Sunday 22 May 2011

Additive Fourier Synthesis


Regular readers (poor things!) might recall that I won a genuine set of Hammond Drawbars on eBay (originally salvaged from a Hammond L100 donor organ)...


I've already explained that "drawbars are gain controls for individual harmonic components" - allow me to be a little more precise...

The nine drawbars in a "set" give partial control over the elements of two harmonic series; one series built on the fundamental (which is associated with the "8 foot" drawbar - the "foot" dimension being a legacy of pipe organs) and the other built on the sub-harmonic an octave below that fundamental. Of course, the fundamental changes with each key that is depressed - but if you imagine we're depressing only "middle C", the following picture is worth several lines of my boring words...


Notice from my graphic that the sub-harmonic series' drawbars are coloured brown, those which sound the same note (in terms of tonic sol-fah) as the fundamental are white, and other elements of the series are black. You will also notice that the seventh harmonic is absent (as it is a de-tuned minor seventh which doesn't sound so good!). Anyway - enough music for the moment - let's get back to some electronics.

The drawbars are a series of multi-way switches, in which each drawbar has a contact able to touch some of seventeen commoned bus bars. The size of the contact is such that two busbars are contacted at any time, avoiding open-circuits during movement. In original use, the busbars were connected using a resistance network, making a series of stepped resistors.

The resistor network was implemented using resistance wire, and the lengths of these wires (proportional to the resistances) reveals that it is a non-linear ladder. The resistance wires were still in place in my unit from eBay...


My interface had been developed for use with a series of potentiometers...


... but this wouldn't work with the drawbars. I had to re-interpret to work with the "multi-way" switches. Fortunately, the change was minor - you can see one drawbar's worth of circuit here, with the common block inherited from the "potentiometer" scheme above...


Notice that, as in the original Hammond organ from which these drawbars were harvested, I needed a resistor ladder. I designed the taper of this ladder to generate a voltage output proportional to the drawbar setting - this one was much higher impedance than the original ladder (by a factor of ~10000).

Here's the genuine Hammond drawbars in place...


and here's a screen-shot of the resulting settings in OrganizedTrio - EUREKA!



Bubbling over with success I went on to try to connect up a pedal drawbar (the L100 had only one (16') pedal drawbar) but hit a problem. The Organized Trio VST plugin doesn't allow external control of the pedal drawbars despite the claims of its documentation -


Oh well - it does sound good and the price certainly was right - cheapskates can't be choosers!

The absence of pedal drawbars left me with lots of "free" analog interface lines, which could be used to connect either analog (i.e. proportional) or "Boolean" (i.e. On/Off) controls. I had planned to do the latter with a separate "switch matrix" scheme, but opted for as simpler approach, using the analog interface, in which the potentiometer used for proportional controls was replaced by a spdt switch, to give "full-on" or "full-off" control...

Here's a temporary control solution, in which switches in the percussion and vibrato / chorus control sections have been implemented using toggle switches...


This gives me pretty much all the controllability of a B3 (bar the pedal drawbars and the restricted lower manual drawbars of the L100). Anybody know where I can buy nice soft action rocker keys???

In the course of adding these "Boolean" controls, I noticed another "funny" about Organized Trio. The percussion "On/Off" isn't a two-state control at all - in fact it is another "level" control, with proportional action. I wonder if the others are like this too.

That's all for now

...-.- de m0xpd

Thursday 19 May 2011

Dippers

A couple of weeks back I won a Heathkit GD-1u Grid Dip Oscillator in the weekly raffle at Warrington A.R.C.


Yes - of course there are the usual raffle prizes - bottles of booze, packs of Kit-Kat - everything you'd expect from the best club (I am wearing my flame-proof suit). But, despite the temptations of beer and biscuits, I prefer to pick up the nice pieces of hardware occasionally on offer - and this dipper certainly is nice!

OK - there was a down-side...

The dipper came without any coils and the meter glass had collapsed onto the movement. A quick disassembly and some suitable glue soon sorted the meter and my friend Albert, g3zhe (who would be called my 'Elmer' if we lived on the other side of the pond) has loaned me his set of coils to copy. Here's one of them...


The coil set came in the nice original fitted box...


First thing I needed to do, having borrowed the cols, was test the dipper. I made up a resonant L-C network using a coil from my g3wpo FET dipper and a 1nF capacitor...


A quick estimate of the coil dimensions and a visit to an on-line coil impedance calculator suggested an inductance of 12.03 micro Henrys, giving an associated resonant frequency of 3.53 MHz.

Here's a video (YES - Shack Nasties is entering the video age!!) showing me tuning the dipper after placing the test resonant network next to the "Band A" coil chosen to place the Heathkit GDO in the correct band...


Here's the frequency dial setting associated with the dip...


The reading of 3.75MHz isn't a bad result given the tolerance of the capacitance and the crude estimate of inductance. To "double check", I tuned the g1wpo dipper, which gave the following result...


AMAZING! 3.75MHz (Scale "B" at the bottom of the photo) - these things actually work!

In fact, the old Heathkit dipper seems rather more sensitive than the g1wpo - here's a(nother) video clip, showing the Heathkit meter as I move the test resonator up to the coil...


The same effect on the g1wpo produces a rather less distinct "dip"...


All I have to do now is copy the coils for the Heathkit - no mean task - I'll tell you how I get on.

What a fantastic raffle prize - sure beats a packet of KitKats!

...-.- de m0xpd

Saturday 7 May 2011

Progress on the Virtual Organ

Quite a volume of water has flowed under the bridge on the virtual organ project...

Firstly, the MIDI bass pedal electronics have been cast in stone (or, at least, cast on a small PCB)...


Next, I purchased a bunch of linear potentiometers from a trader on eBay, and made up a PCB to take nine of them (enough for a single manual). Here's the set...


Readers may recall that I've explained how my drawbars are scanned - but I needed to make up an interface (having previously only breadboarded the solution for 5 drawbars and a single additional analog control). I designed an interface to service a set of 9 drawbars and another set of 5. Two of these units will be needed - one for the swell manual & pedals and another for the great manual and miscellaneous analog controls.

Here's my schematic...


I elected to use the A733 PNP transistors I salvaged from the "Donor Organ" - not only as a signature "cheapskate" joke, but also to ensure that authentic Hammond sound!!!

Here's the finished unit, made up on some nice white PCB board I picked up at the NARSA rally...


So - here's a shot of the "organ" development to date (sans pedals, which are hiding out of view)...



You can even see the lunatic Leslie Switch!

I have the new drawbars AND the original 5 "test" drawbars running (the latter now configured for the lower manual).

Just as I finished the new set of "potentiometer" drawbars, a genuine set of Hammond drawbars came up on eBay. I snapped them up and will be fitting them to this project. Real drawbars are nine-position switches (rather than potentiometers), so an entirely new interface methodology will be required - but it have been fun developing this one (and I can still use it to implement general analog controls).

Watch this space to see the real Drawbars.

...-.- de m0xpd

The COMMONER Beacon

After a week of fairy-tale operation with the Royal secondary locator, my multi-mode beacon (Blogs passim) has reverted to type and abdicated the Royal moniker.

The mr0xpd beacon's FSKCW and S/MT Hellschreiber modes usually were visible on Johnan, on5ex's grabber in Zevergem, Belgium...


but the best reported QRSS performance was the 2706.6 miles to Nova Scotia, where loyal Knight Vernon, ve1vdm, reported seeing "MR0XPD - still at the wedding LOL!!!"

(In fact, you can just see the closing "pd" of the royal FSKCW call and the opening "mr" - look just above the QRM at 7000808Hz - the S/MT Hell is faintly visible, but only if you know where to look!).

The WSPR emissions were spotted in the US several times, including in Illinois by Steven, k9an...


in Michigan by David, w8fgu...


and in New York by Daniel, kc2sta and Craig, kc2lfi...


Having abandoned the royal status, the commoner beacon was immediately spotted this morning on WSPR...


and is now back on Johan's Belgian QRSS grabber.

The Notice-of-Variation is valid until Monday, but this marks the end of all my "Royal" beacon working,

...-.- de m0xpd