Broadband Mosfets / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Broadband Mosfets

Move Thread

LAN_403

MRacerxdl

Mon Jul 20 2009, 05:56PM

Registered Member #989 Joined: Sat Sept 08 2007, 02:15AM
Location: SÃ£o Paulo, Brazil
Posts: 476

After a long time with no projects, I am thinking in a new one.

I have here two Broadband mosfets MRF6VP3450HR5 ( Vdd = 50V Idq = 1.4A ), They are made for broadband transmitters (860Mhz)

And I was thinking about in make some type of small high frequency SSTC's (not much as 800Mhz =P ), in Class B Push-Pull Mode, these mosfets have two mosfets on one case and have a good base for heatsink ( I can adapt it to run on a good heatsink)

My Only question is about the work frequency, its about 860Mhz, and I plan to run on a maximum of 4Mhz... That is a problem for the mosfets or something so? I never used such VHF Mosfets, but I think that is the recomended frequency for work...

I have two of it, but I plan to use only one, that mosfets are a little expensive =P
(If someone have another ideas of projects that I can made with that, I want to hear )

Hazmatt_(The Underdog)

Mon Jul 20 2009, 08:57PM

Registered Member #135 Joined: Sat Feb 11 2006, 12:06AM
Location: Anywhere is fine
Posts: 1735

I've been working on a solid state drive to drive coils like an antenna and the biggest thing you need to consider is that the coil is such a bad mismatch ( ~ 200 ohms in my cases ) and the reflectivity from the coils is so bad, that your mosfets will die the first time you go to draw an arc.

If you still plan to build a SSTC transmission line driver, I'd recommend building the amp first, characterizing your coil and find its load impedance with a signal generator, then building a matching transmission line transformer so you don't kill the mosfets.

Other then that I would build a broadband amplifier for testing purposes because its a good exercise and if you look up the prices for a good wideband power amp, its multiple thousand dollars.

Marko

Mon Jul 20 2009, 09:15PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

MRacerxdl wrote ...

After a long time with no projects, I am thinking in a new one.

I have here two Broadband mosfets MRF6VP3450HR5 ( Vdd = 50V Idq = 1.4A ), They are made for broadband transmitters (860Mhz)

And I was thinking about in make some type of small high frequency SSTC's (not much as 800Mhz =P ), in Class B Push-Pull Mode, these mosfets have two mosfets on one case and have a good base for heatsink ( I can adapt it to run on a good heatsink)

My Only question is about the work frequency, its about 860Mhz, and I plan to run on a maximum of 4Mhz... That is a problem for the mosfets or something so? I never used such VHF Mosfets, but I think that is the recomended frequency for work...

I have two of it, but I plan to use only one, that mosfets are a little expensive =P
(If someone have another ideas of projects that I can made with that, I want to hear )

One problem I can think off is immense cost per watt of power of RF mosfets compared to readily available polysilicon gate mosfets. 50V at 1.4A is also really useless for a tesla coil, despite being great for microwave transmitter usage which the transistor is intended for.

At just 4 Mhz, I don't believe RF mosfets would offer any advantage over polysilicon ones. And either would blow up just equally if used incorrectly, as RF mosfets really don't magically lack output capacitance, and losses about as high as with polysilicon mosfets are expected if no zero-voltage switching is employed.

Marko

Sulaiman

Mon Jul 20 2009, 09:39PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

The mosfets should be 'better' at 4 Mhz than at 860 MHz
but the input/output impedances will be different
so a direct copy of stiplines etc. may be less than ideal.(useless)
I suspect that the power output available will be rather disappointing arc-wise
experiments with Transmission & reception of power by weakly-linked resonant circuits
(TC transmitter & TC [backwards] receiver)

Study the output circuitry of the 860 MHz transmitters to see how they protect against SWR
which is the same result as a sstc arcing.

At 4MHz there's a good chance that you'll radiate rf, since it's small & unintentional it's ok
BUT you'd be wise to ensure you're not near an emergency services frequency.

MRacerxdl

Mon Jul 20 2009, 10:12PM

Registered Member #989 Joined: Sat Sept 08 2007, 02:15AM
Location: SÃ£o Paulo, Brazil
Posts: 476

Yeah I was think first to made a RF Amplifier to educational purposes at frequency near than the mosfet expected. But I live near an Airport and that can be a problem, I didnt see what is the frequency that airport comunications is, but I know that is near the mosfet ratings, so I want to make a think that works far from that frequency.

I was thinking about using a tesla coil to transmit data over-air (sstc one), but I think a common transmitter will be better in this case... I will study the circuit that is on the datasheet to see how it works (its actually complicate to build one of that, smd parts sucks for soldering =P )

I tested the mosfet at audio frequencies, and I dont know if its the Gate Sensibility that is higher or something else, but the audio is MUCH better than a IRFP260/460 ones. The circuit is very stupid, only a voltage divider at 5V Supply to get the 2,5V Quiescent voltage, and a 10uF decoupling cap for the audio input.

GeordieBoy

Mon Jul 20 2009, 10:30PM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

Keep those UHF push-pull MOSFETs for an RF power amplifier or sell them on Ebay to someone who wants to build a UHF linear amplifier. The 100W or so of power they produce would barely be enough to overcome conduction and radiation losses, so you'd be lucky to see any breakout.

You can easily produce hundreds of watts of power at frequencies up to around 16MHz with standard switched-mode power MOSFETs. These are also a lot cheaper and more commonly available than dedicated RF modules.

Regarding Hazmatts comments about Impedance matching and reflected power: Linear amplifiers are far more tolerant of load mismatch, or reflected power (high VSWR or whatever you like to call it,) than the high-efficiency switching amplifiers like Class-C, Class-D and Class-E. It stands to reason that a linear amplifier that already wastes 80% of it's input power in the heatsink can easily dissipate a little more under load mismatch conditions. Conversely, a 95% efficient switching amplifier that is only designed to dissipate 5% of input power as heat, will die very quickly if this figure changes to 100% due to load mismatch. That's why RF labs and EMC test houses use huge linear amps for compliance testing where the load is "badly behaved".

-Richie,

Hazmatt_(The Underdog)

Tue Jul 21 2009, 11:13AM

Registered Member #135 Joined: Sat Feb 11 2006, 12:06AM
Location: Anywhere is fine
Posts: 1735

Yea I did read about Class A vs. Class C, and I agree with that Richie.

My comment was to point out with mainly the load matching for this reason:
When you design your RF amp or broadband amplifier and do your load testing, you find that it has a characteristic impedance output that it readily matches to, which will deliver maximum power.

Consider the broadband power amp I'm working on right now.. I was intending for an amp that goes from audio to several hundred KHz, but it only goes up to about 100KHz. It's load matching output is roughly 8 ohms. Well the scenario is that if you hook the base of the TC up to that expecting a load match and a sine wave, you're sadly mistaken. You will not have either by any stretch of the imagination. Insted you will get some horrible looking ringing poor excuse for a waveform, and nothing on the TC side, why? Load mismatch.

My comment was mainly to point out that if you want to deliver the power into the coil you are first going to have to use a balanced transformer so that the amp is properly loaded to delivering maximum power without ringing, and then match it up to the ~200 ohms of the prospective coil in my case.
so I would need a 8 ohm to 200 ohm impedance match to meet that goal.

Secondarily, you need about 15 Watts of RF drive power to get a 1cm breakout. I have done this a few times from two crummy amps. One was my GR tube type amp with 50 ohm output, and the other times have been a mosfet transformer output line driver. Both have seriously high voltage ratios, so basically what is required is ~200 Volts and 15 Watts.

To base-drive a coil you need over 50 Volts or you're not going to see anything at all, but with 200 V you have something to ring-up the coil with.

I should have made a 1:4 Balun for the GR amp to deliver the proper load match, but something opened up before I could get to that. Had I done this, I probably would have seen an inch or so of output from the TC. To date I have only gotten half that. But you can get output with ~15W, I have done it, is it spectacular.. NO.. but this is for characterizing the coil .. not for impressing onlookers.

Anyways, if you build the amp and load it you won't interfere with air traffic. You would have to do this regardless of what you plan on doing so you can test the amplifier, otherwise it may burn up.

Hold onto the transistors when you feel like making your pirate radio station, and use a cheaper part in the meantime, it'll be worth risking a cheap part for education. Blowing $$$ parts on simple mistakes always eats you up on the inside.

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob