MOSFET Question

[Onelasttime]

You might check and see what is causing this or you will just keep smoking 520's too. There might be some kind of a mis match between the radio and Texas Star that needs corrected.

Well if they work into a dummy load with good solid coax and well-installed connectors it is not the transistors.

Think about it! If it only blows on your Sweet 16 than that should be telling you something! If you do not get it then that is on you. Changing to the 520's will at best be a band-aid fix.

 

[nomadradio]

My stryker 655 keeps popping them every time I run it with my texas star sweet 16 amp. I'm changing to 520's.

A SWR meter and jumper between the radio and the amplifier will probably reveal a shockingly high SWR when the amplifier is keyed up.

This would be the cause of the actual problem.

The radio blowing up is only the symptom.

The amplifier's input circuit should fool the radio into thinking that it's driving an antenna with a low SWR.

Sounds as if your Sweet Sixteen is not quite doing that part of the job properly.

Worn-out contact points on the amplifier's relay can do this, too.

73

 

[Handy Andy]

Greetings!

Take this for what it's worth...

IRF520 versus 13N10's (FQP if being picky)

First off, if the original was a BIPOLAR - then all bets are off...

IF it used IRF's - then the OUTPUT and coupling networks are tuned best for the IRFs' output...

IF if used AN2030 (EKL) then presume it (output and coupling networks) were tuned for that....

I've found 13N10FQP work best for the drop in bipolars with a 1N4148 is series with a 390ohm and use a 470ohm shunt to ground (across) on R56 - I presume AM only...

It's a this or that sort of thing....MOSFET are chips with legs, while the Bi-polars have Knees in their middles - so they can bend a bit...

So for best results - use the one that is the closest to the original design...

It may require that you try both...and don't be a bit suprized if swapping to a 2078 gets you magic watts - if only for a moment...

But I digress, lots of the issues around "mating pairs" depends upon the rituals used for making them "mate" in the first place...

IRF520 needs a little more current drive in the same voltage range due to a integral reactive component that is hard to keep simple, in simple terms...(stray capacitance looking for stray inductance)

The 13N10 needs less current and is very picky with voltage ranges.

So, as you approach the runway, best to have all your wheels locked down with resistors and caps properly set - because with any of the ones I've mentioned above, all will need a larger cap value that will prolly need to be equal owhen the drive (gate) voltage goes up - so be careful with them - because they'll latch faster than you can remove the power and discharge the caps - then the jobs already finished.

Theres' a lot to be said about moss-fed radios'....

:+> Andy <+:

 

[807]

Agreed. Here is a link to the IRF520 that seems to be so popular. Look at the applications. Anyone see "RF amplifier" listed?

http://www.google.ca/url?sa=t&rct=j...xHNRVd1QsCJkeYg&bvm=bv.58187178,d.cWc&cad=rja

Pretty much the same thing for the FQP13N10.

Quote: "These devices are suitable for switched mode
power supplies, audio amplifier, DC motor control, and variable
switching power applications"

http://www.google.ca/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0CDwQFjAA&url=http://www.fairchildsemi.com/ds/FQ/FQP13N10.pdf&ei=o3fLUsy8Ns7gsASkgoGoDQ&usg=AFQjCNFg4mWUEV9b_Ve3d0NOI8C2GZ1Kag&sig2=GjHrgDrK32kQpWAOV7rTVQ&bvm=bv.58187178,d.cWc&cad=rja

=================
Well, that's a data sheet not a fully fleshed application note. Also "variable switching power applications"??? (What's a class C-D-E amp other than a power switch of variable pulse duration with a resonant tank?)

I see Ron ~0.18-0.25 ohm, Ton < 20ns, Toff < 40ns Cin < 500pF That tells me it's a good candidate for class D or class E for low HF.

They work just fine in non-linear RF amplifier applications. Another important rating is the VDS. Ideally you want VDS > 10x VDD for class D or E. 100V is a bit close to the wind for VDD=12V

They were certainly not designed with linear RF PA applications in mind, but then neither was the 813. It works then it works.

 

[Handy Andy]

Greetings!

What's a few Nano-Seconds amongst friends? 32nS or 37nS - one limits to 11 meter the other to 10 meter MUF

Yes, the fun with MOS - one aspect is their gate capacitance.

It (Gate design) can be used as a crude but effective way to "take a Knee" when it comes to the region just where the switching occurs from off-to-on saturated - many MOS gate to substrate designs have larger regions and others are smaller - and some are just spot-on that helps with the output as a rise and fall artifact when they "bounce" between states of fully on, then back down to off.

A small tweak in fiddling with the Gate voltage - one way - it's latched - the other - a pulse driven waveform shows up not unlike Class C amps would show. (Or pulse train similar to the output of a 60Hz Dimmer Switch for the Kitchen Chandelier)

Not that it can't be fixed, but a lot of the reason for the '520 as a favorite go-to of choice is due to their rise and fall times - and skew/slew rate when doing so.

The rest is simply the networks design to take a nearly square wave output and attach cap and a cup to it...or crest and trough - depending upon preference.

Regards!
:+> Andy <+:

 

[Captain Kilowatt]

=================
Well, that's a data sheet not a fully fleshed application note. Also "variable switching power applications"??? (What's a class C-D-E amp other than a power switch of variable pulse duration with a resonant tank?)

I see Ron ~0.18-0.25 ohm, Ton < 20ns, Toff < 40ns Cin < 500pF That tells me it's a good candidate for class D or class E for low HF.

They work just fine in non-linear RF amplifier applications. Another important rating is the VDS. Ideally you want VDS > 10x VDD for class D or E. 100V is a bit close to the wind for VDD=12V

They were certainly not designed with linear RF PA applications in mind, but then neither was the 813. It works then it works.

Running class D or E or whatever in PWM is different than using them as an RF amp which is what is being done in CB radios. No SWPS or PM action happing there just straight 27 MHz amplification. THAT is what I was talking about when I said they were not designed for RF applications like that. Will they work? Sure but not great in those type of applications. The 6KD6 was never designed for RF either but even a couple OLD commercial amateur amps used them.

 

[Tallman]

IRF520 needs a little more current drive in the same voltage range due to a integral reactive component that is hard to keep simple, in simple terms...(stray capacitance looking for stray inductance)

The FET is a voltage controlled device not current. There is a small amount of current that charges the gate capacitance but not sustained current.. To reduce latch up a resistor from gate to source will stop that process.

Stray Capacitance looking for stray Inductance? Really?

https://electronics.stackexchange.c...-can-we-not-assume-that-the-gate-current-is-0

 

[Handy Andy]

Greetings!

The FET is a voltage controlled device not current.

We're in agreement here...

Stray Capacitance looking for stray Inductance? Really?

Referring to the "plate charge" transfer rates. IRF520 are just ... slow ... compared to others like the 13N10. I find that the larger drive levels (not just voltage to turn on but current available to "swamp" the gate and allow RF across the Source Drain regions) is not as a "high-impedance" state as the 13N10 - the "13's" just need some voltage with not a lot of current (as much as their IRF counterpart) to work their "magic" but IRF's need resistance networks of values in ranges of 10X to nearly 40X LESS OHMIC value to provide the current in mA and uA "volume" to offset and quicken their on to off times - while they still need voltage.

I'm not saying IRF's are bad, just more difficult to work with due to the slope that you can develop but using such network values - the benefit of this is the IRF 520 will develop a nearly linear Knee like a bipolar that the 13N10 cannot.

The gate surface design makes it (13N10) act more like IGBT designs - heavy capacitance...that "cap and cup" form easier on a gate with some capacitance that works within the MUF as a reactive component - at least the network the Bi-polar has used, can be used with little to no tweaking in putting the (blasted) radio back into service. I haven't even mentioned the slope and skew rates the "13's" have. Oops - Cat's outta' the bag...Gosh Darn It...

At least you can see it in the datasheets.

The article references a "leakage current" - Stray looking for Stray - and latches on...idle currents are like idle hands - can be the "Devil" in the details...

Not quite Zero - and even when trying to restore "static" state, IRF's tend to "bounce".

And remember...

Which FET works the best?

IN CB 11meter, it seems to be the one that barely makes it to the MUF - like the 13N10 versus the IRF520 - which NEEDS capacitance to "round off" their relative (to 13N10) digital response - killing the effect you wish to keep as an internal response to the signal on the gate...

Again, to put this back on the rails - which is the best? When you're looking for a Bi-polar drop in - I choose JIF - er, 13N10s'.

Why - IRF520 is not as simple as a 13N10 - gate bias design aside...output network for Bi-polars favor the 13N10 over the 520's - but as a simpler drop in - not necessarily as a performer - just an easier actor to throw into the movie that better fits the part versus waiting and servicing the versatile, yet needs supporting staff - IRF520... still fits the part but needs a lot more work...

Again...my reference to; stray capacitance to stray inductance - things to look out for when you cross that bridge...MRF477 is a painful example of the very thing I'm referencing - and that was a Bi-polar...

Regards!
:+> Andy <+:

 

[Tallman]

Greetings!



We're in agreement here...



Referring to the "plate charge" transfer rates. IRF520 are just ... slow ... compared to others like the 13N10. I find that the larger drive levels (not just voltage to turn on but current available to "swamp" the gate and allow RF across the Source Drain regions) is not as a "high-impedance" state as the 13N10 - the "13's" just need some voltage with not a lot of current (as much as their IRF counterpart) to work their "magic" but IRF's need resistance networks of values in ranges of 10X to nearly 40X LESS OHMIC value to provide the current in mA and uA "volume" to offset and quicken their on to off times - while they still need voltage.

I'm not saying IRF's are bad, just more difficult to work with due to the slope that you can develop but using such network values - the benefit of this is the IRF 520 will develop a nearly linear Knee like a bipolar that the 13N10 cannot.

The gate surface design makes it (13N10) act more like IGBT designs - heavy capacitance...that "cap and cup" form easier on a gate with some capacitance that works within the MUF as a reactive component - at least the network the Bi-polar has used, can be used with little to no tweaking in putting the (blasted) radio back into service. I haven't even mentioned the slope and skew rates the "13's" have. Oops - Cat's outta' the bag...Gosh Darn It...

At least you can see it in the datasheets.

The article references a "leakage current" - Stray looking for Stray - and latches on...idle currents are like idle hands - can be the "Devil" in the details...

Not quite Zero - and even when trying to restore "static" state, IRF's tend to "bounce".

And remember...

Which FET works the best?

IN CB 11meter, it seems to be the one that barely makes it to the MUF - like the 13N10 versus the IRF520 - which NEEDS capacitance to "round off" their relative (to 13N10) digital response - killing the effect you wish to keep as an internal response to the signal on the gate...

Again, to put this back on the rails - which is the best? When you're looking for a Bi-polar drop in - I choose JIF - er, 13N10s'.

Why - IRF520 is not as simple as a 13N10 - gate bias design aside...output network for Bi-polars favor the 13N10 over the 520's - but as a simpler drop in - not necessarily as a performer - just an easier actor to throw into the movie that better fits the part versus waiting and servicing the versatile, yet needs supporting staff - IRF520... still fits the part but needs a lot more work...

Again...my reference to; stray capacitance to stray inductance - things to look out for when you cross that bridge...MRF477 is a painful example of the very thing I'm referencing - and that was a Bi-polar...

Regards!
:+> Andy <+:

It is true that there is not a ZERO current, but what current is there is in the nano amp range which is negligible as compared to any Bipolar transistor. There has always been differing requirements for biasing transistors. The MRF455 needs different bias as compared to the 2N2290 or the SD1447. Different birds need different care and feeding. Most FET's need a different setup between the models and brands.
Circuit parameters do not seek each other out. They simply exist and they are what they are, Inductive, Capacitive, or Resistive.

 

[Handy Andy]

Greetings!

A good example lies in the past...
http://pdf1.alldatasheet.com/datasheet-pdf/view/115255/MOTOROLA/MRF477.html

When you referred to the 455, I was referring to a Service Bulettin from Uniden a long time ago dealing with rear panel assemblies of the HR2510 that used the once novel MRF477 - now extinct - which had a BEC design versus ECB OR BCE - meaning the TAB of the TO-220 was Emitter as well - making this a desirable trait to have for those that wanted to remove the ceramic mica heat-sink goo impregnated wafer that was needed otherwise for BCE Bipolar designs in RF back panels - it was a way to simplify the installs of RF transistors to heat sink panels ... hence the "seek" metaphor was due to my familiarity with the event of using these Novel transistors and my familiarity with the article.

Perhaps another way to look at this would be to look it as a Classified Ad that says...:"Upwardly Wandering Capacitive Charge seeks Highly Regenerative Inclusive Inductive Effect to Generate A Low-Resistive Effort On a High-Impedance Platform To Produce Parasitics - Resume Available upon request..."
​

When I referred to seeking, I was reflecting back to a time where "If you want to build an Oscillator - make an Amp..." during a time I was cutting teeth on a lot of the stuff we now take for granted...

I've seen a beautifully tuned radio turn into a squealing mess just due to poor lead dress...I guess since I see that result so prevalent in poor assembly so much under the hood even on new radios right out of the box - it seemed a fitting phrase to use because as one becomes more prevalent in the effort to develop power - instead of cutting it down, they add something to "Counteract" or exhibit out of phase characteristics - which is what you need to do - but in the end even the best layouts and PC-board plans using a Mouse done by Man - lie in ruins due to the inconceivable yet possible to be foretold by a simple reference to a Murphy's' Law - parasites LCR's are...

So I was referring to parasitics that exist in the real world because even the best designs cannot handle seemingly quantitative states of activity within somuch as to refine an interest if not a curiosity trait not unlike that of the casual observer looking into the box to see if they need to feed Schrödinger's Cat - I guess that's why it's called the "human factor"...

If you don't understand any of the above - that's ok - I was seeking by open pleading to the powers that be - "Seek and Ya' shall figure it out someday..." - sticks well to this side of the box - says Confuzion....

Oh I wanted to caution those that pass by that may be reactive to allergens...

Ingredients:
CAUTION: Contains Nuts...

 

[bologna_sammiges4me@yahoo]

Has there ever been built before a rf amplifier utilizing all mosfet irf520,s or simular that capable of 1000 watts ?

 

[nomadradio]

The Galaxy DX98VHP uses four push-pull pairs of that kind of MOSFET. Eight of them. We call it the "Firestarter senior".

Unless you put a separate fuse on each pair of transistors, a breakdown runs the risk of eight transistors' worth of damage, since it has to allow enough current to run all eight of them at full power. Fuse won't trip until you exceed 8 transistors' worth of operating current. Naturally, an overload factor this high will fry circuit board traces.

Apparently this is not considered important by the folks who designed it.

73

 

[brandon7861]

Has there ever been built before a rf amplifier utilizing all mosfet irf520,s or simular that capable of 1000 watts ?

I've been wanting to try that for a while too. A long time ago, I ordered 100 520's from mouser and they all test identical in my BSIDE transistor tester. Somewhere I have schematics for an amp using those, but I'd have to dig around in some hard drives for it.