Hello Alex and all
I have had my AN1x open to repair and have played with the internal key mechanism dealing with velocity & aftertouch so I will describe what I have found and maybe you can apply the knowledge to help fix this or other problems. Although I do not have schematics or drawings, I am an EE and have put together how it works from knowledge and assumption of how I would have designed it:
Keyboard Velocity Control:
Velocity is measured by 3 actuator tabs on the middle underside of each key closing 3 contact points on an attached circuit board. The electrical contact is via a rubbery mask with attached carbon like pieces contacting etched switch contacts on a common PC board. As the key travels from the up position to the fully depressed condition, these 3 contacts of the key mechanism sequential make contact and close the circuit paths on the contact board. The time is measured between contact closures and a velocity is calculated. The mechanism and design relies upon several things to provide a consistent velocity from key to key:
- Mechanical tolerances of the contact mechanism controlled by controlled length of the 3 plastic actuator tabs on the key and the position/flatness of the contact circuit board. The contact circuit board is usually mounted with several dozen screws (be sure all the screw are there and tight) and must be held firm and flat to get even velocity.
- Contact condition and cleanliness. Although the circuit will probably detect key closures fine with normal aging and wear, you many times do not know how the previous owner treated his keyboard. I have found in similar keyboards things like Pepsi (or was it Coke) spilled into a keyboard, corroded contacts (sea air), etc. causing intermittent contacts.
- Spring tension of the key which is the force pushing up on the key while you are pushing down. Variances in key pressure will invariably cause the user to compensate and therefore create some variance in the final velocity calculated.
Keyboard Aftertouch Control:
Aftertouch is determined by pressure on a single resistive strip that runs along the entire keyboard length to an analog circuit that measure the resistance of the strip. Pressing a key down to the end of it's easy travel contacts the key with the resistive strip and press a little further puts additional pressure on the resistive strip. The analog circuit creates a voltage proportional to the pressure and the AN1x takes that voltage, digitizes it and creates an aftertouch value. The analog circuit has 2 tiny adjustable pots/controls mounted on the keyboard circuit board, one for offset and one for gain. The offset controls the initial amount of aftertouch applied to the modulation path it is set up for. This must be matched to the initial crossover point when aftertouch is first initialed, otherwise there will be a jump to the aftertouch modulation instead of a smooth transition. The gain control adjust just that, the sensitivity of the aftertouch to pressure. To get these right, I had to set the AN1x to use aftertouch for say pitch or vibrato with say an 80% effect to the modulation, then iteratively adjust the two pots while testing the result until you like what you get for the pressure you apply.
Good Luck
Bob
El Segundo, CA
----- Original Message -----
From: ecf1001
To: AN1x-list@yahoogroups.com
Sent: Saturday, April 28, 2007 12:20 AM
Subject: [AN1x] uneven velocity
Hi.
I noticed that there are a few keys that produce noticeably higher velocities than others when
struck in a fairly consistent way, and I think it's bugging me. Is there a simple cure for this
(like cleaning) or does it need to visit its makers? Could dirt be stuck there and be causing all
this or is it something more serious? How does the velocity sensor work, anyway? Is it an
actual key velocity detector (like on Fatar keyboards) or just a pressure sensor?
Also some of the keys in the vicinty of those with weird velocity responses smack the front
plastic lip on the front when played with only moderate force. Wonder if that's related.
Thanks,
Alex
[Non-text portions of this message have been removed]