Building the Meat Sphere

[giova.b]

Hello there
I'm planning to build this pedal so I'm collecting all possible info before start.

Can someone confirm, all three filter outs, LP, BP and HP, can benefit an inverting buffer?

[idy]

This is reposting to avoid hijacking the new thread by "zoot" (Meatsphere almost there) with my own ramblings.

In that thread AnotherJim explained why I have had better luck with an LM358 than the 1458: the 358 works better closer to 0v and this is converting audio to DC to drive the LEDs. With the 1458 there is a 2v "dead spot" before the opamp starts rectifying. And the LED won't light below some value, my meter shows a fv of 1.7 volts. I breadboarded the detector and tried adjusting the bias, and sure enough, at around 2.4 volts (the LED has two resistors plus a pot, 430ohms plus the "intensity" pot) the circuit becomes more sensitive. It lights up from gentle play with a passive guitar. It can be biased to stay just slightly on, and the intensity pot used to pad it wherever you want it. So the starting point of the sweep doesn't have to always be "total darkness."

My previous builds I find that passive guitar barely drives it, I need Sensitivity all the way up, Intensity all the way up, attack in the first eighth of travel. I usually use guitars, (acoustic and electric) with preamps so I get by.

So I thought to add a simple bias trimmer inside and am wondering about the best way. The original circuit has a 22k to ground on the non-inverting pin. I am thinking 500k trimmer between 9v and ground, maybe a 47uf cap across this. 22k (1m?) from wiper to non-inverting input. Am I overdoing it? Could I damage the opamp monkeying around with the trimmer, setting it all the way up?
Should I nix the trimmer and just use two resistors (and a cap?) to pick a bias point that lets me use intensity to "dial it down"?

I am also thinking about a neat implementation on my (4) existing meatsphere builds. Would be nice not to take the things apart, all those pots and switches, maybe just clip the 22k in half and use its ends to tack a little "sub-assembly" in place.

[anotherjim]

The 22k to the pin3 of the "1458/358" can be kept, but connect the other end to a pot instead of ground. The pot value could be almost anything, but any lower than 10k is only wasting current from the 9v supply. It doesn't need a smoothing cap there, the 10uF envelope cap further on will clean any noise out. You won't destroy the chip if the pot voltage gets set too high and it's unlikely you will leave it like that.
If you raise the pin3 voltage, it can remove most of the advantage of choosing the 358, but it can still be worthwhile as it can produce about 0.5v more positive swing than a 1458. Even more could be got by using a CMOS amp like a TLC2262 as they can work closer to both the 9v and 0v rails.
A little more swing from the rectifier can be had if those 1N4148's were replaced by Schottky diodes (or Germanium!) such as BAT4x series. The opamp pin3 has to go higher by the diodes forward voltage before it can raise the voltage on the envelope cap.

The attack/decay controls are a compromise for simplicity. You cannot expect them to work as well as a synth envelope. If Attack pot is max and Decay to min, they are making a voltage divider set low and the envelope voltage will struggle to swing up by much. Similarly, at the opposite extremes, the envelope voltage sits close to the bias and won't shut off the LEDs fast enough. Choosing the pot and fixed resistor values is actually a bit of a nightmare and the original designer probably lost some hair over it.

The final piece is the resistance in series with the Vactrol LEDs. If you don't have the same as the EHX original, those 100R series resistors could be far too small. With Intensity at minimum and if the opamp gave full swing, a modern LED will destroy Alderaan.

[idy]

Thanks, this is clarifying the mystery meat.

I was also thinking of lower fv diodes, schotky or Ge.

I have been fooling around with a bias pot (one end to 9v, one end ground, wiper to the 22k) on one of my 358 builds. I think it only needs a little bit to get it close to the trigger point, but not so much that it is always on... It does make it respond differently, a guitar with single coils and no boost will drive it to a satisfying range of sounds. But it seems like it has a harder time "closing" and any background ringing keeps the filter open.

The original Meatball, I believe, did not use vactrols but green LEDs stuck to LDRs. I tried the suggested vactrols in the beginning.... and only had success when I "rolled my own."

[anotherjim]

We know the guitar is brighter/louder during pick attack - so if a smaller coupling cap was used into the rectifier, it will only pump up when picked and will ignore the sustain since that contains less treble and the envelope cap can have longer to discharge before the next pick.

[idy]

It has a choice of caps into the rectifier, I never really played with that.

The Meatball has some "truth in advertising" in that it is labeled "triggered filter'; and not "envelope filter."

About the LED brightness: using the diffused green LEDs on the LDRs with the given R values gives a good range. They are a sort of article of faith among Meatball builders, and Lovetone stuff generally. The visible indicator I put in an ultra-bright and the suggested 4.7k. Normally with 9v I would use closer to 10k there.

[idy]

So i now have two of these with little voltage dividers to bias the rectifier. One with a LM358 and one with LM1458. (A "legitimate brand name" 1458, the generic Tayda ones didn't work for me.)

In either case bringing the pin 3 up to around 2v makes triggering with an unboosted guitar a breeze, and the LEDs stay dark when the guitar is silent. There does not seem any virtue in tweeking this to some special bias value or using a pot. I settled on pulling the end of the 22k farthest from the chip out and attaching a 200k to the empty hole and a 680k to the 9v in nearby. The three resistors' loose ends wound together.

I have not tried lower fv diodes. I am thinking that .6v is still too little to "crack" the LEDs, and that charging the cap starting at .3v won't light anything up. The opamp needs to output over 2v for the LEDs to respond.

Having just returned to reread all 30 (often heartbreaking) pages of this thread I keep thinking it could be reduced to one page:
Roll your own vactrols, LDRs dark over 2M, diffused green LEDs.
Brand name 1458 or, better, 358. Others may work.
Biasing pin 3 to above 2v makes it easier to trigger.
Moog mod good for bass. The 10k between pin 2 and 8 of the 074 (directly under the range rotary) is replaced with a switch. DPDT. Two wires from the 10k's empty place to pins 3 and 5 of the switch. a 10k between pins 3 and 6. a 100k and 470n in parallel between pins 4 and 6. You are adding a cap and parallel r in series with the 10k.
Color mod really a necessary correction. Jumper pins 2 and 3 of the color pot. And you can use A100K for the color pot, better range of resonance.
You can add a buffer and then switch "dry" side of blend knob between send and return. Necessary if you like fuzz. If this buffer had a phase invert switch you could blend LP better.
The schematic has an error in VREF, (mislabeling 9v as vref and missing the arrow at the node for 1/2v) but the board is correct.
Several components are reversed in order on PCB/schem, but they are functional equivalents.
Schem is missing a 10uf cap before send jack but board is correct.
You can up the value of the 10uf cap attached to "attack" to 47uf and you can get the longest decay sounds you hear in some demos of the original.
Switched jack on the return. Notice that end and return are in kind of odd places, send is close to output jack, return close to input. Most of use just use longer wires and cross them over so that looking at the pedal for the normal position you would count L to R output, return, send, input.
Yes you can use the board if you have soldered your rotary switches and pots to the wrong side of the board. But please think before you solder.
No there isn't a guaranteed drilling template (correct me?)
When testing you will need: sens max, attack min, decay above min, resonance mid, intensity max, blend mid to max, mode up, bandwidth middle position(which is full) range low (doesn't matter) filter doesn't matter. If you haven't biased the rectifier you will need a booster on a passive guitar.
The original foot control jacks are not very useful but the manual explains them clearly: Both are passive volume pedals (boss FV50 or 60), Pedal 1 is in series with the decay pot. Pedal 2 is in parallel with the intensity. Neither one of these turn it into a wah pedal. You will need to figure that out for yourself. (maybe a 10k pedal in series with intensity, and then set filter to "down" so the LED is normally on, and turn the trigger bandwidth to off...?)

There are a few more, but much of the thread is general troubleshooting....

[taku0319]

Hi,

Yesterday I made meatsphere and it's really nice!
Well designed PCB to fit with the enclosure.
Also I used Tayda's pcb mount pots, so it was even easier to make.

However, it was necessary to set parameters properly to get a nice filter sound.

[MatthewD]

Hi,

I am still relatively new to pedal building (done a couple of easy ones), I have a couple of questions...

My kit was supplied and the BOM indicated that there are two 4.7K resistors, however I can only find one on the pcb (at the top right), am I missing something?

I am used to seeing a graphic with a u shaped cutout indicating which way the op amp socket should be mounted, but on this board there is a square copper pad on one pin location... am I right in thinking that this pad is pin 1 and pin 1 is the same end as the U shaped notch?

Regarding the Optocouplers, my kit from www.musikding.de came with two NSL7053, however I have also bought some 9203 ldr's from small bear and I am considering doing the rolling my own thing. Has anyone had experience with the NSL7053? I am trying to decide if it is worth trying them first. I have just matched my ldr's and I was lucky to get several good pairs from the 20 ldr's that I bought.

[Taylor]

Hi,

I am still relatively new to pedal building (done a couple of easy ones), I have a couple of questions...

My kit was supplied and the BOM indicated that there are two 4.7K resistors, however I can only find one on the pcb (at the top right), am I missing something?

I'm not involved in putting the kits together, but maybe the extra resistor is the LED current limiter (marked RLED next to the up/down rotary)

I am used to seeing a graphic with a u shaped cutout indicating which way the op amp socket should be mounted, but on this board there is a square copper pad on one pin location... am I right in thinking that this pad is pin 1 and pin 1 is the same end as the U shaped notch?

Yes, that's right.

I haven't experimented much with the optos so can't comment there, but maybe someone else can.

[MatthewD]

Thanks for the reply Taylor,

I can now confirm that the spare 4.7k resistor is for the led... I have since learned that this is quite common.

Foolishly I soldered all the resistors before reading all of the pages of this thread, and I have decided I would like to give the Moog mod a try because I am going to be using this on Bass.

So what I am planning on doing is...
Desoldering one leg of the resistor.
Solder wire to free end of resistor
Solder other end to common leg of an on/on switch (not a footswitch, just a SPDT switch)
Solder 470nF capacitor across the other two legs.
Solder wire to one of the two non common legs on switch
Solder other end of wire to circuit board.

I have ordered a "Polyester Film Box Type Capacitor WIMA MKS2" for the project.

Any thoughts or ideas would be appreciated, I am assuming it doesn't matter if the resistor is before or after the capacitor?
Thanks,
Matt.

[idy]

That sounds OK but:
The guy that came up with the "moog mod" (I think) ended up with a 100k in parallel with the cap.

By removing the resistor from the board you don't have a "flying" component that is likely to break when you monkey with it. It is a little neater to put both resistors and the cap on a switch. DPDT.
10k from 1-4
cap and 100k parallel 4-6
wires to 1 and 5

No, the order doesn't matter. Cap first, R first no matter here.

[rectifierfx]

So I build myself the meat sphere. The effect works. Tried Xvive VTL5C3 and DIY Tayda diffused LED's with Futurelecs 2A LDR's (20k - 2000k). Decided to leave use the Green diffused LED's and Futurelec LDR's.

I still struggle with the OP-Amp though. Tried everything from LM1458 to LM358, nothing sounds quite right. Sound wise for me the NE5532 performed best but my envelope curve aint smooth.

Check out the soundsample below

https://clyp.it/o2a5aomw

First example is with Decay on 3, second with decay on 8

Is it the Op-Amp I have struggle with or is something else wrong in my circuit?

ALso I'm a little bit dissapointed the effect is only usable in one single position: Sens on low, Attack low, decay high, colous max, Intens max and blend max

[idy]

The decay does sound ripply and strange. IME all settings work as long as attack is lower than decay. I used a larger cap and got longer decay time, I heard some demos of originals that had super long slow decay.

What happens if you lower attack and bring sensitivity up? In my builds I always had to crank the sens. all the way and intens. pretty far, and then use a booster to get the thing to trigger right. I found by using LM1458 (real brand name and not generic "work alike") and then adding something close to 2v bias to the pin3 non-inverting input the thing would trigger normally on any instrument.

I think I describe that earlier on this page.

[idy]

duplicate post

[MatthewD]

Thanks for the sound clip Rectifierfx and the advice Idy.

I have finished my home brew vactrols, I also found some "cool audio" vactrols on SmallBear so I have ordered those, hopefully in a couple of months I will post some sound samples of my own to compare. Cool Audio is a Behringer owned company by the looks of it, and because the only thing Behringer does really well is make copies (aka ripoffs).. they should work great!

Now... tonight I think I ALMOST made a stupid mistake, I read earlier on the thread that the rotary switches can? be soldered in any way around. But I was about to solder one in when I looked at the board. Correct me if I am wrong but on each switch location on the pcb only a few of the traces are "live". So if I have pre set the switches with the locator tab to the correct number of positions... then I should use a continuity tester and make sure the live terminals match with the live traces/pads on the pcb. I think I may have read "can" a little too literally

This looks fairly straightforward except the range which seems to have 3 live pads, then one not live, then 3 more live pads. How many positions am I meant to set on the switch for this one? Looking at the schematic it looks like 3 positions(3N3, 6N8 and 15N) + a fourth position as an off position. So I just ignore the other pads.

It looked like Taylor posted an image way back on page one explaining the switch layout but it doesn't seem to be displaying for me... sorry guys.

Thanks for any advice.
Matt.

[Taylor]

It's maybe a little hard to explain why, but yeah the rotation of the rotary switches doesn't matter as long as you're using the plastic Alpha type rotaries. Each switch has three poles with four throws to each pole. So for the hp/bp/lp switch for example, whichever pole goes into the set of pads that are wired will make the connections.

up/down has 2 positions
range 4
hp/bp/lp 3
bandwidth 3

[MatthewD]

My Meat Sphere is all together now, but I have been problem solving an issue for a while...

When the pedal is turned on there is massive amounts of noise coming from the pedal (not white noise). I have done some audio tracing and found that the input to the TL074 sounds fine, but at any point after that including the TL074 output or the clipping diodes or at the blend there is noise.

I have tried replacing the TL074, and have checked that the components are orientated correctly but the noise persists.

Something surprising has happened with the power, when I first turned it on the led lit up, but at some point during the testing it has stopped working. Voltage tests show power is getting to the TL074 (approx. 8.7volts) and the Optocouplers, but voltage is not getting to the RLED or the led, unfortunately as a beginner and due to the complexity of the board I am having trouble figuring out exactly where the power break is.

If anyone has any suggestions I would appreciate them.

[Zoot]

Hi MatthewD,
My experience with noise and the Meat Sphere showed that so many connections (in/out/send/return) are often playing a role.
You probably want to check first if all grounds are grounded.
Have you added the send / return parts? If yes, have you used convenient switching jacks? If not, think of doing that. If you read Spanish (or can Google translate), these guys from Pisotones have a good explanation on the switch part with pictures and so on. Take a look: http://www.pisotones.com/Meatball/psst/meatball-psst.html
Led going on and off in the Meat Sphere doesn't mean necessarily it's not working, it's part of how the pedal works. The bitch on this pedal is to get the triggering right, and the LED going on and off, following the trigger, is part of it. Fixing the noise first might help on this as well.
Otherwise, place a good picture of your circuit on your next post, the people here are really good in finding the mistakes we don't see while building.
Good luck with this pedal, which is totally worth the effort.
Cheers,
Zoot

[Jbakelaar]

Hello, I have the board in hand and the majority if not all parts on the way.  I'm excited to get into this project!
Does anyone have a drill template (the one shared in box on the first page(s) of this thread will not load for me). Also anyone feel like sharing graphics they might have used on their build?
I've read the thread and have gleaned a lot of helpful info. 
Building for a bass - any revelations from recent builds, I'm all ears.