MPC1000 button fixing

I replaced all the small buttons on my newly acquired MPC1000 the other night, and it was dull but three hours well spent. None of the smaller buttons were much good, with the cursor keys being particularly shit.

I followed the advice in this video, and bought a bunch of Alps SKHHAKA010 switches from Farnell and warmed up the soldering iron. Credit to Farnell, they seemed to have upped their game since I started buying stuff from them – I ordered these switches at 8pm one night along with a few other bits to bump it over the free postage limit of £20, they were apparently dispatched at 9pm (“yeah right” I thought as I saw the email) and I got them the next day.

Here it is with the case off.

MPC1000 with the top popped off

Desoldering the legs of each button after I’d clipped them all out was a drag with only desoldering braid, I wished I had a better desoldering pump. Here are the dead buttons, grrr.

Dead buttons from the MPC1000

In the end it took me three hours in total, and that’s from taking it all to bits and wondering at the solid metal construction, desoldering the old buttons, fitting and soldering the new ones, and putting it all back together again.

Having spent ages looking at ideas for buttons for making my own sequencers, I liked the arrangement in use on the MPC, despite the switches themselves being rubbish. The translucent buttons have an LED just above the tact switch.

MPC button closeup

Just remains to see whether I get on with the MPC now – it’s been good fun chopping up beats with it so far, would be nice to actually finish a track with one.

No Comments

Distorted output from an Akai S950 sampler

Dscf2070

Mmmm, chips. Barely 20 minutes into using my newly-purchased S950, the sound became extremely, digitally-ish distorted. So before I start crying, it’s a good chance to whip the top off the thing and see what we’ve got in the grey box. Mmm, circuit board porn.

Dscf2066

…and it’s a lot busier in here than I expected; the weight of the box (almost 11kg according to the manual) should have been a clue here. But then this thing did come out in 1988 – my (also broken…) E-MU ESI 2000 is virtually empty by comparison.

So you can see the power supply on the left, with a large heatsink. There’s at least two big main circuit boards, the lower one hidden by top one, which I’m guessing is the voice board, because of the eight identical columns of chips, one for each voice. I imagine the main out is on the right-hand side somewhere. Underneath this I imagine we have some sort of CPU board. There’s a large trapdoor underneath for adding two 750KB expansion cards. If your S950 says 512 Kwords on bootup, it’s unexpanded – 1024 Kwords 1536 Kwords means it’s fully expanded to 2.25MB.

I was hoping that there was some sort of loose connection rather than a blown opamp or power supply problem, which I’d be rubbish at fixing and tricky to track down. There are wires connecting the circuitboard to the audio out board – waggling and pushing the wires down into the socket the wires in this area (top-right as you look at the picture above) fixed the problem.

Dscf2074

So I’ve not had a great chance to play around with it, but here’s a not-particularly indicative scratchy loop of a TR-606 and Boss handclap, with heavy spring reverb, sampled at 5KHz, sat next to an MS20 bassline recorded straight into Ableton Live.

2 Comments