My Motors got me to Tap into Radiant Energy

[prochiro]

Lm339

Hi UFO and all,
You may not hear from me a lot but I have been here and building and running this fine circuit. I wanted to note fore some who are building this that we, Bob and I, have had a small problem getting a second system running and have found that not all LM339 IC's are the same. We both put in Texas Instruments LM339 and got a short with heave draw. Most all other 339's are correct and remember that datasheets should always be checked when using a chip by another builder. Radio Shack saved our butts again, there LM339 is correct.
Happy Building
Dana

[bobfrench@fastmail.fm]

Lm339

UFO et al,

As Dana said, different pin outs for different manufacturers. Radio Shack ones that should work (even though they are made by TI. TI spec sheets show a different pin out. Maybe they make more than one LM339, ?).

As drawn, the pins should be as follows:

Pin 1=OUTPUT 2
Pin 2=OUTPUT 2
Pin 3=+V
Pin 4=INPUT 1-
Pin 5=INOUT 1+
Pin 6=INPUT 2-
Pin 7=INPUT 2+
Pin 12=GROUND
(other pins not used)

Either find LM339s that match that or re-draw to accommodate what you've got. Now that we've (hopefully) got the screw-ups behind us, we can have even more fun. I've got an NE-30 to light up!

Bob

PS -UFO, I just realized that the drill motor that I ran (slowly) was an AC motor! No wonder it didn't like it. I'll try a cordless one that I've got.

[bobfrench@fastmail.fm]

circuit still not working

Everyone,

We have not discovered what is going on yet. I am not sure of what pin out the LM339s I have are set for. I have an assortment of them.

I will keep you posted,

Bob

[torpex]

Relative error in LM339

Bob, Dana,

It seems that the difference in Motorola is that exchanges the relative values of the comparators 1 and 2. But should not affect the circuit in practice. Right?

I think it's the same pinout.

Pins 1, 6, 7 : One Comparator.
Pins 2, 4, 5 : Other Comparator.

In your post pins 1 and 2 are the same ?

[prochiro]

Lm339

The Texas Instrument ic has pin 1 is one output. Other pins are also different. If you just watch your data sheet, things will go well. We soldered a plugin base into the board and when we changed to TI, well, lets just say we need to replace a few things.
Dana

[torpex]

@Ufo

Resistor test:
10k or 5k stays cold, does not influence in brightness, but volts in output drop.

With bulb+capacitor (20mF-750v), 10k or 5k heated, voltage output is held.

[Ufopolitics]

Thanks Larry!!

Hi UFO
I am in no way disagreeing with you, but interjecting some facts.

I think considering you have 2 different type of loads (one inductive and one resistive) that you would even get this effect with heavy/hot side. If the resister value is considerably larger (more resistant) than the lamps impedance, the current will take the easiest route (through the lamp) and very little will go through the resistor thus lamp not dimming and the resistor not getting warm. Now as the resistance approaches the impedance of the lamp you will see dimming. If you have 120V across the lamp and it is a 35W lamp then you will have less than 0.32 Ohm impedance. Try a 2 Ohm or smaller resistor and see what happens.
Also there hasn't been any discussion about the amount of current that will flow through the fets, but I am pretty confident that this circuit will not generate any where near enough current to heat up the fets to any extent if at all. The only ways the fets could be damaged would be for the voltage to exceed 400V or the driver circuit getting stuck on the on cycle then I believe the coil will melt before the fets are effected.
I am starting to build my circuit today so I will be able to answer some of these questions for my self and be able to analyse for my self what is going on. I am doing the LM339 circuit.
Remember I have a very open mind, but I also have a very strong need to know what if happening.
I have been to the sym vers asym site (very enlightening) and watched your last video... Thanks for all your trouble, your time, and your patients you are going through in order to present this. When we get to the full build and I can replicate it, I will be one to spread the word and help others to get there.

Regards
Larry

Hello Larry,

First, I am glad you read the Asymmm versus Symm thread and also the video, and that you liked it, it is all my pleasure.

Larry, you are right, and it does not bother me at all that you have your opinion on this, on the contrary, this helps to enlighten Us all.

I agree with you on the resistance values, versus watts and voltages...however, Larry, I have done this test with just the Hot side,
meaning No Diodes, just coil (without the coil and just oscillator-output to CFL...NO LIGHT at all, I have also tried that) ...and the resistor burst on fire (same value resistor as I tried before on Radiant or "Inductive" load...) and Mosfet's got super hot to blow out...I turned off everything on time...it was just to film it...

(NOTE: I do not want to suggest-encourage other members to do it, because if you do not act fast, you loose some good components)

Now, I do not agree on your Amperage opinion related to Mosfet's...and you will be able to check that out yourself, when you have your circuit running...They do get very Hot, if you are using a heavy load and/or your Coil resistance is too low...

Regards


Ufopolitics

[Ufopolitics]

Hello Torpex

@Ufo

Resistor test:
10k or 5k stays cold, does not influence in brightness, but volts in output drop.

With bulb+capacitor (20mF-750v), 10k or 5k heated, voltage output is held.

Great,

Well, with Cap on load you are storing the charges there (making them double) ...so that intensifies the charge to load resistor...Voltage will not drop...(you are using the Cap's V, at the low stage [resistor consumption])

Regards


Ufopolitics

[torpex]

The Texas Instrument ic has pin 1 is one output. Other pins are also different. If you just watch your data sheet, things will go well.

Sorry Dana, I do not see. Is the same (for my)

Typical pinout:


Motorola pinout:

[Ufopolitics]

@ Bob and Dana...

Everyone,

We have not discovered what is going on yet. I am not sure of what pin out the LM339s I have are set for. I have an assortment of them.

I will keep you posted,

Bob

Hey Bob and Dana,

I believe , not completely sure though, that we originally had a circuit with just one LM393...and pulsing just one channel (Negative)...
Then I asked Mad Scientist if he could make it for dual channels in Anti-Phase...then He included either two LM393 or just One LM339...

Now, the reason I wrote all this, is because if you are using just one channel pulse (N)...than you could do it with just one LM393..since the LM339 has so many variances in the legs according to manufacturer's...

LM339 is a DUAL Comparator...it has Two LM393 inside...

Anyways that was just my opinion trying to help here...but I really have not make any LM's circuits as of yet...been dedicated to Lab tests and Videos lately...MAYA is very good program, excellent, but consumes awfully big time...

Glad you liked the video Bob

Regards

Ufopolitics

[torpex]

Now, the reason I wrote all this, is because if you are using just one channel pulse (N)...than you could do it with just one LM393..since the LM339 has so many variances in the legs according to manufacturer's...

LM339 is a DUAL Comparator...it has Two LM393 inside...

Sorry Ufo, LM339 is a Quad Comparator, LM393 is Dual Comparator.

Hey Dana, Bob,

The LM339 is a very common circuit, would be very strange to have different pinouts.
There may be other differences in behavior as an oscillator, but not in the pinout.

Comparators name (1-2) are relative.

[Ufopolitics]

Thanks Torpex..

Sorry Ufo, LM339 is a Quad Comparator, LM393 is Dual Comparator.

Hey Dana, Bob,

The LM339 is a very common circuit, would be very strange to have different pinouts.
There may be other differences in behavior as an oscillator, but not in the pinout.

Comparators name (1-2) are relative.

Well I said I was not sure...

But is about same...Two channel pulsing needs an LM339...
LM393 Dual Comparators
LM339 Four Comparators...the thing is this circuit will work also (for one channel, Negative) with just One LM393...

Am I Right?

Regards

Ufopolitics

[larryross]

I agree with you on the resistance values, versus watts and voltages...however, Larry, I have done this test with just the Hot side,
meaning No Diodes, just coil (without the coil and just oscillator-output to CFL...NO LIGHT at all, I have also tried that) ...and the resistor burst on fire (same value resistor as I tried before on Radiant or "Inductive" load...) and Mosfet's got super hot to blow out...I turned off everything on time...it was just to film it...

The problem with your experiment here is it isn't equal to the one you were comparing it to mainly because, In the first test you had an alternitive circuit (through the CFL) which is much lower in resistance so the current went through the CFL not the resister. In the next test you gave the current only one path and ya if you try to push 1200W (at 100% duty cycle basically DC) through a 1/4W resister it will fry very quickly.

Now, I do not agree on your Amperage opinion related to Mosfet's...and you will be able to check that out yourself, when you have your circuit running...They do get very Hot, if you are using a heavy load and/or your Coil resistance is too low...

I will have to bow to you here for my stupidity. I failed to consider the power dissipation spec on the fet which is only 125W. I am not sure why a device is rated capable of 4000W would only handle 125W. Anyway so if you were at 100% duty cycle (basically DC)for the out put of the fets you would have potentially 36 amps out (short circuit). So if you had say a 2 Ohm load that would be 16 amps which would be 576W on your fets and with 6 in parallel you should have the ability to handle 750W with a derateing of 1W/degreeC above 25 degree C at Tc. This would cause the fets to get kind of warm to begin with then the heat will snowball quickly as it gets warmer. The power dissipation can be raised with heat sink and fan. A fet in a TO3 package might be a better choice than the TO220 or TO224.

BTW the power dissipation of my 2SK2837 is 150W at Tc 25 degrees C, but doesn't give me a derateing spec. Think I will look for Fets with a higher power dissipation as this seems to be critical to endurance of this system.

Check out the Fet in the file below. One of these would equal 30 of the NTE2397 Fets power dissipation wise.

Hope I don't make too many more mistakes.

Regards
Larry

[bobfrench@fastmail.fm]

LM339 problem

Everyone,

i do not think that the problem is with the LM339 or the board at all now.

I got to thinking about the fact that this seems to have a short. so I just started taking things off until the short stopped. I took the pots off one at a time, then I took the whole board off and I still had the short. SO...it seems to me that the MOSFETs are blown and shortage the thing! Duh!

I had already ordered more, so when they come in we'll try try again.

I have looked at the circuitry, the borads, the solder joints...everything and I think we're OK once the MOSFETs are replaced.

Bob

[Ufopolitics]

Hello Bob

Everyone,

i do not think that the problem is with the LM339 or the board at all now.

I got to thinking about the fact that this seems to have a short. so I just started taking things off until the short stopped. I took the pots off one at a time, then I took the whole board off and I still had the short. SO...it seems to me that the MOSFETs are blown and shortage the thing! Duh!

I had already ordered more, so when they come in we'll try try again.

I have looked at the circuitry, the borads, the solder joints...everything and I think we're OK once the MOSFETs are replaced.

Bob

Hey Bob,

Yes, MOSFET's are the first to go on this circuit...and most of times it is just one that shorts out Drain-Negative Source, making the whole circuit be in short, so you should make an easy way to take the heatsink bolts out and check them independent one from the other, BUT FIRST, you should check if there is continuity between Drain-Source,before taking them apart, and making sure you switch the meter's probes according to be in "Forward Bias" with the Intrinsic Diode on the Mosfet's...

Regards


Ufopolitics