Minima - Full Circuit TX - Take II

Well,

Following on from my last post, having looked at this spectrum plot:

We needed to understand what the RF peaks on the left were; there seems to be one at 5.something MHz and another at 8.something.

The first thing I did was throw the cable I was using to connect the audio generator to the mic amp in the bin - it was clearly a pile of dingos kidneys. I then connected the audio sig gen to the 'scope and set it for about 40mV peak to peak at 1.2 (ish) KHz.

I then repeated the sweep above:

So, this tells us a few things - primarily that these RF peaks are related to the audio and the drive level - I was probably overdriving the audio amp in the earlier scan.

Now, if I then focus in on the peak we can still see at 5.something MHz and exaggerate the problem by increasing the audio drive I see this:

Now, these multiple peaks look very suspiciously like they are spaced by the audio frequency to me, so I increased the audio sig gen frequency to 3.5KHz:

I can reduce the amplitude of these peaks by setting the balance pot on the bias for the KISS mixer - but this is about as low as they will go.

This will all mean something to those far cleverer than I, but I actually suspect that there isn't really a problem as long as the audio levels are sensible.

What do you recon, fun, egh?

Minima - Full Circuit TX

Well,

I finally got all the bits n bobs together and was able to try the full circuit TX. There's something odd going on with my PTT as the software is going into TX mode without me grounding the Arduino digital pin 3, but the relay isn't changing over until I do so. Anyhow, here the bench (read complete mess) right now:

I've tried to annotate the image above and label the various bits and pieces to help understand the mess.

Here's the output of the LPF on the SA:

This is at the switch on frequency of 14.200MHz. I'm just injecting an audio signal into the mic socket (well, there isn't a socket - just a capacitor lead!) but it looks quite clean and sounds fine on my bench RX.

It's time to fiddle as my LPFs aren't switching when the frequency increases and there are some other odd things going on, but that will be down to my Muppetry I am sure.

Good start though, egh?

Minima Carrier Adjustment

Well,

Been fiddling a little more this morning, and have dug out an Arduino Mega2560 board - this doesn't contain the processor utilised in the Minima but Arduino allows you to compile for the target board that you have so I am going to use this for the time being.

The first thing I did was to attached a 16x2 LCD display and test that it was working OK using a very simple piece of code that uses the same pin assignments as the Minima software:

The only change I made to the Minima LCD wiring was to include a 10R resistor in the +ve line to the display backlight (LCD pin 15).

But once I was confident I would be able to "see" if the Minima code was running I compiled and loaded that into the Mega2560 board, and hey presto:

So I know that the software has loaded and is running OK, I can now convert the pin assignments from the target processor to my Mega2560 board and wire up the various external bits and pieces. There is also a 3.3V line I can hook into on the board to power the Si570 thingamabob.

I then thought I would take a look at the BFO and carrier adjustment - more as an initial experiment with my Spectrum Analyser to see if I could do what I thought I could do....

So, the first thing I did was do a sweep of the crystal filter using the tracking generator (as previously), then I froze that trace (which appears in purple below) and coupled up the crystal filter into the circuit. I then adjusted the USB BFO so that the carrier appears just down the left hand side (because I want to maintain the Upper Sideband which is higher in frequency) of the crystal filter skirt:

Now, I am pushing the capabilities of the instrument a little as the bandwidth is at it's absolute limit, however, you can clearly see the carrier inside the crystal filter and inside the left hand side slope. If I now add an audio signal into the mic amplifier, you can see that the upper sideband is now well within the passband of the filter, and will be from about 300Hz up to about 3000Hz, the lower sideband doesn't feature as that has been supressed by the filter as it is as far to the left of the carrier as the USB is to the right:

It does look like the instrument is struggling with bandwidth a little and I **think** that's why there is a pile of stuff between the carrier and the USB, but visually you can really see what's going on - which is great!

So, for the LSB BFO setting, I need to do the same thing but putting the carrier on the right hand side of the crystal filter skirt so that we maintain the Lower Sideband and then I suspect Bob will indeed be my Uncle.

Interesting, egh?

** UPDATE **

NOTE: Because the second mixer mixes with a Local Oscillator that is HIGHER than IF, the sidebands become inverted. If we start with a 20MHz BFO with a 1KHz audio frequency the output of the 1st mixer will have the USB at 20.001 and the LSB at 19.999. Once this is then mixed for a second time with the LO at 34MHz we will have the previous USB at 20.001 moved to 34-20.001 = 13.999 and the previous LSB at 19.999 will now be at 34-19.999 = 14.001. The sidebands have swapped places!!!!

So the carrier adjustment above needs to be the other way round and the illustration above showing the USB above the carrier will be the LSB after the final mixer stage.

I definitely need to get out more.....

So here are my final adjustments (I hope), I have stored the sweep of the Crystal Filter, this time in yellow in the images below; also in each image the marker is on the carrier and the frequency displayed.

This is the image with the relay engaged so we will eventually have LSB out of the TX:

and here with the relay disengaged so we will eventually have USB out of the TX,, here the adjustment has pushed the carrier to its maximum available frequency which is not quite high enough as ideally the carrier needs to be further down the right hand filter skirt:

and here I've labelled the various bits and bobs: