Madbean Current Lover (Flanger) Sweep cut out

[Mcentee2]

Hi, I hope somebody can chip in and help with this one; I posted the following in the Madbean forum but responses are a little slow:

http://www.madbeanpedals.com/forum/index.php?PHPSESSID=bfea08e1c0be7961799a7ce77155de0b&topic=22654.0


Basically, it is the 2015 version of the current lover - and with the Rate at slowest and Range on widest the sweep seems to cut out at the top end, with a small "pop" at cut out and cut back in.

Details are in that linked thread, biasing seems "ok" otherwise, and the behaviour is audibly there (listening to the noise sweep) even if the MN3007 is biased "off".

Changing the clock trim makes no difference to the cutout itself apart from the rate at which it sweeps through.

IC3 (MN3007) voltages all check out fine except during "cut out" (see below)

IC4 (4013) Voltages are where I think something may be amiss, poss with the feed from the LFO/VCO  - with output Pins 1 and 2 (going eventually to the MN3007) misbehaving during the cutout (Pin 1 goes from 4.5v to zero, and Pin 2 leaps from 4.5v to 8.8v during the cutout) as a symptom.

IC4 Pin 3 (the clock) varies between 0.6v and 8v when Range=max, Rate=min

The minimum voltage at IC4 Pin 3 (the clock) where the cutout stops is about 2.3v, as I increase Range/dec Rate then that low voltage seems to go below 2.3v (ish) and the cutout starts, the lower the vltage towards 0v then the more the sweep peak is cutout.

Is that IC4 Pin 3 voltage swing "normal " (0.6v on the low side, corresponding to highest part of the LFO sweep when range maxed/ rate min) ?

or whether it is the 4013 that itself is misbehaving ?

I have no way of checking the clock frequency at Pin 3 though to see if the clock is actually misbehaving over and above that voltage swing


I don't know a lot about the flanger circuit itself, so am really lost as to where to start with this one in the LFO/VCO, but can read and act on schematics if I can work out (or be taught what each bit is actually doing in the overall scheme!)

Otherwise, in the range/rate settings that work, everything sounds brilliant, flanging is really nice !

[DrAlx]

I have the EM3207 which is a similar EM clone to the current lover.
The VCO can only go so fast.  If it tries to go too fast (which can happen when the control voltage for it is too low) then you will get the drop outs. 

The thing to focus on is Pin 2 of the 4013. EDIT LM311.
That is the control voltage (CV) that sets the target voltage for the Pin 3. (i.e. roughly the max voltage that the clock cap charges up to).   Pin3 minimum will be around 0.6 because the cap is discharging through a diode, so it never fully discharges.

I have a similar clone (the EM3207) and found the clock cuts out when the control voltage (CV) at pin2 goes much lower than 0.85V.
(I am using a 47pF clock cap by the way rather than 22pF and would recommend you do the same... 
See http://www.diystompboxes.com/smfforum/index.php?topic=91981.msg1010635#msg1010635).

One thing to note about the range control of the 9V EM is that if you use slow sweep and increase the range, then that not only increases the maximum value that the CV reaches in a sweep, but also decreases the min value that the CV reaches.

So try this...
1) Put the thing into filter matrix mode.
2) Put color at maximum and range at maximum. That should give bell like sound.
3) Measure the CV at pin 2 of the 4013EDIT LM311.  Now decrease the Range pot.  it will make the CV at pin 2 go lower.
At some point you may find that the bell-like sounds disappear (due to the clock cutting out).  Find what the voltage at pin2 is when that happens.  Report that voltage here, and also confirm what value clock cap you have,
4) Then go to slow sweep mode and see if you can find the minimum value that Pin2 goes to in that mode.

The point of this test is to first of all see if the control voltage is the problem (i.e. if it's  going too low) or if the problem is that the clock cap is being charged too fast.

[DrAlx]

Sorry.  I made a big screw up.  I was trying to reply from memory and got the part numbers wrong. The 1's and 3's got me all muddled
The chip to focus on is the LM311 (the comparator) not the 4013. So I meant pin2 of the LM311 for the CV measurement.

[Mcentee2]

Many thanks for the detailed reply   No worries re the typos, I do it all the time and have to go back after head scratching

CV Voltage checks:

Range Max, Matrix mode, Pin 2 varies from 7.36 down to 1.05v as Range is turned CCW (reduced.

1.05v is the lowest Pin 2 at min (CCW) Range - the bell tone reduces as range reduces but it never "cuts out" or really disappears, just a gradually emphasis until barely there.


In Flange mode, Rate at slowest, Range being adjusted, the lowest Pin 3 is down to 0.57v (Range fully CW) - I can hear the cutout happening at about 0.8v on the way down and back up (same as your experience!).


The Cap is indeed a 22pF.


(I tried to follow that other thread, but is completely over my head at the moment, so will re-read in a while, re Cap value change!)

(Also -I did come across that thread whilst trying to work out the best way to set the clock and found your great answer re tuning the "bell" to C#/D  !!)

[DrAlx]

Forget about the 47pF cap, it won't solve the problem. 
The problem is the CV is going too low.  It should definitely not be going as low as 0.54V.

This is a design flaw in the Current Lover 2015.   In the schematic I see R30 is 15k. 
Compare that to the R29 in this circuit...

http://thmq.mysteria.cz/em3207/build/EM3207_v1.1_Build_Instructions.pdf

The value if 39k.  That is the correct value (i.e. it matches the original 9V EM) and will solve your problem.

I think the value was lowered to 15k in order to allow a bigger sweep but as you have seen, if you lower the value too much you get drop-outs.

[Mcentee2]

Many thanks

relieved it may just be a limiting resistor value, as that means a change doesn't affect any "core" sound in the circuit - the sweep is what it is regardless of when it cuts out, much better to stop it cutting out

I will swap out R30 (15k) later and see what happens - not sure i have an actual 39k to hand but I certainly have a couple of 20k resistors!

Presumably my-mileage-may-vary on that actual 39k figure and I can find out works out the lowest without popping.

Out of interest, what would the 47pF cap give me in terms of sound - I definitely want to stay in the DEM style  ?

[DrAlx]

The LFO/VCO is similar to the Deluxe EM and the 9V EM.  (Check out the "Electric Mistress Mystery Page" for schematics).

The 9V EM had 39k for that resistor (for the reason you have seen).
The Deluxe EM (which runs of 15V supply) has 22k for that value.

If you are running off 9V supply then go with 39k resistor.  If you are running on 15V then go with 22k.

Regarding the "design flaw". 
If the current lover is using an MN3007 BBD and running off a 15V supply, then the 15k value might actually just about work with no dropouts (depending on how much bigger the 100k range pot actually is).
If the current lover is using an MN3207 (which is limited to 10V) or if the circuit is running off only a 9V supply then 15k is too low a resistor.

If you want to get the maximum possible sweep out of the flanger then you use the lowest possible resistor  that doesn't give a drop-out.  Remember to "re-tune the bell-sound" after any resistor change.  Clock trim might be able to remove drop-outs for resistors bigger than 15k.

Regarding changing the cap to 47pF, this is all to do with linearity at the top end of the sweep. Both the Deluxe EM and the original 9V EM used 47 pF.   These clones use a BBD that needs to be clocked at double the rate of the original circuit, so the thinking was that the way to do that was to roughly halve the value of the clock capacitor from 47pF to 22pF.  That doubles the rate but has the drawback that the sweep isn't "right" at the top end.  The better way to double the clock rate is leave the cap at 47pF and just tweak the trim pot.  You may not notice the difference (given that what is changing is the nature of a small portion at the top of the sweep), so if you are happy with 22pF then stick with it.

[Mcentee2]

Aha!

This is the 2015 Current Lover with MN3007, and not a MN3207 - as it happens i am running it at 12v in "normal use" as it is a much cleaner/sprightlier, smoother flange effect, but it still cuts out at 12v too.

I forgot the CL is in fact based on the 9v v6 EM,and not the DEM, but seems closer to the latter with my ears - no bad thing.

Yes, I have been all over the EM Pages re schematics and info etc and saw the 22k versions on the orig 18v EMs.

I'll check out what size R30 works best, working from 20k upwards.

One day I'll truly understand how flangers work in schematics - basics are fine re "English words", but I haven't yet put that into "LFO / VCO / buffer / clock / delay" talk yet

[Mcentee2]

Brilliant - all fixed.

R30 is now 37k (15k orig + 22k in series)- control flexibility sits nicely in the middle of the clock trim, too low and the cut-out comes in, but with the clock trim set to that "C#-D bell" then there is now no cut out with Range max and Rate min".

12v flanging bliss

Very happy - thank you, Dr Alx, very much appreciated!

[StephenGiles]

Dr AIx - did you ever get anywhere with your adventures in Barberpole Flanging?

[DrAlx]

Dr AIx - did you ever get anywhere with your adventures in Barberpole Flanging?

No Stephen. Not managed to find the time or motivation. It would be a lot of work to build and I wasn't really blown away with the effect.  I did go back and look at what I'd suggested for generating the interleaved  ramps using one-shot multivibrators triggered by a regular triangle/square generator.  There might be a problem with that approach because although you can trim the interleaved ramps to rise at the same rate, I think they only ever rise at that rate.  So you dont have independent control over sweep range. i.e if you make the triangle/square wave faster rate, the ramps go at faster rate because they are triggered more often but they will have the same slope as before and so the overall sweep range is lower.  Haven't thought about it more since then.

What I have been wondering about though (nothing to do with barberpole) is if anyone has tried a more complex LFO modulation for a flanger. I was thinking of taking a regular slow large amplitude triangle wave and superimposing a much smaller faster ripple on it. Sort of like applying vibrato to a slow flanger sweep using a second LFO. Ever come across something like that?

[Mcentee2]

Out if interest dows anyone know the max clock rate in the Current Lover/MN3007 build ?

Intrigued as to how far it is pushed with the 4013 flip flo land the 4049 buffer into the BBD.

Min delay must be short for it to be a flanger, and am assuming (and reading from other threads) the Range and sweep work out similar at audible "guitar frequencies" to the EM ?

[DrAlx]

EM3207 is pretty much same circuit and if I remember correctly I couldn't push the delay time much below 0.7ms - 0.8 ms.
I have the timings in a notebook somewhere.  It's not as short a delay as a DEM.

To be honest though I actually prefer the sound and sweep of the original 18V EM which goes from about 1ms (actually slightly over ) to around 10 to 12ms which is closer to chorus territory.
One way that the 9V EM and DEM differ from the original EM is that as you increase the range pot, the max delay attained in the sweep increases and the min delay attained decreases.  (i.e. to get the shortest delay times for the 9V EM/DEM you have to put range at max).  The original 18EM isn't like that.  The min delay time stays pretty much where it is or even increases slightly.
I changed resistors in my EM3207 to give me that behaviour and I also tuned the clock trim using a lower "bell note". ("A" rather than "C#/D").


If you don't have a scope and want to get an idea of minimum delay time in the sweep you can sometimes do it by ear using "bell-note"s provided the sweep doesn't go too high or too fast to hear properly.
Put color at max, range at max, sweep at lowest.  Repeatedly slap the strings down on the pickups to give you bell-notes which will go up in pitch at the flanger sweeps up.  You need to match the highest note achieved with a note on your (in-tune) guitar.  You then invert the frequency of that note to give you the delay time.

e.g. if you find the bell note goes as high as 1st string 14th fret then that has a frequency of 740 Hz
(see table here  http://www.jameco.com/Jameco/workshop/MyStory/arduino-guitar-table.jpg )
so that gives min delay in the sweep is 1000/740 = 1.35 ms.

[Mcentee2]

Thanks again, I will dig out your EM3207 schematic and compare with the CL re that behaviour, and see if I can spot which resistors you changed to make it more EM-like.

Cheers

[StephenGiles]

Dr AIx - did you ever get anywhere with your adventures in Barberpole Flanging?

No Stephen. Not managed to find the time or motivation. It would be a lot of work to build and I wasn't really blown away with the effect.  I did go back and look at what I'd suggested for generating the interleaved  ramps using one-shot multivibrators triggered by a regular triangle/square generator.  There might be a problem with that approach because although you can trim the interleaved ramps to rise at the same rate, I think they only ever rise at that rate.  So you dont have independent control over sweep range. i.e if you make the triangle/square wave faster rate, the ramps go at faster rate because they are triggered more often but they will have the same slope as before and so the overall sweep range is lower.  Haven't thought about it more since then.

What I have been wondering about though (nothing to do with barberpole) is if anyone has tried a more complex LFO modulation for a flanger. I was thinking of taking a regular slow large amplitude triangle wave and superimposing a much smaller faster ripple on it. Sort of like applying vibrato to a slow flanger sweep using a second LFO. Ever come across something like that?

Only just read your reply! Your idea may be found in 2 places that I have in mind - 1. EH Polyphase and 2 EH Super Space Drum where each has LFO modulation of the main control voltage, not flangers of course but worth a gander at the circuits.