Author Topic: To Double or not to double?  (Read 2095 times)

0 Members and 1 Guest are viewing this topic.

Offline NivagSwerdnaTopic starter

  • Super Contributor
  • ***
  • Posts: 2507
  • Country: gb
To Double or not to double?
« on: July 17, 2019, 03:46:31 pm »
I built myself a 35V Boost regulator but have discovered I would really be much better off with 70V...

I came across this (from Pile o Poo)...



Now my boost regulator looks very much like...



i.e. it has an inductor and a diode and a mosfet switch.... So with 2 Capacitors and 2 Diodes I could get my 35V to 70?

I need around 3W... i.e. 50mA at 70V

Or should I go back to the drawing board and design a 70V regulator?

Thanks in advance



 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: To Double or not to double?
« Reply #1 on: July 17, 2019, 03:49:35 pm »
Can you not just make a few tweaks to the inductor size and feedback loop to get 70V? Boost converters often give you quite a bit of wiggle room.
 

Offline MagicSmoker

  • Super Contributor
  • ***
  • Posts: 1408
  • Country: us
Re: To Double or not to double?
« Reply #2 on: July 17, 2019, 04:31:00 pm »
I built myself a 35V Boost regulator but have discovered I would really be much better off with 70V...
...
I need around 3W... i.e. 50mA at 70V

You can do this easily enough with a boost, and you can use your existing circuit if you change the feedback resistor ratio, the timing capacitor value and the boost inductor. See OnSemi AN920 for a comprehensive explanation of how to do the above.

Oh, and be wary of saddling the FB pin with too much capacitance - the '34063 is a hysteretic controller and does not need frequency compensation, so adding a capacitor to the FB pin just slows down its response to voltage over/undershoot. That said, 47-100pF on the FB can help suppress early pulse termination from noise, so don't delete it entirely.



 

Online SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15797
  • Country: fr
Re: To Double or not to double?
« Reply #3 on: July 17, 2019, 04:37:06 pm »
 
The following users thanked this post: NivagSwerdna

Offline NivagSwerdnaTopic starter

  • Super Contributor
  • ***
  • Posts: 2507
  • Country: gb
Re: To Double or not to double?
« Reply #4 on: July 17, 2019, 05:26:10 pm »
Sadly I opted for a chip TPS61175 with an inbuilt switch rated at 38V. (I said it was alike not the same  :) )

I can redesign but just wondered about the merits and weaknesses of doubling.
« Last Edit: July 17, 2019, 05:44:55 pm by NivagSwerdna »
 

Offline MagicSmoker

  • Super Contributor
  • ***
  • Posts: 1408
  • Country: us
Re: To Double or not to double?
« Reply #5 on: July 17, 2019, 08:01:00 pm »
Sadly I opted for a chip TPS61175 with an inbuilt switch rated at 38V. (I said it was alike not the same  :) )

Yeah, not really what I would describe as, "very much like..."

The switch has to withstand the output voltage, at a minimum, so you have no choice but to use a doubler or tripler. Ripple goes up dramatically, of course, but that's rarely a problem at high frequency and low load current/power. One advantage is that the coupling capacitor in the multiplier automatically provides short circuit protection, which is otherwise impossible to do in the conventional boost (you can protect the switch from overcurrent, but not the output diode).



 

Offline NivagSwerdnaTopic starter

  • Super Contributor
  • ***
  • Posts: 2507
  • Country: gb
Re: To Double or not to double?
« Reply #6 on: July 17, 2019, 08:03:58 pm »
Yeah, not really what I would describe as, "very much like..."
The principle is the same as the MC34063... and the question of the doubler?
 

Offline Buriedcode

  • Super Contributor
  • ***
  • Posts: 1718
  • Country: gb
Re: To Double or not to double?
« Reply #7 on: July 17, 2019, 08:16:38 pm »
You can certainly add a charge pump voltage doubler, but its voltage will sag under the 50mA so you won't get 70V.  But its worth trying it, and seeing what boost voltage you'd need to get 70V out the charge pump (it'll be over 35V), then if its low enough, you can just connect the feedback to the charge pump output. It only costs a couple of diodes and a couple of caps.
I would add a low value resistor in series with the fly capacitor, as sometimes hooking a charge pump to a boost converter can stop it switching.

Alternatively, coupled inductors (1:1) are easily obtained, and will give you double the voltage output (or rather, half the peak switch node voltage) if you use it in an auto transformer config.  You could also try other ratios, but these are harder to find. A 1:2 for example would mean the peak switch voltage is ~23V.
 
The following users thanked this post: NivagSwerdna

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: To Double or not to double?
« Reply #8 on: July 18, 2019, 02:05:25 am »
Charge pumps work great when driven by a boost switching regulator.  Move the feedback network to the output of the charge pump to compensate for diode and capacitor losses.

Beware however when operating close to the switch's maximum voltage.  Making up for losses in the charge pump will raise the voltage seen by the switch.
 
The following users thanked this post: NivagSwerdna

Offline NivagSwerdnaTopic starter

  • Super Contributor
  • ***
  • Posts: 2507
  • Country: gb
Re: To Double or not to double?
« Reply #9 on: July 20, 2019, 10:43:10 am »
Total success and a nicely glowing IV-18.  I am getting 69.7V out of the doubler.  Perfect!

Thanks
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf