Troubleshooting Heathkit Analog Trainer Oscillator Bleed?

doobedy · Joined: Jan 12, 2009 Posts: 19 Location: Los Angeles

Hello!

I just bought a Heathkit Analog Trainer on Ebay, much like this one:

Posted Image, might have been reduced in size. Click Image to view fullscreen.

Posted Image, might have been reduced in size. Click Image to view fullscreen.

I built up a CGS 52 wave folder to test it out, but when I plug the power in I get really bad oscillator bleed happening from the power supply.

I checked the four electrolytic caps, one was bad, I replaced them all, but it didn't fix the problem. Now, what I really can't figure out is why the position of the voltage pots affects the loudness of the bleed. I can actually make it go away entirely if I get + and - in just the right positions. Not far apart, either, like 12v versus 13v. I checked the pots with a DMM and the resistance seems to sweep smoothly through the range.

Here is a link to the schematic:

http://www.heathkit.nu/ET-3100_schematic_1600.jpg You can see the 10uf and 1000uf capacitors I replaced.

Any hints would be appreciated, qs my troubleshooting skills are nearly non-existent, and I'd like to be able to use it without disconnecting the function generator Smile

RF · Posted: Fri May 29, 2009 8:07 pm Post subject:

I think you might need some power supply decoupling caps in there.
Try some .01 ceramics to ground on the power supply circuit, from the fixed + 15 and - 15 volt lines and also after the adjustable voltage regulation.

Here's a nifty article on decoupling caps
http://www.intersil.com/data/an/an1325.pdf

good luck - hope that does it for you.

bruce

doobedy · Joined: Jan 12, 2009 Posts: 19 Location: Los Angeles

Tim Stinchcombe · Posted: Sun May 31, 2009 1:07 pm Post subject:

Hi doobedy,
I'm hardly an expert on power supplies, but it looks to me that that design is just asking for trouble! Since the signal generator is powered from the (supposed) fixed +/-15V 'references' at the two zener diodes (ZD101/102), then any fluctuation in current from the signal generator (and I imagine there is quite a lot!) will cause a fluctuation at the zeners, despite the 10u caps there, and even though they are going to be on the 'steep' bit of their curve. This fluctuating voltage is then used directly as the reference point for the variable supply (at pots R106/116), so it looks like it gets straight onto the variable rails.

Increasing C102/112 looks like it should help, but you might need to go many times bigger, perhaps 470u, or even a 1000u. I'd be tempted with a more radical approach, like completely disabling the sig gen when not needed, or even introducing another 220ohm/15V zener in parallel with the existing to power the sig gen from, and so better isolate that from the reference used for the variable supply.

I'm guessing that when you manage to make it go away that you have just come across some 'sweet spot' where everything plus and minus just happens to cancel out!?

Tim

doobedy · Joined: Jan 12, 2009 Posts: 19 Location: Los Angeles

Tim Servo

Too funny, I've got one of those Heathkit gizmos too. I don't use the internal oscillator, as the sine output doesn't work, but the power supply has supplied the juice for a whole bunch of Magic Smoke prototypes. It's a handy little thing, although the breadboard space in kinda limited. Also, congrats on finding the schematic! Hope you get all the bugs worked out, but I'd consider disabling the internal signal gen if nothing else.

Tim (handy little thing) Servo