AMZ Mini Booster based circuit mobile phone interference

[chobot123]

Hi all,
I built a booster/buffer based on the Mini Booster followed by a simple FET buffer.
Schematic here: http://gallery.orosz.sk/main.php?g2_itemId=14413 It works fine, but picks up unusually loud interference from mobile phone when switched to boost mode. When only the buffer is active (it is always active) it doesn't do it.
I tried to put a 10k resistor between R2 and C2, but did not help (the value may be wrong).
The enclosure is Hammond 1590B and I tried it both grounded at input and floating, still the same noise.
PCB here: http://gallery.orosz.sk/main.php?g2_itemId=14418
Does anyone have an advice what could cause this interference and how to fix it? I would like to understand the theory behind it to avoid the same mistake next time. I'm pretty sure that this is my mistake, because the mu amp idea is widely used and my circuit is practically useless with such noise.
Thanks for any hint
stefan

[duck_arse]

your circuit shows only 220uF supply bypass. if you look at the
http://www.muzique.com/amz/mini.htm
diagram, jack shows R4//C4 as the filter. you might try isolating your boost section supply (Q2 and R5) from the rest of the circuit with an R//C filter like shown. it might be the thing.

otherwise, look-up ferrite beads.

[chobot123]

Thanks for the input.
I thought that the R4/C4 filter is a low pass filter with corner frequency around 16Hz to filter out power supply noise (mainly the 50Hz hum). Of course, I will try your suggestions and report back.
Some time ago I built a modified OKKO Diablo Boost+ clone with similar topology (mu amp) which does not pick up the mobile phone noise and guess what... it does have the low pass filter in power supply, but does not have the overdrive and boost sections separated by RC filter.
I'll also look up the ferrite beads, do they go to the DC IN cable (because everything else is PCB mounted).

[duck_arse]

a -3dB low pass at 16Hz will reduce mobile phone freqs by how much?

I'm not sure about the using of beads on this type circuit, but RG has suggested them for troublesome mosfet boosts, with the higher Zin. they would maybe go onto the gate leg of the first 2 fets .....

it may also be your second stage is the problem, but only when fed the mu-amps leavings. maybe it needs the filter, maybe both!

[chobot123]

UPDATE
I remembered that I should have a prototype (made a few years ago) of this circuit somewhere. The schematic is here: http://gallery.orosz.sk/main.php?g2_itemId=14422
You can see that I used the low pass RC filter in both parts of the circuit then.
Found the actual prototype and it does exactly the same. It also pops when I switch it on/off, but that was cured by adding a resistor to both sides of coupling caps.
So I guess that it is not the absence of RC filters on power.
Now the ferrite beads ... and some more reading
thanks

[Seljer]

Electrolytic capacitors have enough inductance that at high frequencies, it cancels out all it's capacitance. Try wiring a 10nF to 100nF film or ceramic cap in parallel with the filter cap to improve it's high frequency filtering capabilities.

[PRR]

> unusually loud interference from mobile phone

[chobot123]

Thanks Paul,
I tried 120p and 39k (I had those values at hand), but didn't work. I also tried the same filter at the input of the buffer stage, but didn't help either.
Anything else I can try?
I even don't understand how that signal gets audible. Mobile phones work somewhere around 900MHz if I'm not mistaken, so is it rectified on some non-linear device and getting de-modulated? Sorry for my ignorance, I'm totally lost here

[karbomusic]

Sorry if already discussed and I missed it above. Are you sure the guitar isn't picking it up and the booster isn't just amplifying what you sent it? I know my guitar loves cell phones. If that is the case, have a talk with your cell phone and let it know it is no longer allowed near your guitar.

[Seljer]

Use a shorted 1/4" jack (or cable) plugged into the input to check if the noise is internal from the circuit or already present in your signal and just being amplified.

[anotherjim]

Try that 120p cap across the booster output pot, 1 to 3. Could be the noise is coming in via the output!

[chobot123]

Thanks Jim, I tried the cap across the output pot, it didn't help.
Then I tried to connect as short cable to the amp as possible and to short the input (I used a short patch cable with a shorted phone jack on one end).
The noise is still there and it changes it's volume by turning the boost pot.
Also tried to battery power the box to eliminate the power supply cable.

Could that be my crappy layout and routing?

[PRR]

> Mobile phones work somewhere around 900MHz

I thot there were higher bands now; but whatever.

AM radio is ~~~1MHz, *far* above the audio band. But if you live near a big AM transmitter, you get reception on *everything*. Crystal diode. Transistor. Even dirty copper contact.

Any rectification will turn the AC into a DC. If the radio carrier is steady, it makes steady DC. That might mis-bias an amplifier.

But many radio carriers are *not* steady. AM has the speech/music as the AC amplitude so you pick-up speech/music directly. Celfones are digital but the digits must come in bursty chunks because it sounds like a disposal eating a cat.

> Could that be my crappy layout and routing?

I *assume* your rig is in a closed metal box. If not, it's like saying "Kick Me!!" to any passing electric wave.

It is possible that the fone is on a higher band, shorter waves, and that one of your wires is exactly that length so is super-sensitive to that wave band.

Specific problem: a MOSFET is very-very good up into the GHZ bands (most BJTs and JFETs not so much). Layout around a MOSFET is VERY critical. Even a few inches of wire laying the wrong way will make it sing GHz, or be ready to at the least provocation. Small, short, tight, and put a 1K resistor in series right _AT_ the Gate.

C2 should maybe be put Gate-Source instead of Gate-Ground.

[chobot123]

I thot there were higher bands now; but whatever.

You're right, higher bands are used for data transfer. In Europe 900MHz is primary and around 1800MHz for UMTS

I *assume* your rig is in a closed metal box. If not, it's like saying "Kick Me!!" to any passing electric wave.

Yes, standard Hammond 1590B type.

C2 should maybe be put Gate-Source instead of Gate-Ground.

Didn't help either

While I'm testing this, I found that almost all my pedals are doing the same but much less noticeable. Unfortunately I don't have any commerial boost or overdrive pedals to copare with.
Anyway, I'm pretty sure it's not my location, amplifier, cables, etc, because this box was built for a friend, who uses it in many different setups and has also a MXR Micro Amp, which does't do this.
So I have to accept the status quo as a failed project for now.
I'll try to experiment with another well regarded boost pedals ant try to learn more.

What would you recommend as a best transparent boost pedal?
I may try Merlin's Glass Blower ...

[blackieNYC]

If you figure out how to keep cel phone noise out of a high gain circuit, please notify every recording studio and radio/TV broadcaster in the world.  they wreak havoc with mic preamps.  Shutting them off is going to be the best solution.  Airplane mode, if you're looking at your guitar tab app.  Does the situation improve with distance if you move your phone away?  Or are you getting every phone in the neighborhood?
I just built the AMZ mosfet booster with just the one mosfet - no fets or buffers.  I've upped the input cap a little, and I currently have no resistors to ground at the input, though I will have to put in a switch depopper. Otherwise it is stock. It is still on the breadboard - a real mess - and surprisingly it is not picking up cellular. Neutral.  I use it to drive the hell out of a fuzz, so it's working quietly with 35dB of gain. Might want to try it.

[chobot123]

The difference is who is using the effect. I have the prototype of this circuit and I didn't even notice that it picks up the cellphone signal. The final product is for a friend who is a real musician and it's difficult for him to tell 100 people in a pub to switch off their phones.
He still uses the booster and likes it's sound, but the noise is annoying.
Another booster (the Micro Amp) behaves much better in the same pedal board (but doesn't sound so good).

The MOSFET booster is a good idea, I'll try it. Maybe I'll be able to put it into the same box.

[blackieNYC]

Here's an idea - you have the gain wide open and then you attenuate at the end.  Would it help if you had a "real" gain control, by replacing R3 with a 10k pot?  I think you can leave everything else as it is including the big cap.  Anyone?

[chobot123]

Makes sense to me. I guess I will not be able to do it with the current PCB, but I can still breadboard it. Not sure that the difference will be noticeable on breadboard though (I mean if it's not in metal box).

[bool]

You need to debug RFI with the circuit as-is in the final form, ie. "in the box".

It's OK to initially debug "RFI immunity" at "circuit" level, ie. the PCB itself, but the real thing should be debugged as-is "on the road".

[blackieNYC]

If you can't remove the 10k resistor and tack on a pot in its place, add a 100k pot in parallel to it. It will have an odd curve to it but it's just an experiment. I don't know if this will make a difference or not, but I would try. See what kind of boost you are getting out of it now - the output volume you will use it for. (For a solo boost it's probably 6-10dB) Keep the newly added pot at the full 100k end for this.   Then max the output volume and start turning your new pot until you've matched that gain approximately. Gain goes up as resistance goes down in this application.  I think.  Listen for your noise and see if it is less.     (Does anyone have a thought on whether or not this might help?)
And yes, the only way to go about this is with the box closed.  See if you can tuck this new pot inside. Remove the battery and use the power supply perhaps.