Tuesday 24 May 2016

MXR Smart Gate

Another eBay purchase. MXR have done a few noise gate pedals but this is one is fairly recent (2008? 2009?) so I'm expecting something reasonably complicated, similar to the ISP Decimators. As I got it nothing happens in effected or bypass mode, just dead.


The small MXR enclosure is quite packed on the inside, four quad opamps and a lot of passives that look like 0805 packages. Interestingly these are TL054, precision versions of the commonly seen TL074 & TL084, so that suggests some actual DC instrumentation is going on instead of just amplifying and clipping guitar signals. I think the construction style is pretty standard for modern MXR/Dunlop/Way Huge stuff, red PCBs, surface mount, PCB mounted hardware.



There are actually two boards, the second has all the hardware (PCB-mounted) and a Coolaudio V2164M, a quad VCA which is sold as an equivalent (clone of) the SSM2164, which must be doing all the downwards expansion of the gate.

Looking this thing over I found that only one opamp was getting power when I plugged it in. All the VCC pins of the other ICs had continuity with each other, but not with the 9V supply or the one powered chip. I would normally be tempted to just run a jumper wire to 9V and see if that fixes things, but I wanted to see where the power supply connection was lost.
From the reverse polarity diode and main bypass capacitor in the lower right hand corner, the 9V trace runs to one IC and then underneath the input jack. I couldn't follow it anywhere else, so I decided to desolder the jack.

Open via underneath input jack.

Here's the problem, the underside of the jack and the PCB were quite corroded. I would guess that some liquid may have gotten trapped in there after reflow soldering the board and before the large through-hole jacks were soldered. I had already cleaned away all the green crap in the above picture, but one via is visibly corroded and green and does not connect to the other side of the board. This one was carrying 9V to the rest of the pedal.

Corrosion on the input jack



I could have scraped away at that via and tried to solder it through, but I didn't trust that, especially underneath a jack. In the end I ran a wire replacing the track, there was a convenient tooling hole in the PCB I could run it through to connect on the other side. The gate works like new again.

As to what's learned, I guess that's not a very reliable track to run power on. Looking at the board it seems like the trace could be considerably wider, which would not have corroded as quickly. What is surprising is that none of the vias on this PCB are tented, which is usually a very cheap/free option with most manufacturers. If that power via had been covered in soldermask the pedal would likely have never failed.