开云体育

I modified the 1.05 Raduino sketch to monitor D7 and flip the state of the "isUSB" variable so that the tuning function would change the injection frequency to allow USB detection. The firmware change seems to work - i.e. I see the display show "USB" instead of "LSB", but the RX sensitivity dropped into the basement. The usual hiss was almost totally gone, and I could barely hear the birdies & QRN I normally hear.

Is this typical of using this method to listen to USB? If so, what is the preferred method? Changing the BFO oscillator?

There is something that took a while for me and others to understand and accept is that the BITX40 is a lower single side band phone mode transceiver. Making it do anything else is impractical and maybe even be impossible. The functions in the raduino code are leftover from the BITX20 which is upper sideband phone mode transceiver. Which can do digital modes because digital modes are on USB. CW on sideband might be legal somewhere, not in U.S.

Getting the BITX40 to do something else other than simple phone mode on 40 meters would require you to rebuild it in a different way. Such as the row of four toroids is a passband filter for LSB. You would need to change them and some other parts to work on USB. And then not on LSB anymore. Not just change it in the code.?Rebuilding it is doable. But if you can do that, then you can make it that way to begin with.

If the uBITX is a buyable product, it might be able to do the digital stuff so many of us are wanting to do. I am aware that I might be wrong about something I said here. And I am sure someone here could say it better.

Indeed, I wanted to use USB for digital, and so I changed my QRP Labs VFO from offset from - to +. I could hear signals, but *really* quietly. I'm just going to jumper over c91 and c92 rather than remove them.?

--
Ryan Flowers KC7RYS?

Driving the BFO from a spare Si5351 clock is the ideal way to do it. ?Makes it trivial to flip between USB and LSB without affectng drive level of the VFO, and also allows an easy choice of 5mhz vs 19mhz for the VFO. ?We badly need this integrated into the "standard" firmware. ?

Here's the thread where Pete posted how he did it. ?/g/BITX20/message/24270? ? Some issues left unresolved there. ?What's the best way to drive the BFO (and VFO), should we all move to the resistor network from the uBitx? ?

I only mentioned CW because it is considered a digital mode. And digital was the context.

I wasn't condemning anyone for it. You seem a little barky in the morning Jerry. Ruff.?

First off, I like that you have a sense of humor. ?I was going to say "Bull ?;-)", but replaced it with "No" at the last minute. ?Figured you'd take it the right way, but that was a bit coarse for the refined tastes prevalent in this forum.

But you are wrong at least as often as you are right, and somebody has to throw something back at you before the entire discussion runs off the rails.

You did say: ?"CW on sideband might be legal somewhere, not in U.S". ?Nothing illegal in the US has been proposed, assuming it was done correctly. ?An audio tone into the microphone input can create legal A1A transmissions, but is difficult to do correctly as Collins Radio found out 60 years ago: ?/g/BITX20/message/22372. ? (The crystal filter must also be perfect, by the way.) ?I much prefer something like what Farhan is doing on the uBitx. ?

> Ruff.?

A bit off topic, but these were our birds: ?? ?And we now have yaks about to calve. ?And I'm several months behind on an engineering project. ?So unfortunately, I haven't gotten back to tinkering with my Bitx40 for over a month now.

Jerry, KE7ER

@Tim

You don't need C91 and C92 if you use the Raduino. They are a part of the analog VFO originally designed. As long as you use Raduino you can take them out for good.

The change from USB to LSB can be done by changing the VFO (Raduino) frequency range from (19MHz - 19.3MHz) to (4.7MHz - 5MHz)
--
Ion

VA3NOI

Switching between LSB and USB by moving the VFO from 5mhz to 19mhz works, and is the way that Farhan wrote the sketch. ? ?Disadvantages are: ?More trouble with harmonics and mixer products when at 5mhz, such as those SW Broadcaster images and the 7.2mhz birdie. ?Also, drive level out to the final changes some between 5mhz and 19mhz, could be due to how the diode ring and transformer respond, in addition to ?C91,C92.

The other method is to move the BFO by a few khz using the Si5351's CLK0. ? ?Leave the VFO at either 5mhz or 19mhz, your choice. ?(19mhz is preferred to reduce images.) ? ?Disadvantage is that the crystal filter skirt is not as steep on one side as the other, so LSB with a 19mhz VFO and USB with a 5mhz VFO will have less rejection of the residual carrier and opposite sideband. ?A better crystal filter such as on the uBitx would help here. ?

I cranked up the Raduino drive level when selecting USB, and that helped. I think I'll pull the un-needed caps, also. The final thing is to tweak the firmware so that the "center" frequency is the same for USB/LSB.

Finally - I have seen some of the ubitx schematics - is there any reason I couldn't build an equivalent filter & drop that into the bitx40? I see that the filter schematic shows transformers on the input/output, with one winding connected to, what? Nothing? Could you just use a 1:4 balun for this application?

The transformers are to transform the 50 ohm bidi gain blocks of the uBitx to the 200 ohm impedance of the crystal filter. ?The 2-to-1 winding ratio gives 4-to1 impedance ratio. ?The spare winding is just that, as on the Bitx40 they wind all those transformers trifilar but not all of them make use of the third winding.

Slipping in the uBitx filter might not be trivial. Google "Dishal Crystal Ladder Filters" and read up. ?The Bitx40 does not have 50 ohm bidi gain blocks. ?Likely on the order of 200 ohms around the filter, but note that the Bitx40 gain block's input port impedance depends on output port impedance and vice-versa. ?They are what they are, and it is not obvious exactly what that is. ?

Jerry

I never said anyone proposed it. But I am now saying it has come up several times. After the earliest conversations of operating CW, on this BITX someone had looked it up and provided greater detail on it. And now we know we can't, and why we can't.

Now for a serious request of you Jerry. Between several to many times you have offered your opinion of running the VFO at 19Mhz instead of 5Mhz. But I don't recall you ever detailing how to do it. You know I can follow along and change the code. But what else is involved. Any other specific changes? As much as you seem to think it would solve a lot of problems and make some hacking easier. I would have expected you to have written a paper, or constantly instruct people how to do it, or upload a sketch with it all done for us by now.?

Is this where the VFO get's set at 19Mhz? This is just a cut and paste.

#define INIT_BFO_FREQ (1199800L)

unsigned long baseTune = ?7100000L;

unsigned long bfo_freq = 11998000L;

Finally, I think your ragging on me for being wrong now is because I said the toroid band pass filter put's it on LSB. But it's the crystal filter that does that. Right? I would rather be corrected and learn, than get a blow up about it later. Or continue the same debate without ever making progress. We are licenced communicators. We can do better.

John Smith,

Here's your paper.

Assume for now that we are transmitting an AM signal at 12.000mhz with a 1khz tone from the microphone. ?The two diode modulator get's a 12.000mhz carrier from the BFO and mixes that with the 0.001mhz audio from the microphone. ? The modulator gives us a signal consisting of the carrier, the carrier plus the audio or 12.001mhz and the carrier minus the audio or 11.999mhz. ?If we send that straight out to the IRF510, that's AM phone on 12mhz.

For LSB, all we want is the 11.999mhz signal. ?We ?can get that by having ?very sharp bandpass filter (the crystal filter) that allows through the 11.999mhz modulator product but not the 12.000mhz or 12.001mhz components. ?Sending that on to the IRF510 would give an LSB signal at 12.000 MHz. ?(The number you read on the dial is the frequency of the suppressed carrier.)

For USB, all we want is the 12.001mhz signal. ?We ?can get that by having ?very sharp bandpass filter that allows through the 12.001mhz modulator product but not the 12.000mhz or 11.999mhz components. ?Sending that on to the IRF510 would give a USB signal at 12.000 MHz.

It's really hard to move the crystal filter. ?Easier to build the crystal filter for only 12.000mhz, and place the BFO at either 11.999mhz (for USB) or 12.001mhz (for LSB).

If we speak into the microphone, most of us give more than just a 1khz tone, there's a range of audio frequencies of interest for intelligible speech from around 500hz to 3000hz. ? ?So that crystal filter must allow a band of frequencies about 2500hz wide through, not just a single frequency. The Bitx40 has a balanced modulator, so the carrier is mostly suppressed. ?It's good enough if the crystal filter does a good enough job of knocking out the far sideband, which starts at 500+500=1000hz away from the low audio frequencies that we are trying to transmit.

We want to transmit at 7mhz, not 12mhz. ?The diode ring mixer combines the above 12mhz sideband signal with the VFO. ?If the VFO is at 5mhz, then the 7mhz bandpass filter (the three torroids L1,L2,L3) will allow through the 7mhz difference signal from 12mhz-5mhz. ?If the VFO is at 19mhz, then the bandpass filter will allow through the 7mhz difference of 19mhz-12mhz. ?Note that 19-12.001 = 6.999mhz, so in the 19mhz case what was 12mhz USB has become 7mhz LSB.

The sketch has been written. ?Several here are successfully using a 19mhz VFO to transmit 40m USB. ?Ask them. ? Only physical change is to remove C91 and C92, the Raduino doesn't need them for 5mhz or 19mhz. ?Pete is successfully using the alternate method of moving the BFO from the Si5351.

Don't take personally my attempts at guiding this conversation in the direction of physical reality. ?I don't want 1000 lurkers coming away with the wrong idea of what works. ?I will occasionally try correct wrong information. ? Some may find me a bit snarky at times, that's my attempt at humor. ? Some may find you a bit snarky too.

> I would have expected you to ...

Tough. ?(See, I can be kind of snarky.)

Jerry, KE7ER

OK, here is where I am so far. I removed C91 and C92. I changed the 1.05 Raduino sketch by? adding code to read pin D7. In the original sketch, this was reserved for the TX/RX line monitor, which I don't use. You can use pretty much any other unused Arduino pin if you want. You also need to got through the code and remove references to that particular pin. Make sure you save your sketch under a new filename so as not to lose the original!!!

In the "setup" area, I made sure that pin D7 had the pullup enabled: pinMode(ULSB_SELECT,INPUT_PULLUP); (I #defined D7 as ULSB_SELECT previously)

Then, I added this code in the "loop" function:

if (digitalRead(ULSB_SELECT) == 1) /* if U/LSB switch is high, do LSB,
?if low, do USB */
{
? if (isUSB ==1)
? {
? isUSB = 0;
? setFrequency(frequency);
? updateDisplay();
? }
}
else
{
? if (isUSB == 0)
? {
? isUSB =1;
? setFrequency(frequency);
? updateDisplay();
? }
}

Finally, the little trick that allowed the displayed frequency to be correct for both LSB and USB: I modified this line in the setFrequency() function:

? if (isUSB){
??? si5351.set_freq((bfo_freq + f - 3100) * 100ULL, SI5351_CLK2);

The addition is the "- 3100" after the "freq + f". This keeps the displayed frequency the same, but shifts the actual VFO frequency to put the passband in the right spot.

Now - the caveat! I didn't have any definitive way to test this, as right now, I hear nothing on 7.076 (JT65 freq for 40m), even on the "big" radio. However, I took off the antenna connection of the bitx40, set my IC-718 to the lowest output I could, keyed it up in 7.150 MHz AM, and then tuned up & down from there on the bitx while trying both modes. In either mode, the frequency shift and volume of the received signal seemed to be pretty much the same, as I tuned up & down from the "center" frequency, certainly it should be pretty darned close! Hopefully, I'll see if I can hear some signals later & confirm the operation was a success!