On my set up another curious result is that TIP 31 works with or without the toroid, when the CFL/neon is either near L1, or touching L1.
With FETs I get NOTHING without the toroid, even if i connect the CFL/neon to L1. They only work with the toroid, and they work even better if i increase capacitance with a metal sieve under the toroid.
Anyone else see anything similar?
Ive just bought 20 FETs (530, 630, 730, 830) and 10 new TIP 31s, and 20 IN4148 diodes - this weekend ill be joining you woopy on trial and error
Announcement
Collapse
No announcement yet.
Joulethief SEC exciter and variants
Collapse
X
-
Wrapping your head around "theory"...
Originally posted by woopy View PostI have begin to study all data sheet, so i become slowly more and more intelligent, but not enough yet to get all the subtility of those info.
The spec sheet idea came up when I was studying a basic book of electronics... I'd read parts of it over an over again but never could grasp it until I started working with SECs. Working with the SECs introduced me to an area of electronics I had no concept of. So when I was reading my book I began to understand better what was being done because I'd already had some practical experience in the area -- just no professional terminology or theory to relate it to. It related the concept that FETs were better than transistors because of their high-voltage rating and low-current rating. I was thinking, "I wonder what FREQUENCY they operate at?" so I was looking for frequency in the datasheets. There isn't a lot of data about frequency but the switching times really indicated to me.
Studying this stuff is definitely hacking knowledge. You just have to keep doing it over and over again until it finally "clicks" and you really get it.
My current "beasts" are inductance and resonance... I think those could be the key to OU...
Take Care,
Mussle
Leave a comment:
-
hi all
going on working and studiing the MOSFET
and good results
so confirming that, with my set up, only the BUZ 11 works (i have not yet IRF 530)
i have also tried a big IRF P 450 and no chance.
I have begin to study all data sheet, so i become slowly more and more intelligent, but not enough yet to get all the subtility of those info.
So no other way to go on the terrrrrifffffic but fantastic way of trial and error
But i like, because it is so surprising and motivating.
OK so it seems to work with the Xee2 circuit and different capacitors and BUZ 11 and different resistors with very different results.
1- if you put no capacitor it works too. But in this case, if you approach the cfl very near from the L1 coil the lighting goes down , and if i touch the L1 with the cfl , the lighting stops and the current jumps as if it is a short cut in the L2.
2-Higher the resistor value = smoother the reaction.
E.G with arround 500 ohms i can get good lihting up to 12 volt and low A draw, but no or very low plasmajet at the end of the free L1 wire.
and at 100 ohms resistor the power is huge (see pix 1 and 2) where i put a 100 ohm and (probably 5 watt ) resistor., the lighting is very good and the plasmajet is at max for what i have got until now.
so for 14 volts, and 1.2 A (that is to say 16.8 watts), i can get the lighting of the pix ,
plus the heating of the fet and resistor and in bonus a strong and noisy 2 cm strong plasmajet.
So far very happy
and good luck at all
LaurentLast edited by woopy; 01-26-2011, 10:48 AM.
Leave a comment:
-
Originally posted by xee2 View PostI assume you mean 1.5 volts in a cell phone charger. I have never been able to get my FET circuit to work with just a single AA battery unless it was in the cell phone charger.
But i'll keep trying.....
Leave a comment:
-
Hi SeaMonkey
Thanks for your advice, I have a scope. What are your recomendations for the probes place.?
@ Xee2
Good idea to test the BUZ 11. We will see if the results is depending of the fet or the complete config or probably both.
good luck at all
Laurent
.
Leave a comment:
-
Originally posted by seth View Poston any batteries above 1.5V!!!! But they can light CFLs easily and on low watts.
Leave a comment:
-
Originally posted by xee2 View Post@ Woopy
I have put over 1 amp through the IRF530 FET using a 1/4 watt 51 ohm resistor. The resistor did not get hot but the FET would get hot if I did not use a heat sink (was OK with a small heat sink). FETs are strange. I ordered some BUZ11 FETs to see what happens when I swap the IRF530 for a BUZ11. I will post results.
My IRF 730 over heated a few days back on my 6V battery at 0.2A . Im using a quarter watt resistors and sometimes half watt resistors, and they never get hot. My FETs get hot on any batteries above 1.5V!!!! But they can light CFLs easily and on low watts.
Leave a comment:
-
So far i know the FET are piloted by voltage and not by current , so why do i need to to have so high wattage in the gate?
I have put over 1 amp through the IRF530 FET using a 1/4 watt 51 ohm resistor. The resistor did not get hot but the FET would get hot if I did not use a heat sink (was OK with a small heat sink). FETs are strange. I ordered some BUZ11 FETs to see what happens when I swap the IRF530 for a BUZ11. I will post results.
a MosFet is to "scope" the Gate signal.
Without an oscilloscope it is very difficult to diagnose
the source of "strange" performance.
Leave a comment:
-
Originally posted by woopy View PostHi xee2 and all
My last test shows that i can get very good lighting of CFL with the BUZ 11 under 12 volts.
But i have to mount 6 x 0.25 watt 2200 ohm resistors in parallel. That is to say the equivalent of one 367 ohm and 1.5 watt resistor. And those resistors get hot anyway , so more would be better
So far i know the FET are piloted by voltage and not by current , so why do i need to to have so high wattage in the gate?
Thanks for answering
good luck at all
Laurent
I have put over 1 amp through the IRF530 FET using a 1/4 watt 51 ohm resistor. The resistor did not get hot but the FET would get hot if I did not use a heat sink (was OK with a small heat sink). FETs are strange. I ordered some BUZ11 FETs to see what happens when I swap the IRF530 for a BUZ11. I will post results.
Leave a comment:
-
Hi xee2 and all
My last test shows that i can get very good lighting of CFL with the BUZ 11 under 12 volts.
But i have to mount 6 x 0.25 watt 2200 ohm resistors in parallel. That is to say the equivalent of one 367 ohm and 1.5 watt resistor. And those resistors get hot anyway , so more would be better
So far i know the FET are piloted by voltage and not by current , so why do i need to to have so high wattage in the gate?
Thanks for answering
good luck at all
Laurent
Leave a comment:
-
Originally posted by xee2 View PostMy understanding was also that FETs did not need much gate current to work. However when woopy is buning up gate resistors that would not seem to be true. And, in my circuits there is a definite increase in drain current when the gate resistor value is lowered. These FERs seem to act more like transistors than a classical FET.
The IRF 530,630,730.830 work while the IRF 640,740,840 do not work. I have looked at the data sheets and I do not see what it is that is different between the 30s and 40s except the power. But it might be a gain difference. Please let us know if you figure it out.
discharge current associated with the Gate Capacitance.
To turn the MosFet "ON" the Gate Capacitance must be
charged to a voltage higher than its Threshold rating and
to turn it "Fully ON" will generally need to be charged to
10 Volts. There are exceptions in that some are "L"
or Logic devices which will turn nearly fully on with 5 Volts
of Gate charge.
To turn the MosFet "OFF" the Gate Capacitance must be
discharged to 0 Volts.
Internal to the MosFet is a protective diode between the
Gate and Source leads to prevent the Gate Voltage from
going too high (about 18~20 Volts) or from reversing polarity.
It is very difficult to drive a MosFet with any appreciable
series resistance in the Gate connection which would
excessively prolong the Gate Capacitance charge/discharge
time constant. A Gate resistance of about 50 Ohms is often
used at high frequencies to limit Gate current magnitude and
to dissipate power that would otherwise have to be dissipated
within the Gate Driver chip.
For best performance a MosFet should be transformer driven,
with adequate protection against polarity reversal, or driven
by a Gate Driver Chip. It is also possible to use discrete
transistors configured as a Totem-Pole for the needed
Pull-up and Pull-down of the Gate input.
Once you learn the "tricks" of getting the MosFet to
respond to its input signal you'll be able to make it do
everything that you want.
The higher the Drain Current rating of the MosFet the
higher the Gate Capacitance. High Gate Capacitance
demands more charge/discharge current to turn the
MosFet ON and OFF.
Leave a comment:
-
Originally posted by anonymussle View Postwith almost no current
The IRF 530,630,730.830 work while the IRF 640,740,840 do not work. I have looked at the data sheets and I do not see what it is that is different between the 30s and 40s except the power. But it might be a gain difference. Please let us know if you figure it out.
Leave a comment:
-
Transistor and FET datasheets
Hi all...
I was doing a little research on the transistors and FETs mentioned in this thread to get an idea why some work better than others. I get the idea that FETs operate faster than transistors with almost no current which would make them good for high frequency/low current applications. I was looking at datasheets for various FETs and transistors and I noticed a particular difference between the IRF640 and IRF730. The 640 switches at 13-17ns. The 730 switches at 11.5ns. I noticed a number of other timing features that would seem to make the 730 better than the 640.
I imagine if we get an idea of what kinds of characteristics are needed for an optimum SEC one could use the specs to get exactly what's needed.
I could be talking below the education level here but I thought it might help to mention this info.
Some datasheets I've checked out (all PDF files) from Datasheetcatalog.com:
IRF730 FET
IRF640 FET
TIP31 Transistor
MPSA06 Transistor
2N3904 Transistor
Best,
Mussle
Leave a comment:
Leave a comment: