PCM Hawk Arduino VFO.

V

VK3ZYZ

Guest
On snooping in a second hand shop, I spotted a lonely looking HF transceiver priced at $60, marked down to $30. Not too bad for a 100W SSB set!
This started my exploration of HF.

This is the first Arduino VFO I did. It ended up looking like this.
Front View.jpg

This set was only fitted with 2 channels so more coils are needed to add more bands. As yet, I do not have them.
After discovering the Si5351 tripple VFO board, and playing with that for a bit, I decided to start the conversion. The crystals have been removed and the PCB I drew up a to hold an Arduino Nano and the Si5351 VFO board is mounted on a 3D printed spacer screwed to the side frame.
CloseUpTopView.jpg


A string of resistors fit nicely between the old crystal switch terminals to produce a resistive divider that feeds a stepped voltage to the Arduino so it knows what band is selected. The PCB has a couple of cuts to suit.

BandSelectConnections.jpg


As I did not want to cut large holes in the front, a 3D printed display box fits over the speaker and is held in by magnets. Just a small hole is needed to feed the 4 wires out to drive the I2C LCD.
3D front1.jpg

The above is a quick introduction to this project.

Here below is a bit more detail, a pdf of the build, some Arduino code and the service manual.
And the info of #1 VFO.
This set is calling out for the Adaptive VFO talked about elsewhere in this forum.
 

Attachments

V

VK3ZYZ

Guest
Just out on interest, I hooked the "VFO Delux" to the Hawk for the club net tonight (20210106) and it worked well :)
VFO_Delux_Hawk.jpg
 
V

VK3ZYZ

Guest
The code now includes a calibration factor. It is nice to have a good frequency counter!
Also, the USB/LSB would work if I replaced the crystal with the CLK1 output.
That will get done later.
Here is the new code, as well as the Si5351 calibration sketch.
This code generated 10Mhz from CLK0 and the Arduino Serial Monitor at 115200 Baud used some keys to adjust the 10Mhz.
A calibration number is produced that is added as below.

This is the Si5351 calibration mods...
.
.
int32_t cal_factor = 19100; // found from Si5351 calibrator sketch.
.
.
.
clockgen.init(SI5351_CRYSTAL_LOAD_8PF, 0, 0);
clockgen.set_correction(cal_factor, SI5351_PLL_INPUT_XO); // <<<<<<<this is found from Si5351 calibrator code.

.
.

This code will work with few mods on a lot of old radios that have a 1.65Mhz IF.
 

Attachments

V

VK3ZYZ

Guest
1.65MhzFilter.png

I have removed the 1.65Mhz crystal, and the 2 x 100nF caps from the oscillator, and added a filter to the BFO signal now produced from the Si5351 board. Now, the Hawk has USB/LSB switching.
These values are not exact according to the calculator, but the signal is now a sine wave.
BFO_filter_Connections.JPG

The output from the filter is connected as above.

This is the "Birds nest" band pass filter.
Birdsnest_BFO_Filter.JPG

I may make up a PCB for this later. There could well be a use for handy filter boards. Diode switching on board could be good.
The main Si5351 output having switched filters springs to mind.
 

Attachments

Last edited by a moderator:

VK3ZYZ

Moderator
Staff member
I've started on VK3KAL's PCM Kestrel.
The synth I'm installing is the Tracker Communications Scout board.
Arduino_VFO_top.jpg


The carrier injection, 1.650Mhz/1.647Mhz is fed into the existing 1.650Mhz osc after the crystal and C16 and C17 have been removed.

Carrier_Osc_Mods.jpg

Removing the caps stops the crystal oscillator, so it is now just a buffer. The coax braid goes to what was one end of C116, and the center to C117.


To prevent the channel oscillator running, one end of R89 is lifted from the board, removing the power.
R89_Lifted.jpg

It is a bit hard to see.

Then, the signal from the synth is fed into TP12, with the coax braid soldered to the ground plane.
Channel_Injection_TP12.jpg


More to follow.......
 

VK3ZYZ

Moderator
Staff member
I have it working (sort off) on a couple of bands.
The channel crystals have been removed and that selector switch bank used as the band switch.
PCM_ChanSwitching1.jpg

The osc feed cap has one end lifted to stop 10V pulse driving into the osc transistor, and the switch wiper connected to 10V.
The clarifier pot is just tied out of th way as it is not used now. Its place has the encoder fitted.


PCM_ChanSwitching2.jpg

All the blue wires are from an exiting mod, and I'm still trying to work it out. There is a "funny" with the second half of the channel switch that uses a relay to select it somehow. I think it is to do with a different channel TX and RX frequency option????


One problem is the RF coils supplied, and there is not a full frequency range. Another is the narrow band width of the could, so I may have to use a couple of "channels" per band. Like 80M_Low and 80M_High.

The display mounting method and placement is still to be figured out.

Don't miss the next exciting episode of "Can a Kestrel Fly Again?".
 
Last edited:

BillC

Active member
I wonder could you stagger tune a couple of the coils to widen the B.W. for 80 m ? Would be a bit lossy compared to a single channel tune up, but the RF amp might provide enough gain in practice.
 

VK3ZYZ

Moderator
Staff member
I think the way to go is to connect my Nano VNA up so the band width can be easily adjusted.
But, as there are 10 channels in the set, and I will only use 5 bands, having more than one channel per band is quite usable.
The big problem will be the lack of coil sets that will tune to 10 and 14Mhz.
I have added a USB/LSB switch and tapped it into the Tune.AM.SSB switch via diodes and resisters to feed an analog input so the Tune and AM always switches to USB even if the USB/LSB switch is in LSB,
ModeSwitchingS.jpg

When it is done, I'll post full circuits of the mods and the updated Arduino code.
 

BillC

Active member
Seems to be a good idea as you have plenty of spare switch contacts, watching with interest, Cheers... Do you need a few neosid coil assemblies for winding the 10 & 14 meg coils ?
 
Top