I figured I had trouble in the mic amp when I saw the low voltage gain. ?So I removed C106, and the voltage gain jumped to 10. The two diodes in the modulator give a very low impedance load to the mic amp. ? Still 200mv after the fix. ?I found the voltage gain through those BiDi RF amps to be surprisingly low as well, will have to revisit this. ?Perhaps simulate in LTSpice. ?Am seeing on the order of 5W out, so likely all is well with my rig. ?My measurements should be taken with a grain of salt, they were made rather quickly. ?Should include DC levels too. ?But it's a step in the right direction.

Jerry, KE7ER

Sorry, correct that - your af out of A0 should be 5v. The Si5351 output is 3v.

Brian.

Jerry, thanks for posting some ac/rf voltages. A couple of points.

The output of the Raduino is 3v square wave. The mic amp gain is then somewhat higher than your 200mv at C107 would imply, nearer what one would expect but still a bit low. After cleaning/resoldering C107 do you still measure 200mv there?

Brian VK4BAP.

That linear pot should work fine, just that setting the volume just right will be kind of touchy at the low end when using headphones. ?Measure the linear pot with an ohmmeter, make sure you see 10k ohms across the two outside pins. ?Measure between one outside pin and the middle pin, should vary between 0 ohms and 10k ohms as you turn the knob. ?If that all checks out, then solder it back in place, being careful not to have cold solder joints. ?It should work just fine.

Jerry, KE7ER

Not a bad idea. ?I think what Sandeep is suggesting is to use your fingers to keep the gate grounded to avoid a parasitic oscillation. ?

Instead of using my fingers, I would first set RV1 fully clockwise for minimum bias voltage to the gate (remember, pot RV1 is backwards). ?Then add a wire from the IRF510 gate to ground, to make sure no parasitic oscillation takes off. ?Power up in receive mode, with an Amp meter in the PA 12v supply to the IRF510. ?There should be 0 Amps through the meter. ?If there is current through the meter, check if the IRF510 is getting hot. ?If it is getting hot, then the IRF510 is shorted internally, and must be replaced. ?If there is current and the IRF510 is not getting hot, this may be a shorted transformer. ? If there is 0 Amps, then power down, remove the ground wire from the IRF510 gate. ?Power up again with PA 12v to the IRF510 disconnected, check the gate voltage on the IRF510, verify that it is near zero. ?Now connect PA 12v up to the IRF510 again, if it pulls current through the meter and is getting hot, then we have a parasitic oscillation. ?Perhaps add a 10 ohm resistor in series with the IRF510 gate as per that other thread to cure this oscillation. ?Alternately, it could be that just substituting a different IRF510 will cure the oscillation.

Jerry, KE7ER

You are right, his initial post said 0.98A in RX mode, almost a factor of 10 too high. ?I didn't read that closely enough.

He did say that the IRF510 was getting hot, which means it probably is not the transformer. ?There was another thread here where our best guess was a parasitic oscillation in the IRF510 while in RX mode, which can apparently occur even when the gate bias into the IRF510 is shut down at 0 volts. ? ?/g/BITX20/message/22548

Jerry, KE7ER

With 5 Amps, the IRF510 may have died. ?Even when transmitting loud voice peaks, it should only be around 1 Amp, with no sound into the microphone the current into PA_12v connection for the IRF510 should only be 0.1 Amps.

The instructions on hfsigs.com under WireUp say in part: ?"Keep the PA BIAS preset (RV1) to fully clockwise position. Now when you press the PTT button, the ammeter should read about zero. ?Increase PA BIAS without speaking into the mic until the PA current shows 100 ma"

Note that for zero current, RV1 must be rotated fully clockwise. ?That is opposite most knobs in this world, where fully clockwise would be the maximum setting. ?If you did not read that closely enough and started out fully counter-clockwise, that could easily draw 5 Amps through the IRF510 when you push the PTT button, killing the IRF510. ?I doubt any other parts were damaged. ?

The IRF510 is very cheap, buy several when you place and order. ?If you cannot easily buy them in Chile, perhaps you know some other ham experimenter who has one he could give you.

Jerry, KE7ER

does sound like the IRF510 has?

Hi Mike,
The ear sensitivity is not linear, so, that is the reason the volume pot should be logarithmic.
Now for the tuning, the way the software is written, you have 50 Khz of tuning range on 300 degrees of rotation and 1024 steps in the analog to digital converter (10 bit A/D). A change in position of the tuning pot with 0.3 deg. will change the frequency. Using a multi turn pot will increase the amount of movement required to have a change in frequency.
Another way of improving tuning feel will be by reducing the tuning range to 25 KHz or even 10 KHz by modifying the software, or installing a rotary encoder (also with modified software).
There is another sketch (firmware) that is using the pot in a different way. As long as the pot is in the middle position nothing happens. When you turn the pot frequency starts changing. The speed of change is increased with the angle of rotation. Once where you want to be return the pot to the middle position and you are tuned.
73,
Ion
VA3NOI

VA3NOI

Jack,?

did you enable compiler warnings in the Arduino IDE?

Armin, DJ2AG

Hi Jack,

you are right with what the compiler is doing when calling he function and that the cast on its own does not save anything, but Helmut was replacing

printLine1("Raduino v1.01");

with

strcpy(c, "Raduino v1.01");

printLine1(c);

with the cast in?

printLine1( (char *) "Raduino v1.01");??

the additional strcpy is not necessary. This is where memory is saved. When I was compiling a short sketch with both versions I got different values for program size and dynamic data.

Your version of printLine looks much better to me. ?I would have proposed to add a pull request to github repository, ?but it looks like the code there is not the one which is actually burned into the Raduinos and the other pull requests from December and January have not been integrated.

Armin, DJ2AG?

开云体育