Boost converter woes. / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Boost converter woes.

1 2 next

Move Thread

LAN_403

aahz

Sun Mar 12 2006, 05:06AM

Registered Member #186 Joined: Thu Feb 16 2006, 07:22AM
Location:
Posts: 42

I'm not sure what I'm doing wrong here, but my boost converter setup doesn't seem to be working at all. Here's the first schematic I tried.

http://img105.imageshack.us/img105/1464/boostconverter5kp.jpg

After building it, ripping it apart, and rebuilding it a couple of times (and losing an igbt and a few resistors in the process) with no luck, and screwing with all the values I can vary (r4 and r5 from 10k to 0, C2 from .01 to .068 to .136uf) All I can get out of it is 5.5-6.6v output. I've added resistors, removed resistors, re-wired it, etc. With one of the first builds, I managed to get about 16 volts charged to a cap, but promptly burned out my igbt.

Oh yeah, and I've replaced the igbt, diode, resistors, 555.... well pretty much every componant at least 3 times (except the inductor).

Could someone point out something extremely dumb that I've done, or post a similar design that I could judge mine off of?

Thanks.

williamn

Sun Mar 12 2006, 05:11AM

Registered Member #55 Joined: Thu Feb 09 2006, 04:56AM
Location:
Posts: 149

First, you need to protect the 555. Use a Vreg with some low value ceramics or tantalums across the inputs and outputs of a common 78xx regulator. Do this for sure. How are you going about constructing the inductor? This circuit seems to be hard for new commers to get right, but once you get it working it should be reliable.

aahz

Sun Mar 12 2006, 05:32AM

Registered Member #186 Joined: Thu Feb 16 2006, 07:22AM
Location:
Posts: 42

I'm using a 12v battery for the 555... but I'll try caps across the imput. The inductor is manufactured. A toroid around a ferrite core. 680uf +/- 10%, around 3 amps continuous (looks to be about 18awg wire). Should I wind my own inductor?

Edit: Success! A bad diode was apparently killing off my igbt's... or at least, I replace the diode and IGBT (again) and it works now. Already getting up to 300v... but now I hear hissing/arcing on the breadboard, so it's about time to start soldering it seems.

Edit: Arrggg. I change nothing, and now it doesn't work at all. I replaced the diode and igbt again, now there is a lot of arcing when I contact the battery, and the IGBT heats up immediately. Replaced the 555 again. Nothing. The cap will charge to 12v, equal to the battery, and nothing more.

This is really frustrating. Time to rip it apart again, I guess.

Edit 3: Changed 555 and changed resistors to 100 ohms and 10 ohms... and it started working! Charged up to about 100v. Changed the 100 ohm to 1k ohm and it charged to about 320v (470uf) within seconds! .... and now it doesn't work anymore. I'm guessing I burned out.... hell I don't know anymore. I'm running out of 555's and igbts..

I guess I'll have to wait another week or two for more parts.

hvguy

Wed Mar 15 2006, 06:48AM

Registered Member #289 Joined: Mon Mar 06 2006, 10:45AM
Location: Conroe, TX
Posts: 154

The capacitors williamn suggested are necessary, battery or not. They are there not to filter out 60hz but to â€œbypassâ€ any rf currents on your power supply rails. The capacitor should be anything from .1uf â€“ 4.7uf and placed directly across pins 8/4 and 1 on the 555. It would also be a good idea to put another from the positive side of L1 to ground. The lack of bypass capacitors is most likely your problem. What is you L wound on/with? D1 shown in your schematic is not needed.

aahz

Wed Mar 15 2006, 10:23AM

Registered Member #186 Joined: Thu Feb 16 2006, 07:22AM
Location:
Posts: 42

Thanks for the advice.

My inductor is a toroidal with a ferrite core, wound with what looks to be 18 guage magnet wire. (it's about 1" in diameter). Anyway, I burned out my four 555's, and surprisingly only two igbts (of six), with more 555's on the way (30 cents each makes 'em easy to burn up).

Oh, and I never put D1 in the actual circuit, but thanks for the catch.

Dr. Shark

Wed Mar 15 2006, 12:01PM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

In the meantime, while you are waiting for new 555s, you could try to make the circuit self-oscillating, blocking oscillator style like the joule-thief.

rupidust

Thu Mar 16 2006, 02:28AM

Banned
Registered Member #110 Joined: Fri Feb 10 2006, 12:23AM
Location: Banned City
Posts: 85

When building Boosters its ever so important to view the signal at the IGBT gate. If the gate signal is noisy, triangular, or too high of a duty cycle > 90%, the IGBT will soon degrade to the point of failure. Which IGBT are you using? Which Diode are you using? The resistor at gate can be bumped up to 100 ohms untill the circuit works out ok. At 10 ohms, the IGBT may ring alot. What frequency is the gate signal or goal frequency? For now, keep the two 555's resistors close such as 1.5k, 1k or 4.7k, 3.3k From what I see, you could be pushing pass 85% duty when R4 is 1k and R5 < 215 ohms. If you adjust those pots too low namely R4, the 555 will burn out at pin 7, which could cause IGBT to latch and burn out, any regulators to burn out and so on.... Insert a fixed value resistor in series with R4 and R5 during testing like 500 ohms. There is no need to make R4 or R5 less than 500 ohms. D1 is not needed.

aahz

Thu Mar 16 2006, 04:12AM

Registered Member #186 Joined: Thu Feb 16 2006, 07:22AM
Location:
Posts: 42

I'm using FFPF30U60S for my diode, and HGTP20N60A4 as my igbt.

I'll be using a smaller cap for timing so I can keep the total resistance higher... probably >2k. In order to get 450v from 12, the duty cycle needs to be extremely high... something like 97%. Also, the frequency needs to be pretty high, too, in order to limit current, to an extent, through the inductor (I think)..... I'm planning around 80-100khz.

And again, I understand D1 is pointless (When I first put it together I realized I was in essence protecting the battery, which makes no sense), it never made it into the circuit.

hvguy

Thu Mar 16 2006, 08:55AM

Registered Member #289 Joined: Mon Mar 06 2006, 10:45AM
Location: Conroe, TX
Posts: 154

Those are some nice parts, both of which I have used in commercial desgins. Once you get it working right you should never kill either part. You might want to try ebnay for your 555s... I picked up 250 from a guy for $5.

rupidust

Thu Mar 16 2006, 10:35AM

Banned
Registered Member #110 Joined: Fri Feb 10 2006, 12:23AM
Location: Banned City
Posts: 85

aahz wrote ...

In order to get 450v from 12, the duty cycle needs to be extremely high... something like 97%. Also, the frequency needs to be pretty high, too, in order to limit current, to an extent, through the inductor (I think)..... I'm planning around 80-100khz.

The parts are good. But do not use 97%. Guaranteed thermal run-away failure. As for the frequency: Inductive reactance = 2 * pi * 680uH * 80kHZ = 341 ohms. 12v / 341 ohms = 35mA, which ends up being only 0.420 Watt consummed and at 80% efficiency will be a 0.336 Watt charger. Will be even less if the IGBT voltage drop is included. Base the frequency on the inductors value. No need to go above 80% duty. 12v will charge beyond 450v even at 50% duty.

1 2 next

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