[Mondial]

Check and make sure you have plate voltage on the 6K7 magnetic tuning (AFC) amplifier tube. With no IF signal on its plate no AFC signal will be developed. 

Check the voltage across each side of the AFC defeat switch to ground. As you tune through a station the DC voltage should swing from positive to negative or vice versa. If no voltage is seen there is a problem in the AFC amp or discriminator transformer.

Read the description of the Philco AFC circuit starting on page 93 of Rider's book. Although based on the 37-116 circuit it is practically identical to the 38-690. He goes into a lot of detail on how the AFC voltage is developed and applied to the 6N7G control tube to tune the oscillator frequency.

https://worldradiohistory.com/BOOKSHELF-...-Rider.pdf

[RGK20m3]

I have been reading John Rider’s book and understand the theory, also have been using his narrative to try to come up with a diagnostic process. Have already checked voltages on the 6K7 magnetic tuning amp and have 34v on the plate and 24v on the screen; the voltages given for the 38-116 show they should be 40/20. The voltages that I measured were with the radio tuned to a strong station (I measured before I read the note of no signal). There are no published voltages for the 38-690.
What voltages should I expect to see going positive/negative on the defeat switch? 0-5v?
I am trying to come up with a way to rough tune the magnetic tuning circuit. My understanding is that 85A(the hex) moves the core, and there is a trimmer concentric that tunes the primary- then the other two trimmers tune the secondary (for minimum signal). The two issues that I see is that I can’t seem to tune the primary, which would defeat the AVC, making the minimum signal moot.

[Mondial]

There are only two adjustments on the discriminator transformer, 85A and 85B. 85A is the primary and 85B the secondary. There is no movable core as 85B is an air variable trimmer where the hex and screwdriver slot are the same shaft.

The Philco alignment procedure is a bit confusing, setting for minimum audio output and then maximum. What you are really trying to do is to maximize the signal on the secondary of the discriminator transformer with 77A and 85A adjustments and then setting the secondary 85B for a zero voltage at the defeat switch terminals when a strong signal is exactly tuned in.

I don't recall the exact voltages at the defeat switch, but each side should swing + and - at least 3 to 5 volts as you tune across a strong signal.

If you have a scope with a low capacitance probe, you can look at the IF signal at the plates of the 6H6G tube. You should see an RF signal there when a strong signal is exactly tuned in. Adjust 77A and 85A for max signal on the scope. which should get you into the ballpark. Then disconnect scope and adjust 85B for zero voltage on both sides of the defeat switch.

[RGK20m3]

I am thinking of hooking my oscilloscope to the grid of the magnetic tuning amplifier, and setting the signal generator to 1000kc, then adjusting 77A for max voltage as a rough adjustment. Next removing the 6H6, connecting the oscilloscope to a plate, and adjusting 85A/85B for maximum voltage.  I can then fine tune from there using the Philco instructions.  I can do that much with the chassis in the cabinet......
Your thoughts?

[Mondial]

Yes that should get you close enough to see if the circuits are at least working. The only problem with connecting the scope to the cap of the 6K7G mag tuning amp is the capacitance of the scope probe will detune the 77A tuned circuit a bit. 

If you can see IF signal at the 6H6 plate pin with the scope, then you can tune 77a without loading the 6K7 grid with the probe. Also the same 77A tuned circuit provides IF signal to the AVC rectifier, so you can also tune 77A for max AVC voltage from the 6J5G AVC rectifier. From the top of the chassis you could probably see the AVC voltage at the grid cap of the first or second IF amp 6K7G if you use a 1 meg resistor in series with the voltmeter to prevent loading the tuned circuit.

[RGK20m3]

I am sorry, I don’t understand why it posted again, it did it before and I deleted the duplicate. Must have something to do with using the back button or leaving the page open.
I will try as outlined and post results.

[RGK20m3]

I connected my oscilloscope to the grid of the magnetic tuning amp and was able to easily peak the 77  trimmer for maximum signal (this with signal generator with a medium signal at 1000kc). I then removed the 6H6 discriminator and installed a dummy tube base with leads on the plate pins, to which I then connected the oscilloscope. I was able to easily trim 85A and 85B for maximum signal, also found that 77 could be trimmed better (it had peaked at a different setting due to the scope lead capacitance).
Magnetic tuning still doesn’t seem to function. I am suspecting that the 6H6 May be bad, although it shows good on my emissions tester. When first tested, heaters did not work; but when I measured heater resistance at 7ohms, I reinstalled it in the tester and it checked good. An original Philco tube....
Next step is to remove chassis and measure voltage across the AFC kill switch.

[Mondial]

Yes, now you are making some progress. Next you will have to monitor the AFC voltage at the kill switch. 

With the primary of the discriminator now tuned and strong received signal you should see some voltage at the kill switch. This voltage should change from plus to minus to zero depending on the setting of 85B. As you slowly turn 85B in one direction, the voltage should change from near zero to a positive peak, then drop through zero to a negative peak and then return to near zero. You want to set 85B to the zero point between the plus and minus peaks.

This is the typical "S" curve of a discriminator. See page 26 of Rider's chapter on the discriminator and what the proper tuning curve plot looks like.

If you still see no AFC voltage at the kill switch, check for shorts on the AFC control wires or a bad 6H6. There is not much else in the circuit except for a few resistors and caps in the 6H6 cathode circuit to cause problems.

[RGK20m3]

As a side, I monitored the voltages on the 1st and 2nd IF tube grids; they both were around -2.4v and did not seem to vary with trimming. I kinda thought that 77 should vary the signal to the IF AVC tube, but perhaps the signal was not strong enough.
I might try a couple of diodes in that tube base, don’t know if I have any Schottky type.....

[Mondial]

Yes, you probably were not feeding in enough signal to activate the AVC. You will need a somewhat strong signal as the 690 has a delayed avc circuit.

The 6J5 AVC rectifier cathode is biased a few volts positive, and the grid (used as the plate) is biased negative, so the IF signal amplitude must be high enough to overcome the initial bias before AVC voltage is generated. 

The initial 2.4V negative bias applied to the grid of the 6J5 is what you are seeing on the IF 6K7 grids. It should increase with a strong enough signal.

[RGK20m3]

First, I do not think the AVC is working; I fed a 1000kc signal directly to the antenna terminals via a .1uf capacitor and monitored the bias voltage on both the 1st and 2nd IF grids- no change as I varied the signal from minimum to maximum. I measured +11.1v on the cathode of the 6J5 and -2.8v on the grid (which is about the same as I measured at the grids of the two IF tubes). I verified all the components on the grid and cathode circuits, so maybe a bad tube? Original Philco....
I measured the resistance across the shorting switches for the magnetic tuning; 0 ohms when off and about 1.4meg when on. I checked both to ground at about 855k (which seems to be about right for the various resistors in the circuit). I verified circuit is correct and individual component values to within 10% of specification. I then powered up and had 0 volts across the shorting switches when off, and only .008v when mag tuning on; this voltage did not change as I adjusted the 85B trimmer. I measured one side at +.018v and the other at +.010v. I rechecked for a signal across the plates of the 6H6 and have a good modulated signal of at least 10v. Sounds like a bad 6H6?
I also checked the oscillator control tube; +4.08v on the cathode(s), 215v on both plates, and zero on grids- at least that tube seems to be working, and shows band switch is ok.
Radio receives well. What are your thoughts? I appreciate the help and any advice you can give.

[Mondial]

It is very unusual that you have 10V of IF frequency signal on the plate pins of the 6H6, yet no DC rectified voltage at the cathodes. It is also unusual that the 85B secondary trimmer makes absolutely no difference in the DC output.

The only things I can think of is that either the 6H6 is bad or there is a open section of  secondary winding in the discriminator transformer. First, as you mentioned you can try wiring up your spare tube base with a couple of silicon schottky or high speed switching diodes (1N4148, etc) and see if you get any output.

The other possibility is an open section of the secondary winding. If you look carefully at the secondary schematic, you can see that the resonating trimmer cap 85B is connected to the ends of the winding, while the 6H6 plate connections are tapped down toward the center. So you could still have signal on the plates, yet if one of the winding ends was open 85B would be disconnected and have no effect on tuning for resonance. Unfortunately, the only way to check would be to remove the transformer and check the full winding for continuity.

As far as the lack of AVC voltage, the voltages on the 6J5 AVC rectifier seem about right, and consistent with the ones of the similar circuit of the 37-690. Because the cathode is biased +11V and anode (grid) at -2.8V, it will take a IF peak voltage of 14V or more to generate any negative AVC voltage over the standing -2.8V bias. The 14V of IF  signal seems like a lot, but with two IF amp stages, plus the gain of the RF stage and mixer, it should easily be attainable with a strong signal. If you can't generate that 14V of IF, it could be due to a weak tube in any of the previous stages, or other problem that reduces the overall gain of the receiver. Keep in mind that this radio has a lot of reserve gain when operating properly, so with reduced gain the radio will still receive fine, yet fail to generate AVC voltage. You can try exchanging one of the other 6J5 tubes in the radio with the AVC rectifier 6J5 and see if it makes any difference. There is also a separate independent AVC amp and rectifier circuit for the RF stage, so it would be interesting to see if that circuit is generating any AVC voltage.

[RGK20m3]

I think that I can rule out the open secondary winding as 85B has plenty of effect on the signal at the plates of the 6H6; as I stated before, I could easily trim 77, 85B, and 85A for maximum signal at the plates with the oscilloscope. What I was referring to was that the small voltage across the cathodes did not change with 85B. I think my Hallicrafter S40 uses a 6H6, but I think it is a metal one. At least I can test the Philco 6H6G in it. Maybe I will make an adapter the try the metal tube....

[RGK20m3]

Success!!!  I had the radio on, tuned to a station, monitoring the voltage across the AFC switch, and had the same readings as yesterday (.008v). I got to thinking about what could cause both sides of the 6H6 to not conduct, and thought I would check the heater voltage- had 6.61v at the terminals of the socket. Then I measured the voltage at the tube pin tips and heard a change in sound- I then had .5v across the AFC cathodes. As I tuned off the station the voltage would rise to over 3v in one direction only. I the turned off the AFC, retuned the station, turned the AFC on again and was able to find the 0v null between the voltage peaks and the AFC WORKS!

If you recall, when I tested the 6H6 I initially didn’t have a heater glow on the tester, and after I measured the resistance it tested good. I think there may be a break between the heater lead and the pin of the tube base.

Thank you for your observations and advice.

[Mondial]

Congratulations on finding the problem. 

Try re-soldering the 6H6 heater pin tips, as this is a common problem with intermittents. Sometimes there is a marginal solder connection between the lead and pin tip which only shows up many years later.