[morzh]

Trying to drill the holes in the rivets. I knew it was going to be difficult. Have experience.
The first drill made one hole, then refused to finish the other one. I looked at it - it was literally worn out. CCC. Thought it would at least work on aluminum. I have 10 of them, so I put another one in; it almost finished the hole but broke.
Well, let's just say getting it out was...onerous. I had to start grinding down the back side and then trying to guess where the drill was trying to come out. I thought I'd hit the drill but obviously the grinding stone bit was grinding the drill bit as easily as the aluminum, so I kept ending up with a flat grind with no indication of the drill tip. So I switched to the ball grinder and it eventually bounced off of something. I used a puncher on a hunch and, luckily, from the 1st try I was onto the bit. The rest was knocking it out. In the process of which my box of assorted capacitors fell off my bench.....so I also spent some time doing nice bend over-straighten up exercise. Probably did me good.
Two more holes to go, hopefully only 2 more drill bit with no breaking.

[morzh]

Before I showed the way I cut the cap.
Now I have drilled the holed through the Positive's rivets, and I also drilled one central hole for the common Negative conductor.
I countersunk the central hole with the thought of using a copper rivet as the holder for teh central conductor. The barrel is shorter than the base which is 1/2" and so it would not go through. Also countresinking from both ends will ensure the extra creepage between the central conductor and the Plusses.

   
   

The rivet is in place. As I need the channel, I knocked out the remaining piece of the mandrel with the mandrel itself, and the barrel became a via.

   
   
   

Now I made the central conductor from a Romex wire: it is copper and is very stiff and solderable.

   

Formed a loop and soldered it in place. Now I have the outer negative terminal and the supporting pole inside the cap for the minusses.

   

Now I put the wires for the cap's plusses in, and solder then in place, and also the film cap 1uF as it will not need the support the small aluminum cans will. To explain, I need 3x1uF and one 2 uF caps. I have 2 aluminum cans of 1uF 450V (Good brands and high ripple, so will last), one film cap of 1uF and one can of 3.3uF which is unimportant as the circuitry does not require precise cap in those filters.
   

Now solder all the caps to the central pole after strategically positioning them to the places they will remain in.

   


Now I cut the Plusses wires to size and solder them to all the caps. The resulting construction is very stiff and will not bend on its own.

   
   
   
   
   

[morzh]

Now, to ensure the bare wires will never touch the can I use one layer of fishpaper. The wires are formed such that they would not touch it anyway, but this is an extra safety measure.

   
   

And it is closed.

   
   


And now we measure the caps.

   
   
   
   

As ordered, 3x 1uF and 1x 3.3uF


I plugged the holes in the rivets with the same JB Weld epoxy I glued the cap together with.

[David]

Mike great write up. David

[morzh]

Today I spent close to a couple of hours gutting and rebuilding the 8uF rectifier filter cap.
The stuff inside is poured over with some brittle hard type tar, mixed with paper.
I broke the corkscrew I use to gut these.
So I drilled, heated and chiseled until I had maybe less than 1/2" at the bottom at which point I deemed it harmless, and then put a Solen 600V 8.2uF cap inside, stabilized and insulated it with fish paper and epoxied. This has the same type depression groove at the bottom so I used the same type cut and the same technique to get the two halves back together.
Tomorrow I will rebuild the last one, it is a 2 section one, 8uF and 10uF. I will use 10uF and 33uF respectively as this one is not critical. The one I have to gut is a 3 section type, so it is just as well as I will have more space to put two pairs of wires instead of three.

Both caps are time-period appropriate and look right. The originals were with solder lugs instead of wires but then it is only so far I will go to make a cap look exactly as the old one did. They do not say Philco anyway.

[Brad Winder]

Mike, looks good!

Question: do you put a solid "liner" or something inside the can, before you glue the can back together? I tried a similar rebuild process a couple years ago, and found that putting the can back together with just epoxy resulted in an unstable joint, that came apart when I clamped it back into the chassis. I didn't put in the fish paper...maybe that's the trick. The epoxy would saturated the paper at the joint, and be the backing needed.

[morzh]

Brad

No, I have no problems. I make sure i really clean the epoxied surfaces well of any traces of wax, and then with isopropyl alcohol.
If you mean something that goes between the can and the base.
I do put the fish paper inside but only as an insulator: it does not get into the actual joint.

I use JB Weld.

[morzh]

This cap, same as the 8uF one, was trying to be tough but I locked the wires in the vise and then used two heat guns on "High" to heat it from opposite sides. I heated it for 2 minutes and then used a kitchen glove to pull the can.
It just slid off freely after this.

   


These are the two heroes, with the cap's innards and the glove.
   

Will rebuild tomorrow.

[morzh]

All caps are installed, and I am re-wiring their interconnect as the repair done some time in possibly 40s was using twisted soldered together wires and black tape over the soldering.
I am slowly getting rid of those plus the caps that were installed over the existing backelite blocks or instead of the 4-section electrolytic.


Also I am not sure about the wire-wound divider: the schematic says 20-100-130 and mine is 20-170-200 ohms or close to that. What bugs me is the fact that the wire-wound resistors typically do not age value-wise. They either OK or they are not. If they are, their value is within tolerance and usually is the same as it was the day of manufacturing. So at this point I am not yet ready to add bypasses until I hear from the folks who have experience dealing with this chassis. I only fixed one 16, the 16B cathedral, and the chassis is different. Even if the resistor is the same.

Chugging along.

Finished the recap.
One of the wires that went to the 4 section cap did not belong there. It was supposed to use a nearby backelite blok's free solder lug as a solder post but instead somehow went to the electrolytic cap. I wonder if this radio was indeed working after that repair.

Now the resistors.
Right from the start the 13K-10K divider (the big resistor) is off, being 17K-13K. I replace when I can, else I parallel a resistor to get to the target value. Some values go up twice.
The cap, 1500pF, that couples the oscillator to the 1st detector, where it has two resistor and the cap connected, had another 470pF more modern ceramic cap soldered with the other lead going nowhere.

Well, today has been productive. Main thing, the electrolytics have been rebuilt and wired; this is the operation I like the least: cutting the open, cleaning, etc. Way less fun than restuffing the backelite blocks.

[morzh]

I realized I missed a tubular cap. I thought it was only backelite blocks.
Then I traced the ceramic 0.05uF caps and saw that a couple of more tubulars are supposed to be there; good thing I keep old caps, a bagful, so I restuffed a couple and put them in.

Now, while tracing them, I realized that the volume pot was miswired.
Someone put the detector output, that should go to the top pin of the volume pot, to the wiper pin, and the output circuit that should've gone to the wiper pin - to the top one.

Keep nibbling at it.....

[morzh]

All tubulars have been re-introduced.

A cerendipitous find: while changing a second large 10K resistor, which became 13K, same as the 1st one, from large divider, I habitually checked it (I ordered a few 5W 10K ones) and found it to be 13K. Surprised, I looked at the marking, and it was indeed 13K. The rest is 10K. So, I took the original 13K one, that grew to 18K, and which I parallelled with another one to make it 13K, yanked it out and replaced it with the 13K one. I did not have any 13K. Or so I tbought.
Lucky day!

[morzh]

Have rebuilt the tone control.

This is fairly easy as all three caps are the same, 3nF. The original caps all rose in value to 8nF, all three, so it would give 8-16-24nF, whereas it should be 3-6-9. I have 3.3nF, so it is close.

Extracting it was a bit tough - I had to de-solder one of the mica caps that kept the part from being able to come out.
I tried to remove the shaft but the chassis holding fixture I am using is a bit flexible and so pressing on the retaining clip, in order to remove it, was unsuccessful. So I removed it as a whole.

I was able to remove the clip once it was out, in order to clean it. Lots of black came off.

   


Then I made a new insulating insert from fish paper.

   

Twisted together three 3.3nF caps.

   

Soldered the wires and situated them inside the fish paper cradle.

   
   

Then covered the caps with fish paper caps to protect from hot glue. I always to this when rebuilding caps or tone controls. The hot glue will still penetrate, but very little. But it will be enough of it around to stabilize and hold stuff and the wires.

And then I potted it with brown hot glue.

   

Then reinserted the shaft.
I used a new Jesus clip (an E-clip): I have large assorted kit, and then these are easier to remove and install. Which proved handy when I realized it is easier to install the tone ctl without the shaft and then insert it later.

   


Caps are done.

[Brad Winder]

That looks great! I've rebuilt a couple of those tone controls before...they are quite a bit of work, but they are satisfying to see working again as intended, without replacement caps just hanging out of the terminals.

May I ask where you get the brown hot glue from? I cannot seem to find it here in Canada...

[mikethedruid]

3nF ?    For us old timers who ain't up on all these new fangled terms, is that anything like .003 MFD, or 3,000 MMF ?

[Ron Ramirez]

Yes:

3 nF = 3000 pF = .003 uF

The term "nanofarad" or nF isn't used much in the USA, but is widely used in Europe and Russia, and has been for decades.