AA764 champ proper grounding scheme.

[goldstache]

Hello!
Built a champ.  Sounds great, however upon firing it up there is a slight hum audible through speaker cabinet.  This noise does not get louder with manipulation of volume control.  It just stays the same volume as when the guitar isn't plugged in and input shunts to ground. 
I built it with a JJ 20uF + 20uF + 20uF + 40uf can cap. Which has a shared ground lug for all internal caps.  I am wondering if I'm getting some stray currents or something that is transmitting through the chassis and on to my OT secondaries.  I know typical grounding practice separates grounding points after first filter cap.  Any thoughts on this one? 
Also, I have a scope so I guess I could diagnose that way, if someone would be kind enough to walk me through it! 
I am strong in the ways of the Newb

[goldstache]

Well here's an interesting cross reference.  My original 63' blackface AA764 champ (which is pretty much stock internally, needs gone over but works) has the same hummmmmm.  I must admit in my build its a little stronger (maybe because its all new working parts) but not by much.  It's also the same Frequency range.  So I think this is me just chasing ghosts.  It's amazing how much scrutiny you put into your builds when they are done.  This never bothered my with the old fender. 
Any thoughts or tips would still be greatly appreciated!
Thanks for being there to help everyone!

[sluckey]

How are your heaters wired?

[HotBluePlates]

How are your heaters wired?

I don't think it's heaters. I think it's lack of filtering.


Everyone around here that has built a Champ in the last few years got some hum as you have. Basically, tapping the 6V6 plate voltage from the first filter cap just isn't clean enough d.c. to be hum-free. Push-pull amps get away with using this cap because the push-pull OT cancels hum from the power supply.


The answer has been to add a stage of filtering between the rectifier output and the existing power supply. A 20uF cap to ground and a 150Ω resistor between the + of the new cap and the + of the original fist filter cap. The 2 stages of filter for the 6V6 plate generally kills the noise. If you had the extra parts, you could even rig up a tacked-in version of this to verify it works before committing to installing them.

[goldstache]

Heaters are twisted pair.  And they are flush with chassis.  My PT has no center tap, but adding 100 ohm artificial CT did Nada!
However, they do go around the tube socket exterior against the chassis.  Not exactly a great idea but when on it's really noise free until you are at 80% volume than the noise floor comes up.  Pretty standard from what my ears can discern.

On to more filtering.  I have a spare 40uF left unused in the can but that might exceed my needs, correct?
I will mock up the RC filter prescribed above and see if it helps!  Can I get away with 3 watt?

[goldstache]

And just to clarify:
 
5y3> B+> 20uf> 150 ohm> 20uF and OT CT?

[HotBluePlates]

And just to clarify:
 
5y3> B+> 20uf> 150 ohm> 20uF and OT CT?

Yes, that works.

Can I get away with 3 watt?

I don't know what bias you get across your 470Ω cathode resistor; the schematic shows 19v, which implies 19v/470Ω = 40.4mA. Power dissipated by the 150Ω resistor will be (according to one permutation of the Equation for Power) 40.4mA²*150Ω = ~0.25w. So yes, the 3w will be plenty. Total current draw through the resistor will be a bit higher, but not terribly much more.

If you have other 3w resistors up to aound 220-270Ω, you might try them. Higher resistance increases the filtering effect but also increases the voltage drop across the resistor and power dissipated by the resistor.

[goldstache]

Wow that worked.  About 6 VDC drop.  But took care of most of the noise to my ears.  Gonna write all voltages down and crunch some numbers.  Thanks so much.
Do you think it would be over filtered if I wired it up using the 40uF?
Well back to Math!

[goldstache]

Oops!  Just checked 5Y3 datasheet.  Looks like 20uF!!!!

[PRR]

Use the 40 as the -second- cap, *after* the 150.

Then the 5Y3 "sees" mostly the 20, not the 40.

Also-- 20uFd is NOT a limit for 5Y3. What you need is a combination of PT resistance and filter capacitor. The computation is painful. Anyway at the time the 5Y3 was introduced (and the specs written), anything over 16uFd was absurdly expensive, so they didn't bother listing bigger caps.

Personally, since the Champ is never going to bust the stadium, I'd be generous with the added dropping resistor. 6V is nothing in 300+V. Something the size of your Cathode resistor will filter more; you may already have one. That may drop 20V-30V, but still "small" compared to 300+V. The 1dB drop in maximum output is hardly audible. Still a big bedroom amp. In the studio your engineer can easily cover the difference so it is all the same on the record (except the hum).

[HotBluePlates]

Yep, what PRR said! 

What we have found on this forum over the years is that modern builders assembling their Champ often use speakers that have better bass response than the original 8" speaker (or they use larger cabinets). They are also more critical in their evaluation of the amp's noise, mainly because they built it and might be looking for errors on their part.

It adds up to needing an extra stage of filtering before the 6V6 gets supply voltage to minimize the hum which is now more obvious than it might have been with the old amps.

[clyde]

My Gibson Falcon uses 40uf as the first cap after the 5Y3, has since 1964.  Being as the Champ is class A, is the size of the first filter more of an issue?  Thank-you.

[HotBluePlates]

My Gibson Falcon uses 40uf as the first cap after the 5Y3, has since 1964.  Being as the Champ is class A, is the size of the first filter more of an issue?

No. If anything, Class AB amps probably pull bigger pulses of current through the rectifier, especially at full power output.

[goldstache]

Good stuff here!  Thanks everyone! 
Wondering if over filtering could cause any low frequency oscillations.  All signal coupling values are stock.
Also, I tried a plug in silicon diode rectifier and was a bit surprised that my voltages were higher with the 5y3.  I was under the impression that solid state rectification provided a bit more B+. 
Apparently not the case!

[PRR]

> a plug in silicon diode rectifier

These are often built with added resistors inside, to "approximate" the loss of a hollow-state rectifier.

Evidently their added loss is greater than the loss of your 5Y3.

I would say, generally, we "do not care" what the exact B+ of a Champ is. Should be well over 250V so nothing is starved, and 6V6 should not be red-plating nor dissipating a lot over 14 Watts (measured). Anything in between is fine.

[Fresh_Start]

You also can use a choke between the "reservoir" cap and the cap at the power tube plate. Champs don't draw that much power so a choke isn't all that expensive. See 5E1 Champ power supply. Only 8uf caps but a choke.

It's been a long time since i played around with Duncan's PSU software, but IiRC PRR's suggestion about using a larger cap second in line does a better job of filtering than the larger cap first. I never did figure out how to use PSU for modeling actual component values. However, it was very helpful to see the relative benefit of different options.

Don't forget that more filtering might well reduce "sag" and touch response. This is an essential attribute of the original Champ circuits. It's not supposed to sound hi-fi. But we all want 'em quiet.

[goldstache]

Really loved the way this champ is sounding. 
Until......... today. 
I used to be able to dime the Vol, Treb, and Bass controls without any problems or oscillations.  (keep in mind its a stock AA764 Champ just no NFB at all)


Started hissing cutting out during last 10% of treble rotation with bass and volume maxed!!!!!
I added negative feedback and it still did it at the top of the rotation.


Whas' happening here?????????
Did a lot of circuit analysis and voltage readings over the course of the last few weeks to make sure everything is within constraints (plate current, plate dissipation, output wattage, all voltages, etc.)

[goldstache]

Well I have it isolated to V1B's plate.  Thats where the oscillations are loudest.  The lead dress seems to cure it.  Though even the slightest manipulation brings it roaring back.  The amp didn't do this for about 20 hours of on/playing time.  I even swapped out an entire tube compliment just in case.  My 6v6 is overbiased as i don't have anything higher than 470ohm 10W on hand.  Probably should go higher to reduce current. 


 It is stable currently but only with the lead off the turret board, heading to V1B's pin 6 which is oriented furthest away from turret. It has to float in a precarious fashion.
Any insights?
All ground connections measure 5ohms or less
I just find it strange that it was so solid for so long before developing this issue.  Break in time?????????
Thanks Everyone.

[sluckey]

Post some hi rez pics.

[goldstache]

The Red wire coming off the 100K's is the culprit.  Pin 6.
Thats the node where the oscillation is strong and the lead acts as an antennae

[goldstache]

Here is a crap schematic.  Fender Design with other thoughts
V1 is located bottom right of photo above.
 
Could it be that my B+ for the EM80 was taken off of the (A) power dropdown?????????

[goldstache]

And 1 more

[goldstache]

+1

[HotBluePlates]

You've got all those nice multi-colored output transformer wires right up next to your preamp tube. That seems like a bad idea (output/high level signal next to input/low level signal).

I would get the impedance select switch well away from the preamp tube, maybe between the 6V6 and 5Y3 (if you even keep it at all). Drill a hole in the chassis on the surface the OT is mounted on, near the 6V6 and very close to the sidewall that the speaker jack is mounted on. Route the OT secondary wires to that hole outside the chassis, so they only come inside right near the new impedance switch location, for a short routing inside the chassis well away from the circuit board and preamp.

Hopefully your OT wires will reach. You may have to re-orient your OT to get things to fit. Nothing says you can't use one of those bolts holding the board in place to also hold one side of the OT (assuming the bolt is long enough).

The long wire to the eye-tube may not be ideal, but V1 pin 6's wire runs right under the impedance select switch so it would seem speaker signals don't mix with preamp signals. The high current of the speaker signal at the OT secondary wiring & impedance select switch will set up a field that could couple easily into preamp circuitry/wiring.

[goldstache]

HotPlateBlues!
Thanks for the confirmation of what I already expected!  I built this puppy in a small chassis and expected some problems!  As the wire sits in photo is position where oscillation does not occur.  Originally it was very similar a run to the other plate wire.  I will heed your suggestions and try moving the OT.  I would like to keep the Imp selector, but if it causes problems then out the window it goes!

What boggles my mind is:  why would it perform without any issues for a couple weeks then Voila!   Huge parasitic noise!   

Thanks so much for your time!  I'm learning and it feels great when it makes sense!

[HotBluePlates]

Oscillation is a weird thing... Many factors impact when and how it shows up.

Additionally, you can have oscillation that's not obvious, such as when it occurs at frequencies above the range of hearing. That could show up as a raspiness to notes, or it might not be apparent at all.

Also, if you turn the volume down, power to the speaker wiring drops and so does current on those wires. It's possible to not have oscillation at low volumes, but then it shows up when you crank up the power.

To be clear, the impedance select switch doesn't cause a problem by itself; its position relative to the preamp tube bring the OT wires close to that tube and the preamp section of the board. Moving all that stuff further from the preamp tube should help solve the oscillation.

[goldstache]

Eureka!  There was some top end fizz when cranked!  I thought it was the speaker I was using!
Thanks again!!!!
On to an 18watt