[morzh]

Why it seems like many of you think you have to bake coils? You only have to bake them in some particular applications when they are penetrated with some particular type of warnish, like in some DC motor rotors or transformers - did that myself when I once in my life rewound one, but the very fact that you have lacquered wire in one layer (or even in several) does not necessitate baking. Was will do it for fixing, or some cements if they do not damage the wire lacquer.

[Mondial]

You really need an RF signal generator to do a proper alignment. A function generator, unless its an expensive digitally synthesized one, isn't stable or accurate enough at RF freqs. A frequency counter also comes in handy to set the generator precisely on freq.

There are quite a few inexpensive service grade RF generators on ebay which may require a little work. Heathkit and Eico come to mind and these can be gotten at low prices.

Having said that, you can do a respectable alignment with what you have. In a pinch I have done IF alignments with only known freq broadcast stations and a VTVM or DVM. I believe your transmitter can be programmed to any broadcast freq and your digital receiver can serve somewhat as a frequency counter. Just keep the signal level low enough into the antenna input that you don't generate more than a few volts of negative AVC voltage. Monitor this voltage and peak the tuned circuits for max rectified AVC output at the detector. Make sure that your IF peak occurs at a LO freq of RF in + 260 kHZ and you should be fine. Then adjust Low freq padder for best signal and dial cal at 600 kHz, High freq osc trimmer at 1400 kHz for dial cal. Then peak Ant and RF trimmers for max signal at 1400 and you should be done.

[morzh]

The only problem with inexpensive RFs like Heath (2-tube types, ECC82/12AU7 based) is at least at lower frequencies their wave shape is quite a bit away from sine wave (it's not the oscillator itself, it is the follower that does it).

Like so.

http://vintage-radio.com.au/default.asp?f=2&th=330

And somehow SG-8 will not generate at the lower part of E-band (I think about 26-27MHz).
And they are not precise so you need a counter which is.

I acquired a Leader LSG-11 (same schematic mostly) and two SG-8 both of which I revived this last week and they behave the same, only Leader is actually precise (except the C band somehow is off).

If you chose to get a counter, I got my BK for 25 bucks and it is very accurate and worked right off the start (did not get to do anything to it).

http://www.ebay.com/itm/BK-Precision-180...35d23d9f87

Like this one (I also have the Dynascan but for higher frequency - however this one is enough for AM work. Whether it works or not - I took a leap of faith considering my price was the same).

[skyscraper]

Morzh - I baked it because Mr. Ramirez and others said to. The bakelite absorbs moisture, and apparently it affects the circuit.

Thanks for the info. Is that BK generator capable of modulation?

My 'transmitter' is just a cobbled together circuit I built with spare parts on a breadboard. Basically a clock circuit generates the carrier with a crystal and dividers, that signal is modulated with an audio transformer, then sent through a pot to the gate of a power mosfet that switches 40-90v across a ferrite transformer. The output is tuned with a tank circuit and fed to 20 feet of wire on the ceiling. The antenna pulls about 4 watts when connected, but it's only audible for a quarter mile or so. Still enough rf to stop a trackpad from working

The signal is a fairly clean sine wave, hopefully enough so to use.

The AVC line measured at the end of the volume pot, correct? Or is it before the 2 meg resistor?

[Mondial]

At the top of the volume control is best. This is the detector output after the RC filtering. If you put the DVM probe on the other side of the 2M resistor the meter will load the circuit reducing the AVC voltage to the controlled tube grids.

Sounds like your transmitter is a bit high powered. You may want to disconnect power to the Mosfet output stage to avoid overloading. Remember that you want to keep the input signal to the radio low.

[skyscraper]

Ok, thanks for all the info. Making the plastic tool now.

The signal out of the modulator is a hideous quasi-square wave. I can just turn down the voltage to the amp stage for a smaller signal.

[Ron Ramirez]

Yes, the coil forms should be baked to remove moisture - no higher than 200 degrees F, after the bad winding is removed, and before it is rewound. It is just good practice to do this, since Philco 89/19 oscillator coils have so many problems, moisture absorption being one of them.

[skyscraper]

That was my intention, but in my over eagerness I forgot to bake it before rewinding. Not wanting to mess it up again, I baked it anyway, but the outer coil melted into the plastic. I should have just taken it off at the time.. But again, based on what others have said I think it should work anyway. Maybe a tiny cap between pin 2 and ground would have the same effect as the gimmick wire.

[morzh]

Actually why 200F?
If the idea is to get the moisture out, any raised temp will do it. 100F will do just fine. Just slower. So what. And you could put it on convection to move the air.

I even thought of microwaving (of course without wires) but then I though that entrapped moisture might create micro-cracks.

[skyscraper]

I aligned the IF to more or less 5% of 260kc with my transmitter and several strong local stations. It didn't change much, it oscillates down to 950kc, so a little step in the right direction.

In the process I managed to somehow mess up the dial calibration by 20kc or so. I think I'm going to put this to the side for now, and look for an rf generator to do it correctly. In the meantime there's the cabinet and five more Philcos to work on

[Mondial]

The reason that your dial calibration moved by 20 kHz is because your IF moved up by that amount.

Now that your IF is at 260 kHz, you need to adjust the osc low freq and high freq trimmers to recalibrate the dial and restore tracking with the ant and RF tuned circuits. It should track the calibration of the dial more precisely now with the correct IF. Someone in the past probably aligned the stages with the IF off freq and compensated with the osc adjustments. What you have now (digital receiver) is more accurate in freq calibration than what was available 60 years ago.

When you adjust the osc low freq trimmer, rock the dial around the freq of a station near 600 kHz and find the setting of the trimmer which gives the best signal strength at or near the proper dial calibration. What you are trying to do is get the RF and ant stages tuned to track the osc at that RX freq. Since the ant and RF stages are not adjustable at the low end of the band, you are compensating by adjusting the osc to match their fixed tuning.

If after optimizing the response at 600 kHz the dial cal is slightly off, you can loosen the dial setscrew on the cap shaft and correct for it. Then go up to 1400 or so and set the osc high freq trimmer for accurate dial cal. The low and high freq osc adjustments interact so you may have to repeat them. Then peak the RF and ant trimmers for max signal at 1400. This should get you very close to optimum alignment and performance.

[skyscraper]

I'm sorry for all the stupid questions, I think I almost understand now. Which osc trimmer is low and which is high? On the schematic they are parts 15 and 16. The problem with aligning around 600 is I can't get anything below 950kc to base it off of. At any point do you adjust the compensators on the tuning cap?


Is this a decent rf generator? http://www.ebay.com/itm/131125159686?_tr...EBIDX%3AIT

I know the waves aren't entirely sinusoidal, but is it close enough?

[Warren]

it looks like that Rf Generator doesn't have an internal audio modulator.. If it doesn't you should probably find one that does. Otherwise you would need to have an audio generator to put an audio modulation in the signal.

[Mondial]

Don't be afraid to ask questions, its the only way to really learn.

First remember that the RX freq is determined by both the IF freq and the LO freq. Since your IF is now about 20 kHz higher and your LO is still the same, the dial will have to be tuned 20 kHz higher to receive the same station. That's why the osc trimmers need to be readjusted to restore dial calibration.

Use this Nostalgia Air schematic as a reference:

http://www.nostalgiaair.org/PagesByModel...013876.pdf

The drawings show all the alignment points I mentioned. Trimmer 16 is the low freq padder. It is in series with the osc coil and determines the calibration at 600 kHz. The high freq osc trimmer 17 is located on the tuning cap. It sets calibration at 1400 kHz and is in parallel with the tuning cap. When these two freqs are calibrated on the dial, all others should fall into line.

The other two trimmers on the tuning cap are the ant trimmer 4 and the RF (what they call detector) trimmer 5. These are also in parallel with their respective tuning cap sections and are peaked at 1400 kHz for max signal once the osc is adjustments are done.

Trimmer 15 is the first IF primary tuning. Its location on the schematic is confusing because of the combined mixer/osc circuit.

Put your clip lead on the 36 tube so that you can get the osc working enough at 600 kHz to set the low freq dial cal and then take it off for the high freq adjustments.

The Heathkit in the link is the type of inexpensive generator I was suggesting. Just don't pay too much for one. The waveshape is really not too important as the distortion just creates harmonics of the fundamental which are rejected by the radio's tuned circuits. Some of these generators purposely had distorted RF output so that it could generate useful harmonics for alignment at higher freqs.

[morzh]

Of course it did. It has AF out.
It would not make sense if it didn't.

Here is the same model (this just got bought)
http://www.ebay.com/itm/Heathkit-Vintage...1125159686

Look at the enlarge photo to the left lower part/ - it says it right there.
Another option - I have 2 rebuilt Heaths SG-8. Won't ask much. Plus shipping. If interested, PM.