Star grounding problem

[jpfamps]

Someone above suggested star-shaped solder tags with 6 solder bits and one bolt hole in the middle. They actually exist, I've seen them in the Morgan Jones book. No idea where to get them though.

I've always wondered where they came from too. I expect they may be NOS.

[Enzo]

OOOh, I take that as a challenge, find the 6-taps.


Trying to make a difference where one is not... Look at the common executive desk toy. The thing with five or six steel balls hanging next to each other in a row on strings. Pull the end one back and let it fall against the end of the row, and the one on the other end flys away, then falls back and the whole thing reverses. The end balls click and clack until it runs out of energy. Point being that the balls in the center never move AT ALL, yet the "current" flows through the line of balls.

We can talk that a particular electron - assuming even that is well defined these days - may take a long time to work its way along a wire. But so what? It may make some point about physics, but in our circuits, the MOTION of the electrons is what counts, that is our current, that is our clicking clacking steel balls. The balls may sit there unmoving, but whether they knew it or not, they passed the current.

[Alan0354]

It all boils down to ground current is quite directional even in a flat plane. If you can buy into this, you can forget all the EM and all. That is to convince you guys that because of it's EM that you can predict the direction much easily...........Current follow the path of least resistance and very low frequency, follow the path of least impedance/inductance for moderate to high frequency.

If you look back to my first drawing. If you literally layout the circuit with the filter cap on each stage with the ground close to the cathode resistor, clean separation of grounding points between each preamp stages in logical order from left to right. You should be very safe as the current travel like the greed arrow indicated and the stages don't interact with each other. You can have multiple points of ground like in the drawing and it would not matter. You don't even have to have the low side of the filter cap at single point to the cathode resistor, just close, the ground current will sort itself out. I truly believe this is the easiest way and the safest way to connect ground. It is very forgiving, you can violate a lot and it does not matter in most case.

But if you use star, unless using one star per stage, you have to force ground current between two stages together. This get's critical when the signal path need ground reference, then you have to choose the right star, or else you force the wrong current and cause hum. This is what one of the person here did wrong and I helped him fixing that. I remember now he put the ground reference of the tone stack to the PI star and it hummed. It was two adjacent star, just the wrong star. Just removing that fix most of the problem. Then I advice him to follow the Fender grounding and he responded it became perfect.

To those that praise ONE POINT ground at the input. If that is true, there should be no other connection to the chassis other than that one point. This mean if you open that ONE connection, you should measure high impedance from the circuit ground to the chassis. Do that, lets see you really have a truly one point ground. If you still measure a short from circuit ground to chassis, then you don't have a one point ground. Point is the more you limit yourself, a single mistake might cause you problem.

I only advocates one point for the power amp and rectifier because it is so high current that it might bounce the whole section of ground plane. Fender don't do that, but they put the tube, screen and filter cap grounding concentrated in the power tube area, so the ground current is localized in that area.

Finally, What I propose absolutely sound, whether it is over kill or not, it's up to people to decided. Yes, in audio, it is not that critical, but you want to start with a good platform so you can violate a lot before you see adverse affect. Yes, it sounds much more complicated, but in practice, it is very easy and forgiving. Just layout components logically and ground to the plane( chassis) as needed.

[Alan0354]

OOOh, I take that as a challenge, find the 6-taps.


Trying to make a difference where one is not... Look at the common executive desk toy. The thing with five or six steel balls hanging next to each other in a row on strings. Pull the end one back and let it fall against the end of the row, and the one on the other end flys away, then falls back and the whole thing reverses. The end balls click and clack until it runs out of energy. Point being that the balls in the center never move AT ALL, yet the "current" flows through the line of balls.

We can talk that a particular electron - assuming even that is well defined these days - may take a long time to work its way along a wire. But so what? It may make some point about physics, but in our circuits, the MOTION of the electrons is what counts, that is our current, that is our clicking clacking steel balls. The balls may sit there unmoving, but whether they knew it or not, they passed the current.

Yes, that's how a lot of people in electronics think, but it is well documented that it is EM wave. Also, bouncing balls don't explain why current follow distinct path in a plane. I can make current in the ground plane follow a zig zag pattern, snake pattern, an arc pattern at will. Yes, it does not work as well at very low frequency, but even at low frequency, it will roughly follow certain pattern.

[loudthud]

That's why Fender have no ground noise issue and is easy to implement.

Leo's ground scheme worked ok until grounded power cords were installed (after he left the company) and you tried to use an external reverb unit with a grounded power cord. Leo tied the HV centertap to ground near the transformer and grounded the first filter cap somewhere else. The only reason the amps didn't have objectional hum is that the preamp filter cap didn't share a ground with the upstream filter caps. (I think blackface Princeton with a can cap is the only exception.)

So many of the novice builders that have hum problems have grounded the HV center tap to a transformer bolt and tried to use some kind of (magical) star ground. The only other common fault is no center tap for the heater supply. A distant third is tying all the filter cap grounds together on the eyelet board. Following old Fender schematics and layouts often yields poor results not because Leo did it wrong, but because safety requirements and practices have changed.

[nickb]

Hi dobrota,

Seems we had a bit of a detour. I'd really like to refocus on your original question.

How are you getting on with your hum problem? Have you been able to take any measurements using ACV range of your meter? Just as a reminder look at the the B+ supply and grid bias. Did pulling the PI tube make any difference?

[trobbins]

Good practical physics is also knowing which rocks are observable above the water line, and addressing them first, and then appreciating how low the water line may be able to go, and also having an awareness that there may be rocks below the water line but not spending effort on them unless the operating range may cause them to impact performance. Experience helps us to guage the relative level of many of the rocks, and allows those lower level rocks to be disregarded. Raising the topic of current flow density variations across a chassis plane is imho a red herring here, and an indicator of inexperience. For starters, we aim not to use the chassis for any signal based current flow, and secondly we are not trying to control MHz frequency current paths per se.

[Chuck H]

Leo's ground scheme worked ok until grounded power cords were installed (after he left the company) and you tried to use an external reverb unit with a grounded power cord. Leo tied the HV centertap to ground near the transformer and grounded the first filter cap somewhere else. The only reason the amps didn't have objectional hum is that the preamp filter cap didn't share a ground with the upstream filter caps. (I think blackface Princeton with a can cap is the only exception.)

I'll add that I've never known a BF amp that didn't have a hummy reverb. But there's so much going on in those amps that no one ever worries through the mess and fixes it though.

[woodyc]

I'll add that I've never known a BF amp that didn't have a hummy reverb. But there's so much going on in those amps that no one ever worries through the mess and fixes it though.

I was talking with a guy at accutronics about 10 years ago and was told they developed a humbucker -- but it added something like 10 cents to the cost of a tank and they didn't think it was worth it.

And by the way, the original Hammond tanks are awesome - no hum and huge output., They used a crystal element like the ones in bullet mics.

[Chuck H]

I was talking with a guy at accutronics about 10 years ago and was told they developed a humbucker -- but it added something like 10 cents to the cost of a tank and they didn't think it was worth it.

And by the way, the original Hammond tanks are awesome - no hum and huge output., They used a crystal element like the ones in bullet mics.

Having built a "reverb amp" (an amp dedicated to providing reverb to an otherwise reverb-less tube amp system), and having seen the ground scheme for the BF amps, I'll go out on a limb and say that Fender just did it poorly WRT the grounding for the reverb on those amps. I had some trouble with my design and project and many here stepped up to help. Honorable mention goes to Steve Conner. Once the ground scheme was idealized the unit was silent as a church mouse other than the reverb coming out of it. Looking at the Fender circuit I see no reason it should be so noisy. In one regard the tank itself IS to blame. It has a shared ground for the input, output and tank chassis. But Fender could have ordered the tank grounding any way they wanted it. Perhaps the Hammond tank doesn't have the in and out sharing a ground. Depending on the ground scheme that would explain your experience.

[jazbo8]

What happened to the attachments in this thread? They are all blank, does anyone have them saved somewhere?

[Chuck H]

All the attachments in every thread disappeared in, uh, I can't remember. Maybe two years ago. The whole image cache just vanished. The site was up and down a lot too. tboy had some bugaboos to evict and by the time it was done this place was a desert. Maybe if you ask after a specific image someone still has it. But there's no secret hiding place where the original image cache for this thread is hiding. It evaporated.

[jazbo8]

Maybe if you ask after a specific image someone still has it. But there's no secret hiding place where the original image cache for this thread is hiding. It evaporated.

It's a bit hard to follow the texts when Alan made references to his own vs. Merlin's grounding scheme. So if anyone has an image showing what the differences are, it would be most helpful.

[dai h.]

found these:

(from page 1) :



(page 2) :


(oh, for some reason uploading the png version didn't show up in the attachments option)

[jazbo8]

Thanks! Now I can try to decipher what all the fuss was about...