Heathkit SG-8 Signal Generator

[vortalexfan]

Originally Posted by Electronic M

It's not a bad idea to replace the original selenium rectifier in it with a modern 1N4007 or similar as preventive maintenance or you can be a Daredevil and run it with the original till it smokes.... I've taken both approaches.

What size of resistor should I use if I were to replace the old selenium with a 1n4007 diode?

[Notimetolooz]

Originally Posted by vortalexfan

What size of resistor should I use if I were to replace the old selenium with a 1n4007 diode?

Since there is a 3.3K resistor in the filter circuit, which suggests there isn't much current draw, I ended up not adding a resistor.
I made several modifications to my SG-8 to see if I could improve the modulation, etc., including adding a tube. I won't go into all I did, it was mostly experimentation and probably not worth it.

[vortalexfan]

Originally Posted by Notimetolooz

Since there is a 3.3K resistor in the filter circuit, which suggests there isn't much current draw, I ended up not adding a resistor.
I made several modifications to my SG-8 to see if I could improve the modulation, etc., including adding a tube. I won't go into all I did, it was mostly experimentation and probably not worth it.

OK, so then I can just take out that Rectifier and just replace it with a 1N4007 and be done with it.

[Notimetolooz]

About that 2420 counter, since it is capable of high accuracy there are a few things you should understand about frequency counters.
The oven crystal oscillator is very stable, however there is an adjustment, a calibration, that is critical so that the display is as accurate as it can be.
The oscillator is capable of staying on frequency within better than 1 ppm (one part per million) that is 0.0001 %, much better than any voltage, current or resistance meter measurement. You need that much accuracy if all 8 digits of the display will mean anything.
However if the crystal oscillation is only calibrated to an accuracy of, say, 0.1% then the display will be off that much. So the calibration needs to be checked from time to time against a frequency source of even more accuracy.
One way is by using super accurate frequency radio broadcasts from stations like WWV and WWVH. More modern methods use GPS managed frequency standards.

[vortalexfan]

Originally Posted by Notimetolooz

About that 2420 counter, since it is capable of high accuracy there are a few things you should understand about frequency counters.
The oven crystal oscillator is very stable, however there is an adjustment, a calibration, that is critical so that the display is as accurate as it can be.
The oscillator is capable of staying on frequency within better than 1 ppm (one part per million) that is 0.0001 %, much better than any voltage, current or resistance meter measurement. You need that much accuracy if all 8 digits of the display will mean anything.
However if the crystal oscillation is only calibrated to an accuracy of, say, 0.1% then the display will be off that much. So the calibration needs to be checked from time to time against a frequency source of even more accuracy.
One way is by using super accurate frequency radio broadcasts from stations like WWV and WWVH. More modern methods use GPS managed frequency standards.

So basically the atomic clock stations in Fort Collins, Colorado and Kekaha, Kauai, Hawaii?

[vortalexfan]

I got the selenium rectifier replaced with a 1N4007 Diode in the Heathkit signal generator.

Picture of the repair is shown below.

[Notimetolooz]

Originally Posted by vortalexfan

So basically the atomic clock stations in Fort Collins, Colorado and Kekaha, Kauai, Hawaii?

Right. Probably more than you really need. Just wanted to let you know how involved things get when you deal with things at high accuracy levels.
One way is to tune in WWV at 10MHz using a receiver with an S-meter and zero beating with the crystal oscillator output.
Anyway that would be something for another thread.

[vortalexfan]

Originally Posted by Notimetolooz

Right. Probably more than you really need. Just wanted to let you know how involved things get when you deal with things at high accuracy levels.
One way is to tune in WWV at 10MHz using a receiver with an S-meter and zero beating with the crystal oscillator output.
Anyway that would be something for another thread.

Couldn't I just tune it in on one of my old shortwave radios? Also what's an S-meter and wh as t do you mean by zero beating with the crystal oscillator output?

[Electronic M]

Zero beating: ever run 2 am radios at once and tune 1 through the band and find a point where they both squeal despite not having touched the one you weren't tuning? That's the local osc zero beating with the osc or station of the other set... In Toulouses example you would tune the Chrystal osc to the frequency of the station and listen for the zero beat then tune for minimum signal on the S-meter (exact frequency zero beat).

[vortalexfan]

Originally Posted by Electronic M

Zero beating: ever run 2 am radios at once and tune 1 through the band and find a point where they both squeal despite not having touched the one you weren't tuning? That's the local osc zero beating with the osc or station of the other set... In Toulouses example you would tune the Chrystal osc to the frequency of the station and listen for the zero beat then tune for minimum signal on the S-meter (exact frequency zero beat).

Ok, I get it now. So then somehow I would basically have to somehow tune the crystal oscillator in the frequency counter to the same frequency as the 10 Mhz signal of WWV in Colorado. And then tuning a shortwave radio to that same frequency and adjust the oscillator in the frequency counter until they cancel each other out in a sense on a tuning meter?

[Electronic M]

Originally Posted by vortalexfan

Ok, I get it now. So then somehow I would basically have to somehow tune the crystal oscillator in the frequency counter to the same frequency as the 10 Mhz signal of WWV in Colorado. And then tuning a shortwave radio to that same frequency and adjust the oscillator in the frequency counter until they cancel each other out in a sense on a tuning meter?

Yup.

[vortalexfan]

Originally Posted by Electronic M

Yup.

OK, got it.

[N2IXK]

Originally Posted by vortalexfan

I got the selenium rectifier replaced with a 1N4007 Diode in the Heathkit signal generator.

Picture of the repair is shown below.

Looks to me like the diode is in backwards....

[vortalexfan]

OK so I got my parts in to get my signal generator fixed up the rest of the way (BNC Jacks and a new set of cables for it so I could hook it up to a radio), and I got it all wired up and I went to test it out on an old Zenith Royal 500 Transistor Radio I had laying around and I couldn't get the signal generator to do anything, it wouldn't put out a tone or a signal to the radio or anything, I'm assuming with a transistor radio you don't need a capacitor inline with the test leads like you do an AC powered set.

Also I got tons of line interference through the signal generator similar to the line interference I said I had been getting on my radios.

I had my test probes hooked up the RF output jack which I'm assuming that's the hookup I was supposed to use to test the generator out on the radio.

any ideas as to what the problem could be?

[vortalexfan]

Originally Posted by N2IXK

Looks to me like the diode is in backwards....

You were right it was in wrong, just shortly after I read your comment and I was trying to test it out on an old Zenith Transistor radio I have and I heard a "Pop" noise and I checked it out and sure enough the filter cap that was attached to the diode had vented, thankfully i had a 22 MFD 160 VDC electrolytic in my stash that I was able to replace the vented cap with and now its fine, although I found out that apparently my Zenith radio was way out of adjustment because the 455 kHz IF tone couldn't be heard until around 600 kHz on the radio dial rather than at the end of the dial.