Charge pump for +15v/-15v?

[rullywowr]

I am looking for the most practical way to obtain +15v/-15v off a standard +9V pedal supply.

I understand there are special DC-DC converters such as the Murata NMA0515SC which can take 5V and output 15v/-15v.  http://www.mouser.com/ProductDetail/Murata/NMA0515SC/?qs=%2fha2pyFaduhFiAPVWrTrFQM2YxeR6Wl2ZuHmK8xFWK4%3d

The problem with these is that they are expensive, and require a 5VDC input.

Is there any issues with say using a standard TC1044SCPA to boost the 9V voltage to +18v, regulate the output with a 15v regulator, and then use another TC1044SCPA to invert this result to +15/-15v?

Current use is not of upmost importance, as my intentions are not to use a 9V battery of course.  I may be splitting hairs here but the Murata solution looks pretty easy...could be about the same cost (and area cost) as using several TC1044.

Thanks in advance. -Ben

[RobA]

Check out figure 24 of the LT1054 spec sheet. You can do it that way with one charge pump and two regulators.

[Govmnt_Lacky]

Current use with respects to the draw on the battery is not the concern. The current draw of the circuit on the pump is where you may run into trouble. If your circuit is drawing more current than the charge pump can deliver, you will see degraded performance in the form of distortion or even circuit failure.

Do you know how many mA the circuit(s) will pull? My golden rule (with the LT1054) is 50mA. If the circuit pulls more than 50mA then I need to figure out another way. Of course, I have "heard" that the LT1054 can be used up to 100mA but, I have not tried it. Also, that was purely as a doubler. You are looking for a positive doubler AND a negative doubler. Might need some figuring out and experimentation.

[rullywowr]

Awesome guys.  Looks like this is the one here:

Am I correct that the TC1044SCPA is pretty much the same and can be substituted here?



Thanks for explaining this.  I'll assuredly will have to experiment with it and see what the resulting current draw is.   

[midwayfair]

Current use with respects to the draw on the battery is not the concern. The current draw of the circuit on the pump is where you may run into trouble. If your circuit is drawing more current than the charge pump can deliver, you will see degraded performance in the form of distortion or even circuit failure.

Do you know how many mA the circuit(s) will pull? My golden rule (with the LT1054) is 50mA. If the circuit pulls more than 50mA then I need to figure out another way. Of course, I have "heard" that the LT1054 can be used up to 100mA but, I have not tried it. Also, that was purely as a doubler. You are looking for a positive doubler AND a negative doubler. Might need some figuring out and experimentation.

Output current is 100mA for the regular 1054 and 125mA for the 1054L (but that can only take 7V in). It's in the datasheet, and it also scales with the supply current from the power source up to 120mA. They're not going to overestimate the current delivered because that's a really good way to piss people off (just like you never see a maximum voltage rating underestimated), and you can probably squeeze a little more out of it. Lots of people have used the 1054 in univibe clones, which easily draw far more than 50mA, but there's no reason to rely on hearsay when the datasheet gives you maximum ratings.

The main drawback to supplying as much voltage as possible is that it increases the voltage loss. You lose about a half volt extra of voltage when drawing more current than you do when drawing very little.

[jkokura]

Am I correct that the TC1044SCPA is pretty much the same and can be substituted here?

Yes and no. Understand that the TC1044/Max1044 are different than the LT1054. Largely it's about current draw and voltage input. Again, the question is - how much voltage are you feeding and how much voltage are you wanting out?

Also, you can do it with some voltage dividing, instead of using regulators on the end. For instance, if you need +/-15V, you can take your 9V in, divide it to 7.5V, feed it into the Charge Pump, then you should have +/-15V. Just an alternative method over using the regulators, but that's just a thought, you should do the research as to which method is safer.

Jacob

[Scruffie]

Awesome guys.  Looks like this is the one here:

Am I correct that the TC1044SCPA is pretty much the same and can be substituted here?



Thanks for explaining this.  I'll assuredly will have to experiment with it and see what the resulting current draw is.

Yes that will work, and depending on what you're powering, 2 x 15V zeners could be used as shunt regulators, this was fine for several vintage EHX effects so...

[RobA]

One other point to using a different charge pump is the boost pin. The LT1054 doesn't need the boost pin (or have one for that matter), but if you use the TC1044SCPA, then you'll need to mod the schematic to connect the boost pin.

[rullywowr]

Thanks again.  The documentation on the TC1044S is pretty weak.  It looks like connecting the boost pin to V+ does raise the clock frequency out of the audible range (which should eliminate whine).  It also looks like enabling the boost pin does slightly reduce efficiency until about 20mA as compared to not enabling it. 

[RobA]

Thanks again.  The documentation on the TC1044S is pretty weak.  It looks like connecting the boost pin to V+ does raise the clock frequency out of the audible range (which should eliminate whine).  It also looks like enabling the boost pin does slightly reduce efficiency until about 20mA as compared to not enabling it.

Yep, that's pretty much the trade offs. I've found the TC1044S to be unusable without the boost connected though. Actually, I've had several of them that had the whine audible even when they were used with the boost in place. There is a TC7660H that works better and has a much higher clock to begin with for use in lower current settings. But, I still think the LT1054 is a better part. I would really consider it in this case where you are going to power two rails with the charge pump.

[rullywowr]

I agree, Rob.  The LT1054 seems much more robust, especially concerning the output current.  The TC1044 has a supply current (with boost) of about 350uA versus the LT1054  can do 4-5 mA.

@ Jacob...that is a pretty nifty idea about the voltage divider then boost/inversion. 

[Govmnt_Lacky]

Also, you can do it with some voltage dividing, instead of using regulators on the end. For instance, if you need +/-15V, you can take your 9V in, divide it to 7.5V, feed it into the Charge Pump, then you should have +/-15V. Just an alternative method over using the regulators, but that's just a thought, you should do the research as to which method is safer.

Dont forget that you are going to get a slight voltage drop when using it as a positive doubler (and possibly when doubling the negative voltage)

I know that when I use the LT1054 as a doubler and feed it with a 1Spot. I usually only get ~17V out and some change. Never get the full 18VDC.

Food for thought 

[rullywowr]

Also, you can do it with some voltage dividing, instead of using regulators on the end. For instance, if you need +/-15V, you can take your 9V in, divide it to 7.5V, feed it into the Charge Pump, then you should have +/-15V. Just an alternative method over using the regulators, but that's just a thought, you should do the research as to which method is safer.

Dont forget that you are going to get a slight voltage drop when using it as a positive doubler (and possibly when doubling the negative voltage)

I know that when I use the LT1054 as a doubler and feed it with a 1Spot. I usually only get ~17V out and some change. Never get the full 18VDC.

Food for thought 

Thanks Gov't.  The goal is to get a stable +15/-15v supply so it's oK if it doesn't hit 18 exactly, it just has to be enough for the regulators to work and shave off what is needed for +/-15V.  I am leaning towards LT1054 with seperate TO92 regulators at the moment. 

[jubal81]

Personally, I'd probably go with the Murata. Fewer parts, less complicated and smaller footprint.
It probably adds about $3-4 to the cost over the 1054 solution, but that'd be worth it to me for the other benefits.

[rullywowr]

How important do you feel the 220h inductors are?  I've seen both with and without...