• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

grounding cable

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
hello,

I am working on a separate power supply for a tube headphone amp.
The power supply will have 150vdc and 2 6.3vdc lines. My question is do I run a ground for each DC line in the cable(Aphenol CONNECTOR) or just use one ground and take off from a common point in the amp? I would think with multiple grounds you will have ground loops and with one ground all the B+ current will mingle with the heater current.


signed
drowning in analysis
 
mingle??? Are you insane????

the two currents would never mingle - they barely talk! B+ is a MUCH higher class curent than lowly heater current - the heater current a mere servant to the B+ signal current.

No no no no no, have no fear, the two currents and their associated electrons will divide themselves based on income, accent and breeding, and apart from the occasional Romeo and Juliet-esque situaition, no mingling will occur.
 
Disabled Account
Joined 2006
I assume that your power supply contains more than just the transformer, and has rectifier and smoothing capacitors also. So...

First attach the incoming ground from the mains to your power supply chassis as a safety ground. Take an extension from that point through your umbilical to your headphone amp chassis, also as a safety ground. OK, so now both boxes are grounded for safety. You are not going to use this safety ground for any audio signal, so make sure no signals use the chassis for a return path. (For example, that means insulating the input sockets from the amp chassis).

Back to the power supply: from the safety ground point at the power supply chassis make a connection to the power supply capacitor(s) negative terminal. This is your power supply ground reference point. (Some people like to add a small resistor between the chassis safety ground and the power supply ground reference point to help break any loops... I use a 10R 7W resistor). Now run two wires from your power supply to your amp - one for your high tension, and one from the power supply ground reference point. Once inside your amp the wire from your power supply ground reference point magically becomes your audio signal ground reference point. Do not connect it to the chassis. Use it for all audio signals.

Now to the heaters. This will depend on how they are configured. Some heater circuits need to be referenced to a tapping from the high tension to place them at the right potential relative to the cathode, while other heaters can simply be directly grounded. You need to let us know more before we can advise. Best to scan and post a circuit diagram if possible.
 
hello,

I am working on a separate power supply for a tube headphone amp.
The power supply will have 150vdc and 2 6.3vdc lines. My question is do I run a ground for each DC line in the cable(Aphenol CONNECTOR) or just use one ground and take off from a common point in the amp? I would think with multiple grounds you will have ground loops and with one ground all the B+ current will mingle with the heater current.


signed
drowning in analysis

Each DC supply needs it's own "return line". In other words you make a "loop" with supply and return for each voltage. The return line is NOT a "ground" but inside your power supply there should be a common ground point. These return paths are isolated from ground except at this common point inside the supply.

Then for safety the ground connector from the power cord meeds to connect to the metal chassis and this safety ground should connect to the above "common ground".

Finally this power cable that supplies the positive voltages and the returns for each of them really, really should also contain a large diameter conductor for the safety ground which connects only to the metal chassis. A rule is that this conductors should be at least as large is any other conductor in the cable. (I'd make it one size larger than the wire used for the heaters)

Also your power supply needs to have a fuse on each voltage. The fuse needs to be on the supply end of the cable to protect the wiring in the cable from fire in case of a short in the amp.

If your high voltage power supply is not housed in the same enclosure and the cable runs unprotected, (something like the way RCA cables run in typical consumer hifi rigs) then it really needs to be inside a braided safety ground shield so that if it is ever damaged, bitten by a dog, rat or cat or handled by a small child there is some protection from electrocution. High voltage DC is VERY dangerous, much worse than AC mains power. In my opinion HT voltages needs to remain always inside some kind on grounded conductive enclosure, either a metal chassis, conduit or shielded cable. But not everyone does this.

You add a lot of complexity and safety considerations when you use a remote high voltage power supply, most commercial designs avoid this because of the high cost. But don't skimp. Think about failure modes (a cap explodes and tosses conductive metal fragments all over and then you have 300 volts shorted to a pair of headphones on your head.) It's harmless if the above guidelines are followed, really bad if redundant safety systems were not in place.
 
Last edited:
excelent stuff

Great reply s , thank you. Looks like there will be 3 DC voltage lines. A B+ =150v
and 2) 6.3V DC lines for the heaters. I f i understand correctly I would ground each filter cap from each DC supply to a common point in the chassis of the supply . These grounds will not connect to chassis ground in the amp. kinda like a star ground in a regular amp. I appreciate the safety advice, when you're dreaming of Dark Side of the Moon on the HD 600's You kinda neglect the fact that you have "300V on you're head"


Thanks again , might be time to buy some Iron
 
Disabled Account
Joined 2006
Great reply s , thank you. Looks like there will be 3 DC voltage lines. A B+ =150v
and 2) 6.3V DC lines for the heaters. I f i understand correctly I would ground each filter cap from each DC supply to a common point in the chassis of the supply . These grounds will not connect to chassis ground in the amp. kinda like a star ground in a regular amp.

The safety points made by ChrisA are very sensible.

Regarding heaters; yes, but only if they are not floating / offset at a high voltage (because of heater - cathode ratings). Even after grounding in the power supply I would still tend to run separate out and return lines for each of the heater supplies. In the amp chassis they would touch nothing else except the tube pins.

So, it looks as if your umbilical will have a grounded overall braid and the following connector to connector wiring:
1. Safety ground from power supply chassis to amp chassis
2. Ground from power supply ground reference point (at filter caps) to audio ground in amp
3. High voltage DC supply
4. Heater circuit 1, 6.3V out
5. Heater circuit 1, return
6. Heater circuit 2, 6.3V out
7. Heater circuit 2, return


I appreciate the safety advice, when you're dreaming of Dark Side of the Moon on the HD 600's You kinda neglect the fact that you have "300V on you're head"


That's a very good reason for using a transformer coupled amp!
 
Gordy's advice will get you most of the way there.

I second the idea of using an RF choke between the power supply star ground and the green (or yellow/green in EU) wire safety ground. In a Parasound A23 (sand amp) I took apart, they used a simple ferrite bead style inductor. So a 0.5 sq mm (about 16-18 AWG) bare wire through a 10 mm x 4 mm ferrite sleeve.

The important thing with grounding is to keep track of the return paths for the critical signals. A return path for a critical signal should never share the same path as a high current load. The reason for this is fairly straight forward: The amp will amplify the difference between the input pin and the ground reference for the amp input. If a large current not associated with the signal runs in the conductor providing the reference ground for the amp, it will create a voltage drop in the reference ground conductor. From the amp's point of view, this voltage drop will be in series with the input signal. Thus, any hum or variation in the large current will be amplified by the amp. This is probably not what you want...

If the HT ground is reasonably quiet, you can probably use it as a reference ground, but beware of potential feedback loops. I'd probably opt for a configuration like this:

HT source
HT return
Heater 1 source
Heater 1 return
Heater 2 source
Heater 2 return
Reference (signal) ground
Safety ground

Join the HT return and reference ground at the negative lead of your power supply cap.
Connect the cable safety ground to the power supply chassis (same place as the input power safety ground connects) and run an RF choke from here to the negative lead of your supply cap (star ground).

If you use a multi-conductor cable make sure that it is rated for the HT voltage. Also note that you can potentially have quite a bit of voltage drop across the heater wires. But of course, you can always increase cross sectional area by connecting multiple wires in parallel. If you run multi-conductor cable, I would strongly suggest using DC for the heaters as you'll get quite tight coupling between the wires in the cable. You probably don't want your signal reference ground or HT sources modulated at 50/60 Hz.

Power supply cords are non-trivial if you want high performance.

BTW, a good text on grounding and shielding is: Ralph Morrison "Grounding and Shielding Techniques".

~Tom
 
Disabled Account
Joined 2006
Is there any preference to which chassis the final B+ PS filter cap resides in?

I don't know the circuit to which you alude, however in an ideal world I would do it like this...

Power supply chassis:
  • mains input filter
  • fuse
  • safety ground
  • power switch with parallel RC snubber
  • inrush current limiter
  • varistor
  • transformer (with RC across primary)
  • rectifier (tube or hybrid)
  • RCRC (or your current favourite) smoothing
  • power supply ground reference point
  • ground loop breaker from safety ground to power supply ground reference point
  • fuse on each output
  • umbilical output connector (FEMALE / sockets on chassis)
  • (I might also add something to deal with DC offset on the mains, but I've not enough experience with that yet to make a difinitive recommendation)


Amp chassis:
  • umbilical input connector (MALE / pins on chassis)
  • capacitor
  • voltage regulator
  • parallel combination of capacitor and capacitor + low value resistor (...but refer to regulator designer's notes for application details!)
  • amplification circuit


I've ignored the heaters in the above in order to emphasise the high tension circuit.
I've probably missed something out, so hopefully our 'core faithful' can fill in the gaps and suggest extras and alternatives...

: )
 
Disabled Account
Joined 2006
Excellent thread!! Thanks for the very concise answers/recommendations. I've been pondering these very issues for a two chassis Tubelab SE build. Is there any preference to which chassis the final B+ PS filter cap resides in?

Following on directly from your question, I've just looked at http://www.tubelab.com/TubelabSE.htm ...

If doing this dual chassis I would get the last power supply cap (C5 in the cct diagram) as close to the output transformer as pratical.

If you can afford the (relatively minor) extra cost then consider placing a 100 micro F in the power supply chassis and another (actually one in each channel) in the amp chassis right next to the output transformers.
(Give George /tubelab a quick buzz to confirm).
 
Following on directly from your question, I've just looked at http://www.tubelab.com/TubelabSE.htm ...

If doing this dual chassis I would get the last power supply cap (C5 in the cct diagram) as close to the output transformer as pratical.

If you can afford the (relatively minor) extra cost then consider placing a 100 micro F in the power supply chassis and another (actually one in each channel) in the amp chassis right next to the output transformers.
(Give George /tubelab a quick buzz to confirm).

Great, thanks. Yes the location of C5 is exactly what I was asking about. I'm Doing a CLC with a 100u motor run cap for C5.
 
Is there any preference to which chassis the final B+ PS filter cap resides in?

You ALWAYS want some capacitance on the supply input(s) to the amp to avoid - or at least reduce the risk of oscillations. It also helps in keeping the output impedance of the power supply low as 'seen' from the amp. I use 10-100 uF as my supply input cap. The exact value depends on what I have in my parts collection.

I would not put all the supply capacitance in the amp chassis. You do need some (preferably the majority) of the capacitance to reside in the supply chassis. Otherwise, you'll get the nasty ripple current to run through the supply cord, thus, providing a real good source of interference.

~Tom
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.