"Thumper" / Projects / Forums

Forums

4hv.org :: Forums :: Projects

« Previous topic | Next topic »

"Thumper"

first previous 15 16 17 18

Move Thread

LAN_403

colin heath

Sun Nov 14 2010, 08:24PM

Registered Member #123 Joined: Fri Feb 10 2006, 12:58PM
Location:
Posts: 162

Finn fantastic work as always and the attention to detail is outstanding.

I have managed to steer clear of coils for around 7 months now but seeing this has really got some creative juices flowing!

Once I get my Hillclimb car finished then a rather large coil is in the making I feel:-D

Tesla Fan

Mon Nov 15 2010, 10:19AM

Registered Member #3353 Joined: Sat Oct 23 2010, 11:21PM
Location: Greece
Posts: 90

I would just like to add my compliments, this is one of the best man-made â€œconstructionsâ€ (of any kind) i have ever seen! A truly amazing project, your engineering is above professional level!

Finn Hammer

Sat Dec 18 2010, 01:20PM

Registered Member #205 Joined: Sat Feb 18 2006, 11:59AM
Location: SkÃ¸rping, Denmark
Posts: 741

Thumper may not care much, but *I* never got at ease with the gate drive transformers I had to resort to, in order to get her running in time.
Recently, I have been working on new gate driver boards, and thus this little report.

I wanted to use an all N-channel mosfet half bridge, and to do this with opto isolated input, I had to sequence the mosfets with a half bridge driver chip. Many nice ones to chose from, I selected the LM5104, due to it's inherent suitedness. It is a one input driver with programmable deadtime.

the present schematic looks like this:

As you see, the gates of the stack needed to be calmed down some more, to get nice waveforms all way round.

Here is a late prototype set up for celebrity photography:

"A smd prototype soon aquires the appearance of a sea urchin when equipped with test points at key nodes."

At this stage, I dumped the zener clamps on the board supply, and opted for full linear regulation, which lead to this present prototype:

Not fully populated, it essentially does what the previous versions do, put out nice waveforms..

Yellow trace is gate voltage at top halfbridge mosfet, Q1, 8V/div
Green trace is voltage at gate of bottom halfbridge mosfet, Q2, 8V/div.
Cyan trace is current into gate of CM600, 2A/div.
Magenta trace is volage on Gate of CM600, 10V/div.

You will see that Q1 is in essence performing the double task of switch and gate resistor, most of the current into CM600 gate being delivered while it's gate is still in the miller plateau.

It turns the CM600 on in less than 500nS, without ringing or overshooting the gate voltage, and without hesitating at the miller plateau.

In my opinion, that is the ultimate gate drive.

Whether this is right will become the topic of my next installment.

With this, I wish you all a merry christmas.

Cheers, Finn Hammer

Dr. Drone

Sat Dec 18 2010, 05:23PM

Registered Member #290 Joined: Mon Mar 06 2006, 08:24PM
Location:
Posts: 1673

Dr Hankenstein

Sat Dec 18 2010, 07:03PM

Registered Member #1642 Joined: Sun Aug 17 2008, 11:36PM
Location: Black Canyon City
Posts: 96

Nice job, Finn. Your mind just never stops. I must compliment you on your persistence and innovation; always a cut above the rest!

Finn Hammer

Sun Dec 19 2010, 07:11PM

Registered Member #205 Joined: Sat Feb 18 2006, 11:59AM
Location: SkÃ¸rping, Denmark
Posts: 741

The only parts that didn't make it in time were the rectifier diodes for the input bridge, I guess not much can go wrong there, so the presumption is that it woikd

What else then, than to initiate "big etch evening", where I also had time to do that burst interuptor board for Daniel which you see in the middle.
I use photoresist stock, print artwork on laser transparencies, have a UV lightbox, develop shamelessly with drain cleaner in water, etch with one third each of hydrocloric acid 30%, peroxide 35% and water. Works for me.

Wow, this is good winter hobby.

Cheers, Finn Hammer

Daniel Uhrenholt

Mon Dec 20 2010, 06:26AM

Registered Member #125 Joined: Fri Feb 10 2006, 01:52PM
Location: Aalborg, Denmark
Posts: 155

Looking good Finn.

You have always been productive this time of year, but there isn't much else to do because of the weather...:)

I see that you made the interrupter PCB for me, so you have to drop by some day for coffee.

Cheers, Daniel

Steve Ward

Thu Dec 23 2010, 05:44PM

Registered Member #146 Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055

It turns the CM600 on in less than 500nS, without ringing or overshooting the gate voltage, and without hesitating at the miller plateau.

The gate drivers i designed for CM600s showed <100nS rise time on the gate... its of course bogus because you arent measuring the voltage at the gate, rather you are measuring the voltage at the driver and there is plenty of module inductance and resistance between it and the gate electrode. So unless you measured the collector current change, i wouldnt claim any switching speed based on what the external gate electrodes show.

first previous 15 16 17 18

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