Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: mmack on August 01, 2023, 07:49:28 pm
-
I don't know why this has been bothering me, but I noticed that in many amp builds it is popular to "split" the preamp & poweramp grounds by grounding the preamp at the input jack and the poweramp near the first filter cap. After some study, I understand that, while this isn't the theoretical "perfect" grounding setup, it should work with little noise (at least for established designs such as Rob's 5f1 build which is what I'm hoping to replicate with my champ kit: https://robrobinette.com/How_Amps_Work.htm)
One thing that is bothering me - isn't some of the circuit going to be sinking down the earth ground point? From what I understand, this is likely to happen since the preamp tube cathodes are all grounded to the chassis - sure, the power amp ground is "closer" but from what I understand since the chassis ground is a big hunk of metal the current will go in all directions outward from the input jack ground point, causing at least some of this to go to the earth/safety ground.
From what I understand, this is bad - most everywhere I have read says the earth ground should have almost no current in normal conditions. I'm also wondering if this could cause issues with GFCI since I know they are quite sensitive.
I don't see any theoretical reasons that the earth ground wouldn't be sinking some of that current, since safety ground & neutral are bonded at the panel and theoretically there shouldn't be a whole lot of difference between them.
Any clarification would be helpful. I'm sure I'm just misunderstanding something simple here.
EDIT: I figured this out. In essence, I found the answer by re-reading Rob Robinette's "How amps work" article: https://robrobinette.com/How_Amps_Work.htm (https://robrobinette.com/How_Amps_Work.htm). From what I understand, the ground is connected to the power transformer's center tap and that is where the current flows to - not earth ground or the neutral wire.
-
Hi mmack.
-I'm not sure what you mean by "sinking down the earth ground".
A few things to consider and in no specific order.
-Signal (preamp) ground, chassis ground and earth ground are not all the same, but they can be shared/combined.
-Ask yourself what is the potential between these grounds. Does signal ground want to go to earth ground? Probably not.
-The "chassis to earth ground" is for safety and shielding.
-Signal ground to chassis ground is most often used to complete the circuit and proper locations (do to potential/resistance) can go a long way to preventing ground loops/noise.
-In your amp, try measuring voltage with your DVM between the (preamp ground) and (power amp ground). Get any thing?
-I've never tripped a gfi with signal from an amp, I have tripped gfi with direct line voltage (120V) shorts and overloaded circuits.
-
Welcome. :icon_biggrin:
I don't get it either. :dontknow:
DRAW THE WHOLE CIRCUIT LOOP!! "Current" does not go-away like wine spilled on the beach. It must flow in a loop.
"Ground" is a poor word with too many meanings. Signal return, power return, buzz return, radio interference.
The preamp tubes run at 1mA which is not enough to trip a GFI even if ALL the tube current went through the GFI (which it can't). GFIs should not trip on 120V shorts since equal/opposite current is taken in both legs (yes, older GFIs were prone to be marginal on this spec; why I don't run reefers or furnaces on GFI anymore).
Also standard GFIs do not sense DC like tubes use: the cancellation is done in a magnetic core.
-
I have drawn & analyzed the circuits over and over. I tried thinking about this but couldn't figure it out hence this post.
The main issue I have is the neutral & ground should be at the same potential, hence the voltage will want to go to either of these indiscriminately. I don't see any reason for it to "prefer" the neutral ground point over the earth/safety ground point.
I guess the answer is both these ground points are the same. I suppose that makes sense that it wouldn't trip a GFCI since the current is quite low.
-
The main issue I have is the neutral & ground should be at the same potential, hence the voltage will want to go to either of these indiscriminately. I don't see any reason for it to "prefer" the neutral ground point over the earth/safety ground point.
The neutral is a current carrying conductor. The earth ground is not. These two are bonded together in your power entrance distribution box but nowhere else. The only thing inside your amp that should be connected to neutral is one of the PT primary wires.
-
I understand the safety ground isn't current carrying, I'm just trying to understand why it isn't carrying current. From what I understand neutral & ground are seen as the same potential so I don't see why the current would "prefer" to go to neutral.
Perhaps the answer is because, like you are saying, the amp is grounded at the power transformer secondary and not the house's electrical system's neutral wire?
These two are bonded together in your power entrance distribution box but nowhere else.
That's where the neutral is tied back to the power grid's transformer too from what I understand, so I'm not sure this helps in my understanding.
-
I understand the safety ground isn't current carrying, I'm just trying to understand why it isn't carrying current. From what I understand neutral & ground are seen as the same potential so I don't see why the current would "prefer" to go to neutral.
The green safety ground wire was never designed to carry any current. You will never see that wire tied into any circuitry. Therefore it can never carry any current. ***UNTIL*** something shorts to the green wire (chassis in most cases)! Then it safely carries a dangerous current away from you and back to the main distribution box and on to a ground rod trips a circuit breaker. That's why you don't get the shit knocked out of you if the 120V hot wire shorts to chassis.
Edit... Don't know why I mentioned ground rod. That's not true.
The normal current doesn't prefer to travel down the neutral wire. It has no choice! Think about it. The black hot wire brings voltage into the amp through a fuse and switch until it reaches it's destination which is the PT and sometimes an indicator lamp. Now this voltage needs a return path back to the source, otherwise no current can flow, same as turning the power switch off. The neutral wire also connects to the PT and it's job is to complete the path back to the source. Black and White are the only two wires that ever connect to the circuit. The green simply connects to the chassis for your protection should a fault occur. Green is not part of the operational circuit. You should not think of the ground and neutral as being interchangeable. They each have their own specific purpose. It's true, theoretically, that they are usually at the same voltage potential because they are tied together at the main entrance. However, in the real world you can often measure millivolts and sometimes even volts between ground and neutral simply because of the length of wire in a typical house. And sometimes you may have a loose connection somewhere in the house wiring that makes this even worse.
-
***UNTIL*** something shorts to the green wire (chassis in most cases)! Then it safely carries a dangerous current away from you and back to the main distribution box and on to a ground rod. That's why you don't get the shit knocked out of you if the 120V hot wire shorts to chassis.
That's what I'm confused about, since the amp design in the OP ties the preamp ground to the chassis (which is also connected to the earth / safety ground). Hence, the fact that there may be some current that can flow from preamp ground to safety / earth ground.
The normal current doesn't prefer to travel down the neutral wire. It has no choice! Think about it. The black hot wire brings voltage into the amp through a fuse and switch until it reaches it's destination which is the PT and sometimes an indicator lamp. Now this voltage needs a return path back to the source, otherwise no current can flow, same as turning the power switch off. The neutral wire also connects to the PT and it's job is to complete the path back to the source. Black and White are the only two wires that ever connect to the circuit. The green simply connects to the chassis for your protection should a fault occur. Green is not part of the operational circuit. You should not think of the ground and neutral as being interchangeable. They each have their own specific purpose. It's true, theoretically, that they are usually at the same voltage potential because they are tied together at the main entrance. However, in the real world you can often measure millivolts and sometimes even volts between ground and neutral simply because of the length of wire in a typical house. And sometimes you may have a loose connection somewhere in the house wiring that makes this even worse.
I think I'm starting to understand. I can see how the primary +120v wants to return to the source via the neutral wire connected to the negative side of the primary. So, the secondary positive voltage (i.e. the source voltage for the rectifier) then wants to return back to the "source" of the negative voltage on the secondary?
So in essence, if you were to tie the neutral to the chassis (without the power transformer) it would result in the issue I was alluding to in my OP, from my understanding. Which is obviously not a good idea.
-
So in essence, if you were to tie the neutral to the chassis
Go ahead. I'll read about it in the Darwin Awards. :l2:
-
> can never carry any current. ***UNTIL*** something shorts to the green wire (chassis in most cases)! Then it safely carries a dangerous current away from you and back to the main distribution box and on to a ground rod. That's why you don't get the shit knocked out of you if the 120V hot wire shorts to chassis.
No. Nothing to do with dirt rods.
The "120V Hot wire" comes from the power company. Therefore when it gets loose and hits chassis, it returns TO power company.
You can't keep electrons. Only use them and return them to where you got them.
(Actually there are no electrons (https://www.amazon.com/There-Are-Electrons-Electronics-Earthlings/dp/0962781592) but the truth is unintuitive, stick with the fiction.)
If your amplifier gets hit by Lightning, then you expect current return to a dirt-rod. You may have dirt current if the 20,000V line falls on the 120V line, but that's complicated by company policy.
Your guitar pickup makes electric power! And the return is the shield on the cable and plugs.
-
Maybe you're thinking of an amplifier without a power transformer.
-
No. Nothing to do with dirt rods.
Of course. Don't know what I was thinking when I said that. I corrected my original post.
-
....Nothing to do with dirt rods.....
Just got this YouTube thrown at me. I have not watched all 20 minutes, but the opening scene gets you thinking.
Where Does Grounded Electricity Actually Go?
by Practical Engineering (https://www.youtube.com/watch?v=jduDyF2Zwd8)
-
Cool video.
I think I figured this out. In essence, I found the answer by re-reading Rob Robinette's "How amps work" article: https://robrobinette.com/How_Amps_Work.htm (https://robrobinette.com/How_Amps_Work.htm). From what I understand, the ground is connected to the power transformer's center tap and that is where the current flows to - not earth ground or the neutral wire (at least not "directly" but via the inductor).
If there's any misunderstanding here please enlighten me.
I'd guess this is how "isolation power transformers" work.
-
I don't follow splitting the power amp and preamp. I've read that in fact that's a bad idea or at least similar to what you mentioned it introduces more interactions with other points on the chassis. Like star grounding which is not recommended because it's like a random ground where points in the circuit interact randomly. I always prefer and always ground one point of the circuit to the chassis. And only one. Power and preamp. Which is tied along a thick (12awg) buss wire. Then the mains safety ground tied near the power entry point.
The circuit to my understanding and from what I've read would be best implemented like a flow or in sequence. With all respective grounds stages filtered separately and tied to the ground of the filter caps of their respective plate. So basically all grounds for that stage would be tied together to the plate filter cap of that stage only like a node.
From there each of these nodes run down the "stream" which would be the buss in sequence. This I would imagine would be ideal. But it is hard with some clone builds unless you customize the board layout (pcb or turret). I don't think Marshall or Fender cared too much about this. Because it was more economical and cost effective to build amps the way they did.
Seperating the power amp and the preamp makes sense. But I would think that would introduce another ground point which would interact again randomly.
I've used one point of the circuit to ground only many times and have not had issues with noise. Even on higher gain amplifiers with extra gain stages. So I think it works. At least for me.
-
I don't follow splitting the power amp and preamp. I've read that in fact that's a bad idea or at least similar to what you mentioned it introduces more interactions with other points on the chassis. Like star grounding which is not recommended because it's like a random ground where points in the circuit interact randomly. I always prefer and always ground one point of the circuit to the chassis. And only one. Power and preamp. Which is tied along a thick (12awg) buss wire. Then the mains safety ground tied near the power entry point.
The circuit to my understanding and from what I've read would be best implemented like a flow or in sequence. With all respective grounds stages filtered separately and tied to the ground of the filter caps of their respective plate. So basically all grounds for that stage would be tied together to the plate filter cap of that stage only like a node.
From there each of these nodes run down the "stream" which would be the buss in sequence. This I would imagine would be ideal. But it is hard with some clone builds unless you customize the board layout (pcb or turret). I don't think Marshall or Fender cared too much about this. Because it was more economical and cost effective to build amps the way they did.
Seperating the power amp and the preamp makes sense. But I would think that would introduce another ground point which would interact again randomly.
I've used one point of the circuit to ground only many times and have not had issues with noise. Even on higher gain amplifiers with extra gain stages. So I think it works. At least for me.
I understand, and can see how merlin's or aiken amps grounding strategies would be much preferable. From what I've read though with an amp like the 5f1 the noise in Rob's circuit is very minimal there's not a huge amount to be gained by going to 1 single ground.
I was more focused on understanding the flow of the circuit, and didn't want to turn this into a "perfect grounding" vs. separate grounding thread.