Simple and robust flyback voltage converter / Electromagnetic Projectile Accelerators / Forums

Forums

4hv.org :: Forums :: Electromagnetic Projectile Accelerators

« Previous topic | Next topic »

Simple and robust flyback voltage converter

previous 1 2 3 next

Move Thread

LAN_403

Yandersen

Fri Aug 23 2013, 03:57AM

Registered Member #6944 Joined: Fri Sept 28 2012, 04:54PM
Location: Canada
Posts: 340

The Q2 is essential. The reversed voltage on a non-polar cap may stay infinitely. Due to that problem I have already burned a dozen of parts forgetting to unload the repolarized cap. Right now I'm considering the option of coupling two chargers to use caps in both polarities, reversing it each next shot, this way I can use 1 cap per each stage eliminating the need of coupled stages which reuse leftover energies.

Well, if you want to present a schematic based on other IC - go on moderning the circuit, maybe it will be better then mine. But I will first try to build my idea as I already ordered the parts.

Yandersen

Tue Aug 27 2013, 09:08PM

Registered Member #6944 Joined: Fri Sept 28 2012, 04:54PM
Location: Canada
Posts: 340

Tested some parts of the design in practice. Guys, need help to find IC with some built-in functions.
The ideal IC must have the following functions:
1) 2 FET drivers;
2) at least 1 voltage comparator able to operate near the ground level with <1us response time (including 0-Vcc output swing);
3) undervoltage protection - outputs of the FET drivers must not go high if supply voltage is below 6-10V.

Any suggestions?

Yandersen

Mon Sept 16 2013, 09:51PM

Registered Member #6944 Joined: Fri Sept 28 2012, 04:54PM
Location: Canada
Posts: 340

After the practical tests some updates are made to the schematic:

Two chargers built by the schematic above are planned to be connected antiparallel to charge non-polar caps both directions:

1379368291 6944 FT156318 Commutation 2chrgrs

Yandersen

Wed Sept 18 2013, 12:21AM

Registered Member #6944 Joined: Fri Sept 28 2012, 04:54PM
Location: Canada
Posts: 340

Ordered two EPCOS' p-cores to make a couple of transformers. Here is what I got:

1379463699 6944 FT156318 Fucking Monkeys

The first one who will guess right what is wrong on that picture, will have a candy! :)

Yandersen

Sat Oct 19 2013, 07:15AM

Registered Member #6944 Joined: Fri Sept 28 2012, 04:54PM
Location: Canada
Posts: 340

Finally got rid of all bugs in the concept. The final schematic is the following:

Yandersen

Sat Oct 19 2013, 07:21AM

Registered Member #6944 Joined: Fri Sept 28 2012, 04:54PM
Location: Canada
Posts: 340

My pcb board (7cm x 5cm):

This is a 780V converter version of the schematic above.
It gave me 126W output at 751V with resistive load (light bulbs) consuming 138.6W from the supply (90.9% efficiency). All the features are confirmed to be working fine, no power-up and shut-down glitches.
Cap 60uF charged to 760V in 0.15 seconds. It gives average power for capacitive loads 115W.
Hm, 2 and a half month designing the charger...

Thomas W

Sat Oct 19 2013, 10:49AM

Registered Member #3324 Joined: Sun Oct 17 2010, 06:57PM
Location:
Posts: 1276

Very nice project here!

how did you manage to solder on that cheap chinese matrix board? i have a load of those... couldnt use them as ever time i tried to solder onto them, the copper track pads keep coming off :c

Yandersen

Sun Oct 20 2013, 04:55PM

Registered Member #6944 Joined: Fri Sept 28 2012, 04:54PM
Location: Canada
Posts: 340

Yeah, that is true - desoldering and replacing parts leaves empty spaces underneath. That is why I used a new matrix board for every other version (good I have hundred of those little boards - still didn't ran out).
To make tracks I just bend legs of the resistors under the board forming tracks. At some point it is even better then normal pcb approach as lines are formed with relatively thick copper. :)

Steve Conner

Mon Oct 21 2013, 09:34AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Well done, good efficiency! :) What is the function of Q2 in your circuit?

Yandersen

Mon Oct 21 2013, 02:35PM

Registered Member #6944 Joined: Fri Sept 28 2012, 04:54PM
Location: Canada
Posts: 340

This is a protection from cap repolarization. See, as I am using non-polar caps without reverse diode, when I shot, cap discharges onto coil, then coil charges cap with hundred of Amps in reverse polarity. Standard charger, if not disconnected, will blow. My charger actively determines cap repolarization by the increased output current, and shuts down, closing the Q2. Every 30-40ms it opens Q2 for couple us to check if cap is still repolaried. It draws tiny amount of energy from cap. The Q2 closes permanently when supply voltage drops below 10V or so as when supply disconnected, so it is safe to shut down or turn on the gun with caps in reverse polarity.

Making further testings I discovered that no-cap-in-load protection (when the output wire broken and charger outputs into air) doesn't work that well. Albi recommended me to have cap on charger side, like a minimum load, but it is not that easy to implement in my design due to existence of Q2. Working on it now...

The efficiency in my setup may be couple percent higher when I rewind the trans in more optimal way. Measuring current output instead of input current (which is more common, as I see) is much more efficient and losses on that are negligible. In my setup, ~2V are taken from output voltage for current measuring purposes, which contributes to <0.3% efficiency drop on 780V output. All other losses (output diode, MOSFET, transformer, driver) are the same unavoidable as in any other flyback setup. Using the Silicon Carbide Schottky diode (no reverse recovery current) on output allowed me to implement continuous current mode to lower resistance losses by 25% due to smaller peak currents for the same output power comparing to discontinuous mode.

previous 1 2 3 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob