Eplay

On a sunny day (Sun, 30 Sep 2012 19:01:54 +0200) it happened Fred Bartoli wrote in :

I have been driving the IRLZ34N logic level MOSFET, its Cgs is about 880 pF at Vgs 0V The minimum gate 'on' voltage is 1V, typical 2, my circuit had no problem driving it (just frequency goes way down), but I am 1 volt or so short of 'on'. I could cheat by pre-biasing it with a CR20something cell, the cell would not deliver any power, but it would look strange...

I have also tried to add a voltage multiplier, but the capacitance or load from it killed oscillation, I even tried an OA91 Ge diode.

Also tried some Schottkys, but those were big and leak a lot at room temperature. I have some very nice low capacitance RF diode arrays, but microscopic small, spares left over from the GPS jammer boards for Al Quaida, did not want to use them, you never know if those will be needed, you do not want to mess around with them. So that is up to 0bama. Lighting a small 800 mcd LED should be possible with a few uA.

The other interesting application, just to clear you up a bit, is this, at 22 C environment (cold end), and 37.5 C body temperature, these thermocouples give 0.636843 mV So for say 20 mV you only need 30 or so. These can be fitted to the human body, and the cold end to a heatsink on your head. A green light right on your forehead will then show if you are alive or not, a 'life' detector, for Joerg's dog, if it goes out it knows it will have to find an other person to feed it.

The disadvantage of the heatsink on the head is that in bright sunshine it gets hotter then the warm end, and the voltage reverses, so you need 2 circuits connected in opposite ways, the advantage of the heatsink on the head it that it also shields for EMI, and NSA brain control.

Hope that clear things up and answers Joerg's questions.

Vgs 0V

'on'.

temperature.

them,

them.

thermocouples give 0.636843 mV

head.

gets hotter then the warm end,

NSA brain control.

I was looking at some 0 threshold depletion mode fets, I think those may work much better :)

When they say 0 volts, I can assume they mean anything under a diode drop.

Jamie

For some of them zero really means zero:

Yeah, there is ~ 40uA drain at 0 volts. -1.6 to get it at 1 uA

I wish I knew about those about 2 mouths ago, we needed that type of switch.

Think they'll be some in my goodie box before the end of this week ;)

Thanks.

Jamie

Mouser has them in stock. Not even expensive, around $3 for such a quad-pack when you buy just a few.

Also, when designing switches and stuff check out the depletion mode MOSFETs from Infineon and IXYS. They are kind of a forgotten species among younger EEs.

Yeah, thanks

Btw, I was able to get that oscillator operating. I tried a different JFET and then I notice that my probe from the scope was loading it down..

I got out the active fet probes for the Tek which has 2 pf and that gave me a much better results. I also noticed then, the signal was no longer a sine but a raw pulse up around -10 volts.

have a good day..

Jamie

On a sunny day (Sun, 30 Sep 2012 17:27:53 -0400) it happened Jamie wrote in :

That is an old idea I had, and I almost ordered some. But think, if you use depletion mode, they are 'on' by default and shorting the supply voltage. I looked at the 'on' resistances at zero gate voltage, and it was so low that I think that could be a problem, but maybe it will just start up, but maybe you do not get enough negative voltage to get out of conduction. But sure try some. they are very cheap on ebay.

I reflected some on my circuit, and I think I am in for some more experiments, I understand a bit better now what exactly happens in that circuit I think, and that is the key to the next stage.

I did not want to mention this yesterday, as I was bidding on ebay on solar cells, (I lost the bid), but then got those for slightly more than my max bid, ebay nr 180958528777

Each of these panels can do .5 V 6 A, and that is where the power MOSFET chopper comes in. So I intent to play with those, and all those in series too.

Wow, "Not even expensive, around $3..."

What kind of disease got you?

I meant for a science project :-)

But really, it is not expensive. I remember when I stopped using the SD5400 because of its rather inflationary price trend. At $1.50-$2 it was a bargain because you could do servo stuff that would cost more if done via uC. Monolithic arrays are marvelous devices because almost everything tracks. Then it shot to over $5, game over. Now it's back down to the $2 region.