Author Topic: 12V-400V 10mA HV DCDC boost with softstart into 150uF 400V of output capacitance  (Read 1175 times)

0 Members and 1 Guest are viewing this topic.

Offline PsiTopic starter

  • Super Contributor
  • ***
  • Posts: 10523
  • Country: nz
Does anyone have recommendations for an ultra long soft start mod to get this circuit to work with a huge amount of output capacitance, like 150uF.

The circuit works fine for around 8mA load at 400V out but no matter what mods I try I cant get it to start unloaded with 150uF on the output.

I've tried the R/C + diode trick on feedback pin for adding soft start with no luck.

The soft start that's built into the IC is useless for adding this sort of ultra long soft start because that cap also sets the shutdown retry repeat rate.
If I set SS really long the converter only tries to start every few seconds.  Need both ultra long soft start while keeping retry rate fast.

Datasheet
https://www.ti.com/lit/gpn/TPS40210

 
Greek letter 'Psi' (not Pounds per Square Inch)
 

Offline Fryguy

  • Regular Contributor
  • *
  • Posts: 160
  • Country: de
You could use the old "current limit resistor and time relay" method . . .    :-/O

But i think you need to limit the output current anyway with such a large capacity - or your boost converter will stop working every time your capacitor gets more or less drained .
« Last Edit: December 27, 2023, 01:02:40 am by Fryguy »
Born error amplifier  >.<
 

Offline PsiTopic starter

  • Super Contributor
  • ***
  • Posts: 10523
  • Country: nz
I tired a few current limiting resistors but they either burnt up, or didn't work, or wouldn't allow enough current through once everything was charged.

But yeah, I've not tired a series resistor with timed short circuit.
Will give that a go.
Greek letter 'Psi' (not Pounds per Square Inch)
 

Offline Wolfram

  • Frequent Contributor
  • **
  • Posts: 422
  • Country: no
A current mode boost into a voltage multiplier will have some issues to begin with. The controller assumes that MOSFET current rises linearly during the on-time, and the pulse is terminated once it reaches a certain current as given by the voltage regulation loop. When driving a voltage multiplier off the switching node, the circuit will act as a charge pump when the MOSFET turns on, giving an uncontrolled current pulse that could terminate the on-period prematurely. In your case, the problematic current loop is R2-Q1-C19-D6-C9.

Personally I've had much better success with tapped inductor boost or flyback for converters like this. Standard cheap 1:1 coupled inductors like the SRF0703 work well for tapped inductor boost, with much better efficiency than a simple boost, and cheap off-the-shelf mains flyback transformers like 230 V - 12 V 5 W work very well both for tapped inductor boost and flyback.
 
The following users thanked this post: Someone, prosper

Offline Fryguy

  • Regular Contributor
  • *
  • Posts: 160
  • Country: de
Now i remember , i did a flyback circuit using a 230V / 6V mains transformer to generate a 450V output for a mobile tube psu - more than 30 years ago  :)

Please don't ask me for details - all i can remember is that it had 1 power transistor , a CD4093 ic and was adjustable for about 400-600 V output .
Born error amplifier  >.<
 

Offline PsiTopic starter

  • Super Contributor
  • ***
  • Posts: 10523
  • Country: nz
hm.. i might try remove the voltage multiplier and run the thing at 400V directly, just to see if that works any better.

Will have a look at tapped inductors, thanks.
Greek letter 'Psi' (not Pounds per Square Inch)
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22435
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Dump it, 40210 is trash.  The killer for me is that, fault is triggered by error amp output. And the error amp output range can cross the fault threshold under normal circumstances (high load, transients). Triggering reset and SS cycle.  It's a dumb design.  I wouldn't go so far as to accuse them of letting the appnote authors into the IC design side, but insinuating something to that effect gives you some idea what I think of the thing...

LM3481 is a reasonable upgrade as a more modern UC3843 though.

As for the circuit, you can't pair a current-mode boost with a capacitor-switched (charge pump) multiplier.  You can however stack SEPIC-style stages, but better is to use an offline flyback transformer in reverse.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: Psi, doktor pyta, moffy

Offline Martinn

  • Frequent Contributor
  • **
  • Posts: 364
  • Country: ch
Maybe look for "photo flash capacitor chargers", like
https://www.analog.com/en/design-notes/photoflash-capacitor-charger.html
https://www.ti.com/lit/an/slva330/slva330.pdf

also, if you consider buying a module, here https://www.aliexpress.com/item/1005003152784850.html they advertise it as capacitor charger. Don't know if this is true, I don't have a HV capacitor here to try it out. See also recent discussion here (with link to schematic) https://www.eevblog.com/forum/projects/12-24v-to-400v-dc-to-dc-boost-converter-help/msg5237703/#msg5237703
« Last Edit: December 27, 2023, 09:16:54 am by Martinn »
 

Offline PsiTopic starter

  • Super Contributor
  • ***
  • Posts: 10523
  • Country: nz
Maybe look for "photo flash capacitor chargers", like

That's a good idea!



Greek letter 'Psi' (not Pounds per Square Inch)
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf