Sidebar Sidebar
  • You can now help support WorldwideDX when you shop on Amazon at no additional cost to you! Simply follow this Shop on Amazon link first and a portion of any purchase is sent to WorldwideDX to help with site costs.

Has anyone looked into these SuperFets II from ON/Fairchild?


Greetings!

Attached are Three datasheet screengrapbs as to what may help you find what you're looking for to demonstrate something...

ON Semi -FQP130N10 (Non - L)
On SEMI FCPF400N60 Super FET II
Vishay IRF520

In simple terms...

MOSFET has an inherent bane of slow OFF times - and it can be shown in their input and output capacitance - larger capacitance on the substrate equates to a harder time to remove the charge from the gate region...or change it from on to off and off to on...lag times...

Large amperage units capable of switching heavy loads are not favorites - they are thick layered substrate and large surface area gates which leave a lot of surface to remove energy as well as place a charge on so they take a considerable amount of EMP push to overcome their reactive impedance AGAINST having a charge placed on them

13N10 and the IRF520 are the most ideal due to switching times and capacitance both in and out of their substrates...less delay and less effort to get them to turn on and off - equates to less output coupling "mess" and less chance of input matching issues. Their small chip designs keep capacitance to manageable levels.

Now just having low capacitance just doesn't always mean they're the best suited for the application - others may have a lagging problem in rise or fall times but ARE the ones that closely exhibit a desirable slope and crest (Overshoot) and trough (Dip) that is a lot like the frequency of interest when these devices are used CLOSER TO THEIR MAXIMUM USEABLE FREQUENCY (like the 13N10) or MUF - have that desirable trait - because of the matching output "trait" or power curve tends to mimic the Bi-polars' own "smith chart wanderings"

Before I start to hear cracking knuckles and banging keyboards...
The "Smith Chart" reference is due to the devices electrical traits not because it matches the conjugate....-ed.
​
Just keep gates above their 2V thresholds to turn on and nothing more than 4 volts else you're wasting power.

See attached screengrabs for details of what I'm trying to show you...
13N10Data.jpg
FCPF400N60.jpg
IRF520Vishay.jpg


Hope this helps!
:+> Andy <+:
 
  • Like
Reactions: rabbiporkchop
Most-important numbers I see are the turn-off time. Always slower than turn-on time, and this should put an upper limit on the RF frequency it can amplify.

Haven't figure out how to predict it as such, but the FQP shows 20 to 50 nS. The IRF sheet presents this as just 20 nS.

A sine wave with a period of 20 nanoseconds has a frequency of 50 MHz. This would suggest that it won't amplify a signal that 'turns off' faster than the FET can follow along.

Would also seem that a FET that shows a turn-off of 100 nS wouldn't be much good above 10 MHz.

Might still blow the doors off an 80-meter signal at 3.8 MHz. But not so much at 27 MHz.

The bigger they are, the slower they get. And that's what's got me wondering how fast the ERF9530 is. No switchmode specs have been published, since they are not marketing it for that kind of duty. Makes it hard to compare, like apples to kumquats.

Makes me want to cobble up the "test circuit" shown in the semiconductor data books and see if I can measure the EFR9530's turn-off time with this new-fangled digital 'scope. Might be fast enough.

Gotta learn how to use it first.

73
 
Both of you thanks! Lots to digest here. I read it all but did not take time to properly absorb it all. A terrible headache. I have not slept in the last 3 days insomnia! I will review it all again tomoorow.


Thanks!
 
  • Like
Reactions: rabbiporkchop
Greetings!

(you're not alone on headaches - allergy season is here...it's gonna' be a doozy)

I see IRF 520s as, well - that comment I made about the "knee" goes back to the gate trigger level...the capacitive element is not so present in the "top" part of the slope going up as much as coming down. So the incoming RF tends to push on the gate and the rapid rise and fall give the TOP of the slope (crest of the wave) a nearly triangular appearance at 27MHz. so there's other stuff going on...the IRF puts on a show like Ben Affleck does in acting - versatile for the purpose but you have to apply a lot of makeup to the face to make it work at the frequency of interest...

The 13N10's are slower in some aspects that make them desirable for the 27MHz band. They don't have the best performance - but have the best curves as they react to the RF incoming across their gates at that frequency - it seems that 27MHz is to be the closest to the MUF these "13's" parts can deliver. They absorb more energy to get the rise out of them but the rise slope and fall rate are desirable - they tend to exhibit an "overshoot" near the top that make the crest a lot more like an arc than the IRF provides. The "bounce" is a lot less harsh than the IRF 520 - which lends to the fact that the IRF units were designed for higher MUF's than the 13N10 from Fairchild.

IRF wants more gate voltage presence to offset the bounce so you get that "knee" effect. Fairchild just is slow to wake up and skews when they do, and the fall? Not a cliffhanger, more like a moving arc...

I can't speak on behalf of the Bipolar - but it does put on a good show...

Regards!
:+> Andy <+:
 
Could one use a digital control unit to turn the fets on and off as needed and amplify the data in a sequence then recombine the output in a way that would reconstruct or mimic analog operation? So go digital on breaking the analog voice into chunks amplify it in many individual amp modules then sequence it to put it all back together? In my mind that makes more sense than trying to use a mosfet in place of a bjt directly.
 
If what you're talking about is the "PWM", or pulse-width modulation, that's is how these MOSFETS are used at the on/off switching frequencies under a MHz found in power supplies. The ratio of "on" time to "off" time gets smoothed out with capacitors. This becomes your analog output. You can amplify audio this way, it's called Class D. To get frequency response to 20 kHz calls for a pulse frequency a lot higher than that. There's a rule called "Nyquist" that requires a frequency at least double the one you want to amplify, preferably a higher ratio than that.

To do what you're describing to amplify RF calls for a square-wave "clock" frequency that's a lot higher than the frequencies you want to amplify.

And since the on-off speed of the MOSFET is what limits how high a RF frequency it can amplify, just using it as a conventional analog amplifier is tricky enough.

Gotta build a test fixture.......

73
 

dxChat
Help Users
  • No one is chatting at the moment.


      Top Bottom
      Back