[Stan the Man]

Hi guys,.. someone clear this up for me.. Philco 38-8, all caps and resistors replaced..
This radio picks up stations and has played for hours,.. 6+ hours continuously for one stretch..
So, I go to hook the ground wire to house ground (outlet face plate screw), and it sparked..
My meter shows 60 volts from chassis to house ground.. It is obviously very little current, because I've had my hands all over it.. I've checked and rechecked everything I can think of.. Transformer seems to be fine..
What the hay?? I read somewhere that this can be typical,.. Really??

Stan

[462ron]

Stan, there are a pair of .015mfd caps across the primary of the power transformer so across the line. This pair has a common connection to the chassis and this will pass some AC to the chassis, thus making for a small spark when you try and ground the chassis.

[morzh]

Yes it is the Y-caps leakage. Personally I use 10nF or less.

And, they have to be Y-type safety caps. This is what in old lingo is known as "death cap", as at the time safety Y type was not available and if those broke down, they often shorted, electrocuting the owner if he touched between the chassis and GND (like if the cap is in an amp and the owner plays guitar and touches the amp, or walks on GND-ed surface).

[Kenneth F. Besso Jr.]

   Also use polarized  PLUG,,,,fat plated Plug

[mikethedruid]

   Also use polarized  PLUG,,,,fat plated Plug

Yes, the wider blade goes to ground, and, by UL convention should be connected to the ribbed or white wire in the cord. The narrow blade goes to the smooth insulated, or black wire. The wide blade is the grounded side of the line voltage, the narrow blade goes to the hot side. Of course, this assumes the house wiring is done correctly, which it isn't always.

[Mondial]

In this case, the polarized plug won't help. In the 38-8, there are two line caps to the chassis, one from the hot line conductor and one from the neutral. Regardless of which way the plug is inserted, there will always be 60 volts AC present on the chassis with respect to neutral or earth ground. The two caps form a capacitive voltage divider, with the chassis connection at the center point.

One way around this is to use a 3 conductor cord with a grounded plug. Then the chassis is always connected to safety ground, and the leakage current is bypassed directly to earth.

[Stan the Man]

Just for giggles I took the safety caps out of the circuit.. Now I get 111 volts to the chassis.. What could be leaking to the chassis now? What am I missing?
The radio seems to work just fine although I need to do an alignment,.. and plate voltage is a little high on 6A8G..

[morzh]

Stan

Why don't you first unplug the set and measure 1) capacitance from either power plug pin to the chassis, and 2) resistance of the same.

Also, inspect the tone control switch. It seems coupled with the power switch, see if there is grime or whatever soiling there that can create leakage.

Then we can proceed.

[Stan the Man]

I have checked so many things lately,.. I know I checked resistance at one point.. I will recheck tonight and post my findings tomorrow.. Thanks

Stan

[morzh]

Basically, we need to establish if there is a leakage through either transformer or some other path.
Capacitive is fine, resistive is bad.

[Stan the Man]

Ok,.. no resistance.. Capacitance on both plug prongs, (.015each)..
Stan

[morzh]

After you removed the Y-caps? Still 15nF on each?

[Stan the Man]

Sorry, I failed to mention,.. I replaced the caps..

Stan

[morzh]

Well....we can infer that since with them it is 15nF each, there is no other capacitive source.

Stupid question, the chassis is floating, right? You do not have it connected to Earth GND? ( Center pin in the outlet). And it is not situated on a gnd-ed conductive surface?

There is a parasitic capacitance of the xfmr to the chassis but it should be fairly small.

I would remove the Y-caps and measure the leakage current. If it is small, the question is then largely theoretical.

[Stan the Man]

Yes, the chassis is floating.. I can check w/o caps again this evening.. When I pulled the caps before is when the 60 volts went to 111 volts on the chassis.. maybe I was looking at something backwards,.. I'll check again.. (transformer and the y-caps) What else would leak AC voltage to the chassis??
I don't know if it means anything but, this voltage is on the red and blk (antenna) terminals also..


Stan