Musitronics Micro V sound question

[Phend]

Re: See my Musitronics Micro V in Pictures Pictures.
Hello:
It works like I hear on the utube but :
When I pick a string it produces a, what I consider a "loudish" guess "attack" then reduces in volume. Would call it sharp a sound, maybe a "spike". Not as smooth as the demos on utube.
Recommended IC pin voltages are very good to spot on.
Any ideas ?
Thanks

Circuit:

[antonis]

Does it occur despite Sensitivity pot setting..??

[Rob Strand]

It sounds like there's something not quite right with the filter.  It's like the filter frequency is pushed way up, becomes peaky, and is therefore cutting the bass and mids.

You might check the value of R17 but beyond that check over all the part values and wiring around IC2 and IC1a.

[Jarno]

I have built this ages ago, and it is (at least with bass) very sensitive, so it could really benefit from an input attenuator.
The knob doesn't do that much, I usually have the switch on "low range".

Seems it is a bit of a limited circuit.

[Phend]

Did some checking on this effect today,
Went thru the entire circuit and everything is correct.
All 3 transistors voltages are close,
Q1 Emitter 3.8 instead of 4.1
Q2 Base 5.0 instead of 5.6
All others good.
R17 is 4k7 (measured)
9V is 8.42 after 1N914 and Vref =4.46

Changed R7 to 100k in lieu of 22k, no real change (thought volume?).
Changed R13 pot to 500k using 500k across 1 and 3, no real change

Sound is best with switch set to Low.
And R13 turned down to 22k ohm, (Gets worse soon after turning it up some)
In my case the Sensitivity pot (R13) works well, guess to well creating that quick "bark" sharp note.

Hope I touched on each of your comments, thanks

Any advice from here would help, like which capacitor might change the "sound" ?
I'll try some more hacking too.

[Rob Strand]

I found the schematic ambiguous.

Have you got the two halves of the LM13700 wired in parallel?   

That might do something weird since there's nothing to make the two halves share evenly.

I'd probably lift R14, drive some voltages in there,
- check voltage across R17 = current into IABC pins
- measure the frequency response, or at least the peak frequency and amount of boost,
  with for a few IABC settings to confirm the filter is actual working as expected.

Any advice from here would help, like which capacitor might change the "sound" ?
I'll try some more hacking too.

I did something bit I can't post the corresponding schematic because the image server is barfing.

Here's the notes out of the simulation with the designators translated to those on the ggg schematic:

In my notes (and below this point) LO means switch closed; LO = low frequency filter.

(Musictronics) Mu-tron Micro V
V1.1, Rob S 22 Jan 2022
ggg schematic cap designators

Filter is largely a band-pass.
- Band-pass Q is higher in LO mode.
   At HF sweep, the Q increases and
   and the peak gain increases by about 18dB   
- Band-pass Q is quite low in HI mode.
  At HF sweep, peak gain increases by about 15dB.

LO Mode:
increase C2:  increases gain but boosts HF side more
                      and boosts HF peak more again
increase C3:  increases gain at LF sweep
increase C4:  reduces HF peak (HF sweep)
increase C6:  scales down filter frequency,
                      but also cause drop in LF peak (LF sweep)
increase C10: shifts frequency down
 
HI Mode:
Increasing C2 boost LF and HF gain,
not as aggressive on boosting HF peak as LO mode.
Some drop in filter freq
C3 as above but also shifts down frequency a bit.
C4, C6 as above

C5 has no effect.
I'm not sure why it is there, perhaps to stop oscillations.

[Rob Strand]

Sound is best with switch set to Low.

Another confusing thing about that schematic is the switch is closed on HI mode.
That's not right; same problem on the factory schematic.

In LO mode the filter is at low frequency ranges.   The filter is at low frequencies when the switch is closed
https://www.youtube.com/watch?v=XksgKgwPa_c

In my notes (and below this point) LO means switch closed.

In HI mode and at the top of the HF sweep, the amount of boost is fairly sensitive to the Q1 transistor gain.
In order to knock down the HF peak in HI mode you need to add a resistor along the track between
the emitter of Q1 and where C2 and C3 join.   Adding even 47 ohms will do something.

The impedance looking into the filter  (C2 and C3) is quite low.  I'd be thinking it would
be a good idea to drop the value of R3 to 4k7 perhaps even 2k2.    The maybe add
a deliberate resistance between Q1.e and C2+C3 to set the peak.

[PRR]

I found the schematic ambiguous.
Have you got the two halves of the LM13700 wired in parallel?   
That might do something weird since there's nothing to make the two halves share evenly.

Pin-numbers strongly suggest two pins per connection; full parallel.

"Sharing" may be nearly automatic. Gm is proportional to Iabc, even if Iabc is flowing, even unequally, in two pins and two OTAs. 100uA+100uA is same-as 200uA+0uA. Output current likewise.

The "sharing" problem arises when we mistake the Iabc pin for a voltage control and parallel two to control two signals.

[Rob Strand]

Pin-numbers strongly suggest two pins per connection; full parallel.

Datasheets ocassionally use the same labeling style to mean either part of the chip (but not both in parallel).
Anyway I check the PCB in some of their docs and the two halves are in parallel.

"Sharing" may be nearly automatic. Gm is proportional to Iabc, even if Iabc is flowing, even unequally, in two pins and two OTAs. 100uA+100uA is same-as 200uA+0uA. Output current likewise.

The "sharing" problem arises when we mistake the Iabc pin for a voltage control and parallel two to control two signals.

Yes, when I looked up all the connections it should at least work.   

There is a voltage-based sharing problem at the IABC pins but if the two halfs don't share it should still work with the same gm.  What you lose is the two halves aren't operating in parallel and you don't get the noise improvements of putting the two OTAs in parallel.

I checked a few other things as well but I didn't post them:

- The original schematic has a single 1M resistor for R16.   For two OTA you still only need a single 1M.    I don't think two 1M's in parallel is doing much damage though.  I'd consider it as something to play with or tweak.  It's definitely not causing the problems Phend is seeing as the 1M's don't affect the high frequency part of the sweep.

- The original circuit used a CA3080 which only has one VBE drop on IABC whereas the LM13700 has two VBE drops.
   If the rectifier opamp swings down to around 1V the should still be enough voltage on the collector of  Q3
  that it operates as a current source with it's collector sitting at 2VBE's above ground.
  It's not as much "headroom" as the original but it shouldn't barf.
  Checking the IABC voltage and current source on the CRO might be useful just the same.

I have a feeling the bad sound is something to do with the strong peaks when the sweep is in the upper frequency range.
So at this point I'm sticking with the idea of tweaking a low-valued resistor between Q1 emitter and C2, C3.

Q2 Base 5.0 instead of 5.6

Need to check if that low voltage combined with the 2VBE drop on IABC is enough to keep Q3 working.   Off hand I think so but need to check - it checks out OK, a Vref reduction shifts frequencies down

[Jarno]

I remember building mine with a ca3080, and vividly remember that I forgot to mirror the layout before doing the toner transfer, so I bent all the legs of the OTA the other way and mounted it upside down

So my comments about sensitivity are about the version with the 3080.

[Phend]

Well using a combination of Rob's suggestions (Thanks Rob) I have a much better sounding effect, no more bark at the "top" of the sweep.

These modifications certainly changed the sound for the better:
1) Added a 47 ohm resistor between Q1e and C2.
2) Replaced the 10K R3 with a 2k2.
3) Made C4 0.0068 uF in lieu of 0.0022.

Notes:
1) I agree with Rob about the Hi - Lo switch, from the sound it is backwards on the schematic. Might fix the tight solder job to switch it.
    (Or change my label with duck tape?)
2) Thanks Rob for the utube link, it is different than what I have seen.
3) The amp you play thru, of course, makes a difference in any barking, (which is gone).
4) R16 is 500K with a 100K trim to adjust up down 50K is in place. Have
not tried to adjust.
5) Yes the two halves of the LM13700 are wired in //. ie 5,12 , 3,14 etc are connected with jumpers. This is as the schematic shows.

Thanks again to all for the comments, they really helped.

Before any of these mods here are some Dollar General scope pics,
Measuring at Pin 1 of 4558:
Labeling HI LOW is the switch position.
22K indicates the pot is at 0 and 30K is the pot turned up to about 1 (after that no change in wave shape)
Also Measured at the 22K resistor POT HI POT LO

[Rob Strand]

Cool.  The cap change is thing that provides the main change here.
The R3 + 47R is probably helping signal quality more than anything.

The C4 mod drops the HF peak at the top of the sweep,
which tames the *variation in the filter peak over the sweep.

dB variation in peak (approx)
C4    2n2       6n8 (mod)
LO:  18dB     9.6dB
HI:   15dB     5.9dB

For LO mode you need some increase in the peak to sound
even since the filter Q gets higher at the top of the sweep.

[Phend]

Hello Micro V builders,
After reading the topic from 2012, (Topic: Mutron Micro V (+?) build guide) and seeing another utub video, I am doing some surgery.
Noticeable change was lowering the Hfe of the transistors.
5089 was Hfe 824 switched to 3409 Hfe 282
5087 was Hfe 425 selected another at Hfe 387
Made quite a change.
Here is the video, can't say anything about it other than my Micro V did not sound like this.
But is getting closer, next experiment with some pots in R14 and R15

[Phend]

In summary these are the changes I have made using suggestions from all, thanks all:

1) Added a 47 ohm resistor between Q1e and C2.
2) Replaced R3 10K with a 2k2.
3) Made C4 0.0068 uF in lieu of 0.0022.
4) 2N3904 Hfe 282 (2x)
5) 2N5087 Hfe 387
6) R2 356K
7) D1 and D2 = Bat41
8.) C6 0.001uf or 0.00068uf in lieu of .0018uf (Nice change) 
9) R14 = B500R added pot + 100R (Attack)
10) R15 = B50K added pot + 6K8 (Decay)

Hope this might help others