Alternative to charge pump

[Transmogrifox]

These are an interesting regulator.  If we can find some inductors that are comparable in size to a capacitor, this makes for a very efficient voltage boost circuit (see typical applications & notes near end of document):
http://cache.national.com/ds/LM/LM2587.pdf

We could get away with smaller-value inductors since there isn't need for much power in a stompbox.

I had a professor that once made a doorknob shocker based on this principle.  Need a higher voltage rated switching transistor for the doorknob shocker.

[R.G.]

That's funny - charge pumps were welcomed roundly in the industry because they were a better alternative to inductor-based switching power supplies.

[gez]

Similar 'roll yer own' circuits crop up in EPE magazine from time to time.  Usually to light LEDs from supplies lower than their thresholds...LED torches that sort of thing.  Never tried them though.

[Transmogrifox]

That's funny - charge pumps were welcomed roundly in the industry because they were a better alternative to inductor-based switching power supplies.

The thing that caught my eye is that it operates at 100kHz instead of 10kHz.  I haven't gone charge pump shopping, but it doesn't seem unlikely that some charge pumps operate at a higher frequency.

What is the expected efficiency of a charge pump?  I do realize that real-world inductors generally are much more prone to burning energy in parasitic resistance than real-world caps.

[Sir H C]

Inductor based regulators are always the highest efficiency.  Some of the Linear Tech ones can do 90% over the whole range.  Charge pumps often are 50% efficient.

The big advantage of the charge pumps over the inductor based set-ups is the ease of design and building.  Since inductor based designs are current based you have more poles and it is harder to keep stable.  Charge pumps are very bang-bang and easier to keep stable.  Also layout is less critical.

[R.G.]

As always, easy and cheap may not be the most efficient. Sometimes that happens, but be very, very happy when you find that.

Practicing EEs avoid inductors at almost all costs.  Inductors are not standardized, they are lossy, they are heavy (that core material stuff is usually metal based), and they are expensive, as well as mostly being custom order things. Lots of beginner EE courses teach ways to avoid inductors  - or they used to back when EE taught circuit design, not programming.

That's one reason that charge pumps were welcomed so heartily. One IC, a few caps and you get a negative voltage. In most cases, that's all that's needed.

Like all power supplies, the efficiency of charge pumps varies. They can be upwards of 80-90% in the right ranges. Inductor and transformer power supplies can also be very inefficient at some loads. Saying "charge pumps are more (or less) efficient than inductor based regulators" has no meaning unless you also specify the conditions. Depending on the conditions, it can go either way.

Running switching power supplies at higher frequencies usually makes the LESS efficient. Whenever you switch an active power device, it takes some power to do that. With MOSFETs in particular, there is a power pulse needed to charge the gate and move the drain from off to on, and another one to remove the charge from the gate and turn if from on to off. That power loss is constant. If you do it more times per second, you eat more power. What high frequencies buys you is smaller reactive components, especially inductors. 100kHz and 1MHz switchers can use really small inductors and transformers, as well as smaller caps. So you pay more power losses to get smaller parts. Sometimes that's a winning strategy. But not always.

Highest power supply efficiency is usually at 50 or 60Hz - that is, no switching losses at all. But it costs you to regulate that. If you're willing to pay more money for bigger iron and caps, lower frequency is more efficient in terms of power out divided by power in.

Mother Nature makes the Rules. One of those rules is that the devil is always hiding in the details.

[Paul Perry (Frostwave)]

RG is right, of course. I'd just like to say, the main advantage of switch mode supplies, for stompboxes, is that it takes away the problem of batteries going from 9 to 6 volts. Saves money too, because you can get the last drop from the battery.

[Ben N]

RG is right, of course.

You can post those words right at the gate.