I'm Still Confused About Hammond Drawbar Wiring

[jdeacon]

I've been unsuccessfully trying to get hold of a Native Instruments B4D drawbar MIDI controller. So I decided to build my own.

 

I had a Raspberry Pi lying around, and hadn't come across this site back then, so I decided to use the Pi to control it all. I've tried out IO piggy back boards and ADC piggy back boards. The Raspberry Pi experiments were successful. I have a Java program reading the IO ports and the ADC channels, and sending MIDI events over the Pi's USB to a USB-MIDI adapter and a synth.

 

Then I obtained some Hammond drawbars. They aren't B3/C3. They are from an H100, allegedly.

 

I've read the posts here, about the original wiring of the drawbars, but I'm still confused. I understand that they are giant switch buses and not potentiometers. But all the buses are connected to each other by a little circuit board at one end. I wondered if there were some differences in resistance between the ciruit board connections, but it doesn't seem so.

 

It's not a show stopper as I can unsolder the circuit board and add some resistors as mentioned elsewhere on the forum, but I'd love to understand what I'm disconnecting.

 

Can anyone kindly explain how everything isn't actually connected to everything, please??

[TK.]

Could you please give us a link to the posting "somewhere else in the forum" which explains a drawbar with "a little circuit board" at one end

This would help to understand the missing details.

 

Best Regards, Thorsten.

[jdeacon]

Thanks.

 

One of the Hammond drawbar threads is Unfortunately most of the attachments and links referenced are no longer viable. I thought I understood from the post that the drawbars were not potentiometers as one might have thought but that as moved they closed switches. That would be fine. As the poster discusses, one could add resistors and create voltages that my ADCs could sense. There is also a reference to adding resistors

 

However, and this is me and not something I've seen on the forum, when I look at the drawbars, they are indeed switches, but the busbars are all connected together by a little circuit board.

 

Here is the drawbar pair for a manual:

 

Here is a close-up of one drawbar's contact and the busbars:

 

These busbars terminate here:

 

And underneath we see that all the busbars are connected to one another! I thought those connectors having different sizes might indicate different resistances, but not according to my DVM:

 

So. Completely confused. I can desolder the little board and connect up a resistance ladder turning the assembly into a bunch of stepped potentiometers; but I'd like to understand what I'm disconnecting.

[TK.]

Well, after reading Henrygr posts (-> ) I would say, that you should try to contact him - he is the expert in this area.

 

Best Regards, Thorsten.

[jdeacon]

I think you're right. I checked the last time Henrygr was on though, and QBAS. But they haven't been on in over a year, or two. I will try a PM in case they're still listening.

 

I'm going to breadboard some busbars in the same configuration, add some potientiometric restistors and some loads and see what happens. As those threads point out, the drawbars are make before break. I've tried to work out on paper what that means but I've given up. Much easier just to try it. Before I desolder the weird little circuit boards.

 

Like one of Henrygr's posts, I'm still confused that apparently everything is connected to everything. Looking at some online wiring diagrams however, of fairly poor quality, it seems like the tracks on the Hammond circuit boards might have different resistances, but they're of the order of 1 ohm which my meter wouldn't be distinguishing.

[TK.]

Actually the drawbars (connected to ADC inputs) don't load the resistor chain with a noticeable value, because the impedance of the ADC inputs is much higher, so that they shouldn't have an impact on the "reference voltages".

 

The jitter issue (which probably happens since the drawbars act like an antenna for ambient noise) can be easily avoided in the software by defining threshold levels for the different drawbar stages with a suitable hysteresis.

 

Best Regards, Thorsten.

[jdeacon]

Someone on an organ forum has sent me a better quality Hammond schematic than those I'd found so far. The little board is a bunch of Nichrome wire resistors; but they don't add up to more than one ohm. How the Hammond works is still a complete mystery, however. I can only think it's to do with resonance circuits or something equally exotic.

 

Anyway, I've removed the Nichrom wire, and now I have a set of busbars that are no longer effectively connected to each other.

 

Then comes the next problem. Because any drawbar contact - the wiper - covers two busbars - make before break - and as there are more drawbars than busbars, there is a pathological condition with the drawbars in a ramp pattern/registration, where there's an uninterrupted path from the top rail to the bottom. A short circuit. And even without that, there's the problem mentioned in that historical thread: that when one busbar shorts out a resistor in the resistance chain, the other drawbars start to see different voltages.

 

I think I see a way around it. I'm going to do a few more experiments. And I have a horrible feeling I might have to try to remember Kirchhoff's rules from my school physics of 45 years ago. :shocked:

[ilmenator]

You're lucky that Kirchhoff already wrote it down for you and you don't have to come up with this stuff yourself!