Picking A Mosfet?

Discussion in 'Homebrew and Kit Projects' started by VK4HAT, Dec 3, 2019 at 12:23 AM.

ad: L-HROutlet
ad: l-rl
ad: MessiPaoloni-1
ad: Subscribe
ad: L-MFJ
ad: Left-3
ad: Left-2
  1. VK4HAT

    VK4HAT Ham Member QRZ Page

    Ok, choosing a bjt is an easy task, take the maximum frequency you want to operate at and multiply the amp stage gain and get a Ft high enough to do what you want. Do a parametric search for that Ft or higher and bob is your uncle.

    But what about non RF mosfets? There are like 20,000 mosfets on mouser even if i narrow that down to package type and power dissipation i am left with 300 transistor. So what are the search parameters that are going to narrow things down to something that may work for most of HF? I could buy 1 of everything and then build a test jig and see how each of them goes, but narrowing that list down to say 10 or so potential devices would help a lot here. So what are the most critical parameters that limit frequency of operation in a mosfet?
     
  2. KB7WG

    KB7WG Ham Member QRZ Page

    It will depend on function. Analog or digital? The most important thing about a mosfet, is the gate switching characteristics.

    One needs the right amount of charge at the right time.

    Depending on the function, the gate charge will need to be "tuned".
     
  3. VK4HAT

    VK4HAT Ham Member QRZ Page

    I thought rf amplification was kind of obvious in the question.
     
  4. KB7WG

    KB7WG Ham Member QRZ Page

    My experience with a mosfet gives me a different outlook.

    Tuned current is analog, tuned charge is digital. A slow switch will amplify, a fast switch will pass.

    Is the function output......... rated.....or only stated.
     
  5. KX4OM

    KX4OM Ham Member QRZ Page

    For starters, I'd make a list of the ones that have been successfully used in transmitter/amplifier projects. IRF510 to 530, W7ZOI, Nov 1989 QST + several Feedback articles, 10 to 50W; IRFZ24N, 50W, W6JL, QST Homebrew Challenge; IRFP440 and 450, 300-500W, KE6VWU et al, May 1977 QST; IRF840, VU3PRX, HamRadioIndia; David Cripe (NM0S) - Rockwell-Collins article, Feb 13, 2014; 1kW COTS switching MOSFETS; STP14NF12, FQP13N10, IXTP2R4N50, STW13N60M2, M0RZF's evaluations for a 40W amp. Those are a few notable ones from my files.

    Good luck, should be fun.

    Ted, KX4OM
     
    Last edited: Dec 3, 2019 at 10:07 PM
  6. VK4HAT

    VK4HAT Ham Member QRZ Page

    I do not have a problem with the monkey see monkey do approach, standing on the shoulders of giants and all that. But why did they choose IRF510 over any of the 100's of other fets in the IRF family? What were the parameters that said that these devices might actually be usable in amp applications. What i need is a popcorn fet capable of 1w in an SOT89 SOT223 package, the problem is no one has done that before and I have a choice of about 300 and I cannot buy and test them all, lest i die of old age and boredom trying to characteristic them all LOL So how do we narrow down the list to potential devices? Someone must have a clue about these things.
     
  7. SM0AOM

    SM0AOM Ham Member QRZ Page

  8. K7JEM

    K7JEM Ham Member QRZ Page

    I would look at input and output capacitance, the lower the better, and the rise and fall and delay times. These are all on the data sheets. One candidate might be the ZVN2106, which is available for under a dollar US money. It should be good for a watt or so output, using 12v or more as a supply. Obviously you would need to do all the matching, etc:

    https://www.diodes.com/assets/Datasheets/ZVN2106G.pdf
     
  9. VK4HAT

    VK4HAT Ham Member QRZ Page

    In and out capacitance and rise and fall times, thanks for that guys. That was the info i needed.

    I was pointed by someone to a surplus supplier in China, they had the Rqa0009sxaqs for 50 cents a piece so i just ordered a life times supply. Will do +37dbm dc to 500mhz. If they are legit and not voltage regulators being rebranded then its a steal LOL I will know in a few days.
     
  10. KW4TI

    KW4TI Ham Member QRZ Page

    I don't know if this will precisely answer your question, but I think there's some basic criteria to picking a power MOSFET to amplify RF, and why the IRF510 is so frequently used.

    RF devices tend to be LDMOS (laterally diffused MOSFET) while the cheap devices are power MOSFETs.

    Diagram of LDMOS
    https://www.microwaves101.com/encyclopedias/ldmos
    Diagram of POWER mosfet
    https://en.wikipedia.org/wiki/Power_MOSFET

    The power MOSFETs tend to have very high gate capacitance as compared to the LDMOS mosfets because they are made for SMPS switching applications where they are either on or off and switch in the 10 kHz-1 MHz range, and so linearity is not a concern, only charging and discharging the gate. Because power MOSFETs are usually operated in cutoff or saturation, they are not designed to dissipate as much power as a transistor operating in the linear region does. For example, the new MRF101AN has a gate capacitance of 150 pF, while the IRF510 has a gate capacitance of 180 pF.
    This is main reason I think IRF510 is used, because one can use a relatively low amount of power sources from 50 ohms to drive a few volts swing in the gate voltage.

    Also, comparing to the MRF101AN, the MRF101AN produces 100 W with a 50 volt drain voltage. It can only do this because it is relatively efficient. The IRF510N can have a drain voltage of 100 V, and conceivably such a device might be able to source that amount of power with the same drain voltage, but for example let's say that the IRF510 dissipates 100 watts when it is amplifying up to 100 watts. The ideal thermal impedance of the IRF510 is 3.5C / W, so that to dissipate a hundred watts, the temperature would rise at least 350 C. The IRF510 has a maximum power dissipation of 40 watts for that reason. Compare this to the MRF101AN which has a thermal impedance of 1.1 C/W, so that the temperature would rise by 110 C, and the device can have a total dissipation of 182 W. It is likely the die size is much bigger on the MRF101AN than the IRF510, accounting for the cost differences (as well as volume). Also, the MRF101AN is going to be more efficient to begin with, and so might be more than 50% efficient in a class A-B configuration.

    The problem is that power MOSFETs that can dissipate more power tend to have higher capacitance gates and therefore be harder to drive. For example, the IRFP260 is used a lot by hobbyists for heavy-duty high frequency switching but has a gate capacitance of 5200 pF! So generally only the smallest power MOSFETs can be driven directly by a 50 ohm output source.

    One way I have tried to alleviate this problem is to use a common-drain stage first which is a IRF510 driven by a 50 ohm input impedance, which in turn drives the gates of one or more common-source MOSFETs. The low output impedance of the common-drain stage can do a better job of driving the higher capacitance gates. I have an amplifier design using IRF510s that should amplify QRP-level power to 100 to 200 watts using eight IRF510s, which are together still much cheaper than one MRF101AN. But for higher power probably shelling out the money for a transistor like the MRF300AN is the way to go. You can check out the designs here

    http://www.github.com/profdc9/QRPAmplifier

    73,
    Dan
    KW4TI

     
    Last edited: Dec 4, 2019 at 3:27 PM
    VK4HAT likes this.

Share This Page