Limit power of exciter

[Ohm]

Is there a simple way to limit the power of an exciter before feeding a amplifier e.g a simple resistor? For example how would you limit 1W down to 300mWs?

[rigmo]

https://www.everythingrf.com/rf-calcula ... calculator
https://www.leleivre.com/rf_pipad.html
https://a29.veron.nl/hams/pa1b/pa1b-pow ... alculator/
http://www.engineers.com/calculate.htm
https://chemandy.com/calculators/matchi ... ulator.htm

drop power supply if is possible if you know how... from 12 to 9v,,,

[rigmo]

[Ohm]

Perfect many thanks for the help

[teckniqs]

you shouldn't need any of that, you should just be able to decrease the voltage.

[Albert H]

Unfortunately, decreasing the supply voltage will change the match impedance. If (for example) you have a 1 Watt stage running from a 12V supply, the collector impedance Zout will be 72Ω (using Vcc²/2Po). If you reduce the supply to (say) 9V, and you get 500mW the collector impedance will be increased to 81Ω. This will give you a mismatch into your output network and could impact on the harmonic cleanliness of the exciter.

There are ways of mitigating this kid of issue, but they're well beyond the scope of the level of design skills in pirate transmitters! It's worth bearing in mind that the match will change as you vary the supply voltage.

That said, I tend to prefer to keep my matching network as designed, run at the full watt, and use a resistive attenuator to drop the power to exactly the level I want. I usually use carbon resistors or surface mount resistors for this (they're non-inductive). The only thing to remember is that the power you're wasting in the attenuator is lost as heat - the resistors may get slightly warm.

https://leleivre.com/rf_pipad.html is a useful on-line calculator for unbalanced attenuators. You'll have to make the required values up out of series and parallel resistors, of course.

[BriansBrain]

Here is another on-line calculator

This one you can change the calculated resistor values to the nearest Preferred Resistor Values

https://chemandy.com/calculators/pi-att ... ulator.htm

[Maximus]

As teckinks said. Lower the voltage of final transistor. Using a resistive attenuator will radiate rf unless you’re using the correct resistors.


Sent from my iPhone using Tapatalk

[Albert H]

I'll say it again - reducing the supply voltage will necessitate a change in the output match.

I put my resistive attenuators into little tinplate boxes with the lowpass filter in an adjacent box.

[teckniqs]

I thought it might change a little bit but I didn't realise it would change as muich as that.

....What about if you still kept the same voltage into the 2N4427 at 15 volts but just lowered the voltage into the RF stages before it?

[rigmo]

exsample

[radium98]

In this Albert is correct , was surprised many times when i lower the driver bfr96 in the 300 mW or 1W version of the rdvv it come to sprug and spirius every where on the band , like the shit chinese bh14xxx tx , i dont have a spectrum the see what is going ,or need to put some dump resistor on the collectors etc...bit this what i can listen on the radio at few meters away.

[radium98]

here it is

[Albert H]

You really don't want to mess with the carefully calculated match that has been designed for you! A resistive attenuator is by far the easiest way to tame the exciter power!

[Albert H]

Your other option (of course) is to build a new exciter that delivers exactly the power level you need. I had a need recently for some ~300mW drivers..... No need for a 2N4427 - I used a pair of BF199s for the output I needed. They're really cheap (in quantity, I get them for about $0.04), and so a simple rig with a pair if BF199s for the oscillator, through a bandpass filter, driving another pair to 300mW out, did the job. I used the SAA1057 and a 16F628 for the PLL (sniffing the output) , and the power was precisely what was needed for the job, at little expense!

[Maximus]

In this Albert is correct , was surprised many times when i lower the driver bfr96 in the 300 mW or 1W version of the rdvv it come to sprug and spirius every where on the band , like the shit chinese bh14xxx tx , i dont have a spectrum the see what is going ,or need to put some dump resistor on the collectors etc...bit this what i can listen on the radio at few meters away.

I had the same problem. It turned out to be the pwm power regulator stage. Changing a few capacitors solved the problem.


Sent from my iPhone using Tapatalk

[radium98]

thank you but this is not a pwm it is a pot and a bd139

[Albert H]

It doesn't matter what it is - varying the supply voltage will necessarily vary the match impedance!

[radium98]

Sure , so why Pira 5w in example does not have a variable output ?
then why rdw version from 300 to 5w have a variable output ,i was thinking off adding a pwr ctrl to my pira tx then i stopped ,i can add it then at the final output ,is it better ,or also the matching will vary a lot .

[Albert H]

The equation for the output impedance of a transistor (or FET) is:

Zout = Vcc² / 2*Pout

....where Zout (in ohms) is the output impedance of the device, Vcc is the supply voltage, and Pout is the power in Watts you're sending.

The Impedance varies quite radically with the supply voltage, because it's affected by the Square of the voltage. If your supply is 12V, Vcc² = 144. If your power output is 4 Watts, the output impedance at the collector of the transistor is 144/8 = 18Ω. Obviously, you need to transform the impedance up to 50Ω for a convenient match. (You could build an aerial that matches 18Ω, but I don't know where you'd find 18Ω characteristic coax!).

If you reduced the supply to (say) 8V and expected 1 Watt, your output impedance would have changed to 32Ω.....

There are ways of getting a good match - Steven Moss (NRG) used a "conjugate" match so that the output impedance conversion network would match reasonably well at both 4 Watts and 1 Watt. It wasn't perfect, but it was remarkably close. There were tens of hours that went into developing the PLL PRO III output stage! He wanted to keep it in Class C (for best efficiency) and found that the PA was around 83% efficient at 4 Watts, and about 68% efficient at 1 Watt.

As ever, electronic design is a set of compromises!