I do not leave comments usually, but not this time. I really enjoined your explanation as it perfect balance between humor, detailed explanation, covering all the aspects and math. Thank you!
I was tracing a negative rail on my bench power supply that failed a while back. I managed to fix it today, Lo and Behold youtube is suggesting every charge pump video. Not bad, i like it.
Just yesterday I learned that the charge pumps in RS232 chips are used in the cheap HC-SR04 ultrasonic sensors, to drive the high-frequency speaker :-) Thanks for this great video!
Nice video Pete! Perhaps you could replace the dev board with a simple hex schmitt inverter. Create an oscillator with one gate and buffer it with a second gate for the first clock signal. Also feed that clock into a third gate to create the anti-phase clock. That should improve synchronization of the two signals and reduce cost to almost nothing.
Can someone please answer the following question. Why was a bridge circuit necessary? Couldn't the 12v simply been switched on and of via two NPN transistors connected to the two outputs?
For high thermal mass components or components that require more heat underneath, I use a lab hot pad like the MS-H280-Pro just underneath the board. Make sure the temp setting is below the solder melting point. Then use a soldering iron to raise the temp the rest of the way. This may be a bad method if there are a lot of low thermal mass components directly underneath the component being replaced. - In reflow, bottom components are usually glued to prevent them from falling off during the top layer reflow process when the board is flipped.
Interesting, but it had me thinking: since the first stage here is most effective (relatively) since it boosts the input voltage by 100% (omitting any losses for simplicity), isn't there a way to double the voltage for every stage instead?
but then you get 1Tp out of phase. The right way would be to use a differential circuit so that both inverted signal are as perfectly time aligned as possible.
Oh, i like this. Im gonna build one tomorrow, but instead of a uC, ill use a plain old multivibrator to produce my tens-of-khz squarewaves. And bat85 schottky instead of 4148s
Pete, disabling Timer 0 on arduino gets rid of jitter in the outputs because of the microseconds() update, it uses Timer 0. Also, you could modify the other timers registers to get I/O clock or PWM into the MHz range. Couldn't you just use an Sparkfun H-Bridge driver for your HV driver? usually the outputs are differential with tight timing being driven with one PWM output. Great video though!
Hi, Pete. I am currently doing the charge pump project also. But i am facing the problem about the output current. My output current is only few mili Amps and micro Amps only. May I know, how you maintain the current of the charge pump?
For those kind of freq/curent/voltage, instead of messing with transistor you could use mosfet driver chip, they make some with two time aligned inverted out so that solve your dephasing on the go, they usually goes up to 0.5-1A, just watch the power dissipation cause they come in small plastic package. Talking about product design here, of course i got it that you just did with whatever you had on hand so here comes the 2n2222 ;c) Just my 2cts...
I do not leave comments usually, but not this time. I really enjoined your explanation as it perfect balance between humor, detailed explanation, covering all the aspects and math. Thank you!
I was tracing a negative rail on my bench power supply that failed a while back. I managed to fix it today, Lo and Behold youtube is suggesting every charge pump video. Not bad, i like it.
Just yesterday I learned that the charge pumps in RS232 chips are used in the cheap HC-SR04 ultrasonic sensors, to drive the high-frequency speaker :-) Thanks for this great video!
12:55 -But Pete,the P2N2222 is an NPN device,while you have it drawn as a PNP.
Crap, did I jack that up? Try P2N2222 for the npn's and 3906's for the pnp's. What an amateur r I.
Nice video Pete! Perhaps you could replace the dev board with a simple hex schmitt inverter. Create an oscillator with one gate and buffer it with a second gate for the first clock signal. Also feed that clock into a third gate to create the anti-phase clock. That should improve synchronization of the two signals and reduce cost to almost nothing.
Can someone please answer the following question. Why was a bridge circuit necessary? Couldn't the 12v simply been switched on and of via two NPN transistors connected to the two outputs?
Thanks Very Much for great video
Great explanation!!
Pete,
Good job. Thanks for the info. I always appreciate an explanation of how things work!
For high thermal mass components or components that require more heat underneath, I use a lab hot pad like the MS-H280-Pro just underneath the board. Make sure the temp setting is below the solder melting point. Then use a soldering iron to raise the temp the rest of the way. This may be a bad method if there are a lot of low thermal mass components directly underneath the component being replaced.
- In reflow, bottom components are usually glued to prevent them from falling off during the top layer reflow process when the board is flipped.
wow really wonderful explanation crystal clear. you addressed most of my concerns thank you
congratulations for this video, very clear and well done...p.s. Gaetano Palumbo was my Professor at University...:)
Interesting, but it had me thinking: since the first stage here is most effective (relatively) since it boosts the input voltage by 100% (omitting any losses for simplicity), isn't there a way to double the voltage for every stage instead?
Wouldn't it be better to switch the bjts to mosfets? would that increase efficiency?
(made this comment before he commented about that)
what are we supposed to input to the phase 1 and 2?
same question.If you have the ans pls do reply me
awesome video! I'll try this circuit.
Instead of using 2 digital ports can I use just one digital port and a NOT gate?
but then you get 1Tp out of phase.
The right way would be to use a differential circuit so that both inverted signal are as perfectly time aligned as possible.
These videos make me realize I need to learn more math.
cool lesson
cant I just use two NPN transistors each instead of 2 NPN and 1 PNP?
Oh, i like this. Im gonna build one tomorrow, but instead of a uC, ill use a plain old multivibrator to produce my tens-of-khz squarewaves. And bat85 schottky instead of 4148s
Pete, disabling Timer 0 on arduino gets rid of jitter in the outputs because of the microseconds() update, it uses Timer 0. Also, you could modify the other timers registers to get I/O clock or PWM into the MHz range. Couldn't you just use an Sparkfun H-Bridge driver for your HV driver? usually the outputs are differential with tight timing being driven with one PWM output. Great video though!
Hi, Pete. I am currently doing the charge pump project also. But i am facing the problem about the output current.
My output current is only few mili Amps and micro Amps only.
May I know, how you maintain the current of the charge pump?
For those kind of freq/curent/voltage, instead of messing with transistor you could use mosfet driver chip, they make some with two time aligned inverted out so that solve your dephasing on the go, they usually goes up to 0.5-1A, just watch the power dissipation cause they come in small plastic package.
Talking about product design here, of course i got it that you just did with whatever you had on hand so here comes the 2n2222 ;c)
Just my 2cts...
Wouldn't it be simpler to drive each clock with a mosfet?
Gaetano Palumbo is the best full professor in electronics department of the University of Catania
nice try, Prof. Palumbo!
Power out over power in you mean.