New to Coilguns - Coilgun Buildlog (Charger apparently fixed) / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

New to Coilguns - Coilgun Buildlog (Charger apparently fixed)

first previous 3 4 5 6 next last

Move Thread

LAN_403

Artikbot

Wed Sept 29 2010, 12:57PM

Registered Member #3247 Joined: Mon Sept 27 2010, 09:42AM
Location: Spain
Posts: 137

Okay, so I gotta hook a potentiometer right after my fixed resistor, isnt it? The higher the value, the higher the frequency, therefore the lower the current. sounds plausible, I'll experiment a bit with that.

That is what happens!! My MOSFET and the diode right after the inductance blow up as soon as the capacitor reaches the inductor output voltage. But somehow the 555 doesn't manage to make an oscillating pulse, seems to be in a monostable state instead of an astable... thb, don't know why, I'm following the astable wiring diagram... I think I'll just rewire the whole thing again, the circuit is really messy now.

So, to sum up, my MOSFET should open-close-open-close-open-close etc. the inductor's input , then my diode captures that current, converts it into a direct current and it goes right into my caps at rated voltage.

Okay, so my current problem is fixing the output of the 555 timer. Because my caps get charged to input voltage, what sucks because it is not what I'm wanting to do.

Thanks a lot!

ScotchTapeLord

Wed Sept 29 2010, 01:08PM

Registered Member #1875 Joined: Sun Dec 21 2008, 06:36PM
Location:
Posts: 635

No, the higher value, the lower the frequency. The output is based on how long it takes to charge and discharge the timing capacitor, so the higher the value, the longer it takes, the lower the frequency. This is why the maximum value of the potentiometer PLUS the value of the series resistor must be the correct value for 10kHz switching. That way, you can only turn the resistance lower than that which is required for 10kHz, which will yield higher frequencies.

If your capacitor bank is only charging to the input voltage, then there's something wrong with your astable multivibrator OR your transistor is not turning off/not turning off fast enough.

Artikbot

Wed Sept 29 2010, 01:14PM

Registered Member #3247 Joined: Mon Sept 27 2010, 09:42AM
Location: Spain
Posts: 137

Aha. So lower the capacitor, higher the frequency. Higher the resistor, more time takes up to load the capacitor, I guess... Okay. I understood it the other way >.<

Got the problem. I was messing up with resistors, input-output pins, and overall pins were a bit messed up. I'm rebuilding the circuit and so far the timer works properly. Let's see what happens when I add the rest of the components.

Edit: I got the circuit working!! But I don't reach the needed voltage for the caps. According to visual deductions, my MOSFET is way too slow switching. Let's see if Yenka has fast switching MOSFETs.

This is the working diagram w/o a fast enough MOSFET. I can't test it because Yenka doesn't have fsat enough MOSFETs. It is a P-Channel, N-Channel switches way faster but blows up... So far with the slow MOSFET it charges up the capacitors to 21.4V. I can upload the file too if you want to play with it.

And this is the Yenka file with the completed circuit.

Thanks!

ScotchTapeLord

Wed Sept 29 2010, 02:45PM

Registered Member #1875 Joined: Sun Dec 21 2008, 06:36PM
Location:
Posts: 635

Ahh, P channel won't work with this configuration. The inductor is in the way of the path for turning the FET on. You need to use N channel. It sounds like you've almost got it working properly... I think you can attribute the MOSFET failure to the shortcomings of the program.

A few things I caught in the schematic:

The 1Kohm resistor in series with the antiparallel diode to the final solenoid is too high. It should be lower, like 10 ohms, or omitted completely.

The SCR's gate is triggered by positive voltages, I believe. Close the circuit with the switch on the ground and put the bottom leg of the 47 ohm resistor (this value will almost definitely need to be higher depending on what SCR you are using) and connect it to the capacitor's + side, with the switch between the resistor and SCR (or on the other side of the resistor, doesn't matter).

You need some decoupling capacitors. The simulator probably won't know the difference so you can just put a few dozen uF directly in parallel with the top of the main ON switch and GND, but in reality you need to have the decoupling physically/electrically close to the 555 timer.

Artikbot

Wed Sept 29 2010, 05:14PM

Registered Member #3247 Joined: Mon Sept 27 2010, 09:42AM
Location: Spain
Posts: 137

Ooohmkay!

Yep, the N-Channel does switch fast enough.

But my new problem is that the MOSFET blows up because gate voltage reaches 120ish values, eventually 300ish. How? I don't know, maybe leakage from S to G? At least the diagram of the N channel shows an arrow going from S to G... How can I avoid that uber excessive currents that I wasn't getting with the P channel MOSFET?

Here it goes the new diagram with everything turned in order to place the N channel with more ease.

I suppressed the SCR from this circuit, it was only introducing problems. When everything else is working I'll put the SCR again.

Thanks!

DerAlbi

Wed Sept 29 2010, 06:14PM

Registered Member #2906 Joined: Sun Jun 06 2010, 02:20AM
Location: Dresden, Germany
Posts: 727

Since you are not really the man who finds good searchword for google and keeping your electrical experiences in mind, i would recomend something different.

Its much better if you use well tested circuits than experiementing with ICs and circuits where you "know" what everything does but where you do not _understand_ the electrical behavior. The posted link makes use of an transformer. So your caps are potential free. One reason less you kill yourself

Also _understanding_ the multivibrator with 2 transistors is much easier because google tells you verry much about this circuit if you ask it for 2 blinking leds

Artikbot

Wed Sept 29 2010, 06:38PM

Registered Member #3247 Joined: Mon Sept 27 2010, 09:42AM
Location: Spain
Posts: 137

Oh lol, that's definitely easier than the design I thought about :P

Yes, the principle of it all is much more understandable. It seems more efficient to me too, since transformers are more efficient than just inductors.

So, it will pump voltage up to 300V? Can it be pumped a bit further to 330 to get the maximum juice possible from my cap bank? Maybe increasing the secondary turns nÂº it'll do the trick :)

The main reason why I am experimenting with 555 is because I can get a regulable output, and this way I also make some sort of introdcution to IC based electronics, and to some concepts I never thought about.

I'll simulate the design and see what happens if I play with values a bit :)

Thanks!

DerAlbi

Wed Sept 29 2010, 06:58PM

Registered Member #2906 Joined: Sun Jun 06 2010, 02:20AM
Location: Dresden, Germany
Posts: 727

I see. If you want a regulated really regulated output this design is indeed not good. BUT: there are some things you can do: This circuit will not draw excessive currents. Maybe 10A or so. (?)
Problem is: when your Cap is empty, the transformer is practically shorted. This can really be a problem.
To avoid this:
1) put a big bipolar capacitor in seieres with the transformer. That limits the energy for every cycle.
2) Do 1Ohm in your power supply line. This would also work.
3) As you know

3 = 1+2

Then: you can swith off your power supply automatically when a special voltage is reached.
A simple solution would be a optocoupler with its diode is powerd by the HV-caps. the other sige drives a Mosfetgate that enables or disables the whole converter by cutting the powersupply

Easy

Artikbot

Wed Sept 29 2010, 08:06PM

Registered Member #3247 Joined: Mon Sept 27 2010, 09:42AM
Location: Spain
Posts: 137

Okay I'll do this non-regulated for now to serve the purpose and later make a regulated boost converter. If i'm in need of, call it because I get bigger caps or because I want more power.

Ima do this because I got the transformer right here, the resistors and transistors can be (hopefully) bought by a crappy amount of money in the same place I'm getting my SCRs.

So, if I build it exactly like it's built on the paper, I should be getting the desired 300V output... I figure it will be enough. My current charger barely reaches 290 :P

Shorted transformer with empty caps? Okay, a big bipolar capacitor. Got it.

But hey... Searching a bit further the transistors, are REALLY hard to find and they cost almost as much as my 40A 700V SCR! Holy jeez that's expensive! I think I'll keep going with the timer design just for cheapiness. I need a cheap yet powerful solution, and all 4 transistors cost the same as my 2 SCRs... not cheap at all imo :/

Thanks anyways :)

DerAlbi

Wed Sept 29 2010, 08:23PM

Registered Member #2906 Joined: Sun Jun 06 2010, 02:20AM
Location: Dresden, Germany
Posts: 727

Here is a scematic that shuts down at a predefined potential. So you just need a transformer that makes 500V and adjust the resistors so, that your wanted voltage will generated.

Note than some Transistors become hot. Partnames mean nothing. Its just a simulation with LTSpice.
Hot parts are Q3 and Q4 when charging, and later M1 when in IdleMode wehn everything is charged.

Current resistor config shuts down at exactly 30V.

first previous 3 4 5 6 next last

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