Scary malfunction on my NS3... with a happy ending
Posted: 27 Aug 2024, 16:25
The other night I was fiddling around on my NS3 (OS 2.60) and went to start working on a synth patch but something was wrong. Nothing would happen when I turned the filter cutoff knob or several others.
Initially I thought it must be a software bug, so I rebooted the keyboard, unplugged the MIDI input that can sometimes be noisy, etc. Still nothing. I played around with a few of the other pots and noticed that one of them seemed to change several of the parameters at once, with the display rapidly flashing between them.
I started to think about the numerous electrical issues that could cause this. My initial hunch was a poor solder connection or internal failure causing ground plane noise on the one noisy pot. I opened up the keyboard, removed the panel PCB, and desoldered the noisy pot.
No dice... upon power up, the next pot over just started doing it instead.
I searched around for the pot (it looks like an ALPS RK11K1140AHZ) just in case I would need to replace it, and searched around for the entire PCB. The right side is "Panel 2"... Syntaur has "Panel 1" for an eye watering $450 USD and doesn't stock "Panel 2." It was now about 12:30AM so I slunk back in my chair, fired off a support ticket to Clavia for a quote on a new Panel 2 PCB, and headed off to bed...
I started to rack my brain as to what would cause this. 34 pins shared between two panel PCBs, with at least 80 or 90 pots and buttons on each... and at least 8-10 of those must be dedicated to the OLED displays and a couple more for power and ground. Most pots not working... one pot controlling everything... this must be a shift register and the clock signal is dead-ending somewhere. And it must be on the panel PCB, because the cable is fine, the connections are solid, and the OLED display, most of the pots, and all of the buttons and endless encoders are working.
This morning I had a few minutes to spare before work and decided to have one more look at the board. Maybe I can build a shopping list to start replacing D flip flop and ADC ICs.
My inventory of the board included several of both of these:
Intrigued, I got out the x-acto knife and carefully scraped the cruft off the pins, then gave it a quick squirt of DeoxIT, cleaned everything up, and reassembled. There was some brief glitchiness on startup with one of the encoders sending some spurious input, and then it cleared up. Everything is now working perfectly.
I truly have no clue how those pins got corroded. The instrument has never been exposed to any kind of rain or harsh conditions. My last outdoor gig was nearly 4 years ago. Since then it's been either in my house set up on the stand or a waterproof Pelican road case. So I'm not sure if it was a manufacturing defect, condensation (although it hasn't been subjected to severe temperature fluctuations either) or just a freak occurrence.
Anyway I hope this can help someone in the event this happens to anyone else. Happy playing, folks!
Initially I thought it must be a software bug, so I rebooted the keyboard, unplugged the MIDI input that can sometimes be noisy, etc. Still nothing. I played around with a few of the other pots and noticed that one of them seemed to change several of the parameters at once, with the display rapidly flashing between them.
I started to think about the numerous electrical issues that could cause this. My initial hunch was a poor solder connection or internal failure causing ground plane noise on the one noisy pot. I opened up the keyboard, removed the panel PCB, and desoldered the noisy pot.
No dice... upon power up, the next pot over just started doing it instead.
I searched around for the pot (it looks like an ALPS RK11K1140AHZ) just in case I would need to replace it, and searched around for the entire PCB. The right side is "Panel 2"... Syntaur has "Panel 1" for an eye watering $450 USD and doesn't stock "Panel 2." It was now about 12:30AM so I slunk back in my chair, fired off a support ticket to Clavia for a quote on a new Panel 2 PCB, and headed off to bed...
I started to rack my brain as to what would cause this. 34 pins shared between two panel PCBs, with at least 80 or 90 pots and buttons on each... and at least 8-10 of those must be dedicated to the OLED displays and a couple more for power and ground. Most pots not working... one pot controlling everything... this must be a shift register and the clock signal is dead-ending somewhere. And it must be on the panel PCB, because the cable is fine, the connections are solid, and the OLED display, most of the pots, and all of the buttons and endless encoders are working.
This morning I had a few minutes to spare before work and decided to have one more look at the board. Maybe I can build a shopping list to start replacing D flip flop and ADC ICs.
My inventory of the board included several of both of these:
- NXP Semiconductors 74HC374 D flip flops (labeled "HCT374")
- TI 108S022 ADCs (I think the ones Clavia used are not TI, but pin-compatible)
Intrigued, I got out the x-acto knife and carefully scraped the cruft off the pins, then gave it a quick squirt of DeoxIT, cleaned everything up, and reassembled. There was some brief glitchiness on startup with one of the encoders sending some spurious input, and then it cleared up. Everything is now working perfectly.
I truly have no clue how those pins got corroded. The instrument has never been exposed to any kind of rain or harsh conditions. My last outdoor gig was nearly 4 years ago. Since then it's been either in my house set up on the stand or a waterproof Pelican road case. So I'm not sure if it was a manufacturing defect, condensation (although it hasn't been subjected to severe temperature fluctuations either) or just a freak occurrence.
Anyway I hope this can help someone in the event this happens to anyone else. Happy playing, folks!
