18v Charge Pump circuit question

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

sonolink

Well-known member
Joined
Feb 15, 2010
Messages
1,279
Location
London-Madrid
The 2 attached circuits are supposed to be the same, but the one on the right has an extra cap and a diode in the opposite direction. Can someone explain the reason why and if there is a "correct" one please?



Thanks
Cheers
Sono
 
Those are two different ways to protect against reverse power connection, (like from trying to plug up the 9V battery backwards).

In one the series diode ignores a reverse battery connection, but the cost for that is losing one diode drop in forward direction, so instead of 9V it is only getting 8.5V or less. 

The second approach connects the diode reverse polarity between input and ground so it will clamp a backwards battery protecting the IC. Since 9V batteries are typically less than 300 mAh a 1A diode clamp will serve and protect.

The extra PS capacitor across the battery will deliver a stiffer rail then an almost depleted 9V battery .

JR

PS: I have also used MOSFETs to protect against reverse battery with very low forward drop, but a 1n4001 is cheaper.
 
10kHz oscillator? Nah. There are lots of fancy switching ICs now. Especially if you're ok with SMD (unavoidable nowadays so you'd better be). A higher frequency oscillator will 1) make it easy to filter out switching noise and 2) make your filter caps a LOT smaller.

What exactly are you using this for? Are you trying to boost to 18 or do you want +-9 or what?
 
Thanks for joining in Squarewave :)

I want to use this to boost to 18v an Overdrive Guitar Pedal. My circuit is a clone of the well known King Of Tone by Analogman with some mods. 18v would give it a bit more dynamic range resulting on more transparency, especially for clean or mild overdriven sounds.

I have a working prototype on vero. About 10 days ago I was working on a PCB that I wanted to cut with my CNC when a friend told me: "what about 18v?"  ???
Simply put, I didn't sleep since....soooo, I had to look for it, I found a charge pump circuit by Fuzzdog and another on Tagboardeffects. I noticed the difference and posted here :)

This is the circuit as it is built now:


As you can see it's a pretty simple 2 channel diode clipping opamp circuit based on a JRC4580 IC.

My idea after JohnRoberts' reply was to include the charge pump but your post caught my interest too :)
And, yes, I'm fine with SMD ;)

Thanks for your time :)
Cheers
Sono


 
Well unfortunately I was wr-wr-wrong because all of the charge pumps are 5V in. TC7660S seems to be the one exception. It can accept 10V in. Fortunately I was also wr-wr-wrong about the oscillator frequency. If pin 1 is tied to V+ (and it is in your schem), the frequency goes up to 35kHz (the Microchip version says 45kHz) which is great.

However, for your application, you don't need or want to make 18V and then split that with a virtual ground. It's much simpler and more efficient to make -V. Meaning you can make a proper bipolar supply thats +9V, GND and -9V. You only need 2 caps (and the one supply reversal protection diode you asked about). Look at the "Simple Negative Voltage Converter" example in the Microchip datasheet. Then you can also ditch the virtual ground. Anywhere in your circuit that reads GND becomes -V and anywhere that reads VB becomes GND. Except the 1M on the inputs and outputs - those need to be tied to your new GND.
 
squarewave said:
Meaning you can make a proper bipolar supply thats +9V, GND and -9V.
You mean like setting an offset or placing the 18v range between -9v and 9v instead of going up to 18v from 0v? This is very interesting :)
Why would you do one thing instead of the other?

squarewave said:
Anywhere in your circuit that reads GND becomes -V and anywhere that reads VB becomes GND. Except the 1M on the inputs and outputs - those need to be tied to your new GND.
Like so?


squarewave said:
Except the 1M on the inputs and outputs - those need to be tied to your new GND.
Those are bleeders right? they need to bleed to GND but my new GND is now 9v right? This feels weird in my brain :)
I have to wire them to the jacks' in and out sleeve, right?

Cheers
Sono

 
Pretty close. Except C5 and C14 should go to GND and not -9. Another exception.

The 1M do not need to be wired to the jacks. They can be wired to any ground point.
 
sonolink said:

Those are bleeders right? they need to bleed to GND but my new GND is now 9v right? This feels weird in my brain :)
I have to wire them to the jacks' in and out sleeve, right?

Think Bipolar (not in a mental health way !)
GND is GND and you have supply voltage rails above (+9V) and below (-9V) it.
 
squarewave said:
Pretty close. Except C5 and C14 should go to GND and not -9. Another exception.
Proud and happy I feel now :)
Thanks for your help Squarewave ;)

Rocinante said:
I like these guys pcb's as they make great clones. Here is their utility charge pump
https://www.pedalpcb.com/product/3pdt-chargepump/
I especially like that pedalpcb has a 9v+-.

Amz has another solution here
http://www.muzique.com/news/how-charge-pumps-work/
Thanks for the link Rocinante :)
That PCB looks great actually and both chips seem interchangeable (TC7660 and TC1044)

Newmarket said:
Think Bipolar (not in a mental health way !)
GND is GND and you have supply voltage rails above (+9V) and below (-9V) it.
I know, it just feels weird to conceive negative voltage hehe but I'm adapting my mental frame slowly ;)

Cheers
Sono
 
sonolink said:
I know, it just feels weird to conceive negative voltage hehe but I'm adapting my mental frame slowly ;)

Think of it as actual sound waves if it helps.
When the air is still and your ear diaphragm isn't moving that's 'GND'. When its pushed one way its +ve and when it's pulled the other way it's -ve :)
 
Actually, that's exactly where the weird feeling begins: what you describe is AC for me :)
The bipolar concept is new to me (I'm a total noob in electronics). DC in my head was a steady value, a stiff line at whatever Volts used to feed active components and a source of noise if not carefully taken care of with caps and GNDs.

The idea of "offsetting" 0v to get a bigger range/voltage is very interesting, but does negative voltage actually exist in physical terms or is it an agreed term/concept?
 

Latest posts

Back
Top