Heathkit IG-102: Ever realize something runs on magic?
 

For several years I've had a Heathkit IG-102 that I bought working great....with the exception of the external modulation input (which is un necessary for troubleshooting, but is nice to have for an impromptu AM transmitter).

A few months ago it's RF output dropped to zilch, but the audio output kept on kicking. The tubes tested good, but I noticed there were some original caps that seemed fishy (60's lytics, and plastic cased tubulars), and replaced them....No change.
Finally yesterday I needed it bad enough to get deep into troubleshooting, and after getting it working enough to be useful I'm left with more questions than when I started....
There was osc. output on the grid of the RF output amp, but not on the plate. All the bias voltages to the stage were within tolerance. There would be signal at the plate if the attenuator was in it's lowest setting, but the attenuator ohmed out reasonably. There was also no modulation despite the 400Hz audio osc. putting out a strong signal on it's jack.

http://i1095.photobucket.com/albums/...psfrjawxr9.jpg http://i1095.photobucket.com/albums/...psdrzarror.jpg

As I went through it I found the wiring mistake noted in the schematic pictures I posted....The darn audio section connection to the RF amp where the modulation should happen was not hooked up! Instead the lead of the untouched (by me) ceramic cap that should go to the plate of the audio osc. was connected to ground!.....How in the hell was it even modulating the RF carrier before it failed?...All I can figure was open filter caps may have allowed enough audio osc. ripple on the B+ to modulate the osc, but even that barely makes any sense.

After fixing that connection I had AM, but backwards....The RF was modulating the AF coming out of the RF port (yeesh!). The osc. DC voltages looked fairly reasonable so the only idea I could think of was to try paralleling another cap with the 2.2pF coupling cap. Remarkably that worked!....adding 66pF in parallel restored RF amplitude on bands A-E enough for reasonable AM modulation to be achieved. Not the best repair and band F is still doornail dead (after the unit died on me that band did not want to work) which irritates me, but at least I can use it for signal injection on AM sets again.

If anyone has any ideas on what the deal is with this thing I'm all ears.

The RF output showing AF is expected, as the only filter to remove audio
is that .01 uF cap. A smaller cap and a suitable RF choke across the
variable RF attenuator would fix that, but is not easy to get a choke that
will work on all bands; a different filter for each band is needed. I have made those
for my Eico sweep generator, they are external and go in the output lead.

As to band F ... check for bad ground connections, dirty switch
contacts, and try a new, hot, tube. Also check to be sure that its getting
full filament voltage and current.