star ground for power tubes and ct, or soldered to chassis?

[elipsey]

Yesterday I was looking inside my Fender, which I find amazingly quiet, and meditating on it's grounding scheme. In some ways it seems like they did it wrong. All the high power grounds go straight into the chassis, no star grounding (that I recognize), input jacks are not isolated, carbon comp resistors... but it's quiet. I'm sure the layout was meticulously planned, in spite of it's appearance, and there are details other than grounding that obviously matter for noise, like they used shielded wire for preamp grids, full bypass on all the early stage cathodes, and the layout allows for very short wire runs from the preamp tube cathodes to their bias resistors, and probably other stuff I haven't noticed.

So my next two builds are a 5F2A and a PR, both with steel chassis, and I just bought a 300 Watt soldering gun I'm itching to try out

Should I:
bring my high current grounds to a star?
or solder them to the chassis like Fender?

What's your vote?

[Mick Bailey]

If I'm replicating a Fender design, I wire and ground it exactly as per the original and I've never had any problem at all. With my own designs I use a star grounding scheme. This is because I'm not replicating anyone else's layout and star grounding works best for me in experimental or one-off builds.

I run all the high current stuff to the first star ground point, and then create one or two (depending on the build) secondary star points for the preamp stages. I use insulated sockets for my own builds and overall don't experience problems even with high-gain amps.

Soldering directly to a steel chassis takes a lot of heat and you need a large thermal mass to get a fluid joint. Have you tried your gun out on the chassis?

[R.G.]

Star grounding is not the only way to get quiet ground returns. There is probably an infinite number of ways to get quiet grounds, of which star grounding is only one.

However, it's the only one you know ahead of time that will be quiet.

Quiet grounding requires a few things: (1) knowing what the ground **currents** are in every ground return wire, (2) arranging the ground current paths so that they do not overlap where overlaps in a conductor will cause generation of noise by Ohm's law, and (3) allowing overlaps where they will either not cause problems or even better, cancel out for that particular circuit.

It takes intimate, extensive, and exhaustive knowledge of the circuit and its operation to achieve #3. Or lots of trial and error, plus some luck. Star grounding succeeds by forcing no overlaps at all, thereby sidestepping 1 and 2 - with no overlaps, there can be no issues with overlap/not, and you don't need to know the currents.

[elipsey]

Star grounding is not the only way to get quiet ground returns. There is probably an infinite number of ways to get quiet grounds, of which star grounding is only one.

However, it's the only one you know ahead of time that will be quiet.

Quiet grounding requires a few things: (1) knowing what the ground **currents** are in every ground return wire, (2) arranging the ground current paths so that they do not overlap where overlaps in a conductor will cause generation of noise by Ohm's law, and (3) allowing overlaps where they will either not cause problems or even better, cancel out for that particular circuit.

It takes intimate, extensive, and exhaustive knowledge of the circuit and its operation to achieve #3. Or lots of trial and error, plus some luck. Star grounding succeeds by forcing no overlaps at all, thereby sidestepping 1 and 2 - with no overlaps, there can be no issues with overlap/not, and you don't need to know the currents.

thanks. i clicked you sig, and i'm just wondering if you are, by trade, a fellow programmer, in which case i would like to offer you a virtual fist bump.

[Alan0354]

I gone through this in post #3 and #5:http://music-electronics-forum.com/t36181/

This subject is well covered in the subject of Signal Integrity.

[R.G.]

thanks. i clicked you sig, and i'm just wondering if you are, by trade, a fellow programmer, in which case i would like to offer you a virtual fist bump.

I'm kind of an everything. My batchelor's is in EE, specialization in analog controls. My masters is in computer architecture. I've done most kinds of electronic hardware design, as well as writing machine boot code, detailed control of printer mechanisms, system boot code, applications programming, and for a while I managed the development/maintenance department for a variant of unix. I've also been a lifeguard, a welder's helper and a garbageman.

If any of those qualify, I'll return the fist bump.

[Enzo]

Garbageman? Cool. Much to my mother's chagrin, I thought our trash guys were really cool. They drove up in this huge truck, with all that hydraulic squeezing stuff. It whirred and snorted and did mechanical things. But best of all they got to ride it hanging onto the sides. In my suburb, they went back of the house for trash. They walked down the drive, spread a big burlap blanket, dumped our trash onto it, pulled up the corners into a bag, over the shoulder, up to the truck. They were the only black people I ever saw up close. They were the coolest guys I knew.

I told my ma once that I wanted to grow up and be a trash guy. I think she about fainted.

So bumps to you, RG.

[NateS]

I like to think of the secret to star^H^H^H^H grounding as:

1) Always return current to the same ground it originates from
3) The "star" is only for tying the zero sum loops to the reference 0V potential.

And yeah - not even firefighters get to hang on to the back of the truck anymore. I missed that by a generation.

[Chuck H]

and a garbageman.

Are you speaking of your roll here?

[zbones]

if only we were all waiters or cleaners at one time or another, sh*t would make more sense to more ppl

[elipsey]

Thanks, I think I see it now:

I would tie all the cathode grounding into one star to the chassis in the middle where the power tubes are and put the ground of both the Reservoir and screen cap onto that star. With that, the high current never get onto the chassis. Your chassis essentially become a quiet signal ground.

I did read that thread before posting and I found it to be quite instructive, but it seemed to me that the subject of the thread was input jack wiring, and, more generally, avoiding hum. I really mean to ask a narrower question: Is it a good idea to solder the power tube and HT grounds to the chassis without star grounding them, like Fender does?

So it seems like your vote is no...

...but maybe it's not a short answer kind of question.

EDIT: I couldn't find the phrase Signal Integrity in that thread. Is this a subject or term of art that I should be familiar with?

[elipsey]

That's awesome. Which Unix did you work on?

[elipsey]

If I'm replicating a Fender design, I wire and ground it exactly as per the original

This is what I'd like to shoot for. I think the Fender layouts can be really tidy, and I have played some that I though were amazingly quiet.

Soldering directly to a steel chassis takes a lot of heat and you need a large thermal mass to get a fluid joint. Have you tried your gun out on the chassis?

Well, I thought I'd test the gun on P.R. amp I have that I like to play around with on the bench for testing, maybe try moving it's grounds around a bit and see what happens. I have heard that sometimes joints like this can cause insidious problems because the solder looks good on the outside, but didn't really get hot enough to make a proper connection to the chassis; noise that goes away when you reflow the cap-can grounds, for example.

The gun I got is a Weller D650. At 300 Watts it's the biggest one I could find, and there was an old gaffer in the amazon reviews talking about soldering cap cans into radios with it, so it sounded like a good bet.

[NateS]

My first amp I starred literally everything. It laughed at high power soldering guns. I ended up using a stained glass window iron an old ham friend had. Later i ended up creating a wire ring to solder to, but thats starting to resemble a 1 dimensional chassis ground. I like the idea of local current loops tied together at ground better Its both more to the point and easier to implement

[R.G.]

That's awesome. Which Unix did you work on?

AIX