Question for Reg.

[William]

Can I replace the power mosfet irf640 in Gary's p.i. with the FQI2N30 power mosfet you changed out in Carl's p.i. detector? If so would the gate resistor r11 150 ohms need to be changed in value? Thnaks William

[Gary]

Re: Question for Reg.

Yes you Can. No Change to the resistor

[Reg]

Re: Question for Reg.

Hi William,

I am not that familiar with Gary's PI to know the average current drawn by the FET, but the use of a 640 indicates the current draw could be considerable.

Now if you want to change the FET, I am also assuming that you want to make the change to be able to shorten the delay time, which would also indicate you are trying to improve the detector's gold responses.

If this is the case, then a longer pulse on time is not needed. One can reduce the pulse on time to 50 usec or so and still get great results on a gold target. I would at least reduce it to less than 100 usec. This will reduce the current requirements of the FET needed.

Now, the reason I mention all of this is because the FQ12N30 is rated for much less current than the IRF640. So, one has to take this into account when trying to replace it.

So, if you use the FQ FET mentioned above it would be advisable to reduce the current in the coil via smaller coil wire, shorter pulse on time, and possibly a resistor in series with the coil. A good heatsink could help also.

Another alternative would be to select a different FET with lower capacitance ratings than the IRF 640 but with sufficient current ratings for the application. I would still reduce the pulse on time.

One could select something like the FQU7N20 rated at 5.3 amps might be a better choice. Other choices would be maybe something like FQB5N40, a 400V 4.5A; FQI7N20 a 200V 6.6A.

Even a IRFI620B might work better than the FQ12N30 which is only rated at about 2A or so. This IRF is rated at 200V and 5A. I suspect all of the above mentioned FETs would work ok and reduce the capacitance associated with the original 640 FET.

As for the gate resistor, any of the newer FET types would not require any gate resistor change. In fact, because of their lower input capacitances, most could display a faster rise time, thus making things just slightly faster. Having a lower input capacitance requires less drive signal which is another positive to the newer generation of FETs.

As an example of what one might gain, lets compare the values of the capacitances of a couple of FET's.

An older IRF640 has values such as these listed in an older Power Mosfet data book:
Input capacitance: 1600pf max
Output cap: 750pf max
Reverse Transfer cap: 300 pf max

Newer generation FET's have the following info;

The Fairchild IRFI640B has the following:
Input cap: 1300 pf to 1700 pf
Output cap: 175 pf to 230 Pf
Rev xfer cap 45 pf to 60 pf

IRFI620B has the following values;
Input cap: 300 pf to 390 pf max
Output cap: 50 pf to 65 Pf max
Reverse trans cap: 10 pf to 13 Pf max

Other newer generation FET's mentioned have output capacitances of less than 100 pf max also. Most are in the 50pf to 70pf range. So, the net result should be the ability to reduce the decay time.

I hope I have answered your question.

Reg

[William]

Re: Thanks Reg. and Gary

Thanks Reg. for the information, I guess I was confused about you using a lower amp Fq2n30 on Carl's detector and thinking it would work well in Gary's detector to reduce some capacitance.
Can you limit how much flyback voltage you get
somehow? Thanks William

[Reg]

Re: Thanks Reg. and Gary

Hi William,

Most of the newer designed FET's have much lower output capacitances, which should allow for faster decay times, which will allow for shorter delays to be used. As an example, the one you mentioned, the FQI2N30 has an output capacitance of 25 pf to 35 pf max. This extremely low value is why I elected to use it. This is extremely small when compared to the 640 that could have an output capacitance of up to 750 pf.

Since Gary says FQI2N30 will work, then I would at least try it. I suspect you will have to increase the damping resistance quite a bit to see any advantage.

As for your question about the flyback voltage, this is generally a function of the FET used (if no other limiting is involved). Just look at the FET breakdown voltage as the reference. The FQI2N30 mentioned above has an avalanche breakdown voltage of 300V. So, the flyback voltage should be limited about 300V or so. The subtle hint on the newer Fairchild FET's is the last two numbers of the FET. In this case they are xxxxx30 which reference this as one with a 300 V breakdown voltage, while a xxxxx20 will be for a 200 V, etc.

I personally see no reason to use anything else besides the FET to limit the flyback voltage. Adding anythng else generally will add capacitance which will lengthen the decay time.

Reg

[William]

Re: Thanks Reg.

I'll give it a try. William