Going in loops about grounds

[mbratch]

I am about to embark on a 5e3 build (from kit) and have been researching how to properly make the ground connections in the circuit.

I've read about star versus buss and various mixes. I believe I understand them. But what's really messing me up is that there are some reputable builders who say you should solder a ground buss (heavy wire) on back of the pots and connect pre-amp grounds to that. Then another reputable build who says to absolutely not take this approach.

I've studied pics of vintage amps and they seem to to exactly what everyone says not to do: solder the pre-amp grounds to the chassis and the pot ground lugs directly to the back of the pot.

So what's the right answer? Can someone provide pics of what they've done for grounding that has worked really well?

Thanks! I hope this doesn't start a ground flame war.

[Enzo]

Some people think you should order lobster from the menu, others think steak. SOme people don't think you should eat red meat at all, while others think anything that came from the sea will be poisonous. Salad or cole slaw? WHICH ONE IS RIGHT????

This argument and the one you proposed is based upon the notion that there is a right and wrong, that there is some sort of absolute. There is not. Some ideas are based upon sound principles, while others are based upon, I did this and it worked (or didn;t).

If you chose a "wrong" grounding scheme, the amp can still wind up working just fine. We worry about center taps on the heaters and elevating the heaters, and even DC heaters, yet some older amps simply grounded one side of the heaters at each socket and ran a single 6VAC wire socket to socket, and the amp came out OK.

Ground bus across the pot backs? As opposed to basically the same thing by using the Fender brass sheet? Not all pots are made the same. SOme pots with metal construction will naturally ground their shell to the chassis. This forms a little shield for the pot innards. Other pots do not make a ground conection for thier covers. In those cases, a volume control on a high gain stage can pick up hum from this ungrounded cover. In THAT case, it needs to be grounded. WHether you use that ground also for your circuits is a separate issue.

Just because someone successfully uses one particular method doesn;t make other ways wrong, even if that successful fellow says so.

[mbratch]

OK, so it's just, "pick one and if it doesn't work out, try something else".

[tubeswell]

Originally Posted by mbratch

OK, so it's just, "pick one and if it doesn't work out, try something else".

Works for me (although a little discussion, when it is forthcoming, can help with motivation). This can be a blimmin' tricky aspect of amp building.

There are lots of different forces at work in the ground return path. Some ground return points are referencing high voltages, and others not so high. My theoretical understanding of this (limited as it is - and I welcome discussion) is that this can have the effect of creating localised micro-rises in ground potential, which interact/react with other parts of the amp's ground return path which have slightly lower ground potential, and bingo, you have a source of hum. If you can get all the ground return points at the same potential, you will reduce (or even 'eliminate') the hum. As we know, the things that affect potential, are current and resistance (E=IxR). The best way to manage this equation is by managing the resistance - (if you make R really low, you will keep E low no matter what). Different metals have different resistance, as do different thicknesses and/or lengths of wire. Thick brass, copper or aluminium will provide very low resistance, thereby ensuring that potential differences at different ground return points are minimised. I prefer (and this is where the 'I tried it and it works for me' part of the discussion comes into play ) running each ground return point by its own separate wire to a common ground return point;- but I typically have two ground return points (one for the pre-amp grounds, including the pre-amp filter caps grounds; and the other for the power amp grounds, including the heater CT (in situations where I do not have the heaters elevated)). The separate wires helps avoid creating a situation where you are introducing a margin for localised rises in chassis potential. To-date I have used 1.2mm (16 gauge) aluminium sheetmetal for my chassis. My amps have all been very quiet. Other peeps will say different stuff (as Enzo said).

[mbratch]

Thanks tubeswell.

Both of you guys have given me more education on grounding.

It's a funny thing though: I'm learning more and more about grounding, but still I suppose there will be a level of uncertainty as to where to start with my particular project.

My initial thinking was to take a piece of 14 or 12 gauge copper wire and run it across the back of the pots, tying the pre-amp grounds and pot/jack grounds to it, and then soldering the end near the jacks to the chassis. That more or less lines up with the Hoffman methodology. The power amp and output section I was going to keep separate.

Then I read somewhere that tying grounds busses to the back of the pots wasn't such a good idea, so I thought I'd do something similar, keep the buss, but somehow make it separate from the pots. Perhaps running it along one side of the eyelet board or suspending it with insulated tie points fastened to the chassis. I would in that case otherwise connect the grounds to it as planned.

I thought I'd look at pics of other peoples' builds but the funny thing is, of all the detailed pics out there of the innards of a 5E3, you rarely see how the grounding is done. It's usually obscured in the photos. :-O

I'll give this some further thought.

[dai h.]

basically anything that you do which considers what is happening in the grounds and is wired accordingly regardless of whether you use stars, buss, plane, etc. will work. Lots of info if you search.

[mbratch]

Originally Posted by dai h.

basically anything that you do which considers what is happening in the grounds and is wired accordingly regardless of whether you use stars, buss, plane, etc. will work. Lots of info if you search.

Oh yeah, I've been searching, finding, reading, and studying. That's what led me here.

[Five_E35]

On mine, I used the brass plate, but experimented with the screen, main, and center tap grounds, and found very little difference in hum. I ended up with everything grounded to the brass plate, similar to what Bruce recommends in some posts on here and what Fender does on the '57 RI. BTW - The schematic on the Fender site includes complete wire by wire diagrams.

I suspect the 5E3 isn't that sensitive as long as the cathode resistor for the output tubes and the main cap are tied together. My guess is the Hoffman method probably makes more sense for fixed bias amps like a Bassman or Twin where the power tube cathodes are grounded to the tube sockets on the other side of the chassis. My impression is that more exotic grounding schemes are needed for higher gain amps and although there's probably some people here who can do it, I wouldn't have a clue when it comes to something like a Soldano.

[mbratch]

Thanks. I didn't think to look at the Fender schematic for their re-issue. It is a very helpful drawing.

What assumptions are made about the brass plate? It looks to me that it's assumed the brass plan makes good electrical contact with the chassis, and that the pots and jacks also completely ground to the brass plate by virtue of being bolted to it. I know these sound obvious, but I'm just confirming.

It's also interesting that for the 3 big filter caps, the old Fender layout ties grounds of the cathode bias cap with filter cap 1, and then ties the grounds of filter caps 2 & 3. In their new layout, they tie grounds of filter caps 1 & 2 together, but not 2 & 3.

[Five_E35]

I've read in several places that the brass plate acts like a giant buss ground, and with everything connected to it the mechanical connections to the chassis are less critical. Dave Funk says the same thing in his book and also talks about grounding parts of the circuit near or opposite from their power source - i.e. cathode resistor & input jacks for V1 opposite the 100K plate resistors, screen cap opposite screen supply, and this seems like what Leo was doing on most of his amps. My own feeling is it's also better to rely on soldered vs. mechanical connections.

I soldered up the ground wires, input jacks, and pots with the plate outside the chassis, included a ground wire from the transformer bolt, and used excess cap lead for jumpers between the caps above the board instead of underneath. I experimented with a 5AR4 & 5881 tubes in it, in a back room near an in-ground power company transformer where it's almost impossible to use single coil pickups. I tried both screen cap variations to the brass plate & to the transformer bolt & to a location closer to the power tubes suggested in the Hoffman method, and there really wasn't much difference. I did clean up the back of the brass plate & the inside of the chassis with electrical cleaner before assembly, so maybe with dirt & age there might be a difference.

[dai h.]

it's Bruce I think who said that the brass plate is sort of like a star (not a giant buss). When you star points, the purpose is to try to eliminate common paths, common resistance. If you have a big conductor in a central location which you connect the grounds to it can be like a star in the sense that the bigger the conductor is the lower the common resistance. Star connecting may look more physically obvious in reducing common resistance, but the big plane can be doing the same thing. The back of the pot buss can be the same as the plate in the multiple connections to the chassis and the pot backs (usually brass I think) to each other. You get an area of higher conductivity, lower R. There might, (with the wire across pot backs) be a way(s) of connecting the ground wires to the buss that is more correct or less correct.

[tubeswell]

Originally Posted by mbratch

It's also interesting that for the 3 big filter caps, the old Fender layout ties grounds of the cathode bias cap with filter cap 1, and then ties the grounds of filter caps 2 & 3. In their new layout, they tie grounds of filter caps 1 & 2 together, but not 2 & 3.

FWIW Out of curiousity I have tried both ways, and I found that if I tied '1' & '2''s ground returns together, it was slightly quieter than if I tie '2' & '3''s ground returns together ('3' being the pre-amp cap). Anyhow you can always change it afterwards. I ended up changing it to whatever was quietest

[mbratch]

I have a kit that doesn't come with a brass plate. So now I'm wondering whether I should go to a HW store and see if I can find a brass plate, or go to my plan A which is fuss with a 14 gauge copper wire method (al la Hoffmann).

As an aside, I'm also a little confused that the drawing shows the PT yellow/red strip center tap grounded whilst it also has the two 100 ohm resistors from the pilot light to ground. I thought these were mutually exclusive?

[tubeswell]

Originally Posted by mbratch

As an aside, I'm also a little confused that the drawing shows the PT yellow/red strip center tap grounded whilst it also has the two 100 ohm resistors from the pilot light to ground. I thought these were mutually exclusive?

Yes the heater winding only needs EITHER a 'real' centre tap connected to ground, OR the 2 x 100R 'artificial centre tap' connected to ground. Depending on how accurately the 'real' Heater CT is placed w.r.t. being in the middle of the winding, the 2 x 100R can provide a quieter method if the resistors are 1% tolerant (but it draws a wee bit more current than the 'real' CT method). But if the 'real' CT is exactly in the middle of the heater winding, then it will be really quiet (All other things being equal)

[mbratch]

Thanks tubeswell. I'm not sure why the drawing shows both hookups at the same time. I figured from prior experience working on vintage amps that they're supposed to be mutually exclusive. Unless I'm misinterpreting what the yellow/red wire is.

[loudthud]

The red wire with a yellow stripe is typically the HV winding CT. The heater winding CT is usually green with a yellow stripe. Usually if a winding is one color, the CT will be the same color with a yellow or possibly white stripe. The 5V winding is usually yellow. Somethimes it will have a CT but a yellow wire with a yellow stripe doesn't make sense, Hammond uses a black stripe. (don't ground that one if used for a rectifier tube!)

[mbratch]

Ah.... thanks loudthud. I appreciate that insight. I do now recall that there were two striped wires when I worked on my vintage BF fender and both were grounded. But in my later SF amp, just one was. The PT I have in my kit doesn't have the 5V CT evidently.

[mbratch]

OK, here's what I am thinking for my grounding given that I don't have a brass plate.

I'm going to run a 14 gauge copper wire in back of the 3 pots and then attach to the chassis somewhere near the jacks (probably solder it there). The grounds from the pre-amp cathodes and the input jacks will attach to that ground wire at the nearest point for those items. The ground connections for the pots themselves will attach there as well (or perhaps to the back of the pots, like the old Fenders?).

The grounds from the dual-100ohm "virtual center tap", first two large filter caps (from the power amp section), and power tube cathode circuit will all "star" at the cathode circuit ground on the eyelet board. From there I'll run a ground wire over to the nearest lug on the PT. The HV CT will also be attached to this same lug.

The output jack will have the black ground wire from the OT attached to it.

The AC ground wire will attach to a lug on the PT mount (different from the one above).

I think this is basically the grounding scheme suggested by Hoffman.

Does this sound OK?

[stingray_65]

in my limited experience Ive found the 5E3 is not a very hummy amp no mater which grounding scheme you use.

each grounding scheme seems to have it's own sort of rules, do's and don'ts so try not to mix them

I can say from personal experience on my 5E3 build when connecting the twisted heater wires to the 12XX7's and then the 6V6's, make sure the same wires go to the same pins. use 2 different color wires or one with a stripe to make it easy on yourself.

this isn't nescassary on a single end amp like a tweed princeton

other than the making sure the heater wires land at the same pins, I've found I get most of my "hum" from the input jacks.

when I wired my 5E3 I followed the Fender layout faithfully.
the noise I did have was seriously reduced when I ran a shielded cable from the input to the tag board.
I was careful to attach the sheild at the jack .but you should only attach one end of the sheild to ground.
my noise went away completly when I replaced the cheap jacks with some switcraft 12A's. it doesn't pay to skimp here, it is an important connection.

nowadays I wire my inputs by laying out my input holes 7/8" - 15/16" apart and turn the jacks till the ground lugs face each other then bend those tabs toward each other and solder them together. I mount the 1 meg and 68K resistors to them so that only one signal wire per input pair is needed, and I land the directly on the tube socket.

If you do choose to assemble the inputs in that manor, it is important that you solder the inputs together when they are mounted securely in thier holes, or I will guarentee they wont fit if you solder them on your bench first the try and fit them.

[mbratch]

Thanks a whole lot. That's helpful.

Come to think of it, after looking at as many 5E3 build pics as I could find, they seem to use all manner of grounding schemes and nobody has complained.

[Enzo]

STingray opens with the point I would make as well. Let me say it a different way. Grounding, like so many other things, is a part of a whole amp, not a thing unto itself. SO it matters where and into what your grounding scheme is placed. The 5E3 is not a high gain amp, so there is less problem with hum than there could have been.

REducing hum is about keeping power supply ripple and hum currents away from the signal path. Tubeswell mentioned microbounces in the ground level. That is a good way to think of it. The 5E3 is relatively quiet, since there is not a lot going on with potential ripple currents in the chassis out of the basic power supply area. SO the brass plate or the pot cover bus can work well there. On some other amp it might not be the best choice.