I am restoring a Atwater Kent 84. After replacing three open resistors and cleaned up, the radio worked very wall. After a short time the radio signal started to fade and distort. To bring it back to normal, I discovered that if I momentarily grounded the chassis it would shack it back to normal. There is no lose connections. The antenna is good.

I have not replace the capacitors because of the brittle wire insulation. Could it be a bad capacitor? What should I look for?

Richard

Yes could be a bad capacitor. Could also be another open resistor. An open grid resistor can act like that. Radio plays for awhile then fades out. Shorting a wire may start it operating again. A tube can go into cutoff with an open grid resistor.

Best to replace caps and brittle wire. Replace one at a time, less chance of an error.

There are four capacatiors cases that the capacatiors are embedded. Is there a easy way to replace the embedded old capacatiors using the cases?

Richard

You could remove the original capacitors with heat. The Atwater Kent cans are sealed by soldering. Mount replacements inside. No easy way. A simpler way would be to cut wires and add replacement caps under the chassis.

Schematic shows what is inside these cans:

http://www.nostalgiaair.org/PagesByMode ... 001491.pdf

Simple enough to restuff them. As mentioned, the hardest part is removing the cover, which is soldered on. I have also found that some .1mfd/630 volt caps are too large, and 4 will not fit inside the case. The brand that WILL work is by Richie Capacitor, sold by Mark Oppat.

I wonder if some of the ceramics might fit in there as well? Many of them are quite flat.

There are occasions with "Dry" joints that the spark from the back EMF of a coil may cause it to work temporarily.

Get rid of the brittle insulation, that is a time bomb. Cloth covered wire looks nice. If it looks like it is going to fray on the ends (as it does) circuit board laquer, clear adhesive and a small band of shrink tube, are ways of reducing that.

Marc

The ceramic resistors in that vintage of AK's are notorious for drifting and intermittents. The crumbling wires are even more of a problem with voltage breakdown and shorts. The caps are generally healthier than those of other manufacturers but that doesn't mean that they don't leak. There's really no simple solution to your problem; the 84 is of the vintage that takes quite a bit of time to get working well.

Pete AI2V

I lost my B voltage! I checked the two 8 mf filter capacitors(new), speaker, rectifier 80 tube, and associated resistors. All okay. The last thing I was doing was replacing the wire between the oscillator transformer and the #1 dec. plate choke. Power transformer tests good with proper voltages.

The AF bias resistor, should be 500k, measures only 61k and the oscillator grid leak resistor, should be 100k, measures 40k. The two filter resistors that I replaced with a 7.5k and 5k, now measures 3k each? I feel I have a ghost from the past!

What other checks should I make?

Richard

Is your speaker plugged in? Without the speaker field negative of the supply will be open.

Read voltage across first 8 mf filter cap, next to 80 rectifier. If you don't have voltage here check the wire wound resistor marked 1100 ohms in the schematic. May be open?

Resistors in circuit can read lower, through other parts of the radio.

Yes, speaker is plugged in.

Voltage across 8 mf capacitor #1, by tube 80 is 222vdc, across 8 mf #2 is 16vdc.

Voltage at tube 80
Pin 1 to 4 - --- 5v ac
Pin 1 to ground ---- 16 v dc ( the 16v dc shows up on other tubs)
Pin 2 to ground ---- 497 v ac, 396v dc
Pin 3 to ground -----497v ac, 396v dc

No sound with slight hum?

Since you have some voltage across the second 8 mf cap speaker field is probably ok. Try disconnecting the second 8 mf. Believe it may be shorted?

I do not think it appropriate to re- start a debate on reformers but...............

In supporting Norms veiw. There are two things I can do related to caps that apply here. Aside from the fact that old paper and Electrolytic caps are better off replaced: For leakage testing purposes, a reformer with voltage steps allows you to put increasing ammounts of voltage on a cap (out of circuit) & measure it's leakage (the regulator (LR8) on mine will lock up on short).

This is also handy for "prooving" newer caps. Silly as that sounds, I have seen and had to replace caps, mainly 450V ones that have failed within two years. As a side line, one of the failures 450V (electrolytic), built up nicely until it got to a round 200V and then started to conduct.

For the normal run of tube non polarised caps (Mica & Paper) an insulation tester will prove very useful.

Do not over look that any of that brittle wire can be shorting, especially on holes. I recently had a HV short under a large 8mfd paper cap. It cut into the 5V filament wire and I have found brittle rubber wire going into transformers.

Be carefull: A short can kill the transformer. Pull the rectifier & detirmine which side of it the short is: Do not operate 80 / 5Y3 recifier sideways.

Marc

The #2 cap is good, measures 7.8mf. The two 8mf, 450v caps are new, all other caps have not been replaced. The measured -15vdc on rectifier tube pin #1 and #4 is a puzzlement? What do you think of the voltage values I have on the rectifier tube? Tomorrow I will start all over with my frequency generator to try and isolate the problem!

Richard

Any chance the second 8 mf cap is reversed? Positive up in the schematic. Negative to ground/chassis.

Rectifier is ok but field coil on your speaker may be open? You have 220 volts DC across the first filter cap. The field should read 2000 ohms on a meter. Since the speaker is plugged in this field may be open? There is another path which will give a little voltage, two divider resistors.

For now don't need to use a signal generator. Using your volt/ohm meter will be enough.

Field coil is exactly 2000 ohms. Input transformer is 489 ohms. I have my meter out, now what?

Richard

Field coil completes the circuit from negative of 1st filter cap to negative of second, chassis. Either something is open between the two on there is a short. If a short the speaker field should be getting hot.

There is a cap marked .1 mf between 1st detector tube and IF tube on the B+ line to chassis/ground. Is this cap shorted? I don't see anything else that would cause a short from B+ to chassis. Use an ohm meter between B+ (positive of 8 mf filter caps) and chassis. It should read above 10K. See if you can find what is causing this line to have low resistance to chassis.

That circuit shows the set to be backbiased. Possibly some form of candohm? Measure the voltage between the actual CT, not chassis and a filament of the 80. If you get something around 400+ volts DC the 1100 ohm Backbias resistor is possibly open.

Looking at the rating on the caps I hope the new electrolytics are in the order of 500V minimum.

Marc

While doing voltage testing the #1, 8 mf cap blew! (puff of smoke and gone) I know it was in the right way because the radio worked fine at first. The cap was rated at 450 volts. It might have been weaken with the problem I am having.

Before the BLOW, I checked all caps for shorts, none! Backbias resistor = 1102 ohms. Positive of 8 mf = 13.5k ohms. The .1 mf #8, not shorted. I am getting discouraged!

Just curious who stocks recent manufacture 8 mf 450 volt caps? If they are NOS from decades ago, they'e probably as bad as those you replaced. 10 mf 450 or better caps would do fine here. Don't be discouraged, this is supposed to be fun!

They are new, MIEC, +85C, 8mf, 450v. You can get them at justradios.com.

I am having fun just a little frustrated. Funner if I can get it working!