Champzog Layout & Humdinger Questions

[Cliff Schecht]

Thanks for correcting me. I thought 26 gauge wire could handle a couple of amps, but you got me researching, and I see that it's not a great idea.

I wasn't planning to put it on the B+ connection, though.

Are you saying shielded wire is generally useless on heater supplies?

Yes and if you can manage an ideal circuit/signal wire layout the filaments don't have to be twisted.

http://www.c3amps.com/EnlargementPages/ ... itch4.html

TM

Why would somebody invest a bunch of money in high quality pots, resistors, switches and a nice chassis only to use those shit quality Chinese ceramic sockets?

[diagrammatiks]

Help me out here. I want noisy stuff far from the input region. I know that means keeping the transformers and rectifier at one end. But it sounds like you're saying the caps themselves should be closer to the amplifying circuitry. Can that be right? I just figured all the ripply stuff and diode switching and whatnot should be far from the input area.

I was going to stick a separate board for the filter caps under the power transformer or very close to it. I assumed I would put a ground on one of the PT bolts, as I did with my 5f6a.

hmm I dunno if I wanna say this right now because I don't know if it's going to help you.

but...yes the transformer is noisy and you want to keep the ac power line near one edge of the chassis and have that as far away from the lower voltage heater lines as possible.

yes the rectifiers can be noisy depending on what type of rectifiers you use and should be by the power transformer and the reservoir cap and star ground for that stage should all be spaced tightly together

however after the b+ gets rectified high tension dc is the same as ground as far as your ac signal is concerned. It's better to have them close to the stages they are filtering so that the caps shield the stages, and the ground wires are kept as short as possible to the end of the caps.

I have to ask. How is the hum removed from that amp? It looks like the fields from the filament wires would be so different, they wouldn't cancel very well. And is that really two bare AC-carrying wires, or is there some kind of clear insulation on there?

It's not DC, is it? I found a schematic some guy did for rectifying his heater current.

I'm wondering if the geometry is the secret. But I know the magnetic fields should wrap around the wires, so it seems like it should smack all of the contacts pretty good.

at a certain point everything is a compromise. Twisted heaters work to cut down noise to a certain extent but it also depends on pushing the wires to the edge of the chassis and making sure the other wires coming from the pins don't have to be near the ac heater wire.

You can't push in either direction on the slo. If you push towards the chassis you radiate directly to the pots. if you push towards the middle you radiate directly onto the board.

the parallel wires are the best method of managing heater hum in that particular situation. It's 16-18 gauge silver plated bus wire covered in teflon tubing.

[diagrammatiks]

Help me out here. I want noisy stuff far from the input region. I know that means keeping the transformers and rectifier at one end. But it sounds like you're saying the caps themselves should be closer to the amplifying circuitry. Can that be right? I just figured all the ripply stuff and diode switching and whatnot should be far from the input area.

I was going to stick a separate board for the filter caps under the power transformer or very close to it. I assumed I would put a ground on one of the PT bolts, as I did with my 5f6a.

hmm I dunno if I wanna say this right now because I don't know if it's going to help you.

but...yes the transformer is noisy and you want to keep the ac power line near one edge of the chassis and have that as far away from the lower voltage heater lines as possible.

yes the rectifiers can be noisy depending on what type of rectifiers you use and should be by the power transformer and the reservoir cap and star ground for that stage should all be spaced tightly together

however after the b+ gets rectified high tension dc is the same as ground as far as your ac signal is concerned. It's better to have them close to the stages they are filtering so that the caps shield the stages, and the ground wires are kept as short as possible to the end of the caps.

I have to ask. How is the hum removed from that amp? It looks like the fields from the filament wires would be so different, they wouldn't cancel very well. And is that really two bare AC-carrying wires, or is there some kind of clear insulation on there?

It's not DC, is it? I found a schematic some guy did for rectifying his heater current.

I'm wondering if the geometry is the secret. But I know the magnetic fields should wrap around the wires, so it seems like it should smack all of the contacts pretty good.

at a certain point everything is a compromise. Twisted heaters work to cut down noise to a certain extent but it also depends on pushing the wires to the edge of the chassis and making sure the other wires coming from the pins don't have to be near the ac heater wire.

You can't push in either direction on the slo. If you push towards the chassis you radiate directly to the pots. if you push towards the middle you radiate directly onto the board.

the parallel wires are the best method of managing heater hum in that particular situation. It's 16-18 gauge silver plated bus wire covered in teflon tubing.

Thanks for correcting me. I thought 26 gauge wire could handle a couple of amps, but you got me researching, and I see that it's not a great idea.

I wasn't planning to put it on the B+ connection, though.

Are you saying shielded wire is generally useless on heater supplies?

Yes and if you can manage an ideal circuit/signal wire layout the filaments don't have to be twisted.

http://www.c3amps.com/EnlargementPages/ ... itch4.html

TM

Why would somebody invest a bunch of money in high quality pots, resistors, switches and a nice chassis only to use those shit quality Chinese ceramic sockets?

are there better quality ceramic sockets?

[Cliff Schecht]

Help me out here. I want noisy stuff far from the input region. I know that means keeping the transformers and rectifier at one end. But it sounds like you're saying the caps themselves should be closer to the amplifying circuitry. Can that be right? I just figured all the ripply stuff and diode switching and whatnot should be far from the input area.

I was going to stick a separate board for the filter caps under the power transformer or very close to it. I assumed I would put a ground on one of the PT bolts, as I did with my 5f6a.

hmm I dunno if I wanna say this right now because I don't know if it's going to help you.

but...yes the transformer is noisy and you want to keep the ac power line near one edge of the chassis and have that as far away from the lower voltage heater lines as possible.

yes the rectifiers can be noisy depending on what type of rectifiers you use and should be by the power transformer and the reservoir cap and star ground for that stage should all be spaced tightly together

however after the b+ gets rectified high tension dc is the same as ground as far as your ac signal is concerned. It's better to have them close to the stages they are filtering so that the caps shield the stages, and the ground wires are kept as short as possible to the end of the caps.

I have to ask. How is the hum removed from that amp? It looks like the fields from the filament wires would be so different, they wouldn't cancel very well. And is that really two bare AC-carrying wires, or is there some kind of clear insulation on there?

It's not DC, is it? I found a schematic some guy did for rectifying his heater current.

I'm wondering if the geometry is the secret. But I know the magnetic fields should wrap around the wires, so it seems like it should smack all of the contacts pretty good.

at a certain point everything is a compromise. Twisted heaters work to cut down noise to a certain extent but it also depends on pushing the wires to the edge of the chassis and making sure the other wires coming from the pins don't have to be near the ac heater wire.

You can't push in either direction on the slo. If you push towards the chassis you radiate directly to the pots. if you push towards the middle you radiate directly onto the board.

the parallel wires are the best method of managing heater hum in that particular situation. It's 16-18 gauge silver plated bus wire covered in teflon tubing.

I think it might also help to point out to Steve that the return currents in a specific stage eventually will end up going into the "-" pin of whatever capacitor is decoupling that stage. These currents are what we are controlling when we do a star ground. It's important to isolate the high energy return currents from the low energy (and very noise susceptible) input stage and preamp stages. This means that for each stage it's best to keep the grounds grouped closely together and going as directly as possible to the B+ decoupling cap ground, otherwise these currents can leak into other parts of a circuit and cause issues. This is why some warn against chassis grounding schemes unless you know what you are doing, if you do it wrong the high power return currents leak into everything and turn your amp into a hummy mess.

I'm rushing this post and fear that I might be unclear somewhere. Steve let me know if anything needs clarification..

[diagrammatiks]

Help me out here. I want noisy stuff far from the input region. I know that means keeping the transformers and rectifier at one end. But it sounds like you're saying the caps themselves should be closer to the amplifying circuitry. Can that be right? I just figured all the ripply stuff and diode switching and whatnot should be far from the input area.

I was going to stick a separate board for the filter caps under the power transformer or very close to it. I assumed I would put a ground on one of the PT bolts, as I did with my 5f6a.

hmm I dunno if I wanna say this right now because I don't know if it's going to help you.

but...yes the transformer is noisy and you want to keep the ac power line near one edge of the chassis and have that as far away from the lower voltage heater lines as possible.

yes the rectifiers can be noisy depending on what type of rectifiers you use and should be by the power transformer and the reservoir cap and star ground for that stage should all be spaced tightly together

however after the b+ gets rectified high tension dc is the same as ground as far as your ac signal is concerned. It's better to have them close to the stages they are filtering so that the caps shield the stages, and the ground wires are kept as short as possible to the end of the caps.

I have to ask. How is the hum removed from that amp? It looks like the fields from the filament wires would be so different, they wouldn't cancel very well. And is that really two bare AC-carrying wires, or is there some kind of clear insulation on there?

It's not DC, is it? I found a schematic some guy did for rectifying his heater current.

I'm wondering if the geometry is the secret. But I know the magnetic fields should wrap around the wires, so it seems like it should smack all of the contacts pretty good.

at a certain point everything is a compromise. Twisted heaters work to cut down noise to a certain extent but it also depends on pushing the wires to the edge of the chassis and making sure the other wires coming from the pins don't have to be near the ac heater wire.

You can't push in either direction on the slo. If you push towards the chassis you radiate directly to the pots. if you push towards the middle you radiate directly onto the board.

the parallel wires are the best method of managing heater hum in that particular situation. It's 16-18 gauge silver plated bus wire covered in teflon tubing.

I think it might also help to point out to Steve that the return currents in a specific stage eventually will end up going into the "-" pin of whatever capacitor is decoupling that stage. These currents are what we are controlling when we do a star ground. It's important to isolate the high energy return currents from the low energy (and very noise susceptible) input stage and preamp stages. This means that for each stage it's best to keep the grounds grouped closely together and going as directly as possible to the B+ decoupling cap ground, otherwise these currents can leak into other parts of a circuit and cause issues. This is why some warn against chassis grounding schemes unless you know what you are doing, if you do it wrong the high power return currents leak into everything and turn your amp into a hummy mess.

I'm rushing this post and fear that I might be unclear somewhere. Steve let me know if anything needs clarification..

ya that. it's hard to make sure all the relevant info is in one post I guess.

The best suggestion I have is to read Merlin's chapter on grounding and power supplies in the power supply book.

At the end of the day where you put your caps should be related to what type of ground scheme you implement. There's much more danger of noise induced by improper grounding then there is of noise due to where exactly you are putting your caps.

Also Cliff, what's up with socket quality up above? Are there different grades of ceramic or are you saying some other material is better?

[Cliff Schecht]

I'm specifically dissing the quality of those sockets, if they are what I think they are. The ones I'm thinking of are these: http://www.tubesandmore.com/scripts/fox ... =P-ST9-511

And I've had bad experiences with these as well: http://www.tubesandmore.com/scripts/fox ... =P-ST9-302

With both you have to be careful about using too much solder (it easily seeps through to the socket-side) and also about general usage. The first one I posted actually makes a good seat with the tubes and holds them firmly but the seconds ones will easily ruin the sockets grip if you bump a tube while installed. The first ones have very little pin-to-pin clearance while the second ones are fine here. Also I've had a few of these chip or crack on me both when installing them and in general usage.

I much prefer Belden stuff now, for the marginal price increase you get a much higher quality socket..

[diagrammatiks]

I'm specifically dissing the quality of those sockets, if they are what I think they are. The ones I'm thinking of are these: http://www.tubesandmore.com/scripts/fox ... =P-ST9-511

And I've had bad experiences with these as well: http://www.tubesandmore.com/scripts/fox ... =P-ST9-302

With both you have to be careful about using too much solder (it easily seeps through to the socket-side) and also about general usage. The first one I posted actually makes a good seat with the tubes and holds them firmly but the seconds ones will easily ruin the sockets grip if you bump a tube while installed. The first ones have very little pin-to-pin clearance while the second ones are fine here. Also I've had a few of these chip or crack on me both when installing them and in general usage.

I much prefer Belden stuff now, for the marginal price increase you get a much higher quality socket..

oh ya that's true. they do like chipping. the pins don't really like to be bent either. it's hard to say no to a bag of sockets at a buck a socket though.

I'll have to watch out for the long term wear though...

seems like cheap plastic out performs cheap ceramic in this case.

[Phil_S]

Steve,
I think you'll find folks here are patient and willing to share what they know. That said, I think you are laboring over this build when it isn't necessary. It is hard to mess up a champ. Use 18 or 20 ga for the filaments. Use 22 ga for everything else. Put the rectifier at one end out of the way. Separate the trannies and turn them 90 degrees to each other. Keep signal and filament wires separated. Remember to ground what needs grounding. Connect the parts per the schematic -- there are only a few of them. That's about it.

Less talk, more drilling and building. Get on with it!

Phil

[The New Steve H]

I made the big ol' PT hole, so now I'm ready to finalize the layout and put it together.

[The New Steve H]

I hate to get back to the original questions, since we have learned so much about sockets, but I am still pondering the Soldano heater thing.

Last night I was thinking about it, and it seemed to me that it can't be a question of loops, because the Soldano layout has several long, straight-sided loops formed by the heater wires and filaments, and it sure looks like the changing magnetic fields would pass through them, causing hum. But then I thought about how big the loops are, and I wondered whether flux DENSITY was the thing.

Is the hum reduced simply because the loops are gigantic and the flux density is tiny? If that's true, it seems like it would allow for a lot of different ways of wiring heaters.

I realize I am proving I remember almost nothing I learned in Electricity and Magnetism.

Now I'll go put on my Nomex suit and wait for replies.

[M Fowler]

Equal distance from each other (wire parallel) throughout the filament wiring is why this type of filament wiring works.

[The New Steve H]

So it's just opposing fields cancelling each other due to geometry? Wouldn't you have to be extremely careful about locating the wires with respect to the parts inside the tubes? I know that if you have twisted wires that open up by 1/8" it will cause hum, so I would think all the sensitive stuff would have to be exactly between the wires. Is that wrong?

[The New Steve H]

I found a page about Faraday's Law, and it looks like density doesn't matter. And I guess I have it backward anyway, since the induction doesn't take place in the heaters.

Or...maybe I don't have it backwards, because lower flux density should reduce the amount of flux through any loop which IS susceptible to induction.

Now my head hurts again.

[Phil_S]

Steve, excluding the rectifier, which runs of the 5V winding, you only have 2 tubes. IMO, you are over thinking, maybe even obsessing about heather wiring. Don't be so hard on yourself. Like I said before, it is hard to mess up a Champ. Old fashioned twisted pair will work very nicely here. You have two basic styles. One is up in the air, the other on the chassis "floor" in the corner created by the fold.

If you want to get fancy here, try using one single length of naked solid copper. As you feed it through the pin sockets, sleeve with heatshrink. That also works and is easy to do for just two sockets.

[The New Steve H]

Hey, don't worry. I'm asking largely for learning purposes. I don't need to know all this stuff in order to get this amp done. The problems with the schematic are pretty awful, but I'm not going to let my heater questions prevent me from going ahead.

I'm pretty attention deficient, and I find that it's easier to learn when I have a project in front of me than when I'm sitting in a chair with a book. Having the Herzog parts on the table raises questions and helps me see the gaps in my knowledge.

If I hadn't screwed up the Bassman clone, I wouldn't have learned nearly as much as I did from that project. Fixing it taught me all sorts of things.