Welcome to Solid State Guitar Amp Forum | DIY Guitar Amplifiers. Please login or sign up.

March 28, 2024, 11:02:29 AM

Login with username, password and session length

Recent Posts

 

Gibson G 30

Started by gbono, December 27, 2012, 01:41:35 AM

Previous topic - Next topic

gbono

I've run into a hum issue (60Hz) on this amp. After fixing the blown output section and replacing the leaky electrolytics this amp is working again. I upgraded the line cable to a grounded type and when the ground is attached to the chassis I get a very noticeable increase in 60 Hz at the output.

The ripple on the voltage rails is .025V but I get almost .3V on the output terminals with the chassis grounded. Ground loop?

J M Fahey

Yes.
Probably you have a (new) ground loop between the amplifier and "something else" (pedalboard/PC/whatever) .

gbono

seems odd since I've never seen this hum issue on other amps using the same mains circuit. The outlet goes directly to the main service and there isn't anything new that's been plugged into the circuit.

I noticed the treble control halves the noise voltage when it's turned to minimum - filter action I guess. Also notice the hum signal shows up at the output of Q7 and then Q8?


gbono

See attached - I'm thinking of adding the following circuit to eliminate the noise issue. I can use fiber shoulder washers on the input jacks - anyone try this?

phatt

Can't hurt to try,
Yes I've done it to some gear and it does help a lot but if the circuit is not wired up well you may still have some hum.
Schematics are not much help in this situation as it's likely a layout/wiring issue.
You have to make sure every com thing in the circuit is lifted from the case.
Phil.

gbono

#5
There is a voltage between neutral and ground at the outlet of .120 V and there is 105V between the ground pin and chassis with ground wire lifted. Why is there such a large voltage between the lifted ground pin and chassis? I'm assuming there is a short in the insulation of the PT?

J M Fahey

*Maybe*
Or perhaps it's damp.
With ground wire momentarily lifted, connect main and neutral both ways, and use a neon bulb screwdriver to see which is the position with less AC leak , as was made in the old 2 wire amps which had a 3 position power switch (On-Off-On) and after that re-connect (forever) the ground wire.
I think you'll get the best compromise: a safe amp with the least ground hum possible.
And yes, it is possible that the PT insulation is somewhat leaky.
Usually, as I said (and typical of old amps unused for ages) , because of dampness.
Many times it "cures itself" leaving the amp on (no need to play) for a couple hours every day a few times.
Just don't leave it on and go to sleep, you don't want nasty surprises in the middle of the night.
Not that I expect it to burst into flames or something, but don't leave unattended equipment on if at all possible.

Roly

@gbono - open the circuit pdf (rotate upright for sanity) and look at the mains input bottom-right.

You will notice a) there is no ground wiring shown, and b) there is a mystery cap C12 shown between the fused line and the chassis.  If we presume the fused line is the mains active then finding a highish voltage on the chassis with a high resistance meter is no surprise - it is being coupled through C12.

In Australia we inherited the British LEN or Live, Earth, Neutral three-wire system from the outset, whereas in the US the system was initially an "economised" two-wire system of active and neutral only.  Moreover, for reasons that are obscure to me, the US opted for an un-polarised connector, meaning that the active could turn up randomly on either mains wire in the amp.

Many amps including Fenders were fitted with a three position mains switch that was "off" in the middle and "on" either way, but the difference was that this "killer capacitor" (C12) was switched to one line or the other, allowing it to be selected to the neutral line.  This stopped "noises off stage" and getting "tings", mild shocks, from touching other equipment.

In LEN practice we don't have this "killer capacitor", and in fact amps with three-way power switching like the Fenders are modified by importers to have a three core LEN mains lead with the safety earth bonded to the chassis, and the "killer cap" removed altogether.

So, you go three wire mains lead, securely bond the earth lead to the chassis, and remove C12.

Your stray voltages should vanish, and hopefully your problem with "noises off" as well.

I'm really glad to see the US is slowly moving towards the safer LEN standard, and 'tho 100 years late, better late than never.   :dbtu:

The idea shown in Hum.pdf is known generically as a "ground lift resistor", 'tho this is a fairly fancy implementation of the idea.  This is rarely required outside studio settings where an awful lot of mains powered gear is cross-connected with screened signal cables, and even the slightest hum from a ground loop can be a major problem.

Most techs see ground lifting as a sometimes necessary evil, and power supply authorities tend to be even less impressed, so we try to avoid where possible.  I note the remarks about audio signal isolation transformers, 'expensive, or no good'.

Again that may be true if you are kitting out a studio, but for the individual guitarist I would disagree because I have found signal isolation transformers generally to be inexpensive and of quite satisfactory quality for guitar (remembering that guitar doesn't need extended bass and treble response, 100Hz to 5kHz normally being more than sufficient).

Modify your mains input to LEN as above (using lots of heatshrink and cable clamps) and see how that goes before applying ground loop cures.  Apart from isolation transformers, there are other tricks you can (and should) apply before going to mains ground lift resistors.

Since your are re-doing your mains input, there are truckloads of old computers and dead power supplies being chucked into land fill every day.  All of these have EIA mains connectors free for the recovery (my local computer shops are happy to give me armloads of dead 'puter PSU's, and it's never the mains connector that's the problem).  Some of these have a modular mains filter built in;



... and these are very suitable for guitar amp service - provided you have an earthed chassis. {and you get a lifetime supply of high voltage caps, coloured hookup wire, fans, and project cases thrown in.  :tu:}

HTH
If you say theory and practice don't agree you haven't applied enough theory.

gbono

#8
Thanks JM,

   If I switch the line and neutral won't the fuse and mains switch now be on the neutral side? BTW the hum diminishes with line and neutral swaped but now the mains switch and fuse are on the neutral (L2) leg.  :-\

GB

gbono

#9
Thanks HTH,

  Yes I removed C12 and have a proper grounded mains cable installed. The hum still exists and when I reverse the line and neutral I get a nice drop in noise (hum) level. Also tried the ground lift resistor with no benefit. Issue now is that the neutral leg (L2) is now connected to the mains switch and fuse  :-\

GB

Roly

I'm Roly.  HTH = Hope This Helps   ;)

You have taken out C12 and you still have a sensitivity to which lead is mains active?

This should not be.  Have a careful check of the mains wiring to see if there isn't another "C12"-like cap hidden somewhere.

Otherwise, have a look at how the mains wiring is dressed around the signal areas.  It (and specifically the active) should be dressed well away from any signal leads, for both safety and for noise pickup reasons.

It's possible you have some mains leakage in the transformer contributing to your problems, and while I think it unlikely JM is dead right about just running the transformer so it keeps warm and "bakes out" any moisture.  You can try isolating the mains side, then measuring the resistance from the mains wiring to ground/chassis, but you are unlikely to detect anything other than gross leakage with a normal multimeter - you need a high voltage tester known as a "Megger" (megohm meter) to do this properly.

You're right, it isn't good practice for the mains switch and fuse to be in the neutral side, remember however that when it had a two-wire mains lead (and non-polarised plug) it was being randomly assigned every time you plugged it in.  Again, in the LEN system the active and neutral both have an assigned side.  You do encounter outlets that are cross wired here, but it's thankfully rare, however given the history it may not be so rare in the US.

It may be a silly thought, but you could run the mains active to the switch and fuse as it should be, and try swapping over the transformer primary (if you have enough cable length &c).  If you still have the hum then it looks more like cable dress than the transformer.

Any chance of posting a pic or two of the area? (using lots of light)

HTH,

-Roly

If you say theory and practice don't agree you haven't applied enough theory.

gbono

Hello Roly,

   See picture of the rats nest :-[ The mains switch is ganged onto the volume control so the AC is routed on the PCB (see fuse holder). There is also a run of wire which is used to power the neon light on the front panel.

With the L1 and L2 swapped the hum is barley noticeable.

This amp was stored on a porch so I guessing that the damp PT may be an issue. I do have access to a megohm meter but I don't know if it worth taking the transformer off the chassis.

THX

GB

Roly

Gaakk!  Let's just say your pic has left me speechless, shocked, and appalled.  No wonder you (and likely every other owner) is having hum troubles.  This is a serious example of bad practice and I'm astonished that a name like Gibson would produce such an abomination.   :o

I'm never mad keen on mains wiring on PCB's at the best of times, but to thread it across where the input stage is located is nothing short of stark raving nuts and an open invitation to exactly the problem you have.  Apart from being terrible audio engineering, for safety reasons this arrangement would not get type approval in Australia, and I very much doubt that it complies with Euro C-tick standards either.  Does the UL really accept this kind of crud I wonder?

Fail and -10 for Gibson.   :duh

You say it was stored where it might have got a bit of moisture?  It is much more likely that your problem is tracking through surface contamination on the copper side of the board than the tranny.  Did this happen to be anywhere near the seaside, where any moisture may have contained some salt?

You could just wash the board with some metho, but my inclination would be to get the mains right off the board altogether.  If it were on my bench I would cut the two connections from the back of the pot to the PCB and turn them up into tags, lift the mains and tranny wires from the near side, add a mains rated in-line fuse holder like this;

http://www.jaycar.com.au/ShowLargephoto.asp?id=3987&PRODNAME=3AG%20In-Line%20Fuse%20Holder&IMAGE=

...and loop a pair of fairly stiff mains-rated wires up well clear over the board to the pot switch.  Naturally the whole job should end up with a good application of heatshrink tubing to all otherwise exposed mains connections.  Won't even need to take the board out to do that.

To test your tranny you don't have to disassemble anything, just clip one side of the Megger to the chassis and the other side to the mains plug.  However, after seeing your pic what you may be reading is leakage across the PCB rather than in the transformer itself.  Also that style of fully enclosed tranny is pretty unlikely to have picked up any significant moisture (unless it was allowed to soak in a swimming pool perhaps).   ;)
If you say theory and practice don't agree you haven't applied enough theory.

gbono

Roly,

I toyed with the idea of using a relay to control the mains voltage using the switch on the front panel and an external fuse holder, etc, etc.

The other issue I have with this amp is crossover distortion. There is a pronounced "notch" at the origin when looking at a 1000 Hz sinewave at the input. When playing a guitar through the amp you can hear a raspy sustain almost like a "ghost note" caused by the intermod that I'm assuming the crossover distortion is the source of.

THX

Roly

Yeah, I toyed with the idea of a relay too, but first you need power to turn on the relay to turn on the power.  ???  Unless you put a battery in there.   ::)

Nah, just get the mains active off the board near the sensitive input stage.  My guess is that this will be more surface leakage than radiation, but with the conductors off the board if it is still a problem (how did they get these out the door in the first place?) then adding some sort of screening around the power leads near the board would be the next step.  But I strongly suspect that once it is off the board the problem will disappear.

Of course mounting a fuse holder and whole new mains switch somewhere on the back panel would also get the mains wiring away from the signal section (like it should have been done in the first place ... speaking of which ...)


Crossover distortion - that's not entirely surprising when you look at the output stage circuit.  We have the E-B drop of the two output transistors, Q2 and Q9, two diode drops plus the drop due to the idle current (if any) in the emitter resistors R16 and R16 (sic, yep drafting error, you never find just one roach in a kitchen).

These two bases are biased apart by ... D1 - one diode drop.  That's it.  So the stage is clearly under biased.

Now you could just stick another diode in series with D1, but this is a bit tricky because if you overdo it there is a real danger the output stage pair will go into thermal runaway.

I'll bet if you measure the idle voltage across R16 and R16 you won't see much at all, meaning there is little or no idle current.  This keeps it cool and efficient, but as you have noted it sounds like crap.  Typical idle currents are around 30-50mA (and you will have to work out that as voltage since the circuit is a value-free zone; typically 0.2 - 0.5 ohm 5 watt).

What I do, rather than using normal diodes for biasing, is to use the B-E junction of a BD139 (or two).  These come in a flat pack with a hole which makes them very easy to attach to the heatsink near/under the output transistors so they are in intimate thermal contact and track the transistor temperature.  Sometimes you can even get them on one of the power transistor mounting bolts. (remember, the pad has to be insulated as it is connected to the BD139 collector)

In short, I think you can improve the crossover distortion a lot; but remember you are tweeking the tail of a tiger and to proceed with due caution, alert for any sign of thermal runaway until bench torture tests prove the amp is thermally stable.  You will be giving it a bit of design it missed out on (and may even devise a publishable mod to improve a sow's ear).

I had to do this with a different Gibson for different reasons, but the approach is the same, and discussed in some detail here;

http://www.ozvalveamps.org/repairs/gibson.htm

HTH
If you say theory and practice don't agree you haven't applied enough theory.