Electric Mistress with MN3007

[Fermin Rovira]

Hi guys! My diy small clone sounds great (thanks FP!) and I want to start my next "BBD stomp". I know that it is possible to build an Electric Mistress with a MN3007 BBD, but I have some doubts; SAD1024 is a dual 512 stage bbd, and MN3007 has 1024 stages. SAD could work in serial mode for working like a 1024 stage bbd, but when I read the schematic I found that it not works in this mode.
-In which mode works the SAD1024 in the Electric Mistress?

-It is correct to replace it with a single 1024 stage BBD without changing the clock frequency?

-It would be better to replace it by two BBD of 512 stages for emulate it?

-Does anyone has the original small clone schematic (with SAD1024)? I want to compare it with the MN3007 version.

Thanks!
Fermín (from ARGENTINA)

[christian]

the SAD1024 works so that it has it´s two 512-stages in parallel. I´m not sure if this raises the impedance on the clock inputs so that the chip can be clocked with higher frequency to get almost 0-delay(thats where the "brains-jump-over" effect comes).
Newer chips like the MN3***-series are not that good if compared to SAD1024 IMO (but I guess that this is very common opinion?), you will get a flanger but with higher minimum delay. That´s why most of the flangers just don´t cut it cause most of them use MN3007 or other from the same series.

MN3001 and MN3002 are also dual 512-stage BBD´s but I think you can´t get them so easily

check these all out at Synth-diy

[stm]

Around 15 years ago (during the golden bucket brigade era) I made a chorus and a flanger based on a MN3007.  Chorus sounds were great.  Flanger sounds were OK but not great.  On day a friend came with a Chorus/Flanger device, which according to what I know now it may have been an Electric Mistress clone.

It sounded soooo great that I opened it to find out what made it!  Surprise, the SAD1024.  I took the IC off (it was on a base) and made my flanger circuit gain using this IC.  Voila!  Great sound.  The explanation I can give is that the frequency response of the SAD1024 is way-way-way higher than an MN3007.  It can also be clocked over 1 MHz, while the MN3007 is specified for 100 kHz maximum. Of course most designs use the MN3007 with clocks over 500 kHz, but high frequencies are degraded due to internal limitations of the device itself.

My advice is:  for chorus, go with the MN3007. It has lower noise than the SAD1024.

For a radical and metallic flanger, you should use the SAD1024.  You can get them from Small Bear.

If you still want to try a flanger with an MN3007, a good project is the Ultra Flanger from John Hollis.  I don't know how it sounds, but it is supposed to do its thing reasonably well, and besides, there is already a PCB design for it.  By the way, this circuit puts up to 1 MHz of clock frequency into the MN3007.

Hope this helps.

[stm]

By the way, the Boss Flanger (BF-2 ?) used a MN3007 (1024-stage BBD).  I owned one for about a year, but it wasn't radical.  It was more like and enhanced chorus.

Later Boss released another flanger device, the Hi Band Flanger (HF-2?), whose only difference according to the schematics is the use of a MN3004, a 512-stage BBD.  I think they finally realized their unit wasn't getting high enough!

If you need the schematics I can e-mail them.

[Mark Hammer]

There are two major factors that influence whether a flanger will sound "dramatic": one is the minimum delay achievable, and the other is the ratio between the smallest and longest delay achievable (sweep ratio).  These two parameters are differentiable but intertwined.  If you have a very very short minimum delay, that certainly makes it easier to achieve much wider sweep ratios (e.g., going from 500us to 10msec is a 20:1 sweep ratio which would require 20msec delay to achieve if minimum delay were 1msec).  However, merely having a short minimum delay does not, of itself, achieve a wide sweep ratio.  The LFO and clock circuit still has to produce that.

The shortest delay achievable is partly determined by the input capacitance on the clock pins of the BBD.  The MN3xxx series from Matsushita has enough capacitance on those pins that once you get much higher than 100khz, the clock pulse starts to become "unsquared" and not reliably usable.  The spec sheets on those chips assume you will be using either an MN3101 or MN3102 to deliver clock pulses, so the stated minimum delay time is what is achievable in the absence of any buffering of clock lines.  As such, the minimum delay time achievable starts to become a function of the number of stages in the BBD (with fewer stages producing shorter minimum delays at the same clock speed), because the 3101/3102 does not provide a lot of current drive.

In contrast, the input capacitance on the SAD1024 clock pins does not provide an obstacle to much higher clocking speeds, so it is often the case that the "really good" flangers from the golden era used the SAD1024 because it could easily be swept to extremely short delays.  That does not mean that MN3xxx *can't*, merely that the Reticon chip can do it without much extra help.  When clock lines are buffered (usually with paralleled CMOS invertors to deliver a clock signal with lots of current), the MN3xxx series can be clocked out well past 1mhz.  A friend adapted an A/DA Flanger to use an MN3007, by buffering the clock lines, and clocked that sucker out to darn near 2mhz.  It sounded wonderful.

[mikeb]

check these all out at Synth-diy

For me almost the entire site seems to be broken with M$ OLE DB errors, unfortunately.

Mike

[puretube]

same problems here...

[stm]

Mark, your very right about the sweep ratio!  This is also applicable to phasers.

I didn't thought that a MN3007 could be forced to go as high as 2 MHs and work, and above that, not die!   I checked the Ultra Flanger schematic from John Hollis and this guy uses a CD4049 hex inverter to drive the clock of the MN3007 (3 gates per clock), apart from the clock generator IC.  This guy knew what he was doing!

[christian]

I forgot that Ultra Flanger!  I haven´t tried this myself, but if I´d had a crappy flanger instead of DEM, I would try to stuff in an external circuit with some 4049´s or 4069´s.  Is 6 inverters in parallel little too much?? Drain´s batteries?

Sorry, I had problems with the Synth-diy too, I didn´t see that before I wrote the msg.. Some time ago, I found a page that had data-sheets for the whole set of MN3***s, but I can´t find it with google anymore..

[Mark Hammer]

I still have the BBDementia.zip file up at my site (http://hammer.ampage.org).  Some of the information needs updating, and some very generous people have sent me some files to add to it, but despite its flaws, the file has a lot of BBD datasheets for you to sift through.

[StephenGiles]

You might be interested to know that I built one using a TDA 1022, and slung in a bounce circuit from the Eventide Instant Phaser for good measure!
Stephen

[Fermin Rovira]

Stephen: Tell me more about your ElectricMistress with TDA1022 . This BBD is available in my country  


Fermín.

[Thomeeque]

Edit: Original post removed, I've decided to start a new topic 9V Electric Mistress retrofit with MN3007 instead..

[WLS]

...
You can get them from Small Bear.
...

I maybe wrong but Small Bear does not carry the SAD1024A any more. They haven't had stock for quite some time. I believe the reason for this is the spike in price due to extremely low availability.


Bill

[Thomeeque]

 Yep, five years can make huge difference

So it's one more reason to focus on my retrofit!

T.