Mods for MN3007 ADA Clone

[njkmonty]

I currently have 2 completed Moosapotomus MN3007 ADA Clones.
I wish to make one stock and the other modded out.

I've been scouring through many posts and have not been able to get all the information I'm after.

The mods Ive come across ( not to indicate i know what the hell they do!)....... are as follows...

1) Stereo mod -    (one dry signal and the other wet)
2) True stereo -    ( two outputs out of phase to each other??"!)

3) TZF              -       Through Zero flanging     (Through zero flanging lets you get the sound of actual tape flanging
                                                                            with an effected signal that goes in and out of phase on both sides
                                                                                  of the "zero" point your original guitar signal represents)
TZF mod
Use send & return so that any other circuit or pedal could be used for the second delay line. Looking at Jürgen Haible's Storm Tide flanger, he used buffers on both the send and the return. You may not need to worry about buffers if you use another pedal for the second delay line, but it maybe something to be aware of if you're going to try putting you own circuit together for the second delay line. So anyway, since I'm not 100% positive of the best places for the send and the return, I added extra pads at the outputs of IC1a and IC1b as potential send outputs, and extra pads at the inverting inputs of IC2c (shown as IC3c on Stephen's scheme) and IC2b as potential return inputs

HOW DOES ONE IMPLEMENT THE TZF ADDON WITH A MN3007 CHIP, AS SCHEMATIC IS FOR THE SAD CHIP???


4) Clock timing Capacitor  -          switchable

5) LFO Timing Capacitor    -          switchable

6) Bounce Mod
Bounce
The bounce circuit (big thanks to Stephen Giles) basically replaces R65, between the range pot and the clock range trim. See notes for Pads H & J, below.

(I found this.... by Stephen Giles
I can't find my manual right now but I found this reference from the manual:

When the BOUNCE control (a knob on the front) is full clockwise, only changes in the control signal (0 - 5 VDC, or what Steve refers to as CV in his post) affect the flanging while at full CCW only the control signal itself affects it. The control signals can come from the oscillator, the envelope follower, a manual adjust on the front panel, and a socket called remote (for a footpedal or other CV source). Unlike the Phaser, the FL-201 allows for combinations of the these four sources - for example the oscillator can be sweeping along at a given rate AND the envelope follower can add transients or longer releases.

At one extreme of the knob, the user brings a signal differentiator, pulse shaper, and an active filter on-line. At the other extreme a VCO for setting the low frequency limit and a full-wave rectifier are in the path. The knob allows for mixing between these two extremes.

They say, "(a)s this control is rotated clockwise, the flanging control signal is progressively substituted for by a damped sine wave initiated by changes in the control signal. This damped signal swings the flanging back and forth by a diminishing amount for each swing, simulating the "hunting" effect created by an AC or servo motor changing speed."

How the electrons flow is described in great detail in the section called "Circuit Description" But without the schematic in front of you, the IC references won't make any sense I'll quote what's relevant. Let me know if you are bored.

It seems IC11 is the differentiator and its output is dependent upon the rate of change of the input. They say, "(t)hus, when the oscillator changes from positive slope to negative slope," (crossing zero, right?), "the output of IC11 changes polarity." It says later, "The effect of all this circuitry is to generate a damped sinusoidal oscillation at a low frequency when the control voltage changes direction" (rising or falling, essentially anything but zero).

Then it says, "(t)his corresponds in classical flanging to removing the load from the reel motor. The damped oscillation in flanging is caused by the mechanical inertia of the motor - when it reaches its proper speed, it cannot stop changing speed instantly, and thus tends to overshoot, causing it to slow down, etc."




has anyone actually done any of these mods to the mn3007 board?
are they correct ?
I ve read a lot of talk about them but cant seem to find anyone who has actually done any of them

[njkmonty]

ARRRHHH IM going TZF Mad!!!!

or do i use this for the TZF??
http://thmq.mysteria.cz/TZF3007/v1_1/TZF3007_v1.1_Build_Instructions.pdf




can anyone help, or nudge in right direction?!!!

[snap]

more zero madness from a year ago: http://www.diystompboxes.com/smfforum/index.php?topic=107353.0

[Mark Hammer]

ARRRHHH IM going TZF Mad!!!!

or do i use this for the TZF??
http://thmq.mysteria.cz/TZF3007/v1_1/TZF3007_v1.1_Build_Instructions.pdf



can anyone help, or nudge in right direction?!!!

The add-on board is well thought-out, and is also, thankfully available for a cheaper-and-more-available 3207.

As thrilling and dramatic as through-zero can be, it's not the sort of thing one uses all that often, and outside of spectrally-rich mixes flanged with very slow sweeps, adds little value as a feature.  I'm glad I have it available, and it's fun to experiment with, but most of us are going to use flangers in subtler ways, such as trying to emulate a slow Leslie, or nail a chorus sound with a teensy bit more harmonic swirl to it.  So don't drive yourself crazy about it.

Stereo can also be fun, but has equally limited use.  In a live context, it is near pointless, unless you're Pink Floyd or similar, although it would be useful for recording purposes.

Perhaps a more consistently useful mod would be "theta processing", which you will find implemented in the Storm Tide, Son of Stormtide and also the Eventide Instant Flanger ( http://www.oldcrows.net/~jhaible/jh_storm_tide_flanger.html ). 

This employs a series of allpass (phase-shift) "lag-mode" filters  The more familiar lead-mode phase-shift stage has a cap feeding the non-inverting input of an op-amp, and a variable resistance from that pin to ground.  This yields more phase-shift, the higher the frequency.  Lag-mode reverses the location of the cap and resistance, such that the cap goes to ground, yielding more phase-shift the lower the frequency.

One of the issues with flangers of any sort is that when it sweeps low, it tends to sound boxey, metallic, and just generally unpleasantly resonant.  Adding several stages of lag-mode allpass to the delayed signal spreads out the location of the notches in the low end in a more pleasing way that will apply equally to faster chorus-type sweeps, as well as slower dramatic ones.  Happily, it's the sort of add-on that can be easily fabricated on a piece of perf or vero with a quad op-amp.

Alternatively, what has worked acceptably well in some instances is to simply insert some crude highpass filtering (bass cut) in the feedback path so that the most annoying aspects of higher resonance settings are not quite as pronounced.  In this instance, it could consist of reducing the value of the cap after the Enhance pot from 100nf down to 33nf or even smaller.  That same trick is used in the PAiA Hyperflange.  I've suggested it to folks who found the boxeyness of their Boss BF-2 irritating, and apparently this improved matters.

[njkmonty]

Every time I do one of these flangers, I get all excited, and start researching how do some mods, then end up getting a little disillusioned!

I think compromising i will..

Not worry about "Bounce Mod"
and just attempt to achieve a satsifying stereo output and TZF

1) Stereo mod,
I have a multi amp setup (not for live) and have read the more desirable stereo, isn't just one wet and 1 dry signal, going to 2 different outputs, but 2 outputs out of phase to each other. Is that correct?
below is the only adaption of a stereo add on I could find,  from Markusw.
But looking at  it looks like a mirror of the existing ouput.? is that correct?
Do I just do it that way? or inverting the add on output recommended? if so how do i do that??? 



Also I noticed on the NZF Flanger by Dr Alx it has an add and subtract switch, with one side going through an inverter stage??
Is that a feature that could be added with the stereo add on inverter section?


Now the TZF ...
adding Thomeeeque's 3207/3007  board, I can see where its send and returns would connect to the electric mistress, but where would i conncet it to the ADA ?
http://www.moosapotamus.net/images/FlangerClone_SCH_rev5_MN3007_jan2010.gif
I'm assuming the send would be at Pad "B"? 
but return would it be Pad Cor D?
or Pad G? or somewhere else?

also the only difference i could see between the TZF add on board (3007 vs 3207) is resistor r10 (mn3207 is 14K  & mn3007 is 1K)
mn3007 runs off 15v, and 3207 runs of 9? is that correct?

also the "theta processing"  suggestion sounds good, but I dont think Im cluey enough to implement it into the ADA without a schematic!
would not know where to start!

sorry for all this dribble,!!

[DrAlx]

Also I noticed on the NZF Flanger by Dr Alx it has an add and subtract switch, with one side going through an inverter stage??

The reason for the complicated looking switching is to keep all impedances balanced regardless of mode:
If you look at the junction of the (C13,R21) and compare it to the junction (C19,R32).
In add mode, both junctions see 10k going to Vbias, and 10k to a "mix resistor".
In subtract mode, both junction see 10k impedance.
You can do a simpler mixer with the same number of components but with only a single pole switch.
That's what I did with this TZF based on Thomequee's add on board. See this pdf...

http://www.aronnelson.com/DIYFiles/up/FlexiFlanger_Build_Instructions.pdf

There are some horrible hacks in the circuit, but it does show a simpler way of doing add/subtract.  The biggest problem I had was heterodyning.  I would strongly recommend skimming through the NZF thread, if only to see the excellent advice given by Ton Barmentloo (aka Puretube).  He basically sorted out my heterodyne problems.  There are a bunch of things you need to do to stop this from happening like make sure you decouple analog from digital sections (and the two digital sections from each other), use star grounding, have good physical separation between the two BBD sections.  The thing that is easiest to overlook is to make sure that the two BBD inputs are not coupled together.  I have seen lots of circuits with separate BBDs that do things in an overly complicated way (that includes the BOSS DC2, the Tantek Infinite Flanger, plus some others that I can't remember).
Those circuits end up using heavy filtering at both BBD input and output to prevent heterodyning. As Puretube points out, you don't need any heavy filtering if you go about things the correct way.  The thing to realise is that BBDs sample the audio at their inputs, and this sampling process actually puts noise onto the input signal to the BBD.  If the BBD inputs are not decoupled properly (i.e. don't have big impedance separating them) then clock noise from one BBD leaks out of that BBD's input and appears at the other BBD's input and you'll then get heterodyning.  All you need is a simple RC LPF at each BBD input with a suitably big R value, and a simple 2 pole LPF on each BBD output.

[DrAlx]

One of the issues with flangers of any sort is that when it sweeps low, it tends to sound boxey, metallic, and just generally unpleasantly resonant.  Adding several stages of lag-mode allpass to the delayed signal spreads out the location of the notches in the low end in a more pleasing way that will apply equally to faster chorus-type sweeps, as well as slower dramatic ones.  Happily, it's the sort of add-on that can be easily fabricated on a piece of perf or vero with a quad op-amp.

I looked into this a couple of years back and did an extensive analysis.  When you put all-pass stages in the clean path (not the delay path) as they do the Eventide and Stormtide, you can indeed cancel lower notches, but it is does not quite have the effect you say.

If you look at the R and C values that both those flangers use for their all-pass filters, you can calculate that they do their maximum notch-cancelling and notch-shifting when the BBDs are giving delays of much less than 1 ms.  That corresponds to the high end of the sweep when there are relatively few flanger notches, not the metallic low-end of the sweep.
For the low end of the sweep (e.g. BBD delay = 8ms) you end up with notches in practically the same place as before (i.e. no cancellation of lower notches). You can however get an interesting result if you also use feedback in the delay line.  What happens then is that the notches vary in depth across the spectrum in an interesting way. So it's more like taking the "mettallic" set of notches and running it through some interesting EQ that varies the pattern of notch depths.  You must have moderate feedback for this to happen though.

If you wanted to do what you describe and cancel the lowest notches at the bottom end of the sweep then you need to make the RC product much larger for the all-pass filters.  It doesn't stop things sounding metallic though, because the long delays give you hundreds of notches, and the vast majority of them hardly move at all in either frequency or depth.  End result:  It sounds just as metallic as before.

It's easy to experiment with RC values for the all-pass stages.  The change in flanger sound will always be subtle (it will always sound like a flanger but with slightly different EQ) and you'll notice the biggest change when the phase shifters are providing relatively small shifts.  The RC values used in the Eventide/Stormtide aren't a bad choice. 

I seem to remember the Eventide/Stormtide having stereo outputs.  I think maybe one channel used theta processing and one didn't.
That apparently does make a noticeable difference to the sound.

[Mark Hammer]

My understanding was that the theta thing did not remove any notches, but simply space them apart a little differently.  Keep in mind that the allpass stages are of the lag kind, where additional phase shift is applied for content below a given frequency.  The net effect may well be tantamount to a delay of 1ms, but that implies a tiny bit more delay for the low end than for the rest of the signal.  And since the spacing of the notches is a function of delay time, having a differential delay time across the spectrum ought to result in a non-harmonic spacing at the low end....in theory, at least.

I can send you a scan of the Electronotes document on theta-processing if you want.  Bernie Hutchins goes into a lot more math than my tiny pointy head can manage.

[snap]

more theta from last year: http://www.diystompboxes.com/smfforum/index.php?topic=106273.5

[njkmonty]

Now the TZF ...
adding Thomeeeque's 3207/3007  board, I can see where its send and returns would connect to the electric mistress, but where would i conncet it to the ADA ?
http://www.moosapotamus.net/images/FlangerClone_SCH_rev5_MN3007_jan2010.gif
I'm assuming the send would be at Pad "B"? 
but return would it be Pad Cor D?
or Pad G? or somewhere else?

[DrAlx]

Think of the junction of R41 and R42 as the"mix point".

The delay signal goes to the mix point via R42.
The clean signal goes to the mix point via R41.

So disconnect the bottom of R41 from B (but leave B connected to R14).

That dangling R41 leg is where you feed in the fixed delay from the TZF line.
But don't connect it directly.  Run it through a large decoupling cap (say 200 nF or more).

IC2b does the mixing.
So you could do a stereo mod by duplicating that block and creating another mix point, but have that second mix point mix different things.

e.g. (Swept Delay mixed with Clean ) for one mix point and (Swept Delay mixed with fixed TZF delay) for the other.

[njkmonty]

is this what you mean?



also could you do and option of the 2nd output  not being inverted?  out of phase?

[DrAlx]

Too complicated. You don't need a two pole switch for the bottom switch.  A single pole switch will do because you can leave the send line permanently connected.  Other than that, the bottom switch looks OK.

The top switch is totally wrong.  It won't mix anything.  You've misunderstood how op-amp adders work.  Read this...

http://www.electronics-tutorials.ws/opamp/opamp_4.html

You'll see that the 27k "mix resistors" are actually part of the "adder" circuit block, not separate entities.
All I was suggesting is that if you take a second adder circuit block and use that for the second channel, then you can play around with what inputs you pass to that second circuit block.  I am not recommending a particular combination of things to mix.

If you are sure you want to try a TZF then I would just focus on getting that working.  You could easily have more than enough problems with just that one board.

[njkmonty]

please dont grill me too hard!!!, your link was pushing the boundaries of my electronics absorbtion!



this way

output 1 :      1)    Stock    50% wet / 50% Dry
                      2)                 50% wet / 50% TZF

output 2 :      1)    stock     50% wet / 50% Dry
                      2)                 50% wet / 50% TZF
                      3)                  0% wet / 100% Dry
                      4)                  0% wet / 100% TZF

[DrAlx]

More like it but get rid of the top switch and leave the red lines permanently connected together

You'll then have two outputs:

Out1 mixes swept delay with either (clean or TZF delay) according to its switch 1 setting.
Out2 mixes swept delay with either (clean or TZF delay) according to its switch 2 setting.

[njkmonty]

how is this?

[DrAlx]

Exactly

[Fender3D]

You should add a recover stage after 2nd BBD.
If you check Charlie's schematic, after BBD there's a x2 gain stage (IC2c)...

Furthermore you won't have 2 "stereo" signals when in no-tzf 'cause the 2 outputs are the same.
You might connect first (stock) BBD signal on +input @ output 2's op-amp thus obtaining usual "stereo" output...
Look for differential op-amp.

[njkmonty]

I'm pretty sure I balls it up !,
how bad did i go?



this is how byoc does its stereo.

[Fender3D]

this is how byoc does its stereo.

yep, and this is how MXR summed 2 signals one of which inverted...