Congratulations excellent construction.
It is possible to publish a diagram of the linear, maybe even a block diagram.
carlo IZ2ABZ
Congratulations excellent construction.
It is possible to publish a diagram of the linear, maybe even a block diagram.
carlo IZ2ABZ
Done plenty of 80m Uk local over the past few days, with the MKARS80 qrp rig connected to the amp. Only got an antenna that works well on 80m at the moment, a 1/4 wave inverted L. When I get bored with 80m, I'll take down the L and put up a 40m dipole :)
Block diagram will be posted soon.
For details of the amplifier module itself, see the EB104 application note
http://www.ab4oj.com/dl/eb104.pdf
Modifications done to amplifier
- Good quality RG58 coax screen used as grounding strap across the FETs
- Thick DC gnd to centre of board
- 4.7k resistor inserted in series with the thermistor to level off the bias temperature compensation
- High voltage large capacitors replace 0.1uF ceramics on PSU feeds to mosfets.
- 820pF caps across transformers
Last edited by mw0uzo; 03-09-2013 at 08:13 AM.
Thanks for the precious info.
Since I'm also building the fates would like a few questions.
The resistance from 4.7 k in series with the thermistor goes to sostiyuire the R10 (8.2 k).
The capacitor 820pF goes put in place of C11 (2 x 1200 pF).
Reading on the Internet I found that the capacitance C1 is made from 68 pF to reduce the input VSWR, and C11 should be placed from 300pF to increase the power on 30MHz.
The supply voltage varies from 45 to 42 V to decrease the ROS, the current BIOS varies from 150 to 120 mA.
Which of these tips did you use?
From your block diagram I saw that you used 12 analog inputs on the PIC (the 16F877A has only 8 analog input) how did you do?
Why are there two 600W SWR Bridge?
Thank you very much.
Carlo IZ2ABZ
Yes it adds 4.7k to R10, for a total of 12.9k.
Yes, replace each one with 820pF.
C1 is not fitted on my amplifier.I have only tried 820pF capacitors on the transformers, reducing them further may increase gain on 10m. However, you may reduce efficiency on 160m or 80m.
My bias is set at 250mA each for the amplifier, but I did experience problems when the amplifier was being tested by itself. I had to set it to 120mA each to avoid spurious output. Always a good idea to start testing with low bias, as it helps reduce gain and avoid death oscillations. My power supply voltage varies from 60V unloaded, to about 47V when loaded heavily. As my supply is linear (40V AC winding) and not easily adjusted, then I have done no tests with lower supply voltage. I chose to go linear to avoid any chance of RFI and because of doubts about the instantaneous current handling of an unknown SMPSU.
I am using 16F887, 40 pin DIL, which has 14 possible ADC inputs.
I use 12 inputs:
AN0 - V
AN1 - I
AN2 - Temperauture
AN3 - ALC level detect in
AN4 - 10W Vf
AN5 - 10W Vr
AN6 - Out Vf
AN7 - Out Vr
AN8 - Amp Vf
AN9 - Amp Vr
AN10 - 100W Vf
AN12 - 100W Vr
with one spare for ALC level detect, which is not implemented.
So that faults can be detected if something goes wrong with the LPFs. With only one SWR bridge a failure in the LPFs could kill the amplifier. The same applies to the 100W and 10W bridges, they are for RF detect and fault detection purposes. It is easy to get something go wrong in a large project like this.
Last edited by mw0uzo; 03-11-2013 at 10:29 AM.
Thank you very much for the useful tips.
Add the resistors and capacitors suggested.
I immediately start to write the code for the 16F877 adding two SWR bridges suggested.
Carlo IZ2ABZ
Sorry,
bat
- High voltage large capacitors replace 0.1uF ceramics on PSU feeds to mosfets.
Are C13 and C14? or C17 and C18?
regards
carlo IZ2ABZ
C13 and C14. I used some large HV caps out of an old TV, I think they were 0.22uF. They have a tendency to explode under fault conditions, and if RF is injected when the amplifier is off, a good proportion of it finds itself across these capacitors, via T2, so they end up dissipating power.
Hi,
What are the characteristics of current and voltage of the relays.
Why in the EB104 uses 5V.
Why LPFs uses bands 30/17/12 and not, 20/15/10?
Best Regards
Carlo IZ2ABZ