Re: 10W PA and Ultimate 3s #u3s #pa

Hans Summers

Hi Thomas

The output of the Si5351A direct is likely a little weak to drive the 10W Linear to full power 10W output. I'd suggest keeping the U3S PA and adjust the U3S bias pot (R5) for around 25mW +/-, i.e. whatever is necessary to get 10W (or the desired output, not exceeding 10W) from the Linear. 

The QCX power meter isn't accurate and particularly not down at 25mW. And maxes out at 5.2W so won't be useful for measuring the 10W output of the Linear either. 

My favored method of power measurement is 50-ohm dummy load and oscilloscope. But yes, unfortunately your 100kHz 'scope wouldn't be up to the job. 

My suggestion just set up the 10W Linear as per the manual, and connect a 50-ohm dummy load and power meter... if you have the QRP Labs dummy load then use a DVM on the rectified output and reference the curve in the dummy load manual... then connect the stock U3S to it, and adjust U3S R5 until you get the desired output power (not more than 10W). Something like that...

73 Hans G0UPL 

On Sun, Jul 14, 2019, 15:12 Thomas Fritzsche <tf@...> wrote:
Hi Hans, Hi OM's,
thanks so much for you help!

I have checked the U3S schematics and notice that the Si5153A is
almost directly connected to PA/LPF. Now considering to buiild PA on a
breadboard first (as I have not yet soldered transformer T1 or Mosfet
Q1). I would take the signal after coupling capacator C1.
Should this work or would I create problems due to longer wires?

Bifilar or "25 turns" winding? The manual suggests "25 turns" for 5W
PW, but I assume this is due to the higher power output (frequency
depended). As I need to adjust bias anyway can I use bifilar winding?

I was thinking about skipping the U3S PA completely and directly
connecting the 10PA kit, but the manual explains the Si5153A can only
have limited current. Do I understand this correct?

One more question: The manual explaines to meansure the power output
while adjusting the Bias. Considering I need to be very accurate: How
to best measure it?
Some power meter you can recommend?

The QCX I build has a simple power meter using a diod detector circut,
but the manual says it's not accurate?

Sorry a little off-topic: I have a Bitscope-Micro occilloscope ( ), but it's not nearly as fast as
advertised, maybe a few 100KHz at max., so not enough for the RF
output in the 20m band.

Looking for a budget friendly upgrade to help me tinkering QCX (and
QSX one available).
Any advice on what to look for?
* Hantek as a 100Mhz version for 200 EUR that be patched to 200MHz.
* UK Company Pico Tech has a 50Mhz model for 350 EUR

73, Thomas - DJ1TF

Join to automatically receive all group messages.