News:

Forum may be experiencing issues.

Main Menu

Electric Mistress: need faster clock

Started by BillyBoy, February 18, 2016, 02:03:33 AM

Previous topic - Next topic

BillyBoy

Hi,

I have this version of an Electric Mistress on my breadboard:  http://www.metzgerralf.de/elekt/stomp/mistress/images/1976-electric-mistress-v2-schematic.gif

I'm trying to stay true to the original, but there are a few differences:

I'm using an MN3007, wired in like Current Lover, including the output filtering and transistor, CD4049, etc.  However, I didn't include the trimpot on pin3 to the BBD.  In this schematic, the signal is already biased coming out of the opamp into pin3. 

I'm also running at about 14.7V regulated.  My power comes from an LT1054 charge pump and 15V regulator.  My bias sits at 7.35V.  I don't need the LM741/BC309 stuff as a result.

In the LFO I subbed in a 3906 for the BC177.  I tried other PNPs and HFEs, but none of that seemed to matter.  Also, in one place there is a 50n cap and in another a 0.5u cap.  I used 47n and 470n instead.  Otherwise, I've duplicated everything else.

It seems to be working.  I have modulated output, controlled by the Rate and Range.  But there isn't enough flanging.  My voltages and behavior are in line with what they should be, everywhere except at the BBD pins 7 and 8.  I have 10V there instead of my bias level of 7.35V like I have at pin 3 - ?   At pin 2 of the MN3007 I get rates of 25-160KHz.  Should be 20-200KHz for the SAD1024, I think.  So I'm a little low.  Plus it needs to be doubled since I'm using the MN3007 with a longer delay.  I'm not familiar with this LFO, but nothing I've tried has adjusted my clock range correctly.  What do I need to do to go from 25-160KHz to 40-400KHz?

I'd like to understand what sets the high and low frequencies so I can adjust them to other values, too.  Any help is greatly appreciated!
Bill Gerlt
Gerlt Technologies
Custom Rack Effects

Scruffie

#1
Trust me, I tried, give in, you can get it higher but then you loose the width or LFO shape, it's a weird circuit.

Also it will always tick to an extent without extremely good layout as it uses the 4013 for the LFO, I have an original which ticks and every one i've heard has too.

That long chain of resistors biases the LFO, current source and VCO so it's all a really delicate balance, you're better off just building the current lover version and aiming for the same frequency sweep, it's close enough and much easier.
Works at Lectric-FX

BillyBoy

Yeah, been beating on this one for a while now.  I was already thinking of switching to the one used in the Deluxe, which seems to be about the same as the one in Current Lover.  Plus I'll have a clock trimmer and the world will seem more normal  :)

Have you tried a parallel MN3007? 
Bill Gerlt
Gerlt Technologies
Custom Rack Effects

Scruffie

Yeah I spent a long time a while back trying, the 339 is just too slow and the design too unforgiving. Yeah the deluxe and current lover (which is the 9V) are nearly identical apart from some low pass filtering of the LFO sweep in the range opamp IIRC, which as it's one cap you could easily tack on to the back of the current lover PCB.

No I haven't, but I don't see much point in it, parallel multiplexing is for improving sampling, doubling a BBD clock achieves the same thing effectively as the clock is then twice as high regards the filtering cut offs. Also... i'm not 100% certain it applies to the deluxe but I think you'll find while it looks like the chip is being used in parallel multiplexing set up it's actually just run directly in parallel so if one half the chip fails, the pedal still works.
Works at Lectric-FX

BillyBoy

I'm not a whiz on BBD stuff, so I might be in the weeds on my parallel thinking.

If you put them in parallel, they would probably be just a tiny bit off from each other.  I thought that might make it a little more "lush" or "swirly" when you put them together.  Could take one clock output from one and the other clock output from the other one (maybe this is the electric mistress way), or just mash all 4 channels together.
Bill Gerlt
Gerlt Technologies
Custom Rack Effects

Scruffie

Sadly, it doesn't work like that, you can improve noise and sampling putting them in parallel but they're always going to be 1024 stages and all getting the same clock frequency.
Works at Lectric-FX

lars

I've found that with the electric mistress circuit, the only way to really get a flanger out of it is to not use an MN3007. Something about the extra stages makes it not sound right for flanging, even with the clock way up, IMHO.

Look at the big box reissue Electric Mistress from around 1999-2000. Since SAD1024s were way out of production, EHX used a 256-stage BBD (the RD5106a) to approximate the original electric mistress sound. So the key here is to keep the circuit the same and reduce the number of stages of the BBD if you want faster delay times.

Right now I run my Electric Mistress clone (Current Lover) build with an MN3006, and it works awesome. I would definitely recommend an MN3006, if you want true flanging-type sounds. Otherwise you will spend a lot of time trying to design a clock circuit to overclock a chip into delay times it was never designed for...

Jules

Is the mn3006 a drop in replacement or other parts that need to be changed?
I always assumed the mn3007 version was close enough, I never have tried an original.

Scruffie

Quote from: Jules on March 28, 2016, 09:00:33 AM
Is the mn3006 a drop in replacement or other parts that need to be changed?
I always assumed the mn3007 version was close enough, I never have tried an original.
This thread was originally about the 18V mistress which is a different circuit to the 9V mistress but, The MN3007 is close enough to the original, I have to disagree with Lars, while I see where's he's coming from he's making assumptions and i'd say the MN3006 isn't long enough for an electric mistress sound.

The 9V mistress like the 18V model was never a big jetty flanger, that'd be more the deluxe, it should have quite a chorusy edge to it and a MN3006 can't get long enough delays to give that.

EHX used the RD5106 for other reasons over its length (they used the MN3009 in the reissue polychorus, if it was about BBD length, they'd have used it for both, no?) namely that the reissue was a copy of the deluxe model and needed a wide sweep which the RD5106 could give without the need of a 4049 buffer as it has lower clock pin capacitance, among other things.

The MN3007 is perfectly happy being swept over its boundaries with higher clock rates as long as there's a buffer, but you have to get the wet/dry mix right (very important and yes it does vary slightly with clock frequency but that's true of the 3006/9 etc. too) and the clock range right or it wont sound good. The quantity of stages isn't really the important thing, when you're designing something you use the most suitable part yes, but that doesn't mean the others can't be used.

The MN3007 has been happily and successfully used in A/DA flanger builds for over 5 years now, it's not the BBD.
Works at Lectric-FX