Wednesday 26 October 2011

Shelving Audio Filter

I was fortunate to pick up a Leslie 825 speaker for my virtual organ project (Blogs passim) for a song on eBay...

Part of the explanation for the attractive price was that - as the seller put it - this is "a Leslie 825 speaker with a difference".

I'll let the seller go on with his description:

"the difference is that the (previous) owner had a number of keyboards and he modified the cabinet by inserting two other speakers front and back - he drilled holes in the cabinet and placed a grill over the front - the grill is available but is not in the picture so that you can see what lurks beneath".

Well, now YOU can see the ugly array of holes in the picture above. I have fixed the grille on to make thinks look a little less unsightly and, like the seller, I haven't yet bothered to test the additional speakers. But the Leslie itself works well enough (with one trivial exception, to be described below).

For those of you who don't know, a Leslie speaker includes rotating mechanical elements (horns and/or baffles) to introduce a cyclical variation to the radiation from the loudspeaker which, in interaction with the acoustics of the performance / listening space, makes some interesting spatial audio effects involving amplitude and frequency modulation and comb filtering. It was developed by Don Leslie...

in an attempt to emulate the chorus effect produced when ranks of (inevitably somewhat de-tuned) organ pipes speak together, giving electronic organs a "pipe voice". The mechanical components are contrived to rotate at one of three different speeds (slow/stop/fast or, to respect the orthology, chorale/stop/tremolo) under the control of the iconic "half-moon" switch. Paul, g1dva, tells me he has seen rotating speakers with continuously variable speed control - but we'll dismiss these as pathological. Certainly, the classic Leslie only operates at a few discrete speeds, latterly the three described above.

Older Leslies used a horrible electrical interface, in which mains power, control signals and the audio were all applied through a long, multi-way cable terminated with special (i.e. expensive) Amphenol connectors. In the case of the 825, it is a 9-pin system, with the following connections...

3YellowNo Connection
4OrangeStationary Input / Aux Control (not used in 825)
5Greendc Output (+28V)
6WhiteSlow Motor Control
7VioletFast Motor Control
8Gray240V In
9Blue240V In

(The colours are those of the cores of the "official" Leslie Cables).

The seller of my 825 kindly gave me a 9-pin cable to connect up the device - thanks Bob!

Well, I had the cable but no socket for the organ end, so I lashed things together with a very unsafe Heath Robinson connection, involving individual pieces of "Choc Bloc" connector...

on each of the pins of the plug - DO NOT TRY THIS AT HOME, CHILDREN!

All this gave me opportunity to confirm that the speaker was working, to bask in the glorious spatial effects of a real Leslie (emulations only work up to a point) and - after the honeymoon was over - to become disappointed with the sound!

The 825 has no horn (and associated HF unit) - only a large "full-range" speaker firing through a rotating drum. In consequence the top end is hardly what might be described as "sparkling". In fact it is so poor that I began to think there was something wrong with the "key click" function in the fantastic VB3 software and I actually contacted the programmer before realising that the problem lay right there in my new 825 - sorry Guido!

All this led to some creativity - I acquired a 9-pin socket to replace the temporary lash up (thanks George) and I decided to build a filter to boost the HF response and try to rescue some of the sound. The (sound card) output from my virtual Hammond wasn't enough to drive the speaker to satisfactory levels, so a Pre-Amp stage with some gain was indicated too.

Here's my HF shelf filter, seen as an extract from the LTSpice simulation of the entire circuit...

The series combination of R5 and R6 was actually realised with a potentiometer, configured to give me a variable HF lift. Here's the magnitude response simulated in LTSpice for an arbitrary setting...

The lift above 1kHz is the "shelf", made into more of a "bump" by the low-pass corner (deliberately) introduced by R1||C4.

I turned the idea into reality in a nice sloping front RS box, picked up at a rally somewhere, intentionally copying the format of the Leslie "Combo" Preamp. Here are the innards...

and here's a view of the back panel...

In addition to the variable high frequency shelf, there is an internal jumper to select 0, 10 or 20 dB overall gain and an external volume control. I haven't had chance to fit knobs to the controls yet!

There is (as is usual) a fly in the soup - the "d.c. output" from my new Leslie (pin 5) isn't working. I looked at the schematic and there is little to go wrong (I suspect a dead diode, D14, or an open connection), as the d.c. comes straight from the power supply to the amplifier driver stage, which most emphatically IS working! Still, at the moment, my new PreAmp is powered from a PP3 battery.

Result - FANTASTIC. The 825 is totally transformed and a pleasure to use.

I have some rather more exciting Leslie-related stories to tell - but they can wait for another day.

For now, I'm all in a spin!

...-.- de m0xpd

Update: Fly Rescued from Soup!

This evening I made up a shorter cable for the 825 - having 30 feet of cable coiled up is a recipe for trouble, not to mention hum. I so doing, I discovered that the 28Vdc issue isn't a speaker fault at all - just an intermittent connection in the long cable. The shorter one is fine (with new connectors) so now the PreAmp is powered from the Leslie as planned.

No comments:

Post a Comment