EM3207 (v1.1) - MN3207 based EHX Electric Mistress (9V) clone

[Tompski]

Why?

Howard Davis who designed the Deluxe EM said the warmth and fullness of that pedal was just as much to do with the circuit as the SAD1024.

[12Bass]

The only mod you could do to make them sound identical (both sound like the V2) would be to rebuild the Current Lover to use the SAD1024 chip 

Sorry to say but, you will never get the same sound from it with an MN-series BBD as you would with a Reticon SAD-series.

I'm also curious to know why that would be. 

From what I gather, the SAD1024 has better high frequency response than the MN3007, so I'd expect it to be a little brighter rather than warmer, assuming the before and after filtering stayed the same.  There could be a bit of improvement in fidelity if the Reticon BBD were used in parallel-multiplex, but I'm not sure if that's used in the original Mistress. 

[ElectricDruid]

The only mod you could do to make them sound identical (both sound like the V2) would be to rebuild the Current Lover to use the SAD1024 chip 

Sorry to say but, you will never get the same sound from it with an MN-series BBD as you would with a Reticon SAD-series.

Evidence? Otherwise it's just somebody's opinion I read on an internetz forum somewhere

There *are* significant difference between the two chips, one of the most notable being the clock input capacitance, which is 7pF on the SAD1024 and 700pF on the MN3207. That means you can get some shorter delays from the SAD1024.
Is that what you're claiming here? That the sweep range on the original gets to the short end that the MN3207 can't reach?

Some numbers on it would be nice to enable us to tie this to facts. Enlighten me.

[Tompski]

I asked someone who knows a thing or two about pedals and circuits to compare the Current Lover ( I am using a MN3007 not MN3207) to the circuit diagram of a V2 original and this was his reply

- One thing that jumps out is that it looks like the Current lover has a make up gain stage at the end after the mixer (IC7_A) which provides the provision for the volume control. The EHX V2 not only just has a passive mixer output with no make up gain but it also has a low pass filter (470ohm / 47n) right at the output. So that would make it substantially warmer sounding than the Roxanne. Plus, the V2 would have a higher output impedance loading the amp or next pedal which would also contribute to a warmer tone.

If you want to keep to make-up gain stage at the end I would suggest tuning it to have a warmer output. You could also of course eliminate that final make up gain and have it output passively like the V2.

http://www.metzgerralf.de/elekt/stomp/mistress/images/1976-electric-mistress-v2-schematic.gif

http://www.madbeanpedals.com/projects/_folders/FilterMod/pdf/CurrentLover_2015.pdf?fbclid=IwAR1XVWp8KIj4cwfWdlziBsNNJ9EQ01n7hyNQM9n3VFTzuxC3V1R4QNMrSxk

[Scruffie]

The V2 lacks the input filter of the 9V the current lover is built on (the 1nF cap parallel to the 5k6, although this removes a lot of the aliasing noise in the V2 but it can be reduced to brighten things up a bit) the output filter is different to the 9V (470R vs 510R Edit: Seems the Current Lover used the V2 470R not the 9V 510R) it has a different delay range (20KHz minimum instead of 35KHz, which is 40 vs 70 for the MN3007 which will alter the sound a lot) and a different sweep shape (the LM311 VCO adds a 'lag' to the sweep).

That may not sound like much but it changes the sound quite a lot.

The MN3007 over SAD1024 shouldn't be a huge factor, while true it inherently rolls off a bit more highs than a SAD1024 it's not a night and day difference and you'd be hard pressed to notice it in a mistress, it will have less noise (although sometimes a little noise can be a good thing, I think reduction in noise is part of the reason some say the panasonic chips can sound "sterile") than either electric mistress due to running at twice the clock frequency (plus improved S/N) as the SAD1024 in the mistress doesn't make use of parallel multiplexing and the gain is fairly linear at the clock frequencies the mistress runs which is usually the issue when switching from SAD1024 to MN3007 as the MN3007 drops off gain at higher clock frequencies resulting in a less than 50/50 mix which reduces depth of flanging.

There's also running the pedal on 9V as opposed to the 12-13.5V the V2 provides and yes, the output boost has an effect on things too (although it's certainly useful on the mistress) but I suspect your biggest problem might be the clock frequency, a different delay range gives a very different sound.

Edit: That 220n/10k pair on the output boost input is cutting some bass, might be worth increasing the value of the cap to 1uF.

[Govmnt_Lacky]

Evidence? Otherwise it's just somebody's opinion I read on an internetz forum somewhere

There *are* significant difference between the two chips, one of the most notable being the clock input capacitance, which is 7pF on the SAD1024 and 700pF on the MN3207. That means you can get some shorter delays from the SAD1024.
Is that what you're claiming here? That the sweep range on the original gets to the short end that the MN3207 can't reach?

Some numbers on it would be nice to enable us to tie this to facts. Enlighten me.

Much of it is my opinion based on multiple builds with either chip however, what Scruffie says pretty much covers it. Most importantly (to me) being the flange depth difference.

[ElectricDruid]

Thanks both. That was indeed enlightening.

[Tompski]

What about my friend's suggestion in my quote?

[Scruffie]

What about my friend's suggestion in my quote?

Try my suggestion about the output boost cap first (you could also increase the resistor but you'll need a larger volume trimmer) that bass cut could be your problem.

High output impedance isn't exactly desirable and unless the mistress is last in the chain or has nothing buffered after it, it wont always have the desired effect.

[12Bass]

... the SAD1024 in the mistress doesn't make use of parallel multiplexing and the gain is fairly linear at the clock frequencies the mistress runs which is usually the issue when switching from SAD1024 to MN3007 as the MN3007 drops off gain at higher clock frequencies resulting in a less than 50/50 mix which reduces depth of flanging.

That makes sense.  In my A/DA build I added a trimpot at the output of the SAD1024 in order to tweak the mix; it's set so that the deepest comb filtering occurs in the mid/high part of the sweep where the effect is most dramatic (measured with a spectrum analyzer).  BTW, I've found that the SAD1024's gain also drops off at high clock frequencies, and varies somewhat throughout the range, though perhaps not as much at the Panasonic BBDs (haven't made a comparison).

[Scruffie]

... the SAD1024 in the mistress doesn't make use of parallel multiplexing and the gain is fairly linear at the clock frequencies the mistress runs which is usually the issue when switching from SAD1024 to MN3007 as the MN3007 drops off gain at higher clock frequencies resulting in a less than 50/50 mix which reduces depth of flanging.

BTW, I've found that the SAD1024's gain also drops off at high clock frequencies, and varies somewhat throughout the range, though perhaps not as much at the Panasonic BBDs (haven't made a comparison).

That's interesting, I'm trusting the datasheets to be accurate here (and the Panasonic ones have been in my experience and even a little conservative in some areas) the SAD1024 claims linear response up to nearly 1MHz but I already knew it overstated its abilities a little bit with its claim of a 340mS max delay time... I'm sure it sounds wonderful with 750Hz filtering.

[12Bass]

That's interesting, I'm trusting the datasheets to be accurate here (and the Panasonic ones have been in my experience and even a little conservative in some areas) the SAD1024 claims linear response up to nearly 1MHz but I already knew it overstated its abilities a little bit with its claim of a 340mS max delay time... I'm sure it sounds wonderful with 750Hz filtering.

LOL

IIRC, I measured a 2dB loss at 1MHz along with a slight gain (and noise) increase at the longest delay times (with A/DA clone).  It may well be more linear than the Panasonic chips.  Gain aside, IMO, the SAD1024 has excellent audio fidelity at high clock rates in parallel-multiplex... the delay path is super clear sounding! 

[JC103]

A few quick notes from my experiences in 2018:

I have 2 older MB Current Lovers running with MN3007 chips, both biased for 12V operation. I have an even older Ross Flanger with the 18V Mistress BBD. I have played the new and "improved" MB Current Lover designed only for the MN3007. You can hear the 18V Mistress tone in the Ross, I believe it is the chip's high frequency character that defines the sound that we love from the V2 era Mistress. The newer version of the CL does not sound very good to me. I think the opamp makeup gain on the output is part of the problem. I prefer the older CL version with the transistor make up gain. Thinking out loud... both MN3007 versions of the CL only use a single BBD, where as the V2 Mistress uses both halves of its BBD. I have experimented using my CL's in stereo as well as summing to mono. I think it sounds better/more authentic with two BBD lines, but the high frequency response is still not quite right when compared to vintage units. Would love to figure out a work around in the future considering that the Reticons are unobtanium.

[pelut5]

Here's what I was trying to describe above.

https://soundcloud.com/alex-lawrow/bbd-clock-trimming

Actually according the Electric Mistress Mystery page, the max range delay is more like 7 ms, which if 1000/7 = 142 Hz.
So the note would be between C# and D.  So instead of tuning to the "B" on the 2nd fret, tune a couple of frets further up.

Thanks a lot.
I've adjusted the internal trimmers (as you explain) of my DEM V2 and now it sounds perfect !!!
Best regards.

[DrAlx]

A few years ago I made a lengthy post (page 22 of this thread) looking at how the frequency response of the wet path (i.e. BBD onwards) in the EM3207 compared to an original EM using an SAD1024.

I found that the frequency response in the "wet path" of an original EM drops off more strongly than in the EM3207.
This can explained by the presence of the transistor buffer Q1 at the BBD output in the EM3207, and the lack of such a buffer in the original EM.

I believe Thomeeque put in Q1 because the EM3207 has (47k || 680nF) on its BBD output rather than (10k || 3nF) as in the original EM.  The buffering effect of Q1 causes more highs to feed into the path that follows (i.e. C8 onwards).
That is not what happens in the original EM.  The highs in the original EM wet path get shelved off (i.e. lowered with increasing audio frequency).

I suggested some mods for the EM3207 to make the circuits a closer match (e.g. by adding extra components before/after Q1) but I have now found what I think is a far simpler and better way of doing this.

I asked myself why Thomeeque used the values he did on the output of the BBD in the first place, instead of using 10k like the original EM ?  There is nothing in the MN3207 datasheet saying that a resistor of about 47k is necessary.  The example circuit in the datasheet shows (100k || 100k) on the BBD outputs, so maybe 47k was chosen based purely on that circuit?

Anyway, I got out my old EM3207 build to try something and guess what?
You can change the parts on the BBD output to match the original EM and drop the Q1 buffer completely!!!

• Change (R12, R13, R14, C7) from (4k7, 4k7, 47k, 680nF) to (1k, 1k, 10k, 3n3).
• Remove Q1 and R15 completely.
• Connect the top of C7 directly into C8

The maximum BBD gain of the EM3207 is then slightly reduced by doing this (about 1dB), and this is actually a good thing if you want a closer match to the SAD1024.
Note that maximum gain of the MN3207 is slightly over 2dB (according to datasheet and my measurements on the pre-modded EM3207), while SAD1024 with 10k BBD output resistor has gain of about 1dB.

So to summarise. If you make the changes mentioned here, you'll have a closer match to the SAD1024 based circuit, and lower noise than the original EM3207 too (since the shelving effect in the wet path once Q1 is removed reduces hiss).

[DrAlx]

Correction:  I just compared the result of the above mods and Q1 removal against an SAD1024 based EM, and my EM3207 now sounds too thin and trebly by comparison.
I can't tell if I messed up while modding things (its built on vero with solder all over the place) or if I have overlooked something.
So I don't recommend making the mod until I've got to the bottom of that.

[Scruffie]

Impedance.

[DrAlx]

The problem was with my "original" EM build. I had added an extra cap on the input that was dulling the sound. 

I am in the process of taking lots of measurements of 3207 against an SAD1024 clocked at half the rate and looking at BBD gain dependence on BBD output load impedance.

If all you have is 10k || 3n3 on the BBD output (and no other parts) then I found I was getting identical BBD gain for both chips across the whole audio band for practically the full clock range.  Only when 3207 was at 400kHz (using 4050 clock buffer) was I getting slight gain loss of 0.8 dB at highest audio frequencies, and the audio below 1600Hz still had same gain on both chips.

Biggest difference between the 2 chips when used in the EM circuit is that the 3207 does not have as high gain as the SAD1024 once you lower the load on the BBD output. e.g. using 3k3 || 3n3 on BBD outputs makes the 3207 between 2.3dB and 2.8dB weaker than the SAD1024.

If you look at the full network hanging off the BBD output in the original EM you see the impedance drop from 10k at DC to around 2k at 10kHz.  That impacts the 3207 more than the SAD1024.  I think the drain-source impedance of the BBD output is about 1k for the SAD1024 and and 3k3 for the 3207.

So the changes I mention above (strip Q1 and use same network as original EM) will have the effect of losing too many highs in the BBD path (3 dB more than an original EM by the time you get to 10kHz), while not making the changes will do the opposite and the BBD gain will be about 3dB larger then an original EM circuit for the high frequencies.

[rankot]

Is this the schematic you're talking about?

[DrAlx]

Yes, but with some R and C changes made to the audio path so that it matches this ...
http://www.metzgerralf.de/elekt/stomp/mistress/images/1976-electric-mistress-v2-schematic.gif.