Ensoniq SD1 issue

[alfadawg01]

Good evening, VSE!

I'm having a problem with my trusty Ensoniq SD1. When I play individual sounds in sound mode, they play with aftertouch or pressure (in Ensoniq-speak) full on. The effect is most commonly volume, vibrato or filtering. Here's where things get weird. If I press the footswitch down (which applies sustain) the aftertouch effect disappears when I release the key and the note sustains with it's normal sound (no aftertouch effect).

When I play in program mode or a sequence, I can disable the aftertouch and play notes normally.

Two caveats with this problem. I bought the machine used in 1995 and the aftertouch never has worked very well so I don't use it. Secondly, this problem surfaced after I played it at church and foolishly did not plug it into a surge protector.

Does anyone have any thoughts on what has gone wrong and a possible solution? I have reinitialized the machine several times with no improvement. The preset mode workaround does the job but is a bit inconvenient. The SD1 works flawlessly aside from this problem and I'd prefer to bring it back to health. Thanks for your indulgence.

[Ashe37]

probably need to take the keyboard apart and clean it. also possible that one of the metal panels on the keys has popped off.

[blueknob]

I have a SD-1, being using it for years, even done festival gigs powered from an alternator set by the stage. All worked well, anyway, I've been giving this some thought, at the mo, best I can think of is, indeed the keyboard sensor can get dirty and cause key faults, the SD-1 does a "keyboard calibration -hands off" upon every switch on. I've only had a failure of that when some nasty person did bad things, was an easy fix but does goes to show if there is a problem with the keys this self test should show it up. Just a thought.

When I first got my SD-1, mucho hair pulling went on, one thing I've found is ensoniq really did think these keyboards out, in minute detail, meaning chances are there will a parameter that needs changing, it's just a matter of finding where it is.

The aftertouch ain't that great on mySD-1 either, some keys play well and others not so. I'd be tempted to check the menus for anything aftertouch and confirm exactly where the footswitch control. The SD-1 allows for 2 footswitches, they can be configured. More poking about required. A part of my hair pulling was also I felt the manual wasn't written too well so trying to find the right section was... frustrating.

[alfadawg01]

probably need to take the keyboard apart and clean it. also possible that one of the metal panels on the keys has popped off.

That's going to be my first task in the repair process as soon as I can clear a good chunk of a day to work on it.

[alfadawg01]

OK, I finally had the opportunity this afternoon to set up a table and take the SD1's keybed out for a look. Following instructions I found in several online sources it was relatively simple to extract the keybed and remove keys. Let's see what I found.

Here's the top of the coil boards under the keys where the two board halves join:

Ensoniq SD1 keyboard by Bill Wilcox, on Flickr

And here's the bottom of the same area. Looks like a hardwired connection to me:

Ensoniq SD1 keyboard by Bill Wilcox, on Flickr

Here's a broader view of the circuits on the bottom of the keybed:

Ensoniq SD1 keyboard by Bill Wilcox, on Flickr

After I removed all the keys, saving screws and springs in their own jars, I gently wiped down the coils with alcohol and dusted the keys with a damp cloth. Here's everything air drying and awaiting reassembly:

DSC_0404_2017-03-14_3769 by Bill Wilcox, on Flickr

Tonight or tomorrow I'll reinstall the keybed and see what happens.

[alfadawg01]

Sigh. I reassembled the keyboard and put it back into the SD1. No change, the original problem still exists. At least I didn't screw anything else up and the keys and coil boards are clean again....

[Ashe37]

ok, so it has two boards for the KB and they have not been bridged. I'm wondering if you've just run into the infamous VFX keyboard issue... I'd ask Rhino.

[alfadawg01]

ok, so it has two boards for the KB and they have not been bridged. I'm wondering if you've just run into the infamous VFX keyboard issue... I'd ask Rhino.

Look carefully at my second and third photos. There's a ribbon cable soldered to points on each of the coil boards. No issues there.

[Rasputin]

I don't know if this is the problem, but Q31 on the keyboard looks like it might be cooked. I don't know if that's the lighting or not, but it's a possibility.

[alfadawg01]

I don't know if this is the problem, but Q31 on the keyboard looks like it might be cooked. I don't know if that's the lighting or not, but it's a possibility.

Is that what's circled here in black? I'll open it up and check that again.....might be the lighting but I may have missed it. Thanks for the good spot.
DSC_0402_2017-03-14_3768_edit by Bill Wilcox, on Flickr

[Rasputin]

Nah, not on the KPC board but the actual induction PCBs. In the upper-left of your last photo there are three groups of two resistors. In the center pair of resistors there's a transistor. It looks distinctively more brown than the solid black transistors everywhere else -- almost a deep rust.

Could be a red herring because I don't think transistors commonly burn up in that manner very often, especially without nearby PCB damage. It would make more sense if the nearby resistors looked affected too, but... could be something.

[alfadawg01]

Ah, now I see it. On my close-up of the ribbon cable it's quite obvious. I'll open it up and look again, that does look damaged in the photo.....

[alfadawg01]

Nah, not on the KPC board but the actual induction PCBs. In the upper-left of your last photo there are three groups of two resistors. In the center pair of resistors there's a transistor. It looks distinctively more brown than the solid black transistors everywhere else -- almost a deep rust.

Could be a red herring because I don't think transistors commonly burn up in that manner very often, especially without nearby PCB damage. It would make more sense if the nearby resistors looked affected too, but... could be something.

I opened up the SD1 again and the transistor in question is fine. The lighting for the photo introduced the crispy brown color.

Here's an update: Ran one more test and played the SD1 via MIDI from an Alesis Ion. No aftertouch to be heard so it's definately a problem with the keybed control circuitry. Soooo.....a 68HC11A1P microcontroller is on order.

[Rasputin]

Ran one more test and played the SD1 via MIDI from an Alesis Ion. No aftertouch to be heard so it's definately a problem with the keybed control circuitry. Soooo.....a 68HC11A1P microcontroller is on order.

Just as a heads-up, unless you sourced the 68HC11A1P directly from another Ensoniq (such as a chip pull from Syntaur, etc.) then you'll likely have a completely non-working keyboard when you swap out the chip. The 68HC11 has 512 bytes of on-chip EEPROM which can (and, in this case, needs to) be programmed with custom code.

All three socketed chips on the keyboard processing board have proprietary code of some sort. A generic 68HC11 can be reprogrammed and swapped in if you've got the skills, and the KPC EPROM is very easy to verify or rewrite with a burner, but the other KPC chip [U3] is a one-of-a-kind.

Out of curiosity, does the behavior change if you go into MASTER and change TOUCH to HARD4?

[madtheory]

Out of curiosity, does the behavior change if you go into MASTER and change TOUCH to HARD4?

Yes, this makes sense. It's highly unlikely, I think, that external MIDI input gets routed through the keyboard control circuitry. That's for converting mechanical movement in the Ensoniq's keyboard into control signal. MIDI already has the AT encoded.

So in addition to Rasputin's suggested test- are you 100% certain that AT is working on your Ion?

Looking back at your original description of the issue, it looks like the actual pressure sensors are working fine, but the behaviour with the pedal would suggest some kind of bad ground or inconsistent loading or interaction of CV signals on the analogue side, or maybe at whatever point those signals are mixed, or multi-plexed (don't know if they are or not). So I'd look at everything before the ADC (I assume there is one taking in keyboard and pedal voltages) input.