EMS Synthi A Clone

status:

all pcbs are completed but untested,

case is in progress,

open:  powersupply and mounting of all parts inside the case, testing, calibration

modification later

figured out the VCO shape cv via the patchbay problem I was having, with input from Derek, who's been really nice with helping me troubleshoot stuff. You need to remove resistors R197, R226 and R260 from the Z PCB and replace with wire jumpers. Then remove the wires connected to the wipers of the 3 shape pots, and attach a 68K resistor to VCO 1's wiper, and 36K resistors to VCO 2 and 3 wipers. Then attach the wires that were attached from the pcb, and the wires from the patch bay to their respective resistors. Then you'll have working shape controls via the knobs and the patch bay.

The On and Off pot wiring is corrected as so -


so basically you're just switching the On and Off pots.

Other knobs that are indicated wired backwards according to the wiring diagram - in other words the "1" is on the wrong side are

Attack
Decay

filter response (resonance)

output filters

Channel 1 & 2 output levels

Ok, got my trapezoid working correctly, all I had to do was change R143 to a 750K.

Q145 is a BF245C, Q48 is a 2N5461, C46 is a 35v 1uF tant.

What led circuit are you using and how is it connected??

Thanks, XXX, I’m using Graham Hinton’s circuit. LED with resistor and Zener diode in series. Realised that my Zener is probably 6V, rather than 6v8 as specified. I have some arriving tomorrow, so I’ll be able to tell you if that makes a difference. I might just see if I can order a neon lamp from Robin, as it’s a lot nicer than an LED anyway, which would do away with the “hack” but I’ve no idea if it would solve the issue

Ok, got this going. I had the connection to the -9v to the wrong place. I did a lot of experimentation with resistors and LEDs and zeners, and ended up liking Graham's values best, with a mouser part No. VAOL-3GWY4, which is just a pretty standard white led.

Replace the jfet with one that has a Vp of 3.5-4 or change the value of r148.use a 3.3uf rather than a 2.5uf cap for c48

VCF mod

For anyone interested in completely subjective testimony on the so-called "18db/24db" VCF mod, I can now offer some, after having installed it.

I first replaced the Rainbow Fish caps I initially used with ceramics. I just wanted to see if I could squeeze a bit more stability out the circuit-- it just sounded a little thin and wonky with the rainbow fish. The new caps had the exact effect I was hoping for. It's not a dramatic change, but it sounds smoother and more linear now with the update.

So, then I installed the 100ohm resistor and 0.1uf cap for the filter mod; again, I'm happy to say it seems to have bolstered the overall "stability" of the VCF. It's not dramatic, but the filter sounds slightly deeper, darker and smoother now-- to borrow a cliche, I'd say it's slightly more Moogy.

The differences are subtle, but I think the sound is improved. It's also slightly less nasal, and I could see some people preferring the more band-pass like unmodified sound, but for most uses, I think I'll prefer the modified sound.

Anyone using Derek Revell's PCBs, the modification couldn't be easier to implement as there are dedicated plated thru holes already in place for it on the Y board b/t diodes 16 and 23.

Despite its name, the mod actually ads a 5th pole to what was already designed as a 4-pole 24dB LPF, effectively bumping it up to a 30dBs. The additional pole, aka the mod, was standard on Synthi's manufactured after 1974. Anyway you can read all about it on most of the Synthi sites.

 Graham comment this with:

Swapping caps with 10% or 20% tolerances will move the centre frequencies of each stage around. Arguing that different types change the sound is only valid if they are matched for value, if you replace one that is -20% its nominal value with one that is +20% of course it will sound slightly different.