but still very audible.
4. The shielded wire from RCA input is connected to the junction of 47K to ground and
a 22K grid stopper. The shield is only connected on one side (RCA end) for
shielding purpose.
I am kind of exhausted what to look for, any suggestion will be greatly appreciated.
My problem is kind of the opposite. When I have nothing plugged in the input RCA's jacks, there is no hum. But once I plug in the turntable (yes, this is a phono stage) or simply short the input with a shorting plug, I get a hum which is highly likely caused by a ground loop. This happens even I isolate to only one channel. My grounding look like:
1. RCA female jacks are isolated from the chassis.
2. The RCA grounds are tied together and wired to the star ground of the whole circuit.
3. One 12AX7 per channel. When I had each channel connected to one local star
ground which was connected to the circuit star ground, the hum was loud. I then
split the grounds of the 2 sections and wired them to the circuit star ground
individually, the hum was lower considerably but still very audible.
4. The shielded wire from RCA input is connected to the junction of 47K to ground and
a 22K grid stopper. The shield is only connected on one side (RCA end) for
shielding purpose.
I am kind of exhausted what to look for, any suggestion will be greatly appreciated.
This thread seems to be causing some confusion!
I'm assuming NO balanced XLR inputs - all unbalanced RCA.
You need:
1) Signal
2) Ground
3) Shield
Often 2 and 3 are the same, but you always need 1 and 2, 3 is optional
The signal is then formed between 1 and 2, miss any one of these and you have loads of hum.
Note: Sometimes soldered ground on RCA sockets do not connect - occasionally you need to file down to new base metal and resolder, you can find this fault by bridging things that should be connected with wire.
A turntable arm lead should be twin screened cables (signal + ground) with a wire connected to the metal of the arm tube for the shield - which goes to the pre-amp ground.
Yes, indeed it is unbalanced RCA.
I use shielded input wire just to eliminate the RF noise creeping into the input as it is phono stage which can really amplify te input. From what I have read, shielding should be grounded on one side which is what I did.
The input grounds are connected together and then is connected to the star ground which is connected to the chassis ground via a loop breaker.
Since I have the hum with the inputs shorted, i.e no turntable is connected, I conclude that the ground loop is within the circuit. FWIW, this is part of an integrated amp that I just rebuilt.The line stage works perfectly, but I have been struggle to trouble shoot the source of the ground loop. I am kind of exhausted what to try ...
Thanks!!
I use shielded input wire just to eliminate the RF noise creeping into the input as it is phono stage which can really amplify te input. From what I have read, shielding should be grounded on one side which is what I did.
The input grounds are connected together and then is connected to the star ground which is connected to the chassis ground via a loop breaker.
Since I have the hum with the inputs shorted, i.e no turntable is connected, I conclude that the ground loop is within the circuit. FWIW, this is part of an integrated amp that I just rebuilt.The line stage works perfectly, but I have been struggle to trouble shoot the source of the ground loop. I am kind of exhausted what to try ...
Thanks!!
We should discount the cable then!
Is it the same on each channel?
Does it still do it without the input tubes?
Can you confirm the RCA ground goes to the phono PCB ground?
Are the input capacitors working?
Fred,
I can't quite know the topology of the circuit, well-enough, from your descriptions. So maybe nothing below is relevant. But...
It might not be a ground loop. WHERE are the input grounds connected together? The input ground conductors should each run as closely as possible to their corresponding signal conductors, for as long as possible, to minimize any loop area formed by each signal ground conductor and its signal conductor (see Faraday's Law, Maxwell's Equations, re: currents induced in loops by changing magnetic fields). The same goes for all power/ground (and other) pairs. Do you have all such wire pairs twisted tightly together, for as long as possible? And are all small signal and small signal ground conductors as far away as possible from all power and large signal conductors?
So, IF you have tied the signal grounds together at the input jacks (you said RCA grounds were tied together), and then run that to the star ground, you might have created relatively large loops with each signal and signal ground pair. Those loops could easily have currents induced in them from any AC fields in the vicinity, which could result in AC-frequency voltages being induced at your amplifying devices' inputs.
And, if the signal grounds go from the RCA jack grounds to the star ground, but each ALSO continues to the vicinity of an amplifying device's input, do they then also go to the star ground from there? Or, if not, are the devices' inputs' ground references forced to be looking all the way back through the RCA jack grounds and then from there on to the star ground? I guess I would tend to want to run each amplifying device's input ground reference to the star ground from very near the device (and very near the signal input's connection to the device), by itself.
Keep in mind that all conductors have distributed inductance and resistance. Any current in them will induce a voltage back at the non-ground end. And time-varying currents, even of very small amplitude, can tend to induce relatively-high voltages, since the voltage across an (ideal) inductance is proportional to the time rate-of-change of the current through it, regardless of the current's amplitude. That's one great reason for not letting ground return currents share the same conductor, i.e. star grounding, which is especially good for not getting a "bouncing" ground voltage as the ground reference for an amplifying device's input. But, also, in the case of loops formed between signal and ground conductor pairs, if an AC field induces a tiny AC current in the loop, then even a wire (that is part of the loop) that goes directly to ground can have an AC voltage induced at its non-ground end. (That could also happen in a conductor loop that eventually connects to ground on both ends, and could be a problem if some of the induced voltage happens to appear at the ground reference point for an amplifying device's input.)
Sorry to have blathered-on for so long about all of that. And, as I said, I don't know if I understood how you have things connected. But, even if I didn't, maybe something I said will jog an idea for you.
Cheers,
Tom Gootee
I can't quite know the topology of the circuit, well-enough, from your descriptions. So maybe nothing below is relevant. But...
It might not be a ground loop. WHERE are the input grounds connected together? The input ground conductors should each run as closely as possible to their corresponding signal conductors, for as long as possible, to minimize any loop area formed by each signal ground conductor and its signal conductor (see Faraday's Law, Maxwell's Equations, re: currents induced in loops by changing magnetic fields). The same goes for all power/ground (and other) pairs. Do you have all such wire pairs twisted tightly together, for as long as possible? And are all small signal and small signal ground conductors as far away as possible from all power and large signal conductors?
So, IF you have tied the signal grounds together at the input jacks (you said RCA grounds were tied together), and then run that to the star ground, you might have created relatively large loops with each signal and signal ground pair. Those loops could easily have currents induced in them from any AC fields in the vicinity, which could result in AC-frequency voltages being induced at your amplifying devices' inputs.
And, if the signal grounds go from the RCA jack grounds to the star ground, but each ALSO continues to the vicinity of an amplifying device's input, do they then also go to the star ground from there? Or, if not, are the devices' inputs' ground references forced to be looking all the way back through the RCA jack grounds and then from there on to the star ground? I guess I would tend to want to run each amplifying device's input ground reference to the star ground from very near the device (and very near the signal input's connection to the device), by itself.
Keep in mind that all conductors have distributed inductance and resistance. Any current in them will induce a voltage back at the non-ground end. And time-varying currents, even of very small amplitude, can tend to induce relatively-high voltages, since the voltage across an (ideal) inductance is proportional to the time rate-of-change of the current through it, regardless of the current's amplitude. That's one great reason for not letting ground return currents share the same conductor, i.e. star grounding, which is especially good for not getting a "bouncing" ground voltage as the ground reference for an amplifying device's input. But, also, in the case of loops formed between signal and ground conductor pairs, if an AC field induces a tiny AC current in the loop, then even a wire (that is part of the loop) that goes directly to ground can have an AC voltage induced at its non-ground end. (That could also happen in a conductor loop that eventually connects to ground on both ends, and could be a problem if some of the induced voltage happens to appear at the ground reference point for an amplifying device's input.)
Sorry to have blathered-on for so long about all of that. And, as I said, I don't know if I understood how you have things connected. But, even if I didn't, maybe something I said will jog an idea for you.
Cheers,
Tom Gootee
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.