Hiss on Amz Mosfet Booster

[Atodovax]

Hello everyone. I buit two Mosfet Boosters into one enclosure in parallel si that both could be used indeoendently one at a time and im having a lot of hiss when they are on . The hiss is only when using a psu but the psu is a quiet one. With batteries its dead quiet. Theres something on the circuit producing the hiss. Any ideas?
I tried grounding the input and outpu of one if the effects when the other one is engaged but still the same

[GibsonGM]

How you mean "In parallel"?  Are you using 'recommended' switching for 2 in 1 enclosure, as the GGG pdf suggests?  You don't actually switch one in, IN PARALLEL, do you?   Got a link to the wiring method you used? 


Ok, the main causes for hiss (and SOME will be normal and always there)....unshielded signal wiring...ground not run well (should be sort of a star ground if you can).  Big long swoops of signal wire that should be shorter.   A not-properly grounded enclosure.   Not decoupling the power supply properly (each PCB in this case should have its own "Vref" and decoupling cap.  Buggy solder connections, maybe at the 3PDT.

Now - the fact it'll go fine with battery really says it's the power supply, bub.  That, or there is some odd wiring error with your DC jack or something.  Decouple, decouple...tho these boosters use VERY little current... If you amplify anything, the noise floor 'comes up'...it's just not as quiet as it was before.  ANY noise is amp'ed along with your signal.    So, improve grounding...use shielded wire for the input, run it away from other wires if you can.   Be sure other pedals ARE quiet with that PSU on that wall outlet first.

If nothing 'basic' does anything, can you record the noise and post it?  At least that would let ppl tell you if it's 'to be expected' or a defect.   

[Atodovax]

Thanks for replying so fast. I am using this layout. I want a switch to show that the pedal is on and then another one two swith between one or the other one. I am not using shielded wires.. i will try that. Psu only has this hiss in this pedal. Do you think it may be wires too long? Or both mosfets always on interacting with each other?

[Atodovax]

How you mean "In parallel"?  Are you using 'recommended' switching for 2 in 1 enclosure, as the GGG pdf suggests?  You don't actually switch one in, IN PARALLEL, do you?   Got a link to the wiring method you used? 


Ok, the main causes for hiss (and SOME will be normal and always there)....unshielded signal wiring...ground not run well (should be sort of a star ground if you can).  Big long swoops of signal wire that should be shorter.   A not-properly grounded enclosure.   Not decoupling the power supply properly (each PCB in this case should have its own "Vref" and decoupling cap.  Buggy solder connections, maybe at the 3PDT.

Now - the fact it'll go fine with battery really says it's the power supply, bub.  That, or there is some odd wiring error with your DC jack or something.  Decouple, decouple...tho these boosters use VERY little current... If you amplify anything, the noise floor 'comes up'...it's just not as quiet as it was before.  ANY noise is amp'ed along with your signal.    So, improve grounding...use shielded wire for the input, run it away from other wires if you can.   Be sure other pedals ARE quiet with that PSU on that wall outlet first.

If nothing 'basic' does anything, can you record the noise and post it?  At least that would let ppl tell you if it's 'to be expected' or a defect.

Do you know of a better way to put two of this boosters in one enclosure and have a switch to turn both on or off and the other switch to select from booster A or booster? I donde want boosters in cascade. Just both boosters as if they were in two different enclosures

[idy]

What the guys are getting at is that these single transistor boosters can be very touchy about picking up supply noise, more than other circuits. And your power supply may be "clean enough" for other pedals but not these boosters, unless the boosters have extra decoupling. That means in additon to the normal electro cap from supply to ground things like: a parallel smaller cap; a series resistor from the supply to the "rail" inside the pedal; any other suggestions? Yes good grounding...

Is your supply a "one spot" or other gold standard quiet supply? I have found: Mosfet booster fed by generic cheap power supply= hum. Mosfet booster fed by One Spot= no hum.

[GibsonGM]

I'm with idy on this.   I assume you built the Mosfet boosts two times...each board having its own "Vr" bias supply.   I would place a smaller cap in parallel with the 10uF already there, C3 on the schematic.  .1uf (100nF) is probably a good choice.    Just solder it in parallel with the existing cap, under the board is OK, on each board.    Where the 9V "+" comes into the PCB, you can try adding a 100R resistor in series, too (9V in, 100R >> where 9V used to be).  That gives us some 'noise bypass'.  It is worth trying, what the hay, it will make a nice clean bias no matter.   Make SURE all of your ground connections are present, and are good quality solder joints.  But I'm also thinking supply since...quiet on batteries.  This is a great circuit, and isn't prone to have many problems.

See if that does anything.  If not, can you post a few pics of your build, front and back sides?  The ground scheme may also be playing a part.  The switch wiring diagram you posted looks right, hopefully your wiring is too    LEDs probably should not ground to a PCB; I'd use a jack ground point if nothing else is available.

[Atodovax]

I'm with idy on this.   I assume you built the Mosfet boosts two times...each board having its own "Vr" bias supply.   I would place a smaller cap in parallel with the 10uF already there, C3 on the schematic.  .1uf (100nF) is probably a good choice.    Just solder it in parallel with the existing cap, under the board is OK, on each board.    Where the 9V "+" comes into the PCB, you can try adding a 100R resistor in series, too (9V in, 100R >> where 9V used to be).  That gives us some 'noise bypass'.  It is worth trying, what the hay, it will make a nice clean bias no matter.   Make SURE all of your ground connections are present, and are good quality solder joints.  But I'm also thinking supply since...quiet on batteries.  This is a great circuit, and isn't prone to have many problems.

See if that does anything.  If not, can you post a few pics of your build, front and back sides?  The ground scheme may also be playing a part.  The switch wiring diagram you posted looks right, hopefully your wiring is too    LEDs probably should not ground to a PCB; I'd use a jack ground point if nothing else is available.

Thanks both for everything. Im going to post pictures tomorrow. Actually the sound is like a whistle, its not hum. I tried with the 100nf capacitor in series but made no change at all. im goiung to try the resistor now. I made the pedal on a stripboard and made an extra line for all the grounds so all the grounds are in the pcb. Do you think i should do the star grounding on the jack?

[GibsonGM]

Whistling is not the same as hum.  That means oscillation.   High gain circuits will feed back voltage from output to input if wires are too close together or are too long, or if there are other build problems.   But, if it does not do it on battery, it is almost entirely likely that the power supply is causing this.

The small cap was to be in parallel with C3, not series...simply added across it...if that does not work, pics may help, and even an audio clip of this whistle sound, so we can tell how loud it is compared to your guitar signal.

[duck_arse]

Thanks for replying so fast. I am using this layout. I want a switch to show that the pedal is on and then another one two swith between one or the other one. I am not using shielded wires.. i will try that. Psu only has this hiss in this pedal. Do you think it may be wires too long? Or both mosfets always on interacting with each other?

just an observation, with nothing to do with your whistle - if you wire all three leds to one single current limit resistor, you may find the indicator operation is not what you'd expected. at the very least, you should have another resistor supplying the top switch, and move the 680R [which will be very bright. very.] to the lower switch/led.

[R.G.]

MOSFETs used as amplifiers have high gain, very high input impedance, and in single ended circuits, no inherent rejection of power supply and ground noise at all. That is, as idy says, they are touchy and tricky to set up well. A MOSFET gate can and will amplify noise and hum from the air touching it unless your circuit makes this impossible.

Putting two of them in a single box, sharing a single power supply, with nearby wires to capacitively couple signals from one to the other, makes all this more difficult. It can be made to work, but the mysterious stuff like wire routing, grounding, and power supply decoupling get ever trickier.

Whistling and squealing is, as noted, self oscillation. What's not generally realized is that if the thing is oscillating at high RF frequencies, it can also "detect" the RF and make audio noise out of the oscillation, sounding like and angry, aggressive hiss, not just a squeal.

To fix it, you have to break the ways that signal can cause oscillating feedback by breaking the pathways that the signal at one output can get to the other input. You do this on power supplies by using good, high quality bypass capacitors locally, right at the amplifying devices. In this case, a 0.01uF to 0.1uF ceramic capacitor from + at the drain resistor to the ground at the source of each MOSFET is a good start. Next, make the two MOSFETs NOT share a common wire for either B+ or ground back to the battery/DC supply.  Another good step is to realize that these MOSFETs have amplfying response up into hundreds of MHz, high enough that most oscilloscopes can't even see it. Cut the amplifying bandwidth down by shielding wires going to the grid circuit. The shielding cuts pickup of capacitively coupled signals, and at the same time wipes off some of that RF sensitivity.

Do NOT make them share a common reference voltage unless you are very careful about decoupling the reference voltage for both RF and audio. That might take the form of RF ceramic caps in parallel with the reference decoupler, separate series resistors to each MOSFET gate, and splitting those series resistors into two series resistors and a separate ceramic cap to ground in the middle of each series pair. Otherwise, RF can leak from one device back through the reference to the other.

In high gain, high bandwidth, high impedance circuis like MOSFET amplifiers, little things can matter a lot. I don't know enough about the physical routing of your wiring and so on to tell you much more than these tips.

[Atodovax]

Thanks for all the replies!. Today i rewired everything using shielded wires for the audio pad and the whistle was reduced considerably but it is still there in the background. Specially on high settings. Do you recommend a grounding scheme? I added a little piece of stripboard and added all the ground wires there. Also added al the V+ wires there. Do yiu think its better to do a star griunding on the input jack? Is there any difference from doing it there or in a piece of stripboard near the jack? I have almost 10 wired going to ground so i tjought it would be better to use a piece of sttipboard for a better look. I did the same with the positive wires. So all power conections finish in the same piece of stripboard and the audio path has shielded wires. I cant think of a better way to do it. Maybe two of this boosters are too much for the same enclosure being to near?

[Atodovax]

Does this picture helps? I donde have the pedal right here but thats the desing. The first 3pdt switches gerelan effect on or off and the other 3odt selects from one effect or the other one changing their LED. Thr first Switch also has another LED that shows if the pedal is on or off

[Atodovax]

Thanks for replying so fast. I am using this layout. I want a switch to show that the pedal is on and then another one two swith between one or the other one. I am not using shielded wires.. i will try that. Psu only has this hiss in this pedal. Do you think it may be wires too long? Or both mosfets always on interacting with each other?

just an observation, with nothing to do with your whistle - if you wire all three leds to one single current limit resistor, you may find the indicator operation is not what you'd expected. at the very least, you should have another resistor supplying the top switch, and move the 680R [which will be very bright. very.] to the lower switch/led.

Ok so i managed to get rid of the whine using all of the methods that yiu recommeneed me. Thank you very much!! Now i have it assembled but i have this poping issue when i turn the pedal on or when i change from one two the other one . I added 1Meg resistors on both circuit inputs and outputs to ground .. the wiring i am using puts the input on the effevt A ir B to griund when it is not ON with the first 3pdt but only one effect input is grounded at a time.. not both.. might that be the proboem whith this poping noise? Is there a way to ground both FX inout to ground when turned off ? Maybe using a 4pdt? Right now i am using a 3pdt tl turn everything ON or OFF and anotherone to choose from one FX to the other one. The center of that last 3pdt outs one if the two leds to ground at a time ti indicate witch one is activated.. sorry for so many questions!!