CM300DU-12NFH Full Bridge / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

CM300DU-12NFH Full Bridge

1 2 3 next

Move Thread

LAN_403

Dr. Brownout

Fri Apr 13 2012, 02:10PM

Registered Member #2405 Joined: Fri Oct 02 2009, 12:59AM
Location: Wisconsin
Posts: 140

I need to pick some brains out here, I am running a full bridge of the subect CM300 Bricks. Using Steve wards 1.3b UD. The wave forms on the gates look very good, however after about 25 seconds or so the 24Volt Regulator seems to be getting extremely hot...to the point of shut down.

It appears these bricks have an enormous amount of gate charge. At 24V gate voltage probably around 2600nC according to the data sheet. That would bring the capacitance around 104 nF or so if my calculations are correct. The switching frequency is around 112Khz, which is really more than it should be.

I was wondering if anyone else has ever used these identical Bricks and if they experienced similar results.

I did have the coil running but it did shut down after the regulator kicked out.

Here is a photo if the gate signal at the bridge:

My Gate resistor is 5.1 ohms and they get very hot as well. I am thinking I am probably drawing over 7 watts across those resistors.

But anyway I am looking for some input on driving these CM300DU-12NFH bricks if anyone has any experience with these, I have 6 of them so I thought I would give them a shot and see what happens.

Here is a picture of the gate drive signal at the bricks terminals.

Thanks for any input.

--Brian

1334326254 2405 FT0 Good Gate Drive Signal

Kizmo

Fri Apr 13 2012, 03:26PM

Registered Member #599 Joined: Thu Mar 22 2007, 07:40PM
Location: Northern Finland, Rovaniemi
Posts: 624

Just add cooling to the regulator. Same happened in my induction heater (which is also based on wards drsstc driver). TO220 cased regulator cant deal with much more than 2W of heat.

NFH series is great for tesla coils :)

Dr. Brownout

Fri Apr 13 2012, 04:10PM

Registered Member #2405 Joined: Fri Oct 02 2009, 12:59AM
Location: Wisconsin
Posts: 140

Hmmm I will put a sink on it and see what happens here, just seems like it should not get that hot. But then again these bricks do seem to have a large gate capcitance.

UPDATE:

Well I gave the heat sink and good amount of thermal grease a shot, and still too hot, even the gate resistors are still too hot. The Half bridge fets on the driver board are not over heating FETs are cool so they are not stressed it seems, but the GD Resistors get hot and the regulator still gets to hot.

So I swaped out the 24V regulator and put in a 15V regulator thinking it may drop the Gate charge somewhat and that did not seem to work well either. Still a heating problem on the GD resistors and the regulator.

Its just hard to believe that I am drawing more than 5 watts across those gate resistors at 15V gate drive voltage. I know other people have used this same driver on CM300's ...maybe not the DU-FNH series though, seems like most of the CM300's I have seen in DRSSTCs are the H series which have about half the gate charge as the ones I am trying to use.

I would like to know if anyone has used these FNH Bricks in full bridge and got them to work ok? Like I said earlier the coil was working really well, nice sparks but the regulator problem is shutting it down

--Brian

Adrenaline

Fri Apr 13 2012, 06:37PM

Registered Member #235 Joined: Wed Feb 22 2006, 04:59PM
Location:
Posts: 80

If your gate signal is from 0 to 24vdc, then yes I believe you are drawing 7W per channel. 24V * 112e3Hz * 2.6e-6C ~7W.

15vdc gives 4.3W. What regulator are you using?

Dr. Brownout

Fri Apr 13 2012, 07:35PM

Registered Member #2405 Joined: Fri Oct 02 2009, 12:59AM
Location: Wisconsin
Posts: 140

I am using LM7824, with 24 Volts I did calculate 7 Watts or so across the resistor but at 15V on the gates that should have brought it down enough where the 5 watt resistors should not be getting hot. But they get hot even at 15 drive.

I put a heat sink on the regulator but it still gets to hot, I actually put it on the 15 volt regulator I did not change it back to 24 volts yet. But I know it will get hot with the 24 in there.

I am just not sure how other folks are able to run these big bricks or at least this brick without an isolated supply but there must be a way.

I am almost tempted to make up an isolated supply but there must another way to make these work in here. The controller seems to be functioning just fine I actually built up two of these controllers so I swapped it out with the other one thinking maybe I did something foolish when I assembled it but the signals just look to good and it reacted the same wasy with the other board so I know the board is good and the controller is working ok.

So I am open to suggestions if anyone has any. Maybe someone out here Tesla land has had a similar issue.

Thanks
--Brian

teravolt

Fri Apr 13 2012, 07:46PM

Registered Member #195 Joined: Fri Feb 17 2006, 08:27PM
Location: Berkeley, ca.
Posts: 1111

if you add a cuple thousand uf in prallel it might help average it out. if your transformer is saturating then it will act like a short. another option is to use a k package for your regulator or a emiter follower

Dr. Brownout

Fri Apr 13 2012, 08:09PM

Registered Member #2405 Joined: Fri Oct 02 2009, 12:59AM
Location: Wisconsin
Posts: 140

Thanks for the input, I am thinking if the transformer was saturating I would see the droop in the wave form when that happens, I did though wind aseveral different GDT for it I am using a 12 turn 1:1 well actually five windings but all have the same number of turns. The measured inductance of each winding is ~480uH.
The leads from the GDT to the IGBTs is a little longer than I wanted them however I dont think that is causing the issue I really see no ringing in the gate signal.

I guess what baffles me most is I know others have run these IGBTs with this driver and it does seem to run these power hog IGBTs but it wants to complain about it at the same...:-p

Dr. Brownout

Fri Apr 13 2012, 08:45PM

Registered Member #2405 Joined: Fri Oct 02 2009, 12:59AM
Location: Wisconsin
Posts: 140

Just so you can all see the coil I am talking about here is a picture so you can see the layout.

...

Fri Apr 13 2012, 09:49PM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

What voltage are you putting into the regulator? Keep in mind that for a linear regulator your efficiency is voltage out/voltage in, so if you are drawing 7w and powering from 30v you will be burning 5w in the regulator (which will make it run quite warm!), and the situation does not improve by putting in a lower voltage regulator (at least not much).

Dr. Brownout

Fri Apr 13 2012, 10:05PM

Registered Member #2405 Joined: Fri Oct 02 2009, 12:59AM
Location: Wisconsin
Posts: 140

I am using a 24VRMS transformer so peak would about 33V. If I dont connect the bricks the regulator of course stays nice and cool but as soon as the even one half of the brick is connected it gets hot pretty fast. Yes lowering the gate voltage did help some but not enough to make it a final change. I guess I just wasnt sure if this regulator would handle the CM300, at least not these bricks, I did notice the CM300DU-24H has about half the gate charge of the 12NFH series at the same gate voltage.

However I have 6 of these 12NFH bricks and they are new ones and certainly would like to be able to use them with this driver setup, I have a couple CM600 Bricks as well but I am not going to do anything with those until I get this one up to snuff.

With the knowledge base that is out here I just thought maybe somone has tried these bricks and maybe had a similar situation and what they might have done too correct it, I am obviously going to have to up the wattage on the gate resistors, right now I am using 5.1 ohm, 5 watt but that wattage is not enough, I think once I get this situation under control this coil should be good to go.

1 2 3 next

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