[dconant]

Hello, This Philco 89 code 123 is playing now. Problem is it plays for a long time then the signal drops off quite a bit. Played all day Saturday in the cabinet unbolted. So I decided must be good to go so bolted it down. It played perfectly for about 15 minutes and signal dropped off. If I touch the top of the 77 tube signal comes back some but not all the way. So I thought maybe antenna coil. So pulled it out looked at the coil and all sections ohm out close to what schematic says. Checked for green corrosion finding none.  While I was in there I checked for loose connections and found none. Fired it up again yesterday and it ran about 4 hours and again dropped off. Pulled it this morning checked again for loose connection and also soldered all rivited grounds and replaced the 300 ohm wire resistor to pin 4 with 2 watt 270 ohm metal oxide resistor. Fired it playing for about 5 hours with same problem repeating. I quickly measured B+ at one of the compensators sticking out the back. That bought the radio back to normal signal. I have changed the 77, 75 and both 44 tubes twice. The connections to that compensator are good. I have even covered the radio partially with a towel to see if it is heat related but I doubt it. Changing tubes will the signal is low makes no difference. Let cool down several hours it comes back normal. Other than that it plays quite nicely.

Anyone have any ideas? 

Thanks,
Dan

[David]

My guess for what it is worth a bad resistor has a crack some were take a lens and flashlight and have a look see. Not always easy to find. David

[Arran]

Dan;
It could be as simple as a bad solder join on one of the grid caps, or one of the grid cap leads. Those dogbone style resistors are also notorious for developing intermittent contact, it's basically a wire wrapped around a carbon rod.
Regards
Arran

[MrFixr55]

89s do strange things. I think there is kinda saga of the 89 and 19 written by Ron Ramirez in the Philco Library. I had one that you looked at it wrong, it would stop working and touching the 77 grid cap would restart it.

Do check the Oscillator coil, not only the tickler but the coil under it. I had improved performance when I cleaned the corrosion of the coil under the tickler, glyptoled it and rewound the tickler. I don't remember what version mine (fixed for a friend who sold it) was, but I think it was early with a 36 Tetrode as the converter. I converted (pardon the pun) the converter to the 77 version.

Also try this:

Pull the RF tube and run the set with a long wire connected to the 77 grid cap. If the problem disappeared, then the issue is in the RF section. If the problem still occurs, the issue is elsewhere. Either converter, IF, detector or audio. On the unit I repaired, I added a radio / aux switch If you do that, you can test the audio section in the same manner, however, I think that the issue is not there.

Hope this helps.

[dconant]

Thank you guys for the ideas. I had already rewound the tickler coil and replaced the grid cap wires. I got my signal tracer out when the radio did its thing again but as soon as I touched the ground wire to the chassis the radio came back full strength. I have checked and rechecked the solder joints and even poked pretty much everything while the radio was playing. I guess my next step is replace the remaining dog bones and poke a little more. If that don't solve the issue I'll remove the oscillator coil and check it again. Looks like this is going to be bugger to solve.

If the oscillator coil was the trouble would the radio still play but weakly?

Would the detector coil cause this kind of trouble?

Dan

[morzh]

The resistor #8, is it 8K?
Shields installed?

[RodB]

89s doing strange things; now if only you can get it to wash windows then you could forgive the other sins.

Dan - so, you mentioned solder points but are there any riveted terminal strips or tube sockets? When I was restoring a 1935 chassis, out of curiosity, I measured resistance in all the riveted chassis "ground" points expecting to see zero ohms. Several went from 500k to 2meg. One, on the grid leak to a tube, just floated around. And those dog bone resistor wires can get flaky and disconnect from the resistor body after heating up. Maybe go tap each one with a stick, staying away from the chassis. These intermittent issues are nightmares.

[SRv2]

What about checking tube sockets?  I have seen issues where a tube slips in pretty easily into a socket and with even small amount of heat a pin can lose solid connection.  Minor movement restores operation.  Worth checking maybe.

[dconant]

Just so we know. Tube shields are all in place. The three riveted grounds I soldered to the chassis. I won't be doing much today as outside chores are waiting. Thanks for the ideas guys. Unfortunately it takes time before the problem show up. Hopefully tomorrow while it rains.

Dan

[Arran]

As I mentioned I have encountered problems with the solder joints on grid caps, especially on the older tubes, they seem to oxidize causing open circuits, or high resistance, and heat or vibration can cause it to come and go, the cement holding them on often fails too. Though what you are describing also sounds like the oscillator stopping, it could be the coil, a resistor, or a bad mica cap. One idea I have heard of as well was to bake the coil to drive the moisture out, though soaking it in a pot of hot wax will do the same. These autodyne oscillators were notoriously unreelable, I've been looking at an Echophone S-5 and the early version used a single #24A as a mixer oscillator, in the later ones they added an eighth tube, a #27 as a stand alone oscillator because they were having so many reliability issues. My guess is that with Philco model 89 sets it was a royalty avoidance scheme since they used the same ruddy front end for four years, even though the model 60s started around the same time and it had a 6A7 pentagrid convertor.
Regards
Arran

[morzh]

For this kind of intermittent problems what I do ism using a light (light!!!) plastic or wooden rod, gently tap on things.
See if it restores the operation/stops the operation, thus trying to localize the general location of the potential bad contact/solder spot.

[MrFixr55]

+1 Arran. Having worked on an 80 which also uses an autodyne oscillator mixer but there were differences. In the 80, theplate connects directly to the IF Transformer primary and is capacitively coupled between the plate and primary of the oscillator coil. In the 89, the 2 coils were in series. I don't know why that made a difference, but it did, IMHO. The 80 performed marvelously for what it was, and especially for its pricepoint. It continued through at least 3 iterations, 80, 84 and 37-84

In addition to all the issues that Arran raises, the tuning cap can be dirty, having bad connection between rotor and frame, and also oxide "crystals" for lack of a better description touching rotor to stator. The "compensating" caps, both standalone and the ones that were part of the tuning cap were also notoriously bad, but I believe that on sets that came from the factory with the 77 pentode came with "mods" to those parts.

[Arran]

Speaking of the tuning condenser, I had one once where the set behaved intermittently, there was one section of the dial where you could hear scratching noises, and the reception pop in and out, sort of a dead spot. Naturally I thought, must be bent plates, so I inspected the outside plates of of each section of the rotor, found a few places where it looked too close, gently bent them with the aid of a pallet knife. It worked somewhat, but it still would cut out around the same area, but after handling it I noticed that the chassis, and the tuning condenser was dusty. So I took an artist's brush, and the shop vac to it with the crevice attachment, and vacuumed the top of the chassis, focusing on cleaning the dust out of the tuning condenser. That shop vac is powerful enough to such of chunks of wood, stones, etc., and it must have done the trick because the dead spot, popping, and scratchiness was gone. I'm not saying that is the problem here, but its another thing that can cause this, along with bad solder joints on grid cap leads, which are often attached to the tuning condenser as well.
Regards
Arran

[dconant]

Here is an update.
Today I removed the solder from the grid caps and roughed up the wire sticking up and resoldered them. Cleaned the tuning condenser with post-it-note sticky side. Didn't get much dirt as I had already given it a good cleaning as part of my restoration. Replaced remaining dog bones and cleaned the two compensators on the back of the chassis by backing out the nut and spraying contact cleaner in there and cleaned and tightened tube sockets. Tube pins are clean. Probed just about everything under the chassis. Radio has been playing outside of the cabinet for five hours now without issues. Yes I redid the alignment after messing with the compensators. Forgot to check the resistor #8 is 8K. I'd look now but I don't want to turn it off yet besides the sixties at six is on Zoomer radio (740 out of Toronto). More tomorrow.

Dan

I checked resistor #8. It is an 8.2K measuring 8.1K. That got me wondering if I did cap #9 correctly so I checked and found I did it correctly.

[SRv2]

Hoping for the best, Dan. Sounds like you hit everything. Also nobody should interrupt 60s@6!  I’m listening to one-hit-wonder hour on Legends radio 810. It’s got to be the smallest transmitter ever at 2200 watts. It’s about 30 miles from here and my MLA-30 antenna barely grabs it. The joys of radio restoration! George