tube amp hum........normal or suggested fix?

[sybednar]

do not rely on the chassis for grounding, all power leads should have a corresponding ground return wire....a star grounding connected to the chassis at only one point is also good...

I've read about star grounding schemes but am still fuzzy about putting one into practice. My understanding is that the placement of where ground point is attached on the chassis relative to the signal sensitive circuits is important.

As I am a newbie could you please elaborate on your recommendation to not rely on chassis for grounding and how I would implement running power ground return wires? Also where would you recommend placing the star grounding point?

 

[jrenkin]

Star Grounding

 

[TonyTecson]

I've read about star grounding schemes but am still fuzzy about putting one into practice. My understanding is that the placement of where ground point is attached on the chassis relative to the signal sensitive circuits is important.

As I am a newbie could you please elaborate on your recommendation to not rely on chassis for grounding and how I would implement running power ground return wires? Also where would you recommend placing the star grounding point?

yes, the entire circuit is earthed/grounded at the chassis where signal is lowest such as input jacks......

to get the essence of what i mean, look at the atx psu, each power lead has its own negative return line that goes back to the 0 voltage point on the pcb...

i also tried grounding the 0 volt side of the psu circuit to chassis at only one point and insulated the input jacks, this arrangement also gave very good results....as long as there is no psu current flowing via the chassis, you circuit will be quiet....

 

[sybednar]

I suppose the picture shows an amp that violates all common sense and construction principles.

Power input cable apparantly not secured against yanking. A few pulls and it may come loose inside and put live voltage on the chassis./tri-comp

I will replace with three conductor IEC socket and power cable. Should the earth ground line be directly connected to the “future” star ground between the main caps or elsewhere on the chassis? Any benefits (i.e. lower hum) from using a ground lift safety circuit here?

Obviously the mains circuit breaker should be a dual-pole./tri-comp

Is dual-pole necessary or would fusing the “hot” line (instead of the neutral) be sufficient?

The bridge rectifier in the power-supply should be connected directly with both + and - to the main filter-cap. (Left in picture) and the negative connection of the cap should be taken to the (future) star-grounding lug between the cap's./tri-comp

What kind of rectifier circuit is attached to the left main filter-cap? It doesn’t appear to be the standard bridge type. Rather it appears that the two AC lines from the power transformer are fed into two sets of tandem parallel diodes that terminate at the + side of the filter-cap. Where is the negative that should be going to the (future) star ground? Should I change it to a bridge rectifier with negative connected to star ground point?


(Future) star-ground lug question: There is a blot between the two main caps with a red/yellow line attached. Is this the site where I should create the star-ground? How should I attach lines here? Should the wires be terminated with spade lugs and connected via the chassis bolt or should I attach a grounded terminal strip and make solder connections to the star ground?

All connections that are now on the negative pin of the cap should go directly to the star-ground except the parallel resistor 470K.
Float (hoover) a dual, bare thick copper-wire bus-bar, soldered to the star-ground, above the small-signal tubes and splitting it left and right above the EL34's. Do not connect far ends to any existing ground-lugs./tri-comp

Stumped here! Even after extensive “googleing” I am not certain where to find a “dual, bare thick copper-wire bus-bar,”. What does it look like? How big is this item?

All components and tube connections with separate grounding-lugs should now be moved and soldered to this bar except the electrolytic cap in the upper right of the picture. It's negative lead should be extended and be connected directly with the star-ground./tri-comp

So, I should add two 100 ohm tandem resistors (1/2 W sufficient?) to each of the pre-tube filament (heater) sockets termini and connect the middle of the series resistors to the star ground. Newbie question: what is the purpose of these resistors?

Filament wiring is hopelessly wrong laid-out. Tightly twisted leads should supply the pre-tubes and have a couple of 100-ohm resistors series-connected across the supply with the center of the resistors connected to star-ground (If the filament transformer doesn't have a centered supply with the center connected to star-ground)
Since the input RCA's are so close to the pre-amp tube I suppose the use of shielded cable from them won't make any difference. /tri-comp

RCA jacks are isolated from the chassis. So connect the grounds from the jacks together and one wire to star ground.

Questions: Why/how is connecting the jack grounds to the star ground (which is connected to the chassis....correct?) any different than directly bolting the jacks to the chassis (ie why bother isolating the jacks from the chassis??)?

I have read that some amps place the star ground next to or close to the input jacks. Why or why not is this a good idea?

However, the RCA's should be floating (NOT connected to the chassis) and the ground-lugs should be connected to the star-ground, directly. Strap from one RCA to the next and on the the star-ground./tri-comp

There's still a possibility that this re-wiring will not cure the hum-problem.
The power-supply transformer may couple with the output-transformers and introduce hum. Not likely, as they are orientated correctly against each other but still very close./tri-comp

How do I implement the suggestion from Indianajo to wrap the power transformer with copper? Not sure where to find copper for this purpose. Might be easier for me to build an aluminum enclosure for the main transformer

I adhered to the above described principles when building a tube-amp recently and it has no detectable hum present. You have to put your ear all against the speaker to detect the slightest amount of hum.
When that is said every amp construction is different and it may be a pain to cure any hum-problem.
Guitar-amps may be laid-out very crazy and still present no hum, even with multiple grounding-points. You'll just have to work with it and see what happens.

Good luck,

rgds,

/tri-comp

Thanks for all the suggestions

 

[astouffer]

Install the 3 prong plug first and then see if the hum goes away. Millions of devices were sold without the use of star grounding and don't have hum issues. Same goes for the foil transformer shielding. Right now anything that needs a ground gets tied to the most convenient spot (the chassis). Star grounding picks one point such as the input jacks and everything that needs grounded gets connected directly to that point. People speak of ground loops but they're actually pretty rare.

To practice good troubleshooting only change one thing at a time. That way when the problem is fixed you'll know exactly why.

 

[indianajo]

like the one change at a time philosphy.
Using the chassis as ground has 2 problems 1. Steel is a high resistance metal, and unless you use a stainless star washer, ring terminal, and stainless screw and nut, the connection is not that good anyway. 2. circulating currents are caused by flux (like from the wall wires, a nearby electric clock, or the transformer) times area. So if the signal current goes down one side of a rectangle, and the return currents go around 3 sides of the rectangle including 2 bad connections copper-steel-copper, then the area of the induced current loop is the whole rectangle, that is big. If a black or bus wire goes directly from the place where the signal wire starts (RCA jack ring) to the input of the next stage(grid), right next to the colored signal wire, the area of any ground loop can be very small. Hammond organ put ground rings with tabs on all their tube sockets (H100) to land the return current wires on next to the grid one (in) or plate (out).
Minimizing flux- Peavey and dynaco both put their 60 hz power transformers in a steel cage inside the steel amp chassis to minimize flux in the air. (Dyna ST70 tube amp has steel transformer wrapper, Peavey PV-1.3k PA amp has the transformer in a corner with a steel bulkhead between it and the high gain input boards.) Peavey doesn't have a steel wrapper around the transformer, but they have a copper wrapper around the paper-coil part, continuing around the e-frame, soldered in a circle. I can't show a picture of the Peavey transformer, but apexelectronic.com the LA surplus dealers has some old transformers pictured that have this feature. You can buy copper shim stock from mcmaster.com, and you can buy bare bus wire from major electronic supply houses. The ground returns doesn't have to be bare, just they can be. I use black myself, except the safety ground attached to the wall plug I use green.
If every outlet you plug your amp in has the hot on the small flat blade, then breaking one input, the black (hot) is sufficient. However, about 1/3 of outlets are wired wrong, with bars and churches that use PA amps about the worst imaginable. So a double break power switch is safer. I had a whole rent house wired upside down (big flat blade hot) when I was trying to fix a television. Ouch!

 

[Pano]

Nice ... no ground and a switched neutral ...

I see switched hot, not neutral. Neutral is white in the USA. As long as the plug is properly polarized, not a problem.

FWIW, I have a stock Dynakit Stereo-35 that hums like the dickens if ungrounded. Connect house ground to the chassis and the hum drops way, way down - tho does not disappear completely. It's from 1964 and has neither power switch, safety ground nor polarized plug. I don't know how they ever keep these quiet back in the old days.

 

[Input2]

Inline stereo isolation transformers should solve the problem.

 

[bobrown14]

I see switched hot, not neutral. Neutral is white in the USA. As long as the plug is properly polarized, not a problem.

FWIW, I have a stock Dynakit Stereo-35 that hums like the dickens if ungrounded. Connect house ground to the chassis and the hum drops way, way down - tho does not disappear completely. It's from 1964 and has neither power switch, safety ground nor polarized plug. I don't know how they ever keep these quiet back in the old days.

Did you try and swap the plug/reverse the plug to ac socket?? Hum from inside the chassis can be had with signal wires close to heater or AC wires which I don't see here. You hum it could be simple like external AC cables close to the input cables. I see there is a line filter so that should tame the noise from the AC input. I'd look close at your input cables from the source and be sure there's no AC 120v cables lying parallel to the signal cables. That will induce ac hum at the signal input. Layout doesn't look too bad ... I dont see anything internal layout wise to blame for hum. MHO ... Could be house ground or anything external to the kit.. I'd move some cables around and see if that helps. OP is in the USA; all kinds of skiddish wiring here, specially it its an older home.. 120VAC is nothing like 220VAC. As mentioned we've had 120Vac without ground for A LONG time. Also try the kit on another house circuit. Use an extension cord if you can. Could be house wire related (reverse the plug thats a good way to tell).
Cheers,
Bob

 

[sybednar]

update on hum problem....BUT....more help needed please

On the "sound" advice of astouffer I've made only a few modifications to the amp to try to clear up the hum problem:
1. Installed IEC power connector and connected 120VAc "hot" through fuse and new switch. Earth ground from IEC connected to center star ground between large caps.
2. Disconnected ground from RCA jack inputs that ran to ground on large electrolytic caps and connected them directly to center chassis bolt "star ground" to which earth ground and power transformer center tap (red/yellow striped wire) are attached

These changes did two things:
1. The hum in the right channel is gone (this is wonderful)
2. The left channel now has a buzzing sound (this stinks and is worse than the original hum problem).

The buzz is there with or without any source connected to the inputs. It can only be heard in the left speaker when the neon (I assume current limiting bulbs) stop glowing and disappears the moment I shut off the main power. Also the buzz doesn't appear to be tube dependent. Moved tubes from right channel to left and vice versa and hum stayed in left channel. Replaced all six tubes (4 power and 2 smaller preamp? tubes) and buzz is still in left channel (no change in level)


I think the buzzing in the left channel is 120Hz not 60Hz hum but I don't have an oscilloscope to directly test this. However the fact that the buzzing dies immediately after I turn off the power makes me think it is somehow related to the HV power supply? But why only in the left channel? Also there is a single electrolytic cap in the right channel circuit (upper left corner to pictures I previously posted) but not one in the left channel circuit connected to power tube to ground. All other components with the exception of this cap are mirrored in the right and left channel circuits. Could the lack of this cap in the left channel be related to the buzzing problem?

Help on how to diagnose and correct this new but likely related problem in my amp would be greatly appreciated
Happy New Year and thanks to everyone for your past advice
Sebastian

 

[wiredwithtubes]

2. Disconnected ground from RCA jack inputs that ran to ground on large electrolytic caps and connected them directly to center chassis bolt "star ground" to which earth ground and power transformer center tap (red/yellow striped wire) are attached

These changes did two things:
1. The hum in the right channel is gone (this is wonderful)
2. The left channel now has a buzzing sound (this stinks and is worse than the original hum problem).

Is that all you changed, move the ground wires from the input jacks? If so check the ground wire between the left jack and ground. You might want to check those input jacks, they look like they might be insulated from the chassis, one might be touching the chassis creating a ground loop.