UFO politics Keeps his word 12-12 12 Let the games begin.......

[picowatt]

I just had a scary thought. Looking at the 555-controlled mosfet pulser, and considering the Ohm's Law mashup and some other things I've read..... I just wonder........

Let's say you have a timer-driven mosfet pulser like is shown in the last few posts "over there". A 555 timer sends a pulse to the gates of a bunch of parallel mosfets, which switch a 36 volt supply to a low-impedance (under 1 Ohm) load. The individual mosfets have an on-state resistance of about half an ohm so the overall on-state resistance of the mosfets is about 0.1 ohm.

OK.... so the load is connected like this: +36 V Positive rail > load > D-mosfet-S > negative rail. Right? Or like this: +36 V Positive rail > D-mosfet-S > load > negative rail. Either way, makes little difference for the moment.

And the 555 timer sends a positive pulse to the gates of the mosfets to turn them ON and allow current to flow through their D-S channels and the load. Right? The timer is variable in frequency and/or duty cycle-- the proportion of time the output pulse is "High" or in this case nearly +12 V.

So now we hook a voltmeter in. Positive lead to the Drain of the mosfet(s) and negative lead to the negative rail. When the 555 timer sends a "high" or positive 12 volt signal to the gate of the mosfets.... what does our voltmeter on the mosfet Drain read?
And when the 555 timer signal is Low or off, the "off" time of the duty cycle, what does our Drain voltmeter read? If we scoped the drain and the gate of the mosfet, what kind of relationship should we see in the two signals?

TK,

I believe the circuit is intended for low side switching, hence the indicated wire on the drain/heat sink connections.

When the MOSFETS are fully on and at 50 amps current draw, there will be about +4.6 volts at the MOSFET drains which equates to about 230 watts of heat they will produce (and power wasted).  That is, just before the rather smallish source traces fuse open!!

Additionaly, using the limited current drive of a 555 to switch all those MOSFETS in concert with the 330R gate resistors is going to produce relatively slow rise/fall times, during which MOSFET dissipation will be much greater than 230 watts.  In other words, if the source traces don't fuse open (whch I believe the ground connection side will do immediately upon seeing 50 amps!), there will be a lot of heat to get rid of.  Not very efficient...

A buffer on the 555 or a proper hi-current gate driver and smaller value gate resistors to reduce rise/fall times, much, much larger PCB traces (or a proper buss bar) for the source connections, and the use of lower RDSon MOSFETS are highly recommended.  Also, to acheive fast rise/fall times with a hi-current gate driver is going to require at least the addition of a ground plane to the PCB and lots of local supply decoupling. 

A fast diode snubber across the output and a transorb or MOV for reverse V and spike protection would also not be a bad idea.

PW 

[picowatt]

Also, decoupling the 555 supply and the 555 signal ground to the high current source trace is likely going to produce a lot of ringing or outright oscillation at the 555 output. 

A single point ground should be employed and all traces should be short and low Z.

I'd also put all the FET's in a physical layout more suited to paralleling for hi-current.

Hi-speed and/or hi-current circuits and PCB's require additional considerations when designing the PCB layout and power/ground connections.

I would consider this circuit design and layout totally inadequate to switch 50 amps or greater at a reasonable efficiency and with any reliability.   

PW 

[picowatt]

Just a few questions for you here:
Do you think the fact that a Clark-Hess costs a couple thousand dollars has anything to do with it?

Do you have such a power meter?

Is it borrowed? Or did you purchase it new with your own money?

I believe TK owns his, and he likely bought it used.  But keep in mind, all manner of test equipment can be rented or leased.

Many times, when I have needed a 20K plus piece of equipment for a specific project, I just lease/rent from one of the many companies that do so. 

Right now, however, a fairly accurate shunt, LP filter, a couple DVM's and a decent Prony set up would go a long way in performng sorely missing measurements.

I'd suggest a single arm Prony with a brake set-up appropriate for the expected HP and dissipated heat.

PW

[TinselKoala]

Just a few questions for you here:
Do you think the fact that a Clark-Hess costs a couple thousand dollars has anything to do with it?

Perhaps it does. How much is an electrical overunity machine worth, anyway?

Do you have such a power meter?

Is it borrowed? Or did you purchase it new with your own money?

I wonder why you are asking me these things. (Actually..... I don't, it's part of the obvious and general strategy of attacking the skeptics instead of addressing their points rationally, in an attempt to take the pressure off the outrageous claimant.)

I have access to one when I need it, borrowed, yes. But remind me..... just where did I make a claim of electrical overunity performance? Am I the one who needs good measurements to bolster a claim of overunity electrical performance?

And as picowatt says, it's easy to rent one for a day or a weekend if you actually need one. Is one actually needed in this case? Well, they are a lot easier to use and interpret than an oscilloscope is for straight power measurements: the CH gives you unequivocal numbers in boxes, all you have to do is hook it up properly, push some buttons and take the readings. Whereas to do energy flow measurements with the scope, one requires some knowledge and finesse.


@Picowatt: You bring up some very important points about the mosfet switcher as laid out "over there". I wasn't even going that deeply into the specifics, though.... I am worried about that "fundamental error" of thinking that the drain voltage being high, at or near the supply voltage, means that the mosfet is "on", and when the drain voltage goes low, the mosfet is "off". This flipped duty cycle error was at the root of someone else's mistakes in power calculations in another circuit we know about. I hope it's not being made here as well. A voltmeter on the drain will read the high voltages and consider the "high" times as "on" for the purposes of duty cycle readings.

[picowatt]

TK,

Sorry, thought you owned that beast.

Regarding your prior post, I knew exactly what you were referring to...  no spoilers from me.

Nice shunts in your previous photos...

PW