Category: Hardware

All hardware posts

Posted on

Conclusions on how to eliminate spurs

Alison KB1GMX has advice to constructors on the spurs on the higher bands:

For bands below 20Mhz spurs are NOT an issue as the low pass filters catch it.

Spurs are only a problem for SSB and frequencies greater than 20mhz.

NOTE: due to the way the uBTX does CW it is never an issue on any band.

The short form is when you mix two frequencies you get a third, in a perfect world.

The diode mixers used are handy but they can present conundrums.  If any of the three ports (IF, LO, and RF) are mismatched (think SWR) The signal can be reflected back in.  Since DBMs are omnivorous in that any port can be input or output and if mismatched both!  This does not include effects of distortion in the source signals.But in the real world things like this exist.

Double balanced mixers also suffer from overload, too many and too strong and you get a plethora of signals.   What that means for lots of simple and complex reasons you can get “spurs” or spurious  outputs that are undesired.

Basic math, addition and subtraction:

If you mix 45Mhz with 73mhz you get 28mhz.  We want that  and the radio needs that.   However if any of the 28 gets reflected back into the DBM where it originated it mixes with 45mhz and we get 17mhz.With those four signals you get mixtures of those like:

  • 73-17=56
  • 28-17=11

Those are “first order” as they do not involve harmonics.  They will be the strongest, but not always equal strength.

Both inputs can have harmonics like 90mhz and 146mhz and the 34 and 56 coming out can have harmonics too.  If you add and subtract all the possibles you get an increasing sea of signals some weak some stronger.  We will not cover the possibles as the first order ones are the most troublesome.

The solution traditionally applied is band pass filters or if it isn’t between 28 and 29.9999 the filter strongly attenuates it.  But you need a band pass filter for most every band… uBitx takes the path of below a certain frequency you only need low pass filters and fewer of them.  And it generally works well especially for 80 though 17M…

But at 20mhz and up the low pass filter passes everything below 30mhz and if you overdrive the rig slightly you get a spur on the tech window on 10m where the spur is 16.5 to 16.7mhz and there is no filter for that.  What makes this worse is some radios are very poor at 10M putting out maybe 2W so pushing the audio to get more invites the problem to be greatly worse.

There is no setting we can safely give that absolutely assures there will
be no problem that is consistent with maximum achievable power.

As a licensed amateur radio operators we are responsible for signal quality and also not generating signals outside our assigned bands.

There are two solutions one is bandpass the other is high pass filter.
Either way the rig must be modified to allow those and there are side effects.

One side effect is you need extra switching not provided.  The other is any filter has a loss though it and that would further reduce power out.

Short of that, keep the power right down on the higher bands and go for it…

Reference
Posted on

Nextion Display and KD8CEC firmware

Ian KD8CEC is working on a protocol to allow Nextion LCD touch displays to communicate with the µBITx.

He is implementing this a little differently to most support for Nextion LCDs.

The firmware will handle communication between the uBITX arduino and Nextion LCD using template files.

There are quite a few variables in the Nextion LCD. If the status of any parameters in the uBITX changes, the variables sent to the Nextion LCD will also be changed at the same time (and vice-versa).  This should allow any Nextion display and any configuration of the display’s User Interface to interact with the µBITx.  This means constructors can customise their µBITx display using the Windows GUI used with the Nextion to configure the “look and feel”.

For example, when uBITX’s frequency changes, it is transferred to a specific variable on Nextion LCD.

Example code

 

Below is an example of simple frequency and mode display changes

 

//Frequency

if(vc.val!=-1)

{

nFreq.val=vc.val

vc.val=-1

}

//Mode

if(cc.val!=-1)

{

if(cc.val==2)

{

tMainMode.txt=”LSB”

}

else

{

    tMainMode.txt=”USB”

}

cc.val=-1

}

Reference
Posted on

Spurs found in uBITx that present issues for operation above 18 MHz

Alison KB1GMX has been exploring ways to flatten the TX output across bands over the last couple of weeks, but  says,

“Here is the problem…  Improving the amp leads to acknowledging other issues.   In particular … the spurs!”

When transmitting on 15m and 10m the LPF has a cutoff frequency over 30 MHz, so effectively the amp will pass everything through below 30 MHz.

If you overdrive the mixers especially the 45mhz output mixer you get more than one product from the mixer.  For a desired output at 28 MHz, you actually get two other outputs, one at 11.995 MHz (that is 50 to 60db down on the desired output) and another at about 17 MHz, which if the mixer is not over-driven is around 40dB down.  However, if the audio is pushed a bit harder, it may be only 25db down from the desired frequency.  The amplifier has no filtering below 30MHz and will gladly amplify all of the signals through the pre-driver, driver and finals stages and pass them out to your antenna.

This  is not an issue until you hit around 18 MHz as the output low pass filters that are switched in help attenuate the spurs.

The Maths!

Output is IF(about 45mhz) + LO (48 to 75mhz)= 3 to 30 mhz
The spur case is IF(45mhz) – Output frequency (42 to 15mhz) Its going down while the dial frequency is going up.

For up to 18mhz the output lo pass filters and the inside 33mhz low pass filter are at work doing the job.

However for 21 and 28 MHz both filters are roughly 33-35 MHz cutoff and the spur is now down around 24 and 17 MHz respectively. They will be about 40-50db down until the point where the 45mhz mixer overloads and then the spurs get very much stronger at a frequency well below the cutoff of the low pass filters.   Since power output is low on 10M people will likely push the audio and the spur will get significantly worse.

The is due to the nature of Double Balanced Mixers: they are three port devices and all three can receive input or deliver outputs.   Also inside the mixer all possible sums and differences for the base frequencies and their harmonics will exist.

This was determined, without the amp operating, by breaking the circuit at C200 to allow a spectrum analyzer to be inserted.   We can’t blame the amplifier chain for this as it would just do its designed duty and amplify everything.

There is no obvious solution at present

Right now Allison has no obvious solution to present.

Commercial radios implement a switched filter at the point were L1 though L4 (and C200 to C204) are present and a filter would be switched in for 15M, 12M and 10M.   Allison is investigating other ideas.

This issue does not show up in CW mode as the radio is keyed in a different manner and there is no 45 MHz contribution.

Additional comment

Jerry KE7ER comments that the spur is not a problem from 3.5 MHz up to 18MHz.  So for operators that never use the 15, 12 and 10 meter bands on SSB, this should not of concern.  For many of us, a $129 rig that covers 80-17 meters is still a very good deal.

For those wishing to use the rig above 18 MHz, perhaps some sort of external band-pass or high-pass filter is in order between the µBitx and the antenna.
Some (but not all) antenna tuners might be sufficient.

If the audio from the mike amp is kept reasonable, Allison reports that the spur is 40+ dB down.  Jerry assumes it should be 43 dB below the primary signal to be compliant in the US, so it’s marginal at best.  The worst of this problem could be avoided by somehow monitoring the transmit signal level  at the mixers. Allison has found that gain with the 2n3904’s can vary on the upper bands depending on your particular device characteristics, so it currently is not sufficient to just check the level at one point in the chain. However, if gains can be made more predictable,  perhaps monitoring at the top of RV1 using a diode RF probe would be sufficient?

The µBitx has a 30 MHz low pass filter between the mixers and the power amp.  As Allison says, most multiband rigs would have switched bandpass filters in that position.  Signals are quite low level, so this could probably be done with analog switches rather than relays. So yet another possibility is to add a daughterboard with switched bandpass filters to replace the filter at L1,2,3,4. We only need 3 or 4 filters there and parts can be small and cheap, so not so bad.

The spur is at the first IF frequency of 45 MHz minus the operating frequency. If the first IF is raised from 45 MHz to 70 MHz or more, that should remove the spur for all frequencies from 3.5 to 30 MHz.  That 70MHz IF strip would definitely require something better than the 2n3904’s that are in the µBITx!

Reference
Posted on

Attenuation on TX to flatten the curve (Phone only)

John VK2ETA has uploaded his latest version of a variation on Ian KD8CEC’s software with a default of “stock standard” which means it should run on units “factory” wired.

It includes the “OPTION_ALC” which use a table and interpolation to set the attenuation for 3 levels (Low, High, Max) for SSB power output.

Not a complex code at all (just search for the “option_alc” keywork). This should give up to 50dB of attenuation. This should be plenty to control the power even for WSPR low power users.

You will probably need to attenuate the signal for “Max” power on 40 and 80m otherwise you will get well past the targeted 16-17W on the lower frequencies.

Note that on 10m, you will not get 10w out.   You will need to follow Allison’s work on the power chain to see if a solution can be found to the fall-off in power out on 10m.

John’s software can be found here.

Reference
Posted on

Fan shroud to help cool your finals

Dave K0MBT has been working to keep his power transistors cooler.  Initally he added a 80mm fan to blow across the transistors and board.

When doing ft8 he would get significant heating of the power transistors running on a 13.8v linear power supply.  The transistors would heat up enough  to make them too hot to keep his fingers against the surface.

He has, therefore, added a shroud similar to the one shown covering the stock heatsinks. It did not help that much.  He had a fairly small but 1″ tall computer heat sink, cut it in half and tapped in mounting holes. The shroud had to be redesigned to fit them but now he can only feel a very slight degree of heating wile running the FT8 tuning feature. 

This worked so well, he has added a similar heatsink to his BITx40.

Reference
Posted on

Another board from Nik VK4PLN that does several things at once

Nick VK4PLN has been working on a new board that will give a few extra features to a stock uBitx and plugs into the audio loom socket.

Its an Audio board providing easy access to Audio I/O pins. (add in your own AGC board, External amplifier…)  It includes an area for adding an SSM2167 Mic Pre-amp module (with filtering for feedback and shutdown on TX).    It also includes the simple 4 component PTT POP fix. (BS170) and a switchable 200hz CW filter. (LM324) that Nick already produces as a board for purchase.
The board has a bonus “snap off” section with a 3.5MHZ BCI filter for the RX chain.

Here is a pre-view, NOTE this is a WORK-IN-PROGRESS.

Reference
Posted on

G3EJS add on board including a different anti-pop mod

G3EJS has used a PIC processor to sequence the mute of the RX and change to TX.  The PIC responds to the PTT going low by muting the audio amp, and 100ms later, passing the PTT low state to the raduino.   When the PTT is released, it stops the PTT to the raduino, then 100ms later un-mutes the AF amp.   He then made a breadboard version, which resulted in a silent transition.

G3EJS then made some boards and finally integrated a number of mods onto a single circuit board.   Check out his pop mod circuit below:

Reference
Posted on

JackAl is here!

Jack W8TEE and Al AC8GY have  released details of the JackAl board following FDIM (preceding Dayton Hamvention).

Friday night is a sort of Show-and-Tell at FDIM and they used that opportunity to show their JackAl board in action. The photo above shows a little more about what it is and does.

At the show, Al hooked up a noise generator to the µBITX to show how the filters work. (There are 4 preset filters for CW and 4 for SSB.) In addition, you can customise one CW and SSB filter to the bandwidth you desire. You might be able to see that the skirts are pretty steep for the filters on the scope in the background. The setting of the CW custom skirts are set differently, in that you pick a centre frequency (e.g., 700Hz in the shot below), press the encoder, and then you see this:

In this case, turning the encoder CCW increases the bandspread (i.e., the 440 red number above, although it looks orange in the photo) by simultaneously moving the skirts (480Hz and 920Hz) further apart. If you turn the encoder CW, you narrow the bandspread. Most CW users will probably center the bandpass on their favorite sidetone frequency, which centers the bandpass on that frequency.

The demo used a 5″ display, although a 7″ display is also available. The third knob on the front is for a second encoder that we use for everything from setting the CW keyer speed to adjusting the filter skirts. You can see some of the plots on the panel at the rear of the picture above for some of the board’s features (e.g., filter responses, compression, etc.) Those will be included in the documentation when the (downloadable) manual is finished.

The JackAl board has the following features:

  • 5″ or 7” touch screen 800×480 TFT color display
  • Dual VFO’s
  • RIT
  • S meter
  • RTC
  • CW keyer, 5 to 50wpm (we could go up to 100wpm, but…really?)
  • Up to 50 CW preset messages, selectable at runtime…perfect for contest messages
  • Touch screen function and control selection (e.g., band changes, RIT, mode, VFO, VFO increment, LSB/USB, etc.)
  • Automatic LSB/USB selection based on frequency (overrideable)
  • One touch frequency increment changes (1Hz to 1MHz in multiples of 10…the white underscore in the frequency window)
  • Dual encoders (frequency, features)
  • EEPROM storage of user preferences (one-click reset to “factory” defaults)
  • Uses Teensy 3.6 processor (1Mb flash @180MHz) and companion audio board
  • Support for 3 external CW push button switches (NO) for sending stored CW messages (e.g., contesting)
  • Hardware AGC using IF take-off
  • Audio AGC with adjustable threshold
  • Mic compressor with adjustable threshold
  • 8 band audio equalizer
  • Receive audio filter: 48dB/octave (8 pole equivalent DSP filters)
  • 4 CW presets (150, 300, 400, 600, [or none] Hz 3dB bandwidth) + 1 user-defined knee frequencies (at runtime!) filter
  • 4 SSB presets (1500, 1800, 2200, 3000, [or none] Hz 3dB bandwidth) + 1 user-defined knee frequencies (at runtime!) filter’
  • Variable Notch filter, encoder adjustable, use specified Q
  • 7 watt power amplifier

The board will be distributed with all (surface mounted) parts in place. The user must supply the Teensy 3.6 ($30), its supporting audio board ($15), and the 5″ ($34) or 7″ ($44) touch screen displaying (using the RA8875 controller chip, BuyDisplay.com).

We expect the JackAl board to sell for $50.

We may need to adjust this price as we have only received “ballpark” cost estimates for the board since we only have the Gerber files for the Beta board.

Currently, we are using less than 20% of the available flash memory (out of 1Mb) and less than 15% of the SRAM (256K), so there is plenty of memory resources available for adding “stuff”. The board also brings out a number of I/O pins to help your experimentation. With the exception of removing one SMD resistor on the µBITX board and soldering two wires to those pads, all interconnections are via existing connectors.

Our best guess is that after finishing the modified Gerber files, production, Beta testing, and writing support manuals, it will be probably two months before we begin distribution. We will announce its availability here as soon as we can. BTW, if anyone knows a high-quality PCB manufacturer who also does pick-and-place at reasonable prices, we are getting quotes and would like to know about them.

Reference
Posted on

Upgrading the PA output transformer

JerryW0PWE notes that in the posts on boosting and/or leveling output power he saw pictures of a binocular core being used as the output transformer in place of the toroid that comes with the uBitx board.  He asks “What size is that binocular core?”

Allison plans to use the longer version of the BN43-6802 (or two BN43-302 cores end to end).

John VK2ETA says he went back to IRF510s in the finals and has used a BN43-3312 as the single 302 was getting warm on the low frequencies and he could not put more than 2/3 turns in the core.  He tried both 1/2T and 2/4T in the 3312 and settled on 2/4T due to better efficiency. The 1/2T drew too much current at 80M.

There is plenty of gain and power at 14MHz and below on 16.8V for the PA but has had to resort to a gain control for flattening the power curve despite 330pF caps in all 6 drivers emitter resistors, 22uH inductor in the base feedback of  the pre-driver,  330 pF across the primary of the finals transformer and 820ohms in the finals feedback resistors.

Nick VK4PLN uses the BN43-202 with a winding ratio of 2:4 using 0.64mm enameled CU.  His rig is outputing 20W+ on 80/40/20 using RD16HHF1.  Nick doesn’t detect any heating… Maybe it will at that power level on 10m?

His plan is to now try using:

– the MPSH10 in the pre-driver
– a single 10T bifilar wound FT50-43 for feeding the power to the RD16HHF1s
– removing the PA feedback loop
– and other pre-driver mods as per Farhans suggestions….

This appears to be the common approach on most of the RD16 amp designs out there…and Nick hopes he will get 20w+ output from 80-10m….

Bill K9HZ tried a BN61-002 and it seems to be a winner.  It doesn’t heat up at full power and it seems to be extremely wide band.

Reference