Is this a star ground?

[defaced]

I have never been able to visualize star grounding until I saw some figures in Merlin's grounding pdf. I've been fighting stability issues in a high gain build I've been working on and figured grounding - even if not the problem - might be worth looking at (to date, I have used a buss ground). So since I'm swapping out the preamp board so the suspect sick one can be looked at, I figured the new one should be star grounded. Which brings me to this - is this layout properly star grounded (SLO Overdrive circuit)?

[mark enger]

I dont think so, star ground is just some ground lugs bolted to chassis, 1 set or star towards the inputs for pre amp cathiods, pots, inputs, pi ground, then towards the power trany is anouther star shaped grounding point for main filtercaps, powertube cathiod, screens filtercap. and centertaps, with bussbar grounding some amps work ok but grounding the preamp away from the power section grounds will help if your buss bar amp is humming, oh ya it helps to scrape chassis at any point you ground then use a large iron to solder that point right to the chassis also scrape where pots bolt through the chassis a dremmel works well. mark

[DIY_Guy]

Thats more or less a multi-star ground you have pictured there. Works fine when I've used it.

[dai h.]

for me personally "star ground" is just connecting points to a star in order to avoid common impedances (the aim of a star connection). But you can also avoid them by using a plane (lower the common impedance between the connections to where it no longer has any negative consequence). When Bruce has said before (a few times) the Fender "multiple points to the brass plate" thing is the "same" as star grounding (although I don't recall him expounding on it specifically) I think that is the underlying mechanism why they can equate. I tried measuring a short (5in.?) from a JCM800 2204 preamp tube cathode connection on my LCR meter, and (on the lowest setting, after "self-calibration" which apparently is akin to a "zero adjust") I still could not get a reading (since it was so tiny, evidently). So that seemed to say common impedances can matter less when the currents are low, super low, or close to zero (lots of preamp circuitry excluding heaters falls into this I think), therefore in such cases it seems you have a bit more leeway in ground placement before something becomes problematic. So when trying to figure out whether a connection will work or not, it should be useful to think about what loops exist and what currents (i.e. how much) are flowing in the circuit. Conversely points where currents are high more care needs to be taken with placement (such as early on after the rectifier where where high current pulses are happening and residual ripple returns back to the source winding back through the ground line and causes a voltage drop--so you don't want to tie these early segments of the ground line into more sensitive areas where less, little, or basically no currents are flowing (this has mentioned more than a few times on the forum)). (The residual ripple is AC "hum" which of course you don't want to inject right into the signal.)

For me personally, I really liked this one as an explanation that made sense (although I liked other ones a lot also such as the Aiken article) :

The Amp Garage :: View topic - Tube amp grounding article/info/links

The Soldano SLO100 I liked looking at also to try to figure out the thinking with the grounding and layout. IIRC (I've never tried cloning one but just from memory of the PCB) there is a ground in the center of the PCB which (I think) you can consider as a star, the grounds connecting the pots (grid connections) although you might look at that as a bus(buss?) I think it makes sense to look at that as a star(or "stretched out star"--not sure if I read this or came up with this expression) where not exactly making a star doesn't matter since the currents are so tiny to be inconsequential(or non-problematic).

[tubeswell]

For me personally, I really liked this one as an explanation that made sense (although I liked other ones a lot also such as the Aiken article) :

The Amp Garage :: View topic - Tube amp grounding article/info/links

Wow Dai, that was a mamoth effort! I think I've read that Kimura article somewhere before. Maybe it was you who posted it somewhere else a while back?

[Steve Conner]

So that seemed to say common impedances can matter less when the currents are low, super low, or close to zero (lots of preamp circuitry excluding heaters falls into this I think), therefore in such cases it seems you have a bit more leeway in ground placement before something becomes problematic.

Very nicely put...

[dai h.]

Wow Dai, that was a mamoth effort! I think I've read that Kimura article somewhere before. Maybe it was you who posted it somewhere else a while back?

yeah probably, lol. I was just trying to imagine it as if an English speaker woke up one day and was able to read the original pages. Or "attempted human google translation"--hopefully not causing offense to the author. (For those that already can read Japanese) the author has written a book based on the material on his site:

edit: amazon link doesn't work, but the book is called "jyounetsu no sinkukan anpu" (Passion of Tube Amplifiers, or Passionate Tube Amps) by Kimura Tetsu

(fwiw, some Norwegian(?) fellow who was checking out the grounding article wanted to know if the author was kosher--someone authoritative--and when I googled him, he is apparently the chief consultant for some IT company.)

[dai h.]

Very nicely put...

thanks Steve. I think this is a positive thing for someone attempting to lay out a circuit because it means you don't have to have everything wired to not share common impedances. For example, this is something I did before I had much understanding of grounding:

MIJ ceramic noval socket bayonet base lug side on Flickr - Photo Sharing!

While it did work (pretty much noise free) I bet I could've simplified the wiring if I knew more. Imagine if you had to wire the grounds like this on EVERY piece of electronics (some with multitudes--hundreds or more ground points). (gee, where is the "horror" icon? lol.) Thankfully you don't have to.

[defaced]

Hmm, lots to think about. I think simply keeping the "upper" ground bus for the power and each tube's grounded components on a short buss and having that connect back to the cap's ground lug will be sufficient for what I'm doing. I'll peruse that article when I get home tonight. Thanks all for the feedback.

[Bill Moore]

I have used the buss type grounding on Fender style circuits with good results, but I use 12-14 ga. copper for the buss. I understand the idea of keeping pre, and power grounds separate, as not to create a potential difference (path) between circuits. If the buss is large enough to carry the current, there should be no differential.
I saw another fellows design a while back, that isolates the circuits from the transformer secondary onward, and is not tied back to chassis ground. I have not tried that system yet, but by limiting the potential paths, it may aid in reducing hum.