I have a VINTAGE Heathkit Sine Square Audio Generator Model IG-18, recently purch'd from ebay.

Full manual:
https://www.rsp-italy.it/Electronics/Kits/_contents/Heathkit/Kits/Heathkit IG-18 Sine square audio generator.pdf

Image:
https://postimg.cc/56XbVdVC

All orig components seem to be in mostly good cond, but I replaced the electro caps anyway because origs. had drifted up in value. I went thru the calibration steps with voltmeter and o'scope, as described in the manual.

One mystery is that the square wave, at any level above zero, has a DC offset depending on what the amplitude one sets .So at max 10v, I have an offset of 1.4 vdc. This might be normal. Not sure.
Also, if I adjust the square wave amplitude, say from 0.5v to 1v, only the TOP (positive pulse width) rises. The negative PW edge stays put. Unlike in the sinewave, when increasing amplitude increases both the pos and neg. ends (crest and trough) equally.

The sine wave has no offset.

Both the sine- and square-wave pattern are otherwise clean and on par with other sig gens I own.
My 2023 model Koolertron 30MHz DDS Signal Generator, has none of these "issues" but the cheap Chinese devices like Koolertron are not quite so low in distortion.

I have not tested the IG-18's distortion specs, which are reputed to be good enough for audio use.

 

One mystery is that the square wave, at any level above zero, has a DC offset depending on what the amplitude one sets .So at max 10v, I have an offset of 1.4 vdc. This might be normal. Not sure.
Also, if I adjust the square wave amplitude, say from 0.5v to 1v, only the TOP (positive pulse width) rises. The negative PW edge stays put. Unlike in the sinewave, when increasing amplitude increases both the pos and neg. ends (crest and trough) equally.

That's normal. Maximum squarewave output amplitude is zero...10V peak into a 2k load. Output DC offset depends on the symmetry of the square wave, output amplitude and load.

 

If that's "normal", then why doesn't Koolertron square wave have a dc offset?
Btw: I'm just using standard probes with my Siglent oscilloscope.

 

Hello,

The square wave is coupled with a 250 uF capacitor to Q8, wich has a potmeter in its emittor to have the squarewave output.
What is the condition of this 250 uF capacitor?

Bertus

BTW: more heathkit manuals can be found here:
https://www.vintage-radio.info/heathkit/

 

If that's "normal", then why doesn't Koolertron square wave have a dc offset?

The Koolertron/Feeltech etc. is a modern yet inexpensive AWG and was designed to give the user the choice between having or not having a DC offset.

 

Hello,

The square wave is coupled with a 250 uF [c8] capacitor to Q8, wich has a potmeter in its emittor to have the squarewave output.
What is the condition of this 250 uF capacitor?

I tested the IG-18 after re-capping. The origs had drifted up in value to about 400uF but the ESR was still v. good. I installed 470uF NIchicon's for both C6 and C8 . Not sure if that higher value is what's creating the offset ???

 

I tested the IG-18 after re-capping. The origs had drifted up in value to about 400uF but the ESR was still v. good. I installed 470uF NIchicon's for both C6 and C8 . Not sure if that higher value is what's creating the offset ???

C8 has no influence on the DC offset. The square wave output is taken from a DC coupled emitter follower with a large unipolar voltage supply. An output signal with a DC offset is inherent to such a stage. Everything works like intended. A large cap between the fine amplitude pot and the coarse output amplitude switch would be needed to get rid of the offset. The cap has to be rather large to avoid waveform distortions at low frequencies.

 

C8 has no influence on the DC offset. The square wave output is taken from a DC coupled emitter follower with a large unipolar voltage supply. An output signal with a DC offset is inherent to such a stage. Everything works like intended. A large cap between the fine amplitude pot and the coarse output amplitude switch would be needed to get rid of the offset. The cap has to be rather large to avoid waveform distortions at low frequencies.

I think the manual [p 48] also mentions this somewhere ... that the implementation Heath uses preserves the fidelity of the squarewave ...

CAUTION: The square wave generator output is DC-coupled
to avoid poor low frequency response (see "Square Wave
Testing" on Page 52). The output is a DC signal that varies
from zero to some positive value when measured at the
output terminals. Do not connect this generator output into
DC circuitry without using capacitive coupling. (Observe
proper capacitor polarity.) Do not short the output terminals
at maximum (10.0V ) output.