Tube Amps / Music Electronics
|For current discussions, please visit Music Electronics Forum.|
|5/27/2000 10:52 PM|
||Star grounding - putting it into practice|
OK, I've read as much as I can find on star grounding, and I feel pretty comfortable with where to locate the star points.
Now, my question is : what about the routing of the actual ground wires themselves, and their proximity to other wires ( B+, AC, and signal carrying wires) ? What about running very long ground wires to star points, and crossing over, under, or parallel to B+ and filament wires, etc. ?
What are the rules of thumb for routing ground wires to the star point(s), theoretically and practically ?
Thanks and regards,
|And now, a word from our sponsors:
|5/27/2000 11:12 PM|
Maybe it is overkill but I consider the ground wires to be either signal return leads or voltage return leads, and treat them accordingly. Running leads parallel to each other can have a parasitic effect so I try to keep them separate and have them cross them at right angles whenever possible.
You might think that ground wires are "just" ground wires but they could have garbage on them which could interfere with the audio signal.
BTW for an amp with a PT CT connected to ground, I think that it is very important to run that CT to the initial PS filter cap negative return, and then run a wire from that point (isolated from the chassis) to the star ground point. I've been using 5 lug terminal strips for my star grounds, connecting all 5 lugs together and clamping the middle lug securely to the chassis using star washers. I run the PT CT/PS filter ground wire to the middle lug (which is connected to the chassis) and then run all of the various grounds to the 4 other lugs, pretty much following the signal path. It works for me!
|5/27/2000 11:43 PM|
On my Humble clone I used the stock Bassman board and built it up on that. Some of the grounds were combined but all the wiring was run straight up under the pots and then all the way to the end of the pots where I had terminal strip that everything was soldered to before going to the main ground point. I wanna say the wiring for the pots were stuffed under there also to some degree. I would have to look to be sure. But I got away with it, the amp is very quiet even in the distortion channel. Guess I got lucky, I'm sure Steve's got some valid views.
My 66 I just rebuilt, using the stock Fender ground points/wiring. That amp is also dead quiet.
Just for the record, I used metal films throughout both amps. I don't what this adds to this thread, but thought to mention two grounding examples in the same chassis.
|5/28/2000 1:03 AM|
FWIW, I run ALL of the wires in my amps so that no wire touches another. I try to keep them separated by at least 3/8" minumum...the further apart, the better. Where I have to cross a wire, it is done at 90degrees and the wires do not touch. I bend the top wire to form a bridge over the other one. Of course this wouldnt work unless you are using solid core wire. All of my amps are done this way and they are extremely quiet at all volume levels....I cant say for sure whether or not this is what makes them quiet because I've never tried it any other way, and I do other things in the circuit design intended to make the amps quiet.
|5/28/2000 2:05 PM|
In addition to the above, I have been running my first preamp stage ground directly to the signal input ground. That's at the opposite end of the amp as compared to the power supply. One of my earlier 'goes' at it had the first stage ground crossing over to the power supply ground, and there was tons of hum. Isoloating this reduced the noise considerably.
The other preamp and PI grounds have a common ground, and then the first cap on the power supply has its own ground at the power transformer.
Good luck . . . jrem.
|5/28/2000 5:17 PM|
||Re: Star grounding - putting it into practice|
One of the projects in the queue at GEO is an illustrated guide to grounding practice. This kind of ties into the article I'm about to post on the differences in point to point and PCB based amp construction and *exactly* what and how much the differences are - kind of a set-the-records-straight thing.
To answer some of your questions:
From the preparation I did for the point-to-point article, I calculated the coupling between two wires. For #22 jacketed wire, the capacitance between two wires running parallel with insulation touching is about 0.3pF per inch. It's resistance is about 25milliohms per inch.
So - take your length of parallel run and figure the capacitance and resistance, and that's the coupling. To B+, you can run them freely together with no ill effect - not enough voltage on the B+ line and not high enough frequency to couple well to the ground line.
For signal carrying wires, depends on the impedances and signal levels. A worst case might be for a 50Vac signal at the phase inverter next to the ground line to the input stage, run parallel for, let's say, 5", the coupling is a 1.5pF capacitor feeding a 0.125 ohm resistor. At 5kHz, the 1.5pf of capacitance has an impedance of about 21M, the resistance of the wire is 0.125 ohm, so the amount of signal coupled is 5.89E-9 times 50V or 0.294 microvolts. This is 50db down from a 100mv guitar signal. I would ignore that one.
For AC-carrying wires, the situation is not the capacitive coupling, but the inductive coupling. I would not run signal grounds parallel to AC wiring within one inch at all. In fact, I'd keep all the AC wiring tightly twisted and completely away from signal wires at all.
|Page 1 of 1|