Super Sixteen build support thread

Creator of of the Super Sixteen Sequencer and based in San Francisco. This is a DIY and Finished Good product for Eurorack.

Moderators: Kent, extralifedisco

User avatar
extralifedisco
Learning to Wiggle
Posts: 49
Joined: Thu Apr 05, 2018 5:03 pm
Location: San Francisco, CA

Re: Super Sixteen build support thread

Post by extralifedisco » Tue Nov 17, 2020 12:34 am

titl wrote:
Mon Nov 16, 2020 6:55 am
Hi
The pdf and the video are very clear thanks for that.
I just finish the build, but i have two problems :
-The second number in the screen doesn't work.
-no gate on output.

I think soldering points are all ok, all ics are well powered, all the ground i can check are ok.
Do you have an idea to fix that?
Thanks
Interesting! For the 2nd digit you'll want to have a look at the transistors by the display (Q1,Q2,Q3) and the resistors just below it (R14,R15,R16).

A more likely culprit however is the header pin connections right near those resistors, as the header pins are sometimes quite difficult to solder correctly (particularly the male headers). I would recommend applying some flux and reflowing all the pins on that header if you don't spot anything else amiss. Additionally, the mating surfaces of the pins might have gotten oil or flux residue on them, so it's always a good idea to clean them off with alcohol before trying anything more difficult.

For the gate output, you'll want to inspect the LED, diode, and resistors right near the gate output. Double-check that the polarity of the diode matches the silkscreen. Again, the header pins near the gate output provide connection to the other board, so inspect those header joints and reflow them with some more flux if you are at all in doubt.

titl
Learning to Wiggle
Posts: 6
Joined: Mon Nov 16, 2020 6:48 am

Re: Super Sixteen build support thread

Post by titl » Tue Nov 17, 2020 11:24 am

Thanks for the answer but i think i will need to go deeper.
I check header pins with my multimeter in continuity mode, i replace the Q1,Q2,Q3 3906 transistors, i measure R14,R15,R16.
I reflux all this connections ( including the screen) and still no second number...
Perhaps the screen is broken?
For the gate, i replace the led in the right way, i measure the diode, i also reflux the connection and i have a beep with my multimeter.

I have to say i had to replace the atmega ic because i used one of my own and not your pre-programed one, it was quite long and oainful, i hope i not damaged something.After that i installed socket for every IC.
I also replace N4001 (D103, D104) with N4007 (no more N4001 in my stock).
Excuse my english and thank you for your help.

User avatar
extralifedisco
Learning to Wiggle
Posts: 49
Joined: Thu Apr 05, 2018 5:03 pm
Location: San Francisco, CA

Re: Super Sixteen build support thread

Post by extralifedisco » Wed Nov 18, 2020 3:47 am

titl wrote:
Tue Nov 17, 2020 11:24 am
Perhaps the screen is broken?
It should be easy enough to test this theory. Unplug the module and use your multimeter in diode test mode. Connect the positive probe to the display's top row, 3rd pin from right. Connect the negative probe to the top leg of any of the nearby 7 resistors to the right of the display. You should see one of the segments light up. If it does light up, you may have a short circuit from that pin to ground, or as I mentioned nearby on the transistors on 680 ohm resistors. You can check for continuity from the middle 680 ohm resistor to pin 12 on the atmega IC as well. If it doesn't light up, but the other digits do (you can find their source pins on the top of the display as well).

If you don't have a multimeter with diode mode, you can test it by attaching a +5v source (9v would probably also work) to that same 3rd pin, and putting the ground attachment on the other side of those resistors (the lower side near the shift register IC).
DSC01998.jpg
titl wrote:
Tue Nov 17, 2020 11:24 am
I have to say i had to replace the atmega ic because i used one of my own and not your pre-programed one, it was quite long and oainful, i hope i not damaged something.After that i installed socket for every IC.
I also replace N4001 (D103, D104) with N4007 (no more N4001 in my stock).
Excuse my english and thank you for your help.
Sorry for the slow replacement! Desoldering ICs is really hard, I generally do it destructively and replace the part altogether so it's a mess. Sockets are always a good idea - I will start including them in kits soon I think. I'm not sure what your issue could be - another thing to check for the gate output is the orientation of the zener diodes on the CPU board (d101, d102) and the type (1n4733). If you haven't got those 5.1v zeners, probably better to leave them out altogether as they could cause your issue.

If you post some photos of your PCBs I can help you look more closely at the soldering, orientation, etc.

titl
Learning to Wiggle
Posts: 6
Joined: Mon Nov 16, 2020 6:48 am

Re: Super Sixteen build support thread

Post by titl » Wed Nov 18, 2020 5:34 am

extralifedisco wrote:
Wed Nov 18, 2020 3:47 am
It should be easy enough to test this theory. Unplug the module and use your multimeter in diode test mode. Connect the positive probe to the display's top row, 3rd pin from right. Connect the negative probe to the top leg of any of the nearby 7 resistors to the right of the display. You should see one of the segments light up.
Ok
Yes it lights up.
No short to ground, just a continuity beep with Q2 pin1 emitter.
extralifedisco wrote:
Wed Nov 18, 2020 3:47 am
You can check for continuity from the middle 680 ohm resistor to pin 12 on the atmega IC as well.
Ok that's it, no continuity but i have a beep with other 680 ohm resistors on pin 11 and 13.
extralifedisco wrote:
Wed Nov 18, 2020 3:47 am
another thing to check for the gate output is the orientation of the zener diodes on the CPU board (d101, d102) and the type (1n4733)
Yes they are real 1n4733, i can read the number.

Image
Image

titl
Learning to Wiggle
Posts: 6
Joined: Mon Nov 16, 2020 6:48 am

Re: Super Sixteen build support thread

Post by titl » Wed Nov 18, 2020 5:41 am

Sorry you can find pictures here
Image
Image

User avatar
extralifedisco
Learning to Wiggle
Posts: 49
Joined: Thu Apr 05, 2018 5:03 pm
Location: San Francisco, CA

Re: Super Sixteen build support thread

Post by extralifedisco » Wed Nov 18, 2020 3:06 pm

OK, glad to see you're making progress! Your PCBs look ok as far as orientation and placement, but it would be helpful to see the solder side in a bit better focus if you can. You can also use the forum image attachment to get the photos inline and not use an external host (click on "full editor and preview" when replying).
titl wrote:
Wed Nov 18, 2020 5:34 am
Ok that's it, no continuity but i have a beep with other 680 ohm resistors on pin 11 and 13.
For the display digit path, I suppose it goes without saying that you need to locate the discontinuity from the 680 ohm resistor to atmega pin 12 by buzzing out the traces in between.
  • The 680 ohm resistor should connect to the nearby header pin
  • The header pin should connect to the opposite header (test this! a pin could be bent or cracked)
  • The opposite header should connect to the atmega pin 12.
The final cpu-board trace may be harder to follow - below is a diagram highlighting its path.
dig2a.jpg

For the gate output, you can do much the same thing. However there are some resistors in the path so you'll need to account for them. You can try to measure resistance from the gate output itself to the CPU pin 2 - should be about 2k. Path diagrams are below. Here's the CPU board path for the gate signal:
gate-out1.jpg
And here's the control-board path:
gate2.jpg
If continuity looks good, of course, check for shorts to ground along the path as well. Just to emphasize, I would suspect the header pins before other components as they are mechanical moving parts which introduce the most unreliability and strain on solder joints. Those resistors and transistors are unlikely to have failed. Diodes can sometimes fail short due to overvoltage or ESD, so it could be worth testing the voltage drop on the 1n4733a's and BAT 85.

titl
Learning to Wiggle
Posts: 6
Joined: Mon Nov 16, 2020 6:48 am

Re: Super Sixteen build support thread

Post by titl » Thu Nov 19, 2020 4:20 am

Hi
Thanks again so :
extralifedisco wrote:
Wed Nov 18, 2020 3:06 pm
The 680 ohm resistor should connect to the nearby header pin
The header pin should connect to the opposite header (test this! a pin could be bent or cracked)
extralifedisco wrote:
Wed Nov 18, 2020 3:06 pm
The opposite header should connect to the atmega pin 12.
Yes it's ok .

extralifedisco wrote:
Wed Nov 18, 2020 3:06 pm
The final cpu-board trace may be harder to follow - below is a diagram highlighting its path.
No ! but pins 11 and 13 connect with closest pins.You found the problem, i probably damaged the pcb. I can solder a wire between the ic and the header pin unless you have a better idea.
super16-cpuboard-verso.jpg
I have the exact same problem with the pin 2 gate output, i have no continuity with r103 ( but i have with D101). I tested the diode and the rest of the path which is ok.
I will try to take my time this time so i will wait for your answer before making an other mistake.

User avatar
extralifedisco
Learning to Wiggle
Posts: 49
Joined: Thu Apr 05, 2018 5:03 pm
Location: San Francisco, CA

Re: Super Sixteen build support thread

Post by extralifedisco » Thu Nov 19, 2020 3:42 pm

titl wrote:
Thu Nov 19, 2020 4:20 am
No ! but pins 11 and 13 connect with closest pins.You found the problem, i probably damaged the pcb. I can solder a wire between the ic and the header pin unless you have a better idea.
super16-cpuboard-verso.jpg

I have the exact same problem with the pin 2 gate output, i have no continuity with r103 ( but i have with D101). I tested the diode and the rest of the path which is ok.
I will try to take my time this time so i will wait for your answer before making an other mistake.
Yep, you should just be able to replace the trace by soldering a wire where it would go. Typically this is done with very fine solid-core wire called "mod wire" "kynar wire" or "wire wrap wire", but you can use anything you've got around. If you've only got some stranded wire, it's probably best to clip about half the strands from the stripped part as they can easily splay out and cause short circuits to nearby pins.

Also, it's a good idea to solder the wire as flat against the PCB as you can, and then tape it in place in 2 places so it doesn't get caught and pulled off later. But it's an easy repair so just mark the solder points and use whatever wire you've got so at least you know if that's the right fix!

titl
Learning to Wiggle
Posts: 6
Joined: Mon Nov 16, 2020 6:48 am

Re: Super Sixteen build support thread

Post by titl » Fri Nov 20, 2020 5:29 am

extralifedisco wrote:
Thu Nov 19, 2020 3:42 pm
Yep, you should just be able to replace the trace by soldering a wire where it would go. Typically this is done with very fine solid-core wire called "mod wire" "kynar wire" or "wire wrap wire", but you can use anything you've got around. If you've only got some stranded wire, it's probably best to clip about half the strands from the stripped part as they can easily splay out and cause short circuits to nearby pins.

Also, it's a good idea to solder the wire as flat against the PCB as you can, and then tape it in place in 2 places so it doesn't get caught and pulled off later. But it's an easy repair so just mark the solder points and use whatever wire you've got so at least you know if that's the right fix!

Ok done and it works!
Thanks a lot for your time. The supersixteen is really great.
Bravo !

ub1k
Learning to Wiggle
Posts: 32
Joined: Fri Dec 14, 2018 2:49 pm
Location: Switzerland

Re: Super Sixteen build support thread

Post by ub1k » Sun Nov 29, 2020 3:24 pm

Just finished building mine. The build guide is awesome, thanks a lot for that!

Upon turning the module on, the lights blink corrrectly, everything seems to work fine, except that I don't get CV/Pitch (gate works fine).
I've checked the MCP4822 and there's something fishy, as pin 8 seems to be shorted to pin 7 (GND). The short is in the IC itself, so I guess it means it's gone?
The weird thing is that I am using a socket, so I'm 100% sure I didn't damage the IC myself while soldering (and it came from DigiKey, it should be a reliable supplier). The circuit seems to be feeding it with +5V as expected...
Sadly I only ordered one of those... I'm afraid I'll have to source them from China and wait a few weeks more :bang:

User avatar
extralifedisco
Learning to Wiggle
Posts: 49
Joined: Thu Apr 05, 2018 5:03 pm
Location: San Francisco, CA

Re: Super Sixteen build support thread

Post by extralifedisco » Sun Nov 29, 2020 10:39 pm

ub1k wrote:
Sun Nov 29, 2020 3:24 pm
Just finished building mine. The build guide is awesome, thanks a lot for that!

Upon turning the module on, the lights blink corrrectly, everything seems to work fine, except that I don't get CV/Pitch (gate works fine).
I've checked the MCP4822 and there's something fishy, as pin 8 seems to be shorted to pin 7 (GND). The short is in the IC itself, so I guess it means it's gone?
The weird thing is that I am using a socket, so I'm 100% sure I didn't damage the IC myself while soldering (and it came from DigiKey, it should be a reliable supplier). The circuit seems to be feeding it with +5V as expected...
Sadly I only ordered one of those... I'm afraid I'll have to source them from China and wait a few weeks more :bang:
That's a drag! In-circuit I get about 550 ohms between pins 7 & 8. I assume you have already, but it's certainly worth testing to see if you still have the short without the IC in the socket - very easy to miss a solder bridge somewhere. If it goes away when you remove the chip, time to order another one.

I'm not too familiar with EU distributors - you can check the list on Microchip's website to see if any of them look familiar or near enough to ring up and see if they have any in stock. From what I can tell most of them ship MCP parts from the US anyway, but Avnet europe seems to keep stock in 2 separate places:
https://www.microchip.com/salesdirector ... witzerland

Octopart also lists stock at non-authorized distributors (none of which I know about besides Quest in California). Schukat seems to have stock in Germany but their MOQ is 5 parts - Conrad.ch seems to have some as well. Worth a look at any rate:
https://octopart.com/mcp4822-e%2Fp-micr ... ceAndStock

ub1k
Learning to Wiggle
Posts: 32
Joined: Fri Dec 14, 2018 2:49 pm
Location: Switzerland

Re: Super Sixteen build support thread

Post by ub1k » Tue Dec 01, 2020 6:05 am

I assume you have already, but it's certainly worth testing to see if you still have the short without the IC in the socket - very easy to miss a solder bridge somewhere.
Thanks but I have already checked it without the IC and everything seems fine. Moreover, I have measured the resistance with the IC out of the socket.
New IC it is. Thanks for the links, I'll see if I find something reasonable delivery-wise, but I'm not holding my breath.

Post Reply

Return to “Extralife Instruments”