Charge Pump Vampire!

[italianguy63]

First off-- Happy Thanksgiving!!

I have an interesting problem.  I am hoping someone can help.  I have built a FUZZ FACE with a charge pump power supply.  I won't go into a huge explanation, but I have the "Power on" LED in the charge pump portion of the circuit.  The charge pump is switched by the 3PDT on the positive ground side.  The charge pump side of the ground is isolated from the FUZZ FACE part of the ground (except when it is on).



Anyway-- the problem is this.  Originally when I did this the LED would stay illuminated even with the 3PDT off.  WTF?  Anyway, I figured there had to be some kind of leakage of voltage "backwashing" into the charge pump IC and leaking through.  I added D2 (1N5817) assuming it would stop the probably A/C causing the LED to light.  It basically worked.  But, not completely!!  The LED is still minutely dim with the switch off and the stereo jack occupied.  It wouldn't be a problem, but obviously this will be a battery killer.

Any idea what is causing this, and hopefully someone can give an easy fix.  Otherwise the circuit works stellar.

Thanks!
Mark C.

[armdnrdy]

First question is....why are you switching the power to the charge pump with the footswitch?

Edit: Happy Thanksgiving to you as well!

[R.G.]

I'll have to go through it a bit but I believe you are not switching the power to the LED with the footswitch. Further, if your schemo is accurate, why are you switching the LV pin of the 1044?

And even further back, the MAX1044 has reliability problems because of its intolerance of voltages higher than 10.00000000V.

[italianguy63]

Because I figured it would work.  I can ground the charge pump, and switch the 9V in, which is probably the soluiton... but, I was actually looking for the answer as to what and why the voltage is backflowing.  I am not seeing anything obvious.  Actually, to me, it looks like it "should" have worked.

You can omit the LV trace.  It is on the board, but not used.

[Mark Hammer]

What I'm having trouble with is why, with all the people wanting so desperately to mimic "dying batteries" in Fuzz Faces, and all the commercial power bricks that include all of that "sag" stuff (not to be confused with South Asian "saag"), specifically for powering germanium fuzzes and assorted FF derivatives, you want to use a charge pump with one.  Does it do more of something you like at higher voltages, or is it a matter of using the pump to simply provide -9v and permit use of standard external power?

[italianguy63]

In this case, just so it is compatable with a wall-wort for external power..

[italianguy63]

OK-- the probable fix is this:



But, I would still like an answer to the original problem if somebody can explain it.

And, in case anyone cares, the power for the LED would be better served coming from the downstream side of C6-P.  It would be superior because it would probably filter the noise that the LED introduced.  That is why the "LED Pop" circuit (2 resistors and 1 cap) were added.  The LED caused a whiny-buzzy sound.  I just happened to put it on the existing PCB I was using.

It's Thanksgiving, and I have time to kill-- I'll go build this revision, and see if it works..  MC

[R.G.]

Because I figured it would work.  

Yeah, I've used that one before too. Mother Nature sometimes has other ideas. In this case, it may translate to "the chip internals don't work that way".

I can ground the charge pump, and switch the 9V in, which is probably the soluiton...

A better solution is to take that ground line off the LV pin of the MAX1044 and insert it between the anode of the LED and ground.

but, I was actually looking for the answer as to what and why the voltage is backflowing.  I am not seeing anything obvious.  Actually, to me, it looks like it "should" have worked.

It depends on what the MAX1044 actually does when the LV pin is pulled. LV is used for switching the internals a bit to allow or disallow a lower voltage to be converted. It's not clear to me what it really does when there is a supply in the "high" range and it's flipped. Like most other engineers with product experience, I've used un- or minimally-documented chip options, but it's easy to get burned that way to. Or spend several days saying to yourself "Gee, I wonder why it did THAT."

You can omit the LV trace.  It is on the board, but not used.

Board?

What I'm having trouble with is why, with all the people wanting so desperately to mimic "dying batteries" in Fuzz Faces, and all the commercial power bricks that include all of that "sag" stuff (not to be confused with South Asian "saag"), specifically for powering germanium fuzzes and assorted FF derivatives, you want to use a charge pump with one.  Does it do more of something you like at higher voltages, or is it a matter of using the pump to simply provide -9v and permit use of standard external power?

From the schemo, I think he's just making a -9V, not going to +/-9. I could be wrong.

OK-- the probable fix is this:

I would avoid switching the power on and off to bypass/not an effect. It gives other issues pretty often. Leave the -9V on and switch the audio and LED.

But, I would still like an answer to the original problem if somebody can explain it.

See above

And, in case anyone cares, the power for the LED would be better served coming from the downstream side of C6-P.  It would be superior because it would probably filter the noise that the LED introduced.  That is why the "LED Pop" circuit (2 resistors and 1 cap) were added. 

I'd move the LED over to the +9V side of things and isolate the grounding wire to turn it on back to the +9V source. There's really no reason to have it be on the -9V side. The charge pump is going to work better the less current it provides, and there is less chance of the LED causing unwanted transients, too. Easier to isolate the LED's transients when separated by a switching power supply if the wiring is done right.

[Mark Hammer]

Ooooohhhhhhh....there was an invisible schem?  Well, then never mind.  Carry on. 

[italianguy63]

Thank you for the info. and the help.

I pulled the LV trace, no change.  MC

[italianguy63]

Yes.. I am using an existing charge pump PCB, and another PCB for the FUZZ FACE.  Just splicing it all together. There was a convienient place for the LED on the -9V side. 

In short, it wasn't "designed" this way.  I was doing the "I wonder if it will work this way" game.  I know it wasn't optimal.  Just trying to make it work.  Isn't that half the fun in this stuff?!

[italianguy63]

But-- we then have another issue... don't we?  The initial problem here is that the LED is illuminating indicating some kind of circuit flow backwards across the diode (D2).  Even if I switch the LED, won't the current still be flowing through the MAX??  Hence, still a battery killer if the stereo jack is plugged in.  Doesn't the power NEED to be switched, or is there another way to isolate the current flow (back to the 1st question). 

MC

[italianguy63]

Here is a gratuitous picture of the "patient."

[armdnrdy]

The power does not need to be switched. That's what the stereo jack switch is for. When you're through playing, you unplug the input cable to turn off the battery.

Now if it's inconvenient to unplug the input due to being incorporated on a pedal board.....maybe a small toggle switch can be added to cut the power. I find it very unnecessary as the circuit you are dealing with has a very low current draw.

If you look at the "standard" way that pedals are wired, you will not find one that kills the power when the effect is disengaged.

This would be like your car's engine being turned off each time you come to a complete stop....then restarting when you hit the accelerator!

As R.G. stated....this can introduce other issues. Switch the signal and the LED.

[italianguy63]

So I am chasing a windmill?

[armdnrdy]

It seems as if you're fixing a problem that doesn't exist.

Don't worry...we've all been through it! 

I think that I'm doing that right now with another build! I'll have to post for advise.

[italianguy63]

My day just got easier! 

Thanks guys.  I'll switch the LED in a later revision...   I am working on redoing this all on one board.  I also started playing with Eagle yesterday.. Figured out the schematic drawing, and the PC board layout.  Still haven't figured out how it "draws" the traces part yet....

I learn a little more every day.

[R.G.]

This would be like your car's engine being turned off each time you come to a complete stop....then restarting when you hit the accelerator!

That seems truly stupid, doesn't it? Increases starter motor and solenoid mortality by about 50-100 times, destroys batteries, and makes traffic flow even more slowly than it does now, as well as wasting gasoline when startups are not quick and easy on every car. Absolutely stupid thing to do.

If you think about governments and their internal aims, you may not be surprised. I heard a note on a radio program that says that Europe has a plan to require all cars to do exactly that: shut off the engine at a stop, which is presumed to be a stop *light* and then restart when you press the accelerator. It gives a whole new meaning to "stop and go traffic".

This is another of those little improvements brought to you from the thoughtful folks who not only don't believe that you will do as good a job as you can of using your resources (energy and money) as you can, but that you cannot be *trusted* or *allowed* to do so; that it is better that the better folks in government be allowed to choose what you do.

And generally pay them for the privilege of doing so.

I wish this were a joke. It's not.

[armdnrdy]

This sounds like typical liberal ideology.

We have something similar going on in the US right now.
It started with lies to sell it, it was shoved down the American people's throats, it will hurt millions of people financially, but it's all for the greater good.....cause GOD knows that the people up in Washington know what's best for us.
We are just too stupid to make are own decisions!

I wish this were a joke....but it's not!

I can't wait until the next elections!  

Okay....I got that out......I feel better now! 

[kaycee]

I think these cars are already in production? Sure ive seen adverts for them? I would assume that they are designed to function efficiently in this mode, its not planned to make folks retrofit this to existing cars. I've never owned a car less that 5 years old, most older, and never sold one on, always run them into the ground so it will be quite a while before I get my hands on one of these new fangled jalopies