x0xsh0p square wave far from specs

Discuss x0x construction and related issues

Moderators: altitude, adafruit_support_bill, adafruit, phono, hamburgers

Please be positive and constructive with your questions and comments.
Locked
User avatar
sheled
 
Posts: 4
Joined: Thu May 14, 2015 4:03 am

x0xsh0p square wave far from specs

Post by sheled »

Hi x0xers,

I just got my kit a few days ago from x0xsh0p and built the POWER+VCO+VCF.
I'm having a bit of a problem with the VCO - the square wave doesn't look like a square.
Judging by the looks of it I would say damaged/wrong capacitor somewhere or the op-amp has a different spec.

5.33v coming in through J4 (measured just with 2 decimal multimeter)
The following pictures were taken with 3vdc applied to R88
When I probe R36 I get the following waveform (actually looks more like a saw): https://www.dropbox.com/s/f9br6i0vy9zwe ... e.JPG?dl=0
This is my saw: https://www.dropbox.com/s/f8ixwl9tn967byq/saw.JPG?dl=0
Time divisions on both pictures are 0.2ms/div, Volts/div is 2v/div so:
the square should be around 2272hz at this tuning (I didn't tune it yet of course because its not yet functional).
Vmin saw: ~6.5v, Vmax saw ~7.5v
Vmin square: ~3.75v, Vmax square ~6.7v

board pics (i am aware of the amounts of flux on it):
https://www.dropbox.com/s/axeakqcvr9snf ... 2.JPG?dl=0
https://www.dropbox.com/s/u8a4zuovom6vy ... 5.JPG?dl=0
https://www.dropbox.com/s/w2v4vttbg1o8n ... 7.JPG?dl=0

Cheers,
Elad

User avatar
rv0
 
Posts: 395
Joined: Tue Jul 14, 2009 4:50 pm

Re: x0xsh0p square wave far from specs

Post by rv0 »

slightly off-topic (or maybe not), but I noticed on your board pics that you may have issues later on with upright mounted resistor legs touching..
when a resistor footprint hole has a "circle" printed around it, that means you have to insert the lead there (till the resistor almost touches the pcb) and keep the "lead" side for the other hole.
i've seen this mistake on x0x'es before. the resistors can easily touch when the silkscreen directions are ignored, causing all sorts of faults.

User avatar
mmm
 
Posts: 517
Joined: Tue Apr 24, 2012 1:51 am

Re: x0xsh0p square wave far from specs

Post by mmm »

sheled wrote: Time divisions on both pictures are 0.2ms/div, Volts/div is 2v/div so:
the square should be around 2272hz at this tuning (I didn't tune it yet of course because its not yet functional).
That would be far, far too high. Since the square wave shape actually changes a bit with the frequency it's really hard to say what exactly is wrong here (except that the basic frequency would be already out of range). The amplitude of both waveforms is also too low; but this could also be the result of the way too high frequency.

User avatar
sheled
 
Posts: 4
Joined: Thu May 14, 2015 4:03 am

Re: x0xsh0p square wave far from specs

Post by sheled »

Thanks for the replies guys.

I'm finally back home and can dive into this again.

So I'm gonna go start pulling out the components off the board and re-solder.
Or - how critical is it really that the resistors sit on the board in the correct orientation (the side from which they stand from).
I understand that if the resistor body itself is touching the silkscreen - that might cause problems. I will pull them a couple of mm out and re-solder the joint but do i really need to pull them out and reverse the orientation?
Will also pull out and socket the ICs (already did one but can barely manage to pull them out now - will need to re-order some of them).

Thanks
Elad

User avatar
rv0
 
Posts: 395
Joined: Tue Jul 14, 2009 4:50 pm

Re: x0xsh0p square wave far from specs

Post by rv0 »

Don't resolder all those resistors now
just make sure the legs arent touching
and in the future, follow the silkscreen ;)

User avatar
sheled
 
Posts: 4
Joined: Thu May 14, 2015 4:03 am

Re: x0xsh0p square wave far from specs

Post by sheled »

So,
I'm back with some (hopefully useful info).
Tweaked around with the board, square wave still looks the same. managed to get the frequency of the wave generated with the 3v to R90 to about 250 Hz minimum, couldn't get it to go lower (with the tuning pot centered).
Saw Vmin is ~10v and Vmax ~15v so I guess thats OK.
Square Vmin is ~6.5v and Vmax ~10v

Some measurements I took with the 3v applied to R90:
Q26:
1: 5.33
2: 5.33
3: 4.79
4: 9.72
5: 5.27

Q28:
1: 9.67
2: 9.72
3: 11.85

IC11:
1: 4.70
2: 5.33
3: 5.33
5: 3.05
6: 3.05
7: 3.05
8: 11.85

Q24:
1: 8.93
2: 8.93
3: 8.32

Q25:
1: 9.68
2: 11.24
3: 9.10

Tuning pot:
0 -- 2.88 -- 5.33

I already killed two days on trying to figure out whats wrong on this thing and I can't find out why.
Today I replaced the electrolytic capacitors (C31, C32, C11 and C13), changed and socketed IC11, IC12.
A note about Q28 - x0xsh0p included two options for this transistor and today I did a swap for the second option (first was SK30A-O, now replaced to SK118).
Anyone able to point me in the right direction?
Cheers

Elad

User avatar
sheled
 
Posts: 4
Joined: Thu May 14, 2015 4:03 am

Re: x0xsh0p square wave far from specs

Post by sheled »

Good morning x0xers,
Status report:
I've finished building the kit yesterday and mounted it to the box.
There's still work to be done - x0xsh0p sent 1K resistors for the LEDs instead of 10K and I'll have to pull all 40 of them out since its nearly blinding.
Overall the synth seems pretty detuned - I suspect the trimpots sent with the kit.
Another audible artifact is that the pitch shifts over time - I could see that on the scope too when I powered the synth and gave the 3v to R90 - the frequency of the sawtooth dropped with time (cannot say how much), till it saturated at some point and stopped. the same thing happens now - when sequencing a few identical notes it is audible that the frequency changes over time.

Any ideas would be appreciated :)

Locked
Please be positive and constructive with your questions and comments.

Return to “Making x0x”