Retake GW-GFG8020 Function generator

[fezder]

KISS, i took R313 out from circuit and that offset from CMOS is gone
but, i had to repair PCB so i might have done miswiring that caused cmos offset

 

[fezder]

hehe, i made miswiring in pcb-repair, that cmos-offset issue is now settled....
more info on duty-cycle:
effects TTL's negative, not postive
CMOS's positive, not negative
normal square waves negative, not positive

edit: typos...

 

[KeepItSimpleStupid]

fez:

I think the trigger is flipping on you. You don't have a reference.

What I'd like you to do is, scope signal named B on Q206:Emitter and trigger on it. I suspect it's a sawtooth. Adjust the offset so it's zero. e.g. 0 to 5 and back to 0 again. Also look at the other channel on the TTL output.

i.e. trigger on the triangle wave and look at both the triangle wave and the TTL out at the same time. You'll probably need level and slope triggering. Don't auto-trigger. The function generator lacks a sync pulse.

Vary the duty cycle. I'm expecting the triangle wave to stay the same and the duty cycle to vary like it's supposed to,

 

[fezder]

hmm, i'll see what i can do!

 

[fezder]

well, here's what i got, triangle wave is effected when duty is changed
yellow: TTL output
blue: emitter on Q206 (c1815)
normal duty:

bit changed duty:

even if i pull duty knob so i'd like to change positive portion of square wave, doesn't effect. it does effect on working unit. I can take measurements from working unit too as reference if needed, surely it would help narrowing things?

 

[fezder]

as this is how i can change duty cycle in working unit, shot from TTL output

 

[KeepItSimpleStupid]

Mmm. Why did the frequency change? The top part of the sq wave has the same length. I have to do some stuff today.
Because of the triggering, it didn't flip as expected.

 

[fezder]

No idea why frequency changed, perhaps rigol doesn't measure correctly
i didn't trigger it correctly?

ok, i tested bit frequency counter and for some odd reason frequency count changes on all my scope, as well as both function generators. in my understanding, if frequency is stable and only duty cycle is changed (high-low portion) frequency is still same? like with 50hz where period is 20ms, period is still 20ms if i change duty to say 5ms on-15ms off, totalling 20ms? i test this with one more frequency counter

for some reason this generator does ''duty cycle'' by stretching portion selected, so that is nowtconventional duty cycle (more like pulse-stretching, argh....),why i didn't notice that earlier...then again i haven't used duty much at all with generators, only with 555 and 40106 pwm circuits

so that quite well makes duty-cycle useless in this generator, i don't need pulse stretcher :/

 

[KeepItSimpleStupid]

See page 6 and 7 of the manual. Particularly how it operates. One side fixed and one side variable. Then note the table on page 7.

Do we conclude it's working "as designed"?

Triggering like I did, really shows the one side fixed.

It doesn't look like you can vary the duty cycle 1%-99% (picked numbers) with one knob?

What is currently confusing to me is that you say you have another one, "identical" that operates "differently?

 

[fezder]

hmm, well perhaps i understood how this device works in duty cycle bit wrong, i thought duty could be adjusted in some wahy 1-99 with two know settigns (pushed/pulled)
Only difference is that in working unit both portions of period can be adjusted separately, but even with that frequency changes as it stretches portion selected (high-low) and loses control of pwm-ratio

 

[KeepItSimpleStupid]

Take the "good unit" and trigger the other way using the triangle wave. Measure the frequency of the PWM wave.

Even if one side stays fixed and the other is supposed to move, the frequency should not change, right?

What (designation) variable resistor on the schematic changes the duty cycle?

 

[fezder]

Even if one side stays fixed and the other is supposed to move, the frequency should not change, right?

Yes! this is how i also understand duty cycle, just variable on/off time while frequency stays same

What (designation) variable resistor on the schematic changes the duty cycle?

it's VR204 for potentiometer (knob) and VR203 for trimmer (calibration) located in schematics quite at beginning from VCF input

 

[fezder]

ok, here's what i got, triggered on ch1 rising edge from working unit
ch1 (yellow) ttl output
ch2 emitter of q206

still frequency does change
oops, you asked to trigger on triangle wave, my bad

 

[fezder]

And, forgot to mention, i measured output with frequency counter i have and it too confimed that varying duty effects frequency on BOTH units

 

[KeepItSimpleStupid]

So, both have "bugs" which are considered a "feature" or are they "both broken"?
I realize that for all "intents and purposes" they seem broken.
See PM for an idea.

 

[fezder]

Right, i now have permission from GW to post these as public, so here goes:
Main unit, waveform generators and such:

Power supply section as well as TTL/CMOS:

And counter section:


So, only thing confusing now is that varying duty cycle effects frequency, which is odd and not needed behaviour. I have linked this topic for GW staff, hopefully there is someone who is able to tell what's going on

 

[KeepItSimpleStupid]

Fez:

In CAL mode is the duty cycle 50% like it's supposed to be?

 

[fezder]

Yes, it is 50%

 

[KeepItSimpleStupid]

Was thinking about a possible calibration issue. Guess not.

 

[fezder]

hehe, too late ;D i mean i already calibrated it to be 50%....