High Voltage on Chassis - help me with a clue

Hey y'all,

I've been refurbing a Harmon Kardon all tube mini PA. Haven't gotten to recapping the PS yet, just cleaning up the audio path (taking out old crusty pots, unnecessary coupling caps, etc). I wired up a 1/4" jack for a DI and tried playing bass though it. It worked, but I was noticing that if I slid my fingers down the strings, there was a crackling noise like static electricity coming out the amp.

Suspicious of a ground loop problem, I measured the chassis voltage to ground, AND IT SHOWS A HUNDRED AND TWENTY SOME ODD VOLTS!

I immediately unplugged everything and thanked my lucky stars I didn't fuck myself up.

Here's the technical picture so far. No schem to be found for this beast, sorry:

1. The power cord is 2 prong. I got a 3 prong to replace it with, but I'm wondering now if I ground the chassis if I'll just be blowing fuses.

2. The power supply does use a big transformer. It's not one of those scary old transformerless deals where the chassis can sit at line voltage by design.

3. The PS rectification is done with only 2 diodes, not 4. On one of them, the lead to ground is quite blackened, I'm wondering if maybe the diode is leaking to ground. With the unit unplugged, a meter reading shows a 0.5 V diode drop, seemingly OK, but maybe when the real juice is applied there's a breakdown. ??

4. The PS filter caps consist of 4 in a multi-cap can, and 1 other one that's got a cardboard casing. Could these be leaking line voltage to the chassis?

5. There's no continuity measured between either side of the power cord and the chassis.

Thanks for any tips for dealing with this. I'd be happy to run some further tests if that'll help provide a clearer picture.

Leigh

Look for a capacitor (usually .022uF or .047uF) connected between the AC supply input and the chassis. Remove it.

[quote author="NewYorkDave"]Look for a capacitor (usually .022uF or .047uF) connected between the AC supply input and the chassis. Remove it.[/quote]

Make sure you still go with the 3 prong cord even if you remove the cap.

I'm sure NYdave meant this too.

I´m using this caps in one of my projects.

Should I also remove it?

> a Harmon Kardon all tube mini PA.

As Dave says, the idiots at H-K (and many other older gear) put a capacitor from line to chassis. Usually they told you to reverse the power plug in the outlet for least buzz. Usually this left the chassis capacitor connected to the Neutral power wire, which should be a very low voltage to ground.

But sometimes it ends up the other way, cap to Hot.

The cap hold the maximum leakage current to less than one milliAmp. In the 1950s, this was considered "not dangerous". It usually won't kill a healthy person. I think the acceptable leakage is much less today.

As an "improved" (NOT) version, one such mixer I worked with had TWO caps, one to each side of the line. Mostly it never buzzed, and it sure made no difference which way the plug was. But it ALWAYS had 60V from chassis to power ground!

I was running PA for jazz. My system was properly grounded, because the bassist also signed my paychecks and I didn't want him dying before payday. A film crew came in with this mixer and lashed their mikes to my mikes. The musicians complained. Of course it was "my fault", since they knew they were brushing my mikes (and didn't notice that they had also brushed the film crew mikes set a bit back).

I confirmed the 60VAC leakage with a meter, but the film crew didn't want to hear it. They went out for dinner; I whipped the cover off the mixer and confirmed the two-cap connection. I nearly cut the caps but I hate to mess with another man's gear. Instead I noticed an unused Line Out on an RCA jack riveted to the chassis. I got my best RCA cable and ran their line out to a spare hole on my mixer. I told them I wanted to add their mix to my PA feed. In fact I just wanted a hard-ground between their leaky beast and my solidly-grounded system.

I sure was glad I'd done that. We got Eubie Blake in. He was approaching 100 years old, and obviously this was going to be one of his Last Performances. The film crew was there, of course, with their leaky mixer. His mind and fingers were sharp, but boy did his skin and bones look thin, and there were a lot of miles on his heart. I pictured a headline: "Eubie Blake Killed By Soundman" and made double-sure that wouldn't happen.

Find the chassis cap, cut it out. Then test for AC voltage on the chassis to wall-outlet ground, with a 100K resistor in parallel with your AC voltmeter. It should read very low, less than 10 or 20 volts. Any more suggests you have another chassis cap or a leaky power transformer. If you can't get the leakage low, be very careful with this beast! In any case, put on a proper 3-pin plug, green to chassis, and always use a grounded outlet.

That´s it. I will get the caps out!

I never liked it anyway.

I did it based in the original Pultec MB1 schem. That´s why I´ve included the caps.

[quote author="PRR"]As Dave says, the idiots at H-K (and many other older gear) put a capacitor from line to chassis.[/quote]
Old Danish equipment is the same - I think that cap was required by law...

As an "improved" (NOT) version, one such mixer I worked with had TWO caps, one to each side of the line. Mostly it never buzzed, and it sure made no difference which way the plug was. But it ALWAYS had 60V from chassis to power ground!

Click to expand...

In Denmark you generally have 115V on the chassis of everything. The two caps are often used in the RFI filters (required for CE approval). The center point between those two caps is meant to be grounded, but that's seldom the case in Denmark. Nearly all equipment is sold with German power cords (Schuko plugs) that don't have the ground pins needed in Denmark. All most places don't have grounded outlets anyway - my house has one grounded outlet for instance...

So my computer, TV, stereo and other equipment has 115V on the chassis... I don't think it kills anybody.

Best regards,

Mikkel C. Simonsen

Thanks to all for your helpful and speedy replies.

I apologize for some incorrect numbers earlier, which would have affected your diagnoses. The voltage from the chassis to the wall outlet ground is actually 178 VAC, or 81 VAC if the 2 prong cord is plugged in the other way.

Of course, this tells you that it's not leakage from the mains input (which sits at 123V), but from something post-power transformer.

Tracing the 178 VAC output from the power tranny, I found a couple possible culprits. One is a 40 mfd filter cap, the other a diode which might be damaged (as I mentioned above).

However, the third possible culprit is the strangest: theres a wire that runs with the 178 VAC to the audio output transformer. This is a multi-tap transfomer that outputs at several different impedances for several different speaker loads (these are labeled on the rear of the PA). This transformer has 5 input wires: 4 come in pairs from the 2 output tubes, but the 5th is carrying this 178 VAC.

Why would they do this? Run a high voltage (non-signal) line to an output tranny? Also, I should mention, this tranny is of the paper and wax type, so after 40 (?) years, it could be a complete disaster.

Thanks,
Leigh

errr, you need a B+ feed for a push pull tube circuit.

Sounds like the rectifier went south.

Why, that is the question.
If you replace it, you might just be treating the symptons and not the cause.

I got tired of that cap leakage too, so I started putting the .047 across the pri. only, helps but not as good as a cap to one side. bUT AT LEAST NO SHOCK.

Nowdays, everybody uses wireless for the stage, so what a relief not to have to "lip test" for ac!

Eubie might have gone an extra 10 years if he had gotten a little jingle!

[quote author="cjenrick"]errr, you need a B+ feed for a push pull tube circuit.

Sounds like the rectifier went south.[/quote]

Well, these are the things I don't necessarily know, and why I post here.

But it makes sense... after looking at the amp some more, I was thinking, but my "B+ feed" has all that AC voltage, and hardly any DC... and of course, your next assertion, that the rectifier is crapped out, makes perfect sense.

I don't know why or how, but yeah, the rectifier (which in this case is two discrete diodes, apparently w/o a center tap from the power tranny, so halfwave rectifier?) seems to be shot. The leads from the diodes are all blackened, like there was a blowout at some point.

It's very late, I'm hope I'm being clear. I'll try replacing the diodes and see what happens...

Leigh

[quote author="ebartlet"]Make sure you still go with the 3 prong cord even if you remove the cap.

I'm sure NYdave meant this too.[/quote]

Absolutely. Since the original poster mentioned he'd already bought a 3-wire cord for it, it seemed he already knew it was a good idea...

I sketched the schematic of the power supply. I believe this is called a "half wave bridge" style (?).



After putting new diodes in there, I'm still getting the high voltage on the chassis (also 83 volts AC, but no DC thankfully). It's very low current, so I plan to ground that with a 3-prong cord, though, so that should take care of that.

More troublesome to me is the high AC voltage on the B+ feed. I haven't recapped these big electrolytics yet, that would help, no?

Also, I'm no expert, but the use of a big cap in series in a power supply is new to me. ???

Leigh

[quote author="leigh"]I sketched the schematic of the power supply. I believe this is called a "half wave bridge" style (?).[/quote]
No, it's a voltage doubler.

More troublesome to me is the high AC voltage on the B+ feed. I haven't recapped these big electrolytics yet, that would help, no?

Click to expand...

Yes, it probably would... Replace both the diodes and the electrolytics

Also, I'm no expert, but the use of a big cap in series in a power supply is new to me. ???

Click to expand...

It's not really in series. Here's the way the circuit is normally drawn:


You often use it to get a dual supply (+/-) from a single transformer secondary. But in this case the circuit was used to get a higher voltage from a cheap transformer. So the "0" is unconnected, the "-" is grounded and you have twice the voltage at the "+" terminal compared to a half-wave rectifier.

Best regards,

Mikkel C. Simonsen

mcs,

Thank you so much for taking time to explain that. It's much clearer to me now.

Leigh