Can we talk about Charge Pumps? Design.

[psychedelicfish]

I believe the 7660S can give 10 mA, not 2. Depends also on the configuration (voltage inverter or doubler).

I believe Mark was saying above that the currents in his table were how much each chip needed, not how much each chip can supply.

[italianguy63]

Yes.  It is misleading, I should probably take that column off...

Still waiting on the 1054's.  VERY slow boat from China this time.  I thought they looked great on the spec. sheet.  I then had to make a Mouser order, so I ordered some of the 1044S's at like .92 cents.  Can't beat that price with a stick.   I also made a Pedal Parts plus order for my next set of production Fuzzes.. I have the prototype running.. looking forward to doing some "real-world" testing soon.  I will report back when I know something......  MC

[gena_p1]

cheap chinese sellers with low rating (<1000) usually have no own store,
so You order LT1054 by aliexpress and your seller backorder it from taobao ) That's why 60 day of waiting is normal thing.

[italianguy63]

The 1054's landed in Wisconsin Saturday.  So, only a few more days.  Ordered them on 11/28.  So, about 3 weeks.  Usually only takes about 2 weeks.  I wish Tayda carried them, I love Tayda.

[italianguy63]

A teaser:  Testing coming eventually, I have the next 2 weeks off from work!   

[italianguy63]

Here's a pedal I built today as a "tester."



New board works BRILLIANTLY.  No noise at all (using a MAX1044 right now).  I even purposely used a noisey eBay wall-wort for power and it cleaned that up.  I had never used the "Output Jack Switching" circuit before either-- it is steller as well.  Nice positive switching and no hint of noise.

The ONLY noise was a tiny bit of transistor hiss at full fuzz and full gain.  (I can add the filter caps to the trannys if I want to).  But, this build I used higher gain and higher leak trannys than usual.  I had 2 nice BLACK ones I wanted to pair up.  It actually sounds WONDERFUL, and the bias resistor was spot on spec too.  A successful day!  I will have to reconfigure the board for "No Boost" and try out the 1054 and the 7660S.  Hopefully I can get to that shortly.

MC

[italianguy63]

(removed)

[GodSaveMetal]

PLEASE some body have a alternative to those in comon ICs  cuze in some lands; example PERÚ; we don´t have that doubler´s ICs;   I know there are some alternative circuits; with a 555 or CMOS; there are very chap em!!!!

[midwayfair]

PLEASE some body have a alternative to those in comon ICs  cuze in some lands; example PERÚ; we don´t have that doubler´s ICs;   I know there are some alternative circuits; with a 555 or CMOS; there are very chap em!!!!

Mouser ships worldwide, I'm pretty sure.

I mean, I live in the U.S. and I can't go down to the electronics shop and buy them.

[GodSaveMetal]

In PERÚ is A luxure TO buy BY internet, THERE ARE XTREME RESTRICTIONS TO A PERSON TO DO THAT!!! SORRY THERE IS NO liberty!!!! to buy wetheber you want;   you have to do a permission, pay taxes and other, the persons who have that permissions don´t know what to buy!!!!    that is the enormous problem in my land; you please have somo alteantive circuits whit analog single ICs??? please to many many lands that are in the same problem!!!!

[R.G.]

PLEASE some body have a alternative to those in comon ICs  cuze in some lands; example PERÚ; we don´t have that doubler´s ICs;   I know there are some alternative circuits; with a 555 or CMOS; there are very chap em!!!!

It is possible, but quite complicated to make your own charge pump out of other chips. The 555 can do it, but it has bad noise issues on its own. The CMOS 555 versions may be OK.

555 based circuits need either special switching to make the pump work like the charge pump ICs, or need to be an inverting charge pump doubler (or tripler - or quadrupler) because they can't natively switch the "bucket" cap across the whole power supply. And part of the low-noise of the board is likely the layout and grounding setup as I mentioned earlier in the thread.

A more expedient solution might be to trade a forum-friend something that *is* easily available in Peru for some chips, perhaps exchanging Christmas presents.    

If that doesn't work out, yell, and I'll try to figure something out from CMOS and transistors.

[GodSaveMetal]

Well I found these in INTERNET a charge pump with NE555:

This circuits says that +18V 300mA voltage output at 12V 150mA voltage input

Other with CD4049:

can I use the transistors output to increase the miliamps output of this

[R.G.]

Well I found these in INTERNET a charge pump with NE555:
...
This circuits says that +18V 300mA voltage output at 12V 150mA voltage input

Yep. Notice that it's a doubler, and uses the +12V supply as one of the "doubles". In this circuit, the 555 has the same issues I mentioned - it pulls BIG pulses through its power supply as it switches. These are heard as a whine in the circuits it powers if the wiring is not immaculate.

can I use the transistors output to increase the miliamps output of this

(1) Yes, you can buffer that up if you want. The complexity and circuit difficulty goes up. MOSFETs make nice external switches too.
(2) It's worth asking the question - why? 150ma is much larger than most pedals take. If you're powering multiple pedals from one circuit, use a separate power supply.

There are many, many, many such circuits on the net. Basically, any circuit that makes rectangular oscillations can be rigged up to drive current switches and make charge pumps.  The fine points are making it small, simple, cheap, and quiet. And for all those, the details matter a lot.

[GodSaveMetal]

Thanks R G thanks a LOT!!!!

[armdnrdy]

Here's a circuit from the Japanese stomp box book.

[Govmnt_Lacky]

Looks promising Larry....

Looks like it would only be good for 12V application as the 16~17V is not enough to get a good 15V from a regulator.

Also, I wonder how many mA this would handle? 

[armdnrdy]

Well....my Japanese is a bit rusty these days.... 

but....what I can make of it, I think it says 10-15ma running at 70kHz

Good as is for smaller current requirements, would need some help (darlington) for larger current draw.

But....we can just purchase charge pumps that do the same thing......better.

[R.G.]

I did a bit of thinking about charge pump inverters with everyday household ICs.

It's possible to make an inverting doubler with four caps and four diodes and any square wave source to drive it.

Now the details. The source has to be able to supply enough current. In practice, this means one N-channel and one P-channel TO-92 MOSFET. The oscillator drives the gates of the MOSFET, simplifying things a lot. The oscillator can be anything that gives a good, 50% duty cycle square wave that runs nearly the full power supply (taken to be 9-10V). A CMOS 555 set up to be as close to 50% as possible will do it. So will an oscillator made from CMOS inverters, although the CMOS will come in a 14 or 16 pin package. Using 22uF electros and BAT54 or 1N5819 Schottky diodes gives about -16V at low loads (10K or higher) and no lower than about 11V with a 90ma/100 ohm load on the output. Obviously, the output needs to be regulated, so a 7909 or LM337 would be a good idea on the output so the circuit both gets a stable output voltage and has some ripple rejection.

And again, power, ground, and decoupling has to be immaculate to keep noise from escaping into the circuits it powers.

I'll see if I can draw something up. It is a lot of work to avoid finding a charge pump IC if the charge pump can be had at all.

[italianguy63]

I have completed all the testing I require.  I added a couple 220nF caps to the trannies, and the hiss went away as suspected.  Swapped out the IC for the TC1044S and again experienced excellent results.  I am ready to call this experiment a complete success.  Thank you again to everyone who contributed!!

On a side note, I ordered 50 of these boards to play with (cheaper in bulk).  I probably only need 20 or 25 of them.  If anyone is interested in playing with these, just PM me, and I will pass some on to you for a couple bucks each and actual shipping (to cover my cost).

I have already made a couple slight revisions to the PCB for future use.  I changed the layout slightly:  Rotated C1 to align it with the others.  Changed D1 back to PCB mount.  Grounded the mounting hole with a via.  And, enlarged the holes on the wire pads.

MC

[PRR]

> you please have some alteantive circuits whit analog single ICs???

WWJD?

What would Jimi do?

Jimi Hendrix could not get voltage booster chips.

Didn't stop the music.

Mostly: a 9V battery *should* be ample for guitar-cord circuits. There's only about 1V of signal off the guitar. 9V is plenty.

And I suspect Peru has battery factories.

And if Peru restricts imports, then you probably don't have two dozen pedals and are not depending on pedal-supplies.

If Jimi needed a reverse-polarity supply he'd just use a battery wired the other way.

If Jimi needed an 18V supply he'd use two batteries.


> a circuit from the Japanese stomp box book

That plan is widely used, with more diodes&caps, to make 48V from low-voltage (even USB 5V) to power phantom mikes.