EI9GQ 16 Watt Power Amp - Start to Finish

Discussion in 'Homebrew and Kit Projects' started by KK4DAS, Oct 5, 2021.

ad: L-HROutlet
ad: l-rl
ad: Left-3
ad: L-MFJ
ad: Left-2
ad: abrind-2
ad: L-Geochron
  1. KK4DAS

    KK4DAS XML Subscriber QRZ Page


    [​IMG]

    Hi all - I decided to add some big-boy shes to my homebrew QRP rig. I decided to build the 16 watt power amplifier designed by EI9GQ, Eamon Skelton in his book "Building a Transceiver". The amp used two Mitsubishi RD16HFF1 transistors in a push pull configuration. On the bench I measured about 13.25 dB gain across the HF bands. I can't wait to test it in my rig. This was a fun build an not nearly as difficult as I had imagined,.
    Details on the build are on my blog - check it out and follow or like it if you choose.

    https://kk4das.blogspot.com/2021/10/ei9gq-16-watt-power-amp-start-to-finish.html
    73,
    Dean
    KK4DAS
     
    W0RIO and KV6O like this.
  2. N2CI

    N2CI Ham Member QRZ Page

    Nice project. I built a few amplifiers based on the RD16HHF1. For the output transformer, I experimented with both type 43 and type 61 cores and found that type 61 provided a bit more power and better IMD numbers then the type 43. I'm getting about 20 w out across the HF bands with the 3rd order products down 30 dB (PEP) or more. You may also want to try experimenting with a small capacitor (100pf - 470pf) across the low impedance side of the output transformer for improved performance.
     
  3. KK4DAS

    KK4DAS XML Subscriber QRZ Page

    Hi thanks for the note. I appreciate your suggestions and will give them a try. First test looks pretty good - but I am experiencing some low frequency parasitic oscillations that I need to run down. When you say a cap across the low impedance side of the output transformer do you mean one cap bridging the inputs or one from each input to ground? And what improved performance do you expect?

    Thanks,
    73,
    Dean
    KK$DAS
     
  4. N2CI

    N2CI Ham Member QRZ Page

    Are you seeing the oscillations when testing the amplifier standalone, or when its installed in the transceiver ? Also, what frequency ? I noticed that you have more inductance on your choke (T2 - FT50-43 - 8 BFT) than should be necessary. I am only using 3 BFT on an FT50-43 core for the choke. Also, I used #22 wire for the choke and output transformer.

    The capacitor on the output transformer should be one capacitor across the MOS FET side of the transformer. A small capacitor to ground (10 pf) on the output side of the transformer may also improve your output power a bit.
     
  5. KK4DAS

    KK4DAS XML Subscriber QRZ Page

    20 watts clean!

    I put the transmit amplifier chain together yesterday - two 2N2219A stages followed by an IRF-510 stage to get the drive level where I needed it for the dual RD16HHF1s. After adjusting the RF gain stage I am getting 20 watts clean out - which is a bit more than I expected. The input drive level range seems to be fairly narrow and the amp runs a bit warm. Skelton has the bias on the RD16s at 500mA. I may try reducing the bias and see how things shake out.

    Here is the video showing the result...

    73,
    Dean
    KK4DAS
     
  6. KK4DAS

    KK4DAS XML Subscriber QRZ Page

    I was getting the oscillation when tested in the transceiver at around 7.2 MHz - the oscillation look like they are at audio frequencies. My latest bench test posted a few minutes ago does not have the oscillation. 8 BFT on T2 was what EI9GQ specified in his article and book. I'm an old guy but relatively new ham without formal EE training so am learning as I go. How did you determine that the inductance should be lower on the choke?

    Thanks again for you comments and advice,

    73,
    Dean
    KK4DAS
     
  7. N2CI

    N2CI Ham Member QRZ Page

    Usually you want 5 to 10 times impedance in the winding relative to the circuit impedance, so it does not load down the circuit. The circuit impedance at that point should be about 12.5 ohms. Plugging 16 turns at 1.8 Mhz into one of the online calculators shows an impedance of more then 1200 ohms for an FT50-43 core. The 5 to 10 times is a good starting point. The final value should be determined by making measurements in the actual circuit.
     
  8. KK4DAS

    KK4DAS XML Subscriber QRZ Page

    Thanks so much - that makes pefect sense. It gives me a few more things to look at and tune up.
    73,
    Dean
    KK4DAS
     
  9. N5HXR

    N5HXR Ham Member QRZ Page

    Could you elaborate on what the problem is with more windings than 5-10x would be? I feel like I understand why you don't want it too low, but why is too high bad? Thanks!
     
  10. N2CI

    N2CI Ham Member QRZ Page

    There are copper losses that you want to minimize. More turns = longer wire = larger loss. Also, there is winding capacitance that will impact high frequency performance. More turns = more capacitance. You need to find the right balance.
     

Share This Page