News:

Forum may be experiencing issues.

Main Menu

Glasshole bypass LFO tick

Started by jpier2012, February 26, 2021, 09:11:38 PM

Previous topic - Next topic

jimilee

Noise gate, that'll fix it. A decimator.


Sent from my iPhone using Tapatalk
Pedal building is like the opposite of sex.  All the fun stuff happens before you get in the box.

Zerro

#16
It is clear, that this tick is VF noise - so,

- at first try heavy duty batteries instead power adaptor, to be sure, that it is not problem of that LT1054 - do substitional power. Put off that LT and connect instead it baterries to +Vc and -Vc points. This will give you picture about power strenght of LT... If weak source from TL, power batteries will show it.

- second, paralelly with R40 (68k) try to connect some capacitor, cca 10-47nF. That caps around IC4, I recommended before, let here anyway. It won't hurt :@)

- third, those caps I recommended to connect to LFO IC4, connect the same way to IC1 now. You can let them here anyway too. Use 47nF or something around, you have now, but ceramic types. As for IC1, as IC4.

Some of this must help. Try all, including that IC4, or only some ot this.

p. s. The best against VF noise from triggers is to use at output wires in machine, ferritic pearls. Several turns around it will shut down most of VF. You can try it too. I will attach picture.
"Nudíte se? Kupte si našeho cvičeného ježka! Pobaví vás svými veselýmí kousky!"

geetarm@n1

Gotcha - I'll give it all a try at my next opportunity and will report back in a few days.

Many thanks and enjoy your weekend

geetarm@n1




Update:


Zerro,

I've tried the following but unfortunately there has been no reduction in the ticking (which again is present only when in non-bypass mode and only present on the Six and Eight Stage settings).   Granted, I'm using just my ears for this test, not an oscilloscope, but the fact remains that the ticking remains loud enough to be audible after about 4 seconds into the decay of the ringing out of a struck chord played, with no other FX in line and into a non-overdriven amplifier.  There is no other unwanted noise - a nice, quiet effect other than the ticking.  I still have dead silence (in terms of LFO tick) in the Two and Four Stage settings.



a)  Replaced the LT1054 with two 9v batteries.    This did not reduce/eliminate the ticking, so I reverted to using the LT1054/OneSpot power supply.



b)  Soldered two small-sized 47nf poly caps from pins 4 and 11 to ground on IC4.  This did not reduce/eliminate the ticking, so I left this in place and then:



c)  Soldered two small-sized 47nf poly caps from pins 4 and 11 to ground on IC1.  This did not reduce/eliminate the ticking, so I left this in place and then:



d)  Soldered a small-sized 47nf poly cap in parallel with R40.  This did not reduce/eliminate the ticking, so I left this in place and then:



e)  Harvested a ferrite bead from an old computer keyboard and replaced my shielded output wire (running from switch to output jack) with a longer, non-shielded wire and looped it three times through the bead (the max that the bead's diameter will allow).    This did not reduce/eliminate the ticking.



f)   I re-flowed the solder on all of the LDRs, this did not reduce/eliminate the ticking, so I then:



g)  I went back and double-checked my voltages.   Everything still looks good and matches with a few mv to what Brian's build doc specifies with the exception of IC3's pin 7, which he has listed as being -10mv, but in my case, I read +12mv.  Is this significant?



Note that I've not tried using ceramic capacitors for this test as you suggested.  I can certainly order some if you feel that will be the magic bullet.

Given that none of this has had any audible impact and that (as I understand it) your suggestions are aimed at reducing an induced or indirectly picked up (same thing?) noise, then perhaps the problem is rather something directly injecting it or not filtering out as it is supposed to do?   Is there any particular set of components whose job it is to do so?    Perhaps I have a bad component or component with incorrect value?  Two to four bad LDRs?   A bad solder connection?  Perhaps an incorrectly oriented polarized component?   Given that the thing works and the voltages are all in pretty much in order tell me that the latter is not the case, but at this point I'm open to trying anything and also open to assuming that I could have done with component choices or build/soldering technique could be the culprit.   Any further theories are welcome and will be much closer to the mark than what I could come up with, so bring it on!


Thanks again


Zerro

Hi geetarm@n1, I am sorry, that those enhancements didn't work - my suggestion - raise values of those blocking caps (100n, 470n) but surely try those ceramic caps. It will be about to "fit up" proper timing. It is good to let them there - it can't harm anyway :@)  At g) item, 12mV is ok.

Hi freq noise is very tricky, and many etchers trying rid it off, but it was always about penetrating VF ticks to power. Professional machines doesn't do it, so I must admit, that maybe bad PCB concept is here. Generally, it is good to separate tracks of power lines to those problematic parts of machine, from power tracks for audio parts, and separate those LFO tracks (even at grounding side), by some little value resistors 100-47O ohms or little inductions - few turns around ferrit bar, so that sections with VF noise are slightly "floating". Then those blocking caps will work far better. See my pictures. Hapilly, there is only one source of this all - IC4. Or more preciselly, IC4a. If it is possible, maybe prepare to lift up with resistor ground of IC4, too. But then blocking caps at pins of it must be stronger - uF values. Because of stability. But it is last possibility. I hope it will help, otherwise there is none another solution I can offer now.
"Nudíte se? Kupte si našeho cvičeného ježka! Pobaví vás svými veselýmí kousky!"

Zerro

#20
At real, many years ago, I built mono-phonic guitar synth based on 4046 PLL, and there was non-coherent ring modulation too, by 4011 osc, and even when VCA was closed, under that silent was that VF mod freq clearly at output. I was trying everything to separate all circuits related with it and finally helped this "floating" conception with ceramic caps blocking together - I copied it from my old walkman, where circuit for motor riding was separated at both sides of power line. So, it's noise wasn't almost annoying sound section. I used inductions at ferrite cores instead resistors, and I sucessfully shot down this VF penetration.
"Nudíte se? Kupte si našeho cvičeného ježka! Pobaví vás svými veselýmí kousky!"

geetarm@n1

Zerro - you rock.   Many thanks for the detailed information and know that I appreciate the lessons you're providing.

As for the caps, are you suggessing small MLCC or actual ceramic discs?




Zerro

Not at all...

Use conds that you have - only respect voltage  - minimally cca 16-25V
"Nudíte se? Kupte si našeho cvičeného ježka! Pobaví vás svými veselýmí kousky!"

geetarm@n1

Got it.   I'll reply back in a week or so once I've had time to order what I need.     Thanks again.

Zerro

#24
Hallo, today I was digging in my bearhole and I finally set out this old schematic of phaser with shift 0-720°! Notice down part with LFO modulator. Here I assigned with red circles 2 points coming from power supply. In accompanying text is literally said, that those R-C sets are here to avoid getting trigger clicks to the output signal.

Interesting is double pot riding not only modulation Speed, but Depth too. For low freq, modulation is deeper, and for high is weaker - more like tremolo. Ratio you can additionally set with R37. MOS-FETs was here KF521, which was old czech details. I recommended here 2N4447, but you can try 4448 too. Or another. Chiao.
"Nudíte se? Kupte si našeho cvičeného ježka! Pobaví vás svými veselýmí kousky!"

geetarm@n1

I finally had some time to sit down with this again and tried what you (Zerro) suggested in your last post.   

I lifted the cathode side of D9, attached that to one end of a 120ohm resistor, then attached that where D9's cathode's pad on the board.  That connects to the + of C23 and to pin 4 of IC4. 

To take care of decoupling and creating a reservoir for IC4's pin 11, I added a second 14 pin socket with the pin 11 section removed on top of my original socket.  I soldered a 120 ohm resistor and the (-) leg of a 10uf cap to IC4 pin 11, then soldered the (+) leg of the cap to ground.  Then soldered a 1n914's anode to the other end of the resistor and inserted its cathode into the original socket for IC4 pin 11.  Looked  pretty questionable, but the voltage readings told me that all was well in terms of connectivity.

The result, unfortunately, was the same - tick tock still present at the same volume level it was prior.   

I took a last shot at re-calibrating it and doing the chopstick trick (moving around the now shielded Input/Out leads which run from jack to switch) just for grins, but no go.   I think I'll keep this thing as a great sounding 2 and 4 stage phaser and maybe try building another and see how that goes.

Many thanks for the suggestions and although I wasn't able to resolve it, I definitely have learned a thing or three in the process.  One eye opener was how all of these additional components I've added throughout this process have had so little effect on the core sound - it still sounds just as rich/full as it did originally.

Zerro

Still look at my post 258, I added before to this thread. Try to do it literally this way. Only resistors and caps to both power branches. Nothing more I suggest :@)
"Nudíte se? Kupte si našeho cvičeného ježka! Pobaví vás svými veselýmí kousky!"

danfrank

#27
It's the sudden turn on of the op amp forming the square wave in the LFO. I think it's IC4-1 on the Glass Hole schematic. Take a look at the following link:

https://www.scribd.com/document/515633043/Stompboxology-Mo-tremlo

Look at figure 3, drawing 2. The addition of the 33nf cap and 68k resistor will solve the ticking problem of the Glass Hole. I have modded the Glass Hole the way the article describes and mine has no more ticking. Try it, you'll like it!

geetarm@n1

Zerro - I stripped the new diode out of the equation (the one I tacked on to serve pin 11) leaving just the added 120 ohm resistor and 10uf cap.  The other diode D9 is board mounted (and tied to the 120 ohm resistor I added in line) I've not tried removing it, but I did jumper it.    Neither of these changes affected the ticking.

danfrank - Thanks for the suggestion.  I gave this a try as well on IC4-1 - which I'm taking as being IC4-A (tying a 33n between pins 1 and 2, then a 68K resistor between pins 2 and 5), then unfortunately, this didn't change the behavior.

Given that on this project, IC4-A's pins 2 and 5 are tied to ground, yet the examples shown of this 'soft square' trick from the .pdf you linked don't have these pins grounded, I'm wondering if to implement it here means that I need to lift those pins from ground a bit or something.  I tried disconnecting them completely from ground (again, with a stunted row of sockets which allowed me to not make that connection to those pins of the original board-mounted socket) and that "fixed" the tick - by stopping the oscillation altogether  :)

Also, here is another link to that PDF which doesn't require an account with scribd.com (it's a good article - thanks for that)  -  http://www.moosapotamus.net/files/stompboxology-mo-tremlo.pdf

So assuming that you are correct that IC4-A

geetarm@n1

whoops - I posted before completing my final sentence:

So assuming that you are correct that IC4-A in the Glasshole is actual the target for this fix, it unfortunately didn't work for me.