Re: UBITX Q90 mod with BFR106
|
Glenn, What is the actual rf power being delivered by the 2 ea 2n2222a transistors in the predriver section?? Are you able to measure that?? I wonder if 2 ea 2n5770s would be a better transistor in that location.? Spec sheet shows a Po of 30mw each @ 500mhz...? I have attached a copy of the 2n5770 spec sheet. Do you happen to have this transistor around to try?
Thank you again for all this work you are doing!
Marco - KG5PRT?
On Wed, Sep 19, 2018, 12:43 AM Glenn < glennp@...> wrote: Strangely enough, bypassing had little effect.
I fitted a 150pF across the primary of T11 (output trans. wound 2t : 4t) and this made a huge difference.? 28MHz is now down about 1.8dB. Bottom end is affected a little also as you would expect.
The 150pF is no magic value, just first value pulled from the parts drawer, that that was fairly high voltage and seemed 'right'!? A disc ceramic. Further value changes might be interesting.
As Allison alluded to in some posts, the PA is susceptible to RF feedback with the heatsinks at RF hot voltages. (non insulated) I can see some instability at about 35MHz which goes away when the input RG174 (Chinese) coax cable is removed or some rough shielding of the 1st stage is put in place.
The board values are per the Wiki in this group, by Allison.? I did alter the turns on T9 down to 5 turns at one stage and it's still there. In my build I used a 2 turn bifilar wound on an unknown type balun? core to DC feed the IRF510's. ( ie there are no 60uH chokes)
I think one should not try to get more than 10W out of this PA. As built at least. Input level should not exceed -20dBm. Bias is about 200mA.
Glenn vk3pe
|
