Charge pump

problem: the tripler takes three times the current in as you get out and that current comes in through VDD.

Haha - except it will then try to make 36V - except it won't because the

piglet

Shore

I can use a 74HC series no problem.

Send teh coedz?

One section of a 74HC132 or HC14 should make a mA.

Here's the thing I posted to s.e.b. to educate Jim.

With HC gates, you probably don't need the three buffer sections. I think you said that you only have one gate section available.

Here's one way it could be done as a micropower boost:

And before you fall off your chairs laughing I actually used sucessfully in the 1980s to get a 5V rail from 2.5-3V at low power!

piglet

Spiced it...that should work great. Yeah I only have one section as the others are tied up in a flip-flop sorta deal.

Do they make "small signal" Schottky diodes of that type in a six-pack package? Because I could actually use two more of them for another part of the circuit.

The 4093 has Schmitt trigger inputs, so it actually doesn't need a capacitor to oscillate, right? ; )

Yep, I recall that whole deal...

Oops. I hadn't thought of that. With a tripler, to get 1ma out, the tripler would need to supply 3ma to Vdd. But, since the tripler outputs only 1/3 of its input current, the tripler would need to be supplied with 9ma by the 4093 oscillator. Each time I need some output current, I would need to supply 3 times that in input current. Bad idea. This isn't going to work.

Use 74HC logic, then you don't need to buffer. CD4000 is too soft at 5V, no oomph.

If not regulated John's Cockroft-Walton version will do it but not with a CD4093, should be more powerful 5V logic. If it has to be a CD4093 and you can't use its 5V brother (can't imagine why not) you could add a PNP/NPN follower.

If you need the 12V regulated you can make an oscillator plus boost and have a NPN or TLV431 throttle the input when it gets to 12V. Like a governor. That needs only one inverter section, ideally a Schmitt inverter.

Sure--connect the 4093 output to the enable pin of an SMPS.

:-)

HTH, James Arthur

Did you use schottky diodes? What output voltage did you get at 1 mA out? Post the .asc please.

There are some dual-series-pair parts, and some 3-diode ones too. I use a lot of duals, like the BAT74. Take a look at CMXD3004SR and MMBD3004BRM and HN2D03F. Those are all high-voltage PN parts, but lower voltage/schottky parts are similar.

Given the ten passives hung on that gate to make 12 volts, there is probably a simpler way to do it.

Right, if you don't care too much about the frequency. Just a resistor will oscillate with the gate input capacitance.

I was thinking about using an analog mux instead of the diodes, to make the charge pump. It could run off its own +12 supply. Need more coffee to work out the details.

Connect the 4093 output to the enable pin of a SMPS which drives the startup circuitry for the control system that runs a factory which manufacturers SMPSes to be used in the startup circuitry which runs other factories.

Other factories that manufacture...4093s.

Yep. That works. ...Jim Thompson

Naturally.

Here are a few more serious ideas:

Fig. 1 - OUTPUT BOOSTER +5V === | |/ .--[R1]--+------| Q1 | | |>. | |\ | | +--| >o--+--[R2]--+--- | |/ | | --- C1 | |. | | === ===

Fig. 3 - BOOTSTRAP CONVERTER Here's an adaptation of piglet's bootstrap

Cheers, James Arthur