[RGK20m3]

Dial is keyed to preset drum; before installation I verified that it was in the correct spot relative to the original dial. The presets call letters from its previous life match the dial scale. I just have not aligned anything that far off.

[klondike98]

I merged your two 38-690 threads to keep the discussion on the radio together.

[RGK20m3]

Well, it gets weirder- I checked the accuracy of my signal generator against a known good digital tuner at 1000kc.  I went through the alignment with no difficulty, wasn’t too far off, carefully lining up the dial indicator to the set points (which were very close to the signal generator dial, checked by rocking both signal generator and tuner).  Reception seems good, as I was able to pick up stations 800 miles away last night, and magnetic tuning seems to work on 2 stations. Dial still off by 40kc; condenser trimmers (22A and 22B) are about at max in.

[Mondial]

Are all frequencies up and down the dial off by 40 kc? If so it could be your IF is off by 40 kc. This would make the dial calibration low or high by that amount.

I had a similar problem with an RME-69 communications receiver. Every dial freq was off by 35 kc, but at first did not realize what was the problem. Thought it was just the osc alignment. Turns out a previous owner aligned the IF at 500 kc instead of the correct 465. I finally found the problem when the crystal IF filter did not work since it is fixed at 465 and did not line up with the IF transformers tuned to the incorrect 500 kc.

Aligning the IF to the correct 465 kc cured the dial freq error and allowed the crystal filter to operate properly.

I don't know if you may have the same problem, but it would be worth checking the IF alignment and confirm it is peaked at the correct frequency.

[RGK20m3]

Philco literature says 470kc and that is what I used (twice); perhaps I will check the accuracy of my signal generator at that frequency. Yes, all seem to be off, perhaps by more at higher frequency.
Is there a mechanical alignment of the dial to condenser? I have the dial scale ending at the mechanical stops both high and low.

[morzh]

Let me understand: you use a generator and

1. all is good (IF and the dial) and then it is off when tuning in, or
2. IF is good and then the dial is off
3. All is off compared to the generator in the very beginning.

[Mondial]

You can do a simple test to see if the oscillator is off frequency or the problem is the IF

Confirm the the local oscillator is operating at the right frequency by listening with your digital receiver. For example, set the dial on your Philco to 600 kc. With the correct 470 kc IF, the osc should be running at 600 + 470 = 1070 kc.

Tune your digital receiver to 1070 and place near the osc tube. You should hear a strong carrier. If not tune the Philco dial slightly until you hear the osc signal. If the Philco dial reads at or near the proper 600 kc, then the osc is at the right freq and the IF is off. 

If the dial needs to be tuned 40 kc from the 600 kc dial mark to hear the signal at 1070, then there is an oscillator problem.

[mikethedruid]

To check my IF is right on, I set my digital receiver to exactly TWICE the IF frequency of the set I am aligning. If the IF is 470KC, then I st the digital receiver to 940KC.

I then tune my frequency generator, set to audio modulation internal, to the IF frequency, in this case 470KC,, and connected to a short test work output antenna and carefully adjust it until I get the maximum sound output on the digital receiver. (Make sure your signal generator has THOROUGHLY WARMED UP, at least a half hour, so it won't drift !) Now you know your signal generator is putting out the proper frequency.

I then follow the manufacturers instructions for aligning the IF, or use this general method. If the radio uses a mixer/oscillator tube like a 6A8, I pull off the cap connector to the control grid, and directly feed the output of the signal generator into the control grid through a .001 MFD condenser. Set the output of the signal generator to the lowest setting that gives a clearly audible tone through the set being adjusted. I then peak the first and then the second IF transformers until I get the strongest audio signal through the speaker. You can use a voltmeter set to AC volts across the voice coil for this, or do it by ear if you have good ears. After I peak the second IF transformer, I go back to the first and adjust it again as necessary. I check back and forth a couple of times.

That's it, when both of the IF transformers, or all three if it has a two stage IF strip, are adjusted for maximum peak, you're all set.

[RGK20m3]

I did not disconnect the grid connector clip on the mixer/detector 6A8; I also used a .1 per the Philco alignment procedure- but was able to peak. I did check the signal generator by wrapping a couple of turns of wire around the digital tuner loop antenna and connected the wire to the output of the generator.  Good idea about checking the oscillator frequency.

[RGK20m3]

Excellent suggestion Mondial!  I set the dial on the 38-690 to 600kc, and my digital radio picks up the oscillator at 1070kc, confirming proper oscillator frequency.
Any ideas on what to look for in the IF section? If I am understanding, it cannot be anything past the mixer. All of the IF coils had continuity and I replaced all the paper capacitors (none of the mica), and all resistors are in spec.

[Mondial]

Ok, now we know that the oscillator is running at the right frequency and agrees with the dial calibration, so the dial error must be caused by the IF not being tuned to 470 kc.

I don't think anything is physically wrong with the IF, only that the IF transformers are tuned to the wrong frequency. The compression trimmer caps in the IF cans have a large tuning range, so they are easy to set to a freq other than 470 kc. 

Somehow your generator is off freq or there is some other reason the IF is different from 470 kc. You can try the suggested idea to check the second harmonic of your 470 kc generator signal by listening with your digital receiver at 940 kc.

Another thing to try is to gradually tune the IF transformers toward the correct 470 kc by tuning in a known freq broadcast station, setting the dial to where it is actually received, not where it should be. Then tune slightly toward the correct dial calibration and repeak the IF transformers on the received signal. Repeat this procedure in steps until the station is received at the proper indication on the dial. Since you know your oscillator frequency is correct, this should align your IF to very close to 470 kc. If you can't reach proper dial calibration by this method then something is really wrong with the IF transformers or circuitry.

[RGK20m3]

I went through the alignment again, verifying the signal generator frequency with a digital radio at each step.  Basically same result, dial is 40kc low at 950, and 25kc low at 760.  Couldn’t verify any stations high up on dial, as reception wasn’t good and stations I did pick up didn’t state call sign.
I verified that the oscillator was tracking correctly at 600kc, but found it was 50kc high at 1200 dial setting. I found oscillator at 1670kc when dial at 1150 (so oscillator at 520kc).
Would have to be a condenser issue I would think; this radio has Magnetic Tuning (AFC), but acts the same with it on or off, and I did verify that switch does short the AFC out, as does dial switch.
At least I know where to start looking. I did notice that the mixer 6A8 was microphonic...wish I had some others to try, as oscillator is same type.

[Mondial]

So as I now understand it, your oscillator frequency is correct at 600 kc, 25 kc high at 760, 40 kc high at 950 and 50 kc high at 1200 kc. 

With this further info and the fact you checked and aligned your IF at 470 kc again, looks like the IF is not at fault since the tracking offset is not constant but varies from 0 to 50 kc.

This points to an oscillator error as you tune higher in frequency. The tuning capacitance is too low as you go higher in frequency. Check the connections to the main tuning cap as each section has a split stator where both stators are used in parallel on the broadcast band.

Could still be caused by the AFC circuit since the 6N7 control tube is always in circuit even if you short out the AFC with the switch. I would double check any work you did in the oscillator and AFC sections of the RF section.

[tbone]

As Mondial mentioned, the tuning cap is suspect. The bandswitch plays a role switching the tuning capacitor sections. Make sure it’s clean. Also, are the other bands accurate? This may point you in the right direction.

[RGK20m3]

I verified the band switch is connecting both sections on the condenser, and connections are secure; also no visible misalignment of the blades.  I went through the entire tuner chassis and matched to the schematic; if I made an error, I am sure not seeing it. There was a resistor that was 30% out of spec and I replaced it.  I  tried swapping 6A8 mixer and oscillator tubes(with each other) and realigned the receiver; no change at all.
Can’t tell for sure on the other bands, but they seem off the same way at the higher end of the scale.
It doesn’t seem adding a small value capacitor in parallel with the oscillator section would work well, as the low end is accurate.
Ideas???