[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Terry's DRSSTC - Frequency/Pulse Width Controller



Original poster: Terry Fritz <teslalist@xxxxxxxxxxxxxxxxxxxxxxx>

Hi All,

I didn't get very far today on the Frequency/Pulse Width Controller. I got too wrapped up in the protection stuff.

Here is the IGBT over temperature circuit:

http://drsstc.com/Design/OverTemp.gif

It simply uses a LM35CAZ temperature IC that puts out 10mV/degree C. I compare that with a DC 0 to 1.5 volt signal from the pot for a temperature range of about 0-150C that can be easily pre-adjusted. The open collector output of the comparator is buffered to give complimentary Over Temp output signals. R3, C1, and R5 form a hysteresis effect. If the setpoint is 100C, the circuit will trip at about 102C and not reset until the temperature drops back down to about 92C. Sort of cool and prevents fast oscillations. A simple LED indicates when the circuit is active. I am using little IGBTs with little heat sinks. VERY cheap, but have to shut down if the temperature goes out of control. Best case is they can pump 500 amps just like the big boys and remain happy ;-))

For the Over Current part:

http://drsstc.com/Design/OverCurrent.gif

An off-the-shelf high speed CT (DigiKey) just feeds two comparators with a positive and negative trip point. A LMC555 timer provides a "timeout" if an over current is detected. I have it set to 500mS but it could be just about anything like 500uS if one were to want to restore the signal immediately for the next bang. These two circuits would go together on a card very happily since they could share ICs. The CT signal could also be used to drive the inverter logic.

Just simple goofy logic stuff today, but I thought real hard and long on it %;-))

I know of a digital way to do the Frequency/Pulse Width Controller but it is a little messy. I could use a Basic Stamp and add an LCD display and all too ;O)) A microcontroller could do a lot of stuff here really. But it adds a whole new set of complexities most people are not used to. I'll think more on this. I could not quite get any current source ideas to work yet. Steve Conner has a nice circuit too... I am curious as to how well he like the linear control??? But a basic stamp could just set the BPS rate and has pulse width output too... Could read out IGBT temp which is a real issue... I am very tempted just to use a microcontroller... I have Rabbit RCM3000 modules too for some of the 55MHz IEEE floating point math stuff, but that would be way too much power unless I wanted the "black box" thing to record all the data right before it blew up... Lots of unknowns about what is the coolest way to control things. A fully programmable controller might be the way to go to find out... The "box" could "protect itself" so the user could pick any BPS and pulse width solution they wanted. From Dan's simple LM555 thing to a laptop ;-) I think I need to add a many signal military connector to get to the remote box... I certainly have the armored cables and shielding from my robot stuff... The days of "computer controlled" Tesla coils, may have arrived!!!

I am "really" putting stuff on the drsstc.com site too ;-)) I "need" to move everything there since my DRSSTC notebook is now so messy I can't find stuff in it anymore...

Here is the input stuff and the parts list so far:

http://drsstc.com/Construction/InputBoard/InputSection.gif

Note the blocking inductor between the RF and AC line ground... It "might" be a "good" idea... We always want the RF and line grounds to be the same, but separate...

http://drsstc.com/Construction/MasterPartsList.xls

Yeak!! $725 so far ;-)) But It is all top of the line stuff... Probably hit $1000 easy. The rest of the parts are cheap but the PC boards will add up now still. But the box should be able to run a wide variety of coils and be really cool!!! I am using all the best ideas and state of the art stuff. It should run most lower frequency coils in the 1.5kW range and I can easily lift it with an out stretched arm ;-)

I am sure that DRSSTCs can be extremely reliable. They are a very simple form of switching power supply. There are no great tricks to it... A little harder to come up with at the "amateur level", but we will get there given a little more time and 1/1000th the R&D budget ;-))

Cheers,

	Terry