ad: UR5CDX-1

IRF510 single Mosfet 40-meter amplifier

Discussion in 'Homebrew and Kit Projects' started by W6NFU, Apr 1, 2016.

ad: L-HROutlet
ad: l-rl
ad: Subscribe
ad: K5AB-Elect-1
ad: MessiPaoloni-1
ad: Left-3
ad: L-MFJ
ad: Left-2
  1. W6NFU

    W6NFU Ham Member QRZ Page

    I just homebrewed the NB6M Miniboots 40-meter amplifier using a single IRF510 Mosfet. Driving the amp with a 2-watt output Norcal40A, both inputs were tried, ie a 3db resistive attenuator and also a broadband transformer input. Both methods initially amplify to about 4-watts output and if the key is held down for about ten seconds, the ouput climbs to 14-watts and will stay at that level during a normal cw qso. The heat-sinked Mosfet heats up to around 138-degrees F.

    The goal is ten-watts output which can be attained by adjusting the amplifier Drive adjustment after about ten seconds of operating. But it starts out around 4-watts output. I'd like it to start out at ten-watts output and stay there! Any ideas on the problem and how to solve it?
    Thanks,
    Dave - w6nfu
     
  2. KB1YFD

    KB1YFD Ham Member QRZ Page

    Hi Dave,

    Do you have any test equipment to measure the RF chain?

    Measure the amps input and output to see where this gremlin lives.

    Maybe you have two wattmeters one for input and one for output.

    Check the power supply as well.

    Good luck with your kit.

    John KB1YFD
     
  3. W6NFU

    W6NFU Ham Member QRZ Page

    Thank you John. I will divide up the Driver (Norcal) from the amp and check each stage of the filter input and output with wattmeter, SWR indicator, and current measurements.

    Will get back to this Forum with any found solutions. Thank you for your response and help.

    Dave - w6nfu
     
  4. N7DMA

    N7DMA Ham Member QRZ Page

    Funny thing...

    I just listened to the QSO Today podcast, and Eric interviewed Ashhar Farhan, designer of the Bitx xcvr.

    I looked Ashhar up on the interweb, and found his blog. He addresses this amplifier design as a dual band amp, juiced up to 25 watts.

    See:

    http://hfsignals.blogspot.in/p/25-watt-linear-for-40-and-20.html

    73,
    Karl
     
  5. N7EKU

    N7EKU Ham Member QRZ Page

    Hi Dave,

    Looking at the circuit, the only thing I can think of is that perhaps the drive adjust pot is having trouble handling the power going through it. Since you seems to be saying that the amp can produce no more that 4W from a cold start, can you try just bypassing the pot and see what you get?

    73,

    Mark.
     
  6. W4KJG

    W4KJG Subscriber QRZ Page

    Over the years, I've built up several single band IRF-510 single-ended Class A, AB1 and C amplifiers. I've also built up several push-pull single band Class B RF amplifiers over the years.

    Looking at the circuit used, I'm not sure I see any kind of heat compensation. Maybe that is what the 12 volt Zener is for. However, as an IRF-510 starts to draw current, it will obviously start heating. As it heats, it draws more current. Generally, it will be self-protected -- but its amplification will change with drain/source current draw,which is a factor of the gate voltage.

    If the Zener is meant for temperature compensation, it must be thermally coupled to the IRF510. I've previously used cheap 1N400x series diodes, "globbed" with heat sink compound, and tightly mounted (by pulling the leads tight through the PC board) at the top of the IRF510 and the screw head. These diodes are then used in conjunction with a trimmer resistor to set bias at any reasonable temperature. The diode should change resistance/barrier voltage at about the same rate at the IRF510.

    At the temperatures you mention, an IRF510 should be operating very well. They can provide more power out, but when it is just a 1 or 2 dB, it is better to be where you are operating -- after it warms up.

    Anyway, just those few thoughts.

    One last thought -- but, it doesn't seem like this is a problem. There are a lot of "fake" IRF510s available on the Internet. I'm not sure why I still have it, but I have a pack of 10, from a 12 pack package that I bought from China. They work, but not like real IRF510s. IRF510s are cheap enough to buy from reputable US suppliers like Mouser, DigiKey, etc. Real ones work really well.

    Good luck.
     
  7. N7EKU

    N7EKU Ham Member QRZ Page

    Hi,

    I looked into that circuit a bit more. The trouble is being a MOSFET, the gate turn-on voltages are going to vary a lot from different manufacturers, different lots, and even each device in the same lot. That becomes a problem with this design, because there is no voltage provided for the gate except that which is coming from your signal. If your transistor ends up having a higher turn-on voltage than the original designer's, then the amp will not provide as much gain. You are seeing the lowered turn-on voltage coming from a hot vs cold device, so your amp is only working well when hot.

    So you could try:

    * Randomly changing the device to a different one
    * Measure the turn-on voltage of the devices you have and install the one with the lowest reading
    * Add some kind of gate bias: google around and look at many of the IRF510 SSB QRP amps and use their typical 5V regulator/zener + pot and a few resistors bias (http://f6feo.homebuilder.free.fr/transceiver_simple_20m.html) then bias a bit under turn-on so that the amp stays in Class C even when hot
    * K8IQY also did a nice LED bias system for the Tuna Topper amp and you could try that too

    Oh, and the zener on the miniboots, is not for thermal adjustment or biasing, but for gate protection.

    Cheers and have fun!


    Mark -- n7eku/ve3
     
  8. W6NFU

    W6NFU Ham Member QRZ Page

    Many thanks to each who responded to my Forum post regarding a homebrew IRF510 Miniboots amplifier that takes a few seconds of key-down at 5-watts to reach the desired output of 12-watts. After it reaches that level, it stays there and does not drop off until transmission is stopped during a qso.

    Following advice of KB1YFD, it was confirmed:
    The Norcal 40A by itself into a Dummy Load is rock solid with no change from key down, putting out about 2-watts.
    SWR between the Norcal and the amp is less than 1.2:1 and does not change over time.
    Nothing gets remotely warm to the touch for the first 10-20-seconds of operation.
    Measured Gate voltage is set about 3.86V for an output of 12-watts. It does not change from key down.
    Different IRF510's have been tried, even a IRF530, all do the same thing at key down.
    All resistors and capacitors have been swapped many times.
    The Mosfet is heatsinked with compound between and does not usually exceed 130-deg. F during a qso.
    My Scope confirms power output change at key-down, but the scope reading remains constant at key-down if the reading is taken beween the two filter toroids??
    A RF Meter with short wire antenna out in the front yard under the antenna confirms the power increase problem.

    Mark, N7EKU suggesting bypassing the Drive Adjust pot. The power output still started about 5-w, increasing at about 15-w.
    Ken, K8KJG mentions the cheap-fake IRF510 from China. I admit, that is what I am now using, waiting for 10-each Vishay-brand IRF510's to arrive. But....I have a Ramsey Qamp 40 which started all this, which when first assembled got so very hot. It was modded with most all the mods from the internet, including replacing the two finals Mosfets with RS IRF510's, the last two in our city. The Ramsey now really works well and last evening I removed one of the IRF510's and placed it in my homebrew amp. No change, the power output still goes up with key-down.

    Mark, N7EKU suggested adding a gate bias because "there is no voltage provided for the Gate except that which is coming from your signal". This morning I installed a circuit consisting of a 2.2K resistor from the 12-V input to a 1K pot to ground. The center of the pot went to the gate via a 10-ohm resistor. Gate voltage could then be set anywhere between 0 and about 5+ volts, 3.86 gave the power output of 12-watts that I wanted. Still, the up-swing in power output was there at key-down, the Gate voltage rock solid at 3.86V.

    My next step is to replace the Mosfet with the Vishay brand when they arrive, probably next week. Again, thanks to each who responded with comments. I will definitely post results as they develop, hopefully it will help others also.
    Dave - w6nfu
     
  9. N7EKU

    N7EKU Ham Member QRZ Page

    Hi Dave,

    Was your RF signal test at your antenna with a field strength meter, or a wave meter?

    A MOSFET bias circuit should be regulated so that its voltage doesn't change with voltage sag during transmit. Since you have a QAMP, you can look at their circuit and there should be a zener or three terminal regulator in there.

    Also, the bias should be set by monitoring the drain-source current. Then with Vcc applied to the transistor, very slowly adjust the bias so you can see where it starts to draw current. There's no harm in setting it at an SSB setting (20-50mA or something) even if you are using it for CW. The QAMP manual should also have a procedure for that.

    Ugh. When you devote all that time, thought, and effort into a creation, don't waste it cheap parts from unreliable sources. Digikey, Mouser, Jameco, etc all have reasonable shipping and no minimum orders.

    From that last part it sounds like you get 12W from the start, isn't that good enough? Maybe I read it wrong.

    That's a strange place to take a signal reading (between the filter inductors) but I would expect the 'scope to show increasing signal there too. Do you observe any heating in any of the filter components (inductors or capacitors)?

    73,


    Mark.
     

Share This Page