Re: [softrock40] Re: Chasing a TX problem
- I am away from my bench, but if memory serves me correctly, it was about 22V at the antenna connector into a dummy load. I don't have a micro-watt meter and the 1K slug in the Bird didn't cut it. :)Mike va3mwOn Wed, Jun 12, 2013 at 11:53 AM, yl2qn_andy <yl2qn.andy@...> wrote:
Yes, absolutely agree - measurements at C24 are higher than at the output. For some time it was really gotcha for me :)
--- In firstname.lastname@example.org, "warrenallgyer" <allgyer@...> wrote:
> Picking a nit here Michael: C24 is not a pure 50 ohm resistive load point so your 25 volts does not translate to 1.5 watts. To be accurate you need to make the same measurement across a 50 ohm dummy load on the output. Voltage measurements at C24 are typically higher than at the output. I am guessing you are still above a watt but not at 1.5.
> Warren Allgyer - 9V1TD
- Hi Warren,
> 1) Run the rig at 13.8 volts rather than 12.Absolutely agree. All my experiments show the same - 13.8 is better then 12. Already running on 13V8.
> 2) Remove C22 from the 15/12/10 model. This actually works against your power objective because it reduces the gain of Q6. Q6 however is the primary cause of non-linearity in the RXTX and the additional gain of the C22/R42 combination causes IMD3 products to rise faster than output power.Playing with C22/R42 do not make big difference in gain but have a big difference for IMD3. Additional gain is not big, but IMD products rise really faster. The same result as you report.
Another situation with BS170 instead of 2N2222. More linear, more gain and less IMD3. I'm replace Q6 with BS170 accordance GM3BSC recommendations (thank you Ed !).
> With just these three steps, running at 13.8 volts, I have two RXTX that do better than 1.0 watt on 10 meters with IMD3 at -30 dB and 1.4 watts on 15 for the same distortion.Unfortunately can't got the same results. 15/10 m only up to 0.5W. And these results after only after replace Q6 to BS170 and more audio drive.
> This is with 3.3 volts P-P of audio drive from the PC.Yes, also put 3.2Vpp before change R15/R18 to 3K3. Now need only 1.6Vpp to have clear drive on R25/R26/R27/R28 (max level without signal destruction is 4.7Vpp at oampf output).
> 3) If you have a number of BS170s choose one for Q5 that maximizes the static current drawn by the finals on transmit. The higher the static current, the higher the power out and the better the linearity. Don't worry about the BS170s.... once they are in the circuit they are hard-to-kill beasts! I put a dab of heat sink grease on each of Q5,7, & 8 before installing the heat sink. This is not so much to improve the heat sinking as to make sure all three devices run at the same temperature as this makes the final thermally stable and also stabilizes the IMD3 they produce.Here I have a problems. Yesterday make decision put 4xBS170 instead 2xBS170. Make changes for bias accordance GM3BSC recommendations and put two additional BS170. The transformers left unchanged. Problems start from very beginning - I have a very low voltage across R48/R49 - only 7 mV DC (expected 45 mV). Any way I try to put tone to test output. Out was the same as with 2 BS170 - around 1.5W (80-20) and 0.5W(15-10) but DC current raised to 1A and make the BS170 is too hot. Where is all power :) ?
Andy - YL2QN.