HP 895A power supply

[N4LTA]

Had one of these a few years ago and it worked will for tube type work.

Found for on EBAY cheap except for shipping and bought it.

Plugged it in in my new shop with all GFI circuit breakers  and the GFI breaker tripped immediately. Before it tripped , though the pilot lamp came on and the meter moved.

I ran a drop cord from a part of the house with non GFI breakers and it comes on and seems to work , but likely needs the caps changed and a tune up. It is easy to get into and replacing parts should not be a problem.

I had to order a manual. I found a manual online for a similar power supply but with 1/2 the current rating. This schematic shows a three wire cord with the green wire grounded to chassis  and the black and white wires through a fuse and switch and then directly to the two power transformers. The pilot lamp is directly across the power transformers primary. The supply that I have is wired exactly the same but uses a 10 amp fuse where the half sized schematic shows a 5 amp as you would expect.

The main power cord goes to a terminal strip and has two .1uF 400 volt Sprague Black Beauty caps from black to chassis,  and white to chassis. The green power cord wire is bolted to the chassis.  This looked like a probably cause for tripping the GFI.

I clipped both out and the GFI stopped tripping. I replaced the two Black Beauties with new Cornell Dublier  .1uF  630 volt film caps and assumed it was fixed. Plugged it in and immediately tripped the GFI.

Was this type of line cord connection normal in the 1960s (using  three wire cord as shown in the schematic) , before GFI circuit breakers?

0.1uF is about 25 K ohms at 60Hz, and the pass current from the Black wire to ground would be about 4.8mA pretty close to the 6 mA required to trip a GFI. At the peak of the 120 volt AC voltage it would rise to almost 7 mA. I guess that is what is happening.

I guess I will use a couple of .05 caps and cut the current in half and see if that fixes the problem.

Any comments would be appreciated.

Pat
N4LTA

[w8khk]

I think you have a good plan, Pat.  If it still trips the GFCI device, there is another way to resolve the issue.

Place one capacitor from neutral (white) to ground (green).  Place the other capacitor from hot (black) to neutral (white).  This method should still provide good RFI suppression, but should NOT trip the GFCI device.  

The GFCI is measuring the current on the neutral and the hot leads, and trips when there is a difference between the two.  It does this by passing both the neutral and the hot (1 turn) through a toroid, and measures for a difference on the sense winding.

The original capacitor from hot to ground may cause enough imbalance to cause the device to trip, but the capacitor from neutral to ground should not cause a measurable current on the neutral lead, since both neutral and ground are very close to the same potential.

I normally use disc ceramic capacitors for the power line input filtering, as they typically have lower inductance and provide better RFI reduction, and they rarely get leaky with moisture.  But I still evaluate and replace if necessary when servicing older equipment.

I prefer to place the hot lead capacitor after the line fuse, and I do not fuse the neutral.  Grounded polarized line cord plugs eliminate any issues, so long as the hot and neutral are wired to the correct terminal on the receptacle and the line cord plug.

A little-known fact is that the GFCI protection device will perform its duty, even if there is no ground lead wired to the GFCI receptacle.   Of course it is always safer to use a properly grounded receptacle and line cord, but it is interesting to note that if there is a ground fault, through a person's body, or otherwise, this current from the hot lead to ground, whatever path it takes, will trip the GFCI device, even if the device has no ground connection.  There is no connection from ground to the circuitry in the GFCI device; it does all its magic by detecting a difference between the neutral and the hot, and the cause of this difference is some of the current on the hot is not returning through the neutral lead in the GFCI device.

Hope this helps,
73, Rick