Chargepump/battery switch

[Vendt]

Dear All,

I'm working on a tone bender with Ger PNP transistors.
I now know the input voltage is important for this type of effect.

The pedal I want to build should be able to be powered by both a battery and a Wallwart.
The wall warts I use (and a lot of the players that I know) are approx 12V instead of 9V at least they are inconsistent.

To get a steady -9V from the wall wart I should use a charge pump (Max1044 or similar) followed by a 7909. However if I enter the battery power before the charge pump and the 7909 a lot of power is lost in the circuit, and the 7909 will not work anymore.

I've tried to solve this with a mini relay switch
What do you think? is this safe? Will it be durable? Are there other solutions?

[Aph]

Yeah, a 9 volt battery is not going to get the 7909 (or a 79L09) to regulate. You need almost -12 volts for that.
If you are using an adapter that is stated as 9 volts but is putting out 12 volts, that's an unregulated adapter and I'm surprised you're not getting hum through some of your pedals.
I've seen charge pump circuits included in schematics of tonebenders, but I've never built or breadboarded such a circuit... so, what I say here is assuming that these circuits actually work:
Get a proper regulated 9v adapter (mine actually puts out 9.6 volts). Then you can get rid of the transistor circuit, the 7909 circuit, and the relay... just use the 1044 (or ICL7660S, etc) circuit.

[Vendt]

Thanks for the reply.

I have tried this on the breadboard (without the relay) and it is working. I have not checked what the lowest voltage for operating the 7909 is in this circuit.
I will use a ICL7660S as the Max1044 can break over 10V. I believe an ICL7660S can handle a bit more.

Off course a proper regulated 9V power supply would do the trick and is preferred, but some of the power supplies (yeah, some where cheap) I have bought over the years are said to be regulated to 9V but will actually give 12V.
If I lend out or even sell one of these Tone Benders I would like that the effect can be used with "inferior" power supplies as well.
A lot of the people I know will use the cheapest option for power supplies they can find....

[PRR]

> I have not checked what the lowest voltage for operating the 7909 is in this circuit.

Pencil "3V headroom". It may be a bit less with low load, but 3V is safe.

That means anything over 9V but a wee bit under 12V is a 'grey zone' where it may not work properly.

If you have careless users, you don't want that.

There are several ways to do this. A quick thought is to use a 7805 to take 8V to 18V and get a solid 5V. Then find a voltage doubling inverter (are there such things?) to make that negative 10V nominal. With losses it may wind up very near -9V.

Your LED should be fed the raw +8V/+18V, not load the doubler.

LM2682 Switched Capacitor Voltage Doubling Inverter

[Vendt]

Ha,
That's a complete new way of thinking. I'll look into that later this evening.

Thank you for the input!


Verzonden vanaf mijn iPhone met Tapatalk

[Vendt]

I have had a look

My normal suppliers do not stack the LM2682...

So I went on looking for other solutions and I came across this:

https://electronics.stackexchange.com/questions/143702/why-is-the-lt1054-circuit-providing-far-more-than-double-voltage/143739

This after the 7805 should work wouldn't it?
Than there is no need for the relay...The battery can come before the 7805.

[Aph]

Well, breadboarding it should take very little time. I would try it. I've had a bit more luck with LT1054's than other charge pumps. However, I have built the circuit that you linked to (or something very similar), and it was not able to supply the current I needed (we're talking in the low mA's here). A voltage doubler is going to steal some of your current (as will the 7805 regulator, by the way). Also, I was using a Texas Instruments LT1054... I have some Linear Technology LT's on order just to see if that makes a difference.

[PRR]

Yes. I stopped on the first line of Google with a part that was in-stock at a big distributor. It is really more about the diodes and caps than the converter chip. Losing the relay, and having wide tolerance, is probably worth the small loss in the regulator.

The '7805 will work down to 8V. A tired battery may be lower. An "LDO" 5V regulator may hold 5V down below 6V.

Aph-- note that the StackExchange discussion suggested that "cheap caps" were a cause of quick sag. Low ESR (Tantalum or good Al electrolytic; maybe MLC ceramic these days) seemed to help. If so then diode loss (including speed) may matter. We are taking a lot of leverage here and can't afford much drag in the many pulleys.

[Vendt]

Again something I have to check!

Thanks, much appreciated!

[Aph]

Thanks, PRR. I had just noticed that. I have some tantalums on order now, too!

[Vendt]

Hi all,

I received the parts yesterday....And did some breadboarding.
It looks very stable to me! I got a real neat 9,01 V.
I will do some more testing.

Attached below is a perfboard layout.
It should work between 7,5V and 30V I would not recommend the 30V but it should easily hold the 12V that I have asked for.
Comments and remarks are more than welcome.
Thank for the help/ideas it is much appreciated.