Hi Kevin, hope you are doing well! I've been binge watching your videos of late and really enjoying your work. You always give me something to consider. Thank you and 73 from Tracy VE3TWM.
Great project! Recently found your channel and subscribed. I'm taking my Technician's License test today but I am already intrigued by CW and QRPing. Looks like I'll be studying for my General class exam so I can try some of this stuff.
Great project. I am looking forward to working with this one. Thank You again. P.S. Enjoy those 80’s, it is forecasted to be in the teens here tonight.
Looks like you have a transmitter that can run off a 5 volt phone charger battery pack or a couple 18650 cells. Build a receiver with it using the SI5351(Direct Conversion) and have a low voltage transceiver. Very cool! I can see some fun possibility's with this and SOTA. I see this in my future. Thanks Kevin! Great project. De Troy KF7SEY
Was also thinking of using the SI4732 chip which does AM/FM/SW all in a chip. Would be cool to have a system that can actually run on a couple of AA batteries.
Wow, this is all so very impressive. Thank you for sharing your knowledge, Ham radio has to be the most engaging and immersive hobby I have ever been involved in and people like your good self Kevin make it so much better. The things I am learning are mind blowing to me to be honest. Thanks again for your work I really appreciate it.
I see your posts on Qrz by someone else. Looks like a fun project for a cold winters day. It would be cool to have a keyboard keyer project to go with this. Thanks
Just thinking that an UDOO RYZEN BOLT V1000 Maker Board SBC could be used as there are various hats for the board and you can use the gcc compiler for the 80586
Very Cool! I have an Arduino Uno working with Ada Fruit Si 5351 breakout board. I had to add to the freq variable an offset of long freqcorrect = 002740000 to get the frequency output correct as measured by my Siglent 200 Mhz scope. DigKey and Mouser are out of stock on the line driver chip so looking around.
In the Arduino ide while adding the offset to the variable freq wich is is type long I kept getting an error saying the variable is typed octal..... huh?
Why do you need your frequencies in hundredths of a Hz? Define the initial values in Hz. Set your minimum tunestep to50. If you need hundredths of a Hz somewhere, have a local variable where the frequency in Hz is multiplied by 100. It will make the code easier to follow and also reduces the chance of mistyping the initial values - the fewer the digits in a variable, the better. [edit]If I recall correctly, CW is allowed anywhere in the bands, with a few exceptions. There are just small segments of various bands that are reserved for CW only. As you mentioned, it might be wise to limit QRP operation to the CW segments.- your 500 mw signal is going to be 10 dB below a 5 watt signal. On the other hand, it can be fun trying to dig minute signals out of the noise.
I'm not a programmer, I just figured things out as I go. I'm sure there's always a better way. The software is open source, you can do whatever you like with it.
Hi Kevin, The project is well along. So Harry will be producing a commercial version for sale. Wonder if his circuit could be designed to allow the option for more output using another stage of amplification. Like you mentioned an IRF510, if you add these components you can increase the output to 5 watts. And I like the logo you created. Stay safe. 73 WJ3U
I think, and may experiment down the road, that you could replace the line driver with something like the IRF510 and an inductive choke on the power rail to it. It would need to be fed off the 12 volt rail and heat-sinked. It would also need a driver stage since the output of the Si5351 is not enough to turn on the 510. Just a simple 2n3904 and a couple of resistors with another cap to couple it to the irf510. The choke value I'm unsure of, would need to do some research. It'd be more of a stink bug and less of a flea at that point. ;-)
I like it! Simple use of modern techniques and components. How about putting the tx/rx relay before the LPF so it also works on receive? Some simple receivers will benefit if they have poor receive filtering.
@@loughkb That's true, but in a simple receiver it's worth a couple of dBs loss. Heathkit did it on their HW-7, 8 and 9. There are some more mods for the Pixie in Sprat a year or so ago, and used the slightly lossier BPF which was well worth it to reduce broadcast breakthrough. (Plus the Pixie Pin 7 mod!)
6:20 12 to 50 ohms "4:1 transformer" IIRC, a voltage and turns ratio might be better called "2:1" to achieve a 4x impedance ratio. i.e. A "2:1" transformer, with a 2:1 turns ratio, changes voltage by 2:1 and current by 2:1 in the opposite direction, thus changes impedance by 4:1. But it's still a "2:1" transformer.
Sure, once I complete the LP filter when the parts arrive. You can also go to the zachtek site and look at the page on his LP1 beacon. It's the same output section and he has scope pics of the output and harmonics there.
Kevin this looks awesome! I am a bit intimidated by winding inductors, i never have done that before. Can you point me to some resources or show how you are you doing your windings? BTW built your field day end fed vert for my OM, works super!
I'll show and talk a bit about winding in the build video. It's easy really. Also, Zachtek provided a great illustration in the PDF manual on his mezzanine board. You can download that on his website.
Will be building this project. You said there will be a released product version, they doing a kit version? Otherwise I'll send you the BOM order and a PCBWay or JLCPCB order option for the board you could link the project so people can kit build it without having to the additional steps.
Definitely send me that info if you do a PCB version of your own. I may do another video down the road on what the project's lead to. Zachtek will be doing a commercial version. I believe he will be selling a populated PCB version, ready for air.
No, it's a relatively small sketch. Sure. This method could replace the oscillator on other rigs. Depending on their design, it might be neccissary to include a LP filter to smooth the 5351's square wave to a clean sine wave and reduce harmonics. You might also need a single transistor 'driver' stage to increase the output signal from the chip.
It looks like a fun project. I have a question about the tuning procedure. Do you reduce power (or switch to built in dummy load) when tuning to prevent QRM?
I think pin 3 is programmed to only come on when you "spot tune". This is happening when in receive mode and the only signal is that sneaky little one turn coupling of wire from pin 3 on the receiver input....so not being transmitted as relay is not engaged in transmit mode. So no QRM. :)
I detailed that in the video. The second output of the 5351 is turned on and coupled to the receiver output via the gimmick wire. The power path is not enabled during tuning. It should also be quite clear in the schematic. (shown in the video and also in the blog write up that's linked in the video description. Which also includes an explanation of the tuning process.)
Hey Kevin, great project. Are you planning on integrating all the components on the pcb? Because if you are, you're most likely going to have issues finding sources for the SI5351. Most vendors are out of stock of those ics. I have a friend of mine who was designing an sdr transciever using those ics and he had to halt the project because he could not find mcus or si5351 anywhere. If you are planning on using modules, there might be a limited amount of them out there at this point. I hope chip shortage fixes itself soon because even my day job has been deeply affected by it...
@@loughkb Yes, I had to struggle early this year when they ran out to find some one and I manged to finally get a few thousands in August so I am good at the moment but it is a constant struggle to keep inventory up. There is now a Chinese clone of the Si5351 that is as good or better according to Hans Summer. It is called MS5351M, there are plenty of stock for those at the moment.
Suggestion: replace D1 with PMOS polarity protection, that won't blow a fuse (or melt PSU if there isn't one), and the cost shouldn't rise too much. Also Q1 with NMOS such as 2N7002, lower current consumption and the price of small NMOS should be even lower in quantity.
The additional voltage drop across the mosfet would probably increase the minimum supply voltage from 7.2 to 7.8 and we'd lose the option of running it off a 7.2V lithium pack, I think. Q1 is just acting as a switch, the relay coil is the load, you might shave off a fraction of a milliamp, but hardly substantial. However, you are certainly free to modify it as you like, it's an open design.
I think the amp design philosophy is all wrong. This issue of getting power out of an Si or an AD98xx has been solved before. The CHinese HB 1A solves it by driving the gate of FET , J310 or simlar with NO bias (there ) and the FET has a pair of RF transistors in the Drain line. The second one is pulling enough current to drive out 150 mW or so which is compatible with a home made RF transformer. This latter drives an IRF 510. Given you are getting 40x the power of the current offering why not make the effort. The HB1A is no longer offered and the schematic not easy to find it seems, though I have a copy. .Adam Rong used to sell it and had a part in the design. The other option is what was done by Dave Benson for the 4SQrp group monobander where a 74AC04 drives three BS170s in parallel - no transformer required. This is compatible with drive by an SI 5135 with no blocking capacitor , which is always the source of strife. These simple designs did not happen by accident - they are distilled . Output of this latter - about three watts. TE . Fidler
So go ahead and build it that way if you like. It's an open design. Of course you won't be able to run it off a 9V battery or 7.2V lithium battery. It won't be a 'flea' any longer. :-)
BTW the HB1A uses a low power BPF to achieve a sinusoid before it even reaches the FET I spoke of . If you are lacking power to do that then a 74AC logic 00 or 04 (latter hex inverter - using all six gates is asking a lot of it ) can bring the power up considerably as each gate can nominally drive a 75 R load. You can run that logic at 6 V instead of 5V too. I don't agree with the use of 74LS logic to drive HF ham applications where there is something much better. 74 AC logic will produce useful power at 40 Mhz so in the HF band it has head room. Running a low square wave through a capacitor and expecting enough RF current to drive a home made RF transformer is a big ask. but for people who want a challenge... I'd suggest that a type B design which drives the input oscillator cct of a Diz Kits and Parts organic 5W transceiver to make it frequency mobile would be a lot more rewarding, esp if it included either: audio annunciation of frequency or drove a two line LCD display.
Hi Kevin, hope you are doing well! I've been binge watching your videos of late and really enjoying your work. You always give me something to consider. Thank you and 73 from Tracy VE3TWM.
Great project! Recently found your channel and subscribed. I'm taking my Technician's License test today but I am already intrigued by CW and QRPing. Looks like I'll be studying for my General class exam so I can try some of this stuff.
Neat project. I can already see a ton of variations , even using a second pin of the Si5351 for a RX VFO in a DC receiver.
Great project. I am looking forward to working with this one. Thank You again.
P.S. Enjoy those 80’s, it is forecasted to be in the teens here tonight.
Looks like you have a transmitter that can run off a 5 volt phone charger battery pack or a couple 18650 cells. Build a receiver with it using the SI5351(Direct Conversion) and have a low voltage transceiver. Very cool! I can see some fun possibility's with this and SOTA. I see this in my future. Thanks Kevin! Great project. De Troy KF7SEY
Was also thinking of using the SI4732 chip which does AM/FM/SW all in a chip. Would be cool to have a system that can actually run on a couple of AA batteries.
👍Looking good Kevin and thank you for sharing this as open source. I like that LPF chart - knowing wire lengths for inductors is nice.
Harry over at Zachtek had already done the work putting that chart together and gave me permission to use it as is. Saved a lot of time.
Wow, this is all so very impressive. Thank you for sharing your knowledge, Ham radio has to be the most engaging and immersive hobby I have ever been involved in and people like your good self Kevin make it so much better. The things I am learning are mind blowing to me to be honest. Thanks again for your work I really appreciate it.
I see your posts on Qrz by someone else. Looks like a fun project for a cold winters day. It would be cool to have a keyboard keyer project to go with this. Thanks
Looks like a fun project.
Just thinking that an UDOO RYZEN BOLT V1000 Maker Board SBC could be used as there are various hats for the board and you can use the gcc compiler for the 80586
Very Cool! I have an Arduino Uno working with Ada Fruit Si 5351 breakout board. I had to add to the freq variable an offset of long freqcorrect = 002740000 to get the frequency output correct as measured by my Siglent 200 Mhz scope. DigKey and Mouser are out of stock on the line driver chip so looking around.
In the Arduino ide while adding the offset to the variable freq wich is is type long I kept getting an error saying the variable is typed octal..... huh?
Why do you need your frequencies in hundredths of a Hz? Define the initial values in Hz. Set your minimum tunestep to50. If you need hundredths of a Hz somewhere, have a local variable where the frequency in Hz is multiplied by 100. It will make the code easier to follow and also reduces the chance of mistyping the initial values - the fewer the digits in a variable, the better.
[edit]If I recall correctly, CW is allowed anywhere in the bands, with a few exceptions. There are just small segments of various bands that are reserved for CW only. As you mentioned, it might be wise to limit QRP operation to the CW segments.- your 500 mw signal is going to be 10 dB below a 5 watt signal. On the other hand, it can be fun trying to dig minute signals out of the noise.
I'm not a programmer, I just figured things out as I go. I'm sure there's always a better way. The software is open source, you can do whatever you like with it.
I may be mistaken, but couldn't you use the si5351 output to a mixer for receive?
As a direct conversion receiver, yes.
Hi Kevin,
The project is well along. So Harry will be producing a commercial version for sale. Wonder if his circuit could be designed to allow the option for more output using another stage of amplification. Like you mentioned an IRF510, if you add these components you can increase the output to 5 watts. And I like the logo you created. Stay safe. 73 WJ3U
I think, and may experiment down the road, that you could replace the line driver with something like the IRF510 and an inductive choke on the power rail to it. It would need to be fed off the 12 volt rail and heat-sinked. It would also need a driver stage since the output of the Si5351 is not enough to turn on the 510. Just a simple 2n3904 and a couple of resistors with another cap to couple it to the irf510. The choke value I'm unsure of, would need to do some research.
It'd be more of a stink bug and less of a flea at that point. ;-)
I like it! Simple use of modern techniques and components. How about putting the tx/rx relay before the LPF so it also works on receive? Some simple receivers will benefit if they have poor receive filtering.
That's a thought. Although there is always some insertion loss with a filter, so it might harm the receivers sensitivity on the desired band also.
@@loughkb That's true, but in a simple receiver it's worth a couple of dBs loss. Heathkit did it on their HW-7, 8 and 9.
There are some more mods for the Pixie in Sprat a year or so ago, and used the slightly lossier BPF which was well worth it to reduce broadcast breakthrough. (Plus the Pixie Pin 7 mod!)
@@janetwinslow2039 Thanks for the info. I might experiment with it down the road.
6:20 12 to 50 ohms "4:1 transformer" IIRC, a voltage and turns ratio might be better called "2:1" to achieve a 4x impedance ratio. i.e. A "2:1" transformer, with a 2:1 turns ratio, changes voltage by 2:1 and current by 2:1 in the opposite direction, thus changes impedance by 4:1. But it's still a "2:1" transformer.
I'll insert the word "impedance" in there on the documentation to make it clear.
Any scope/spectrum pictures of what the output signal actually looks like? 2nd, 3rd order harmonics?
Sure, once I complete the LP filter when the parts arrive. You can also go to the zachtek site and look at the page on his LP1 beacon. It's the same output section and he has scope pics of the output and harmonics there.
@@loughkb Thanks!
I am just wondering......... are you considering an electro-mechanical relay for the TR switch or solid state switch?
Mechanical. The published schematic on the blog has a mouser part number.
Kevin this looks awesome! I am a bit intimidated by winding inductors, i never have done that before. Can you point me to some resources or show how you are you doing your windings? BTW built your field day end fed vert for my OM, works super!
I'll show and talk a bit about winding in the build video. It's easy really.
Also, Zachtek provided a great illustration in the PDF manual on his mezzanine board. You can download that on his website.
hi kevin from jim N5BCR is there some way i could contact you concerning a problem with my heathkit SA-5010 umatic memory keyer
My public email is listed on my qrz.com page.
How is Yuma ? It was 81'f here in Mesa today
We hit the mid 80's today. It's looking like we'll still have some 80's through the rest of the month. I'm hoping for a mild 'winter' here this year.
Will be building this project.
You said there will be a released product version, they doing a kit version? Otherwise I'll send you the BOM order and a PCBWay or JLCPCB order option for the board you could link the project so people can kit build it without having to the additional steps.
Definitely send me that info if you do a PCB version of your own. I may do another video down the road on what the project's lead to.
Zachtek will be doing a commercial version. I believe he will be selling a populated PCB version, ready for air.
Is the sketch HUGE for the Arduino code? Could this be a good VFO for other QRP rigs that are less frequency "nimble"? Thanks for sharing this!
No, it's a relatively small sketch. Sure. This method could replace the oscillator on other rigs. Depending on their design, it might be neccissary to include a LP filter to smooth the 5351's square wave to a clean sine wave and reduce harmonics. You might also need a single transistor 'driver' stage to increase the output signal from the chip.
@@loughkb Thanks 🙏 Kevin.
Nice
It looks like a fun project. I have a question about the tuning procedure. Do you reduce power (or switch to built in dummy load) when tuning to prevent QRM?
I think pin 3 is programmed to only come on when you "spot tune". This is happening when in receive mode and the only signal is that sneaky little one turn coupling of wire from pin 3 on the receiver input....so not being transmitted as relay is not engaged in transmit mode. So no QRM. :)
I detailed that in the video. The second output of the 5351 is turned on and coupled to the receiver output via the gimmick wire. The power path is not enabled during tuning. It should also be quite clear in the schematic. (shown in the video and also in the blog write up that's linked in the video description. Which also includes an explanation of the tuning process.)
Hey Kevin, great project. Are you planning on integrating all the components on the pcb? Because if you are, you're most likely going to have issues finding sources for the SI5351. Most vendors are out of stock of those ics. I have a friend of mine who was designing an sdr transciever using those ics and he had to halt the project because he could not find mcus or si5351 anywhere. If you are planning on using modules, there might be a limited amount of them out there at this point. I hope chip shortage fixes itself soon because even my day job has been deeply affected by it...
I'm building mine as a hobbiest and using the modules. Zachtek will be using the SMD components and has a source lined up I think.
@@loughkb Yes, I had to struggle early this year when they ran out to find some one and I manged to finally get a few thousands in August so I am good at the moment but it is a constant struggle to keep inventory up. There is now a Chinese clone of the Si5351 that is as good or better according to Hans Summer. It is called MS5351M, there are plenty of stock for those at the moment.
Suggestion: replace D1 with PMOS polarity protection, that won't blow a fuse (or melt PSU if there isn't one), and the cost shouldn't rise too much. Also Q1 with NMOS such as 2N7002, lower current consumption and the price of small NMOS should be even lower in quantity.
The additional voltage drop across the mosfet would probably increase the minimum supply voltage from 7.2 to 7.8 and we'd lose the option of running it off a 7.2V lithium pack, I think.
Q1 is just acting as a switch, the relay coil is the load, you might shave off a fraction of a milliamp, but hardly substantial.
However, you are certainly free to modify it as you like, it's an open design.
Great stuff OM , 73 K9UT
Idea: support a iambic keyer ;-)
I'm using a stereo jack for the key input. Plenty of spare I/O pins on the arduino. I'm sure someone will shoehorn in keyer code. :-)
Looks interesting to try. If you hear me at 10 wpm be kind!
I think the amp design philosophy is all wrong. This issue of getting power out of an Si or an AD98xx has been solved before. The CHinese HB 1A solves it by driving the gate of FET , J310 or simlar with NO bias (there ) and the FET has a pair of RF transistors in the Drain line. The second one is pulling enough current to drive out 150 mW or so which is compatible with a home made RF transformer. This latter drives an IRF 510. Given you are getting 40x the power of the current offering why not make the effort. The HB1A is no longer offered and the schematic not easy to find it seems, though I have a copy. .Adam Rong used to sell it and had a part in the design.
The other option is what was done by Dave Benson for the 4SQrp group monobander where a 74AC04 drives three BS170s in parallel - no transformer required. This is compatible with drive by an SI 5135 with no blocking capacitor , which is always the source of strife. These simple designs did not happen by accident - they are distilled . Output of this latter - about three watts.
TE . Fidler
So go ahead and build it that way if you like. It's an open design.
Of course you won't be able to run it off a 9V battery or 7.2V lithium battery. It won't be a 'flea' any longer. :-)
BTW the HB1A uses a low power BPF to achieve a sinusoid before it even reaches the FET I spoke of . If you are lacking power to do that then a 74AC logic 00 or 04 (latter hex inverter - using all six gates is asking a lot of it ) can bring the power up considerably as each gate can nominally drive a 75 R load. You can run that logic at 6 V instead of 5V too. I don't agree with the use of 74LS logic to drive HF ham applications where there is something much better. 74 AC logic will produce useful power at 40 Mhz so in the HF band it has head room. Running a low square wave through a capacitor and expecting enough RF current to drive a home made RF transformer is a big ask. but for people who want a challenge...
I'd suggest that a type B design which drives the input oscillator cct of a Diz Kits and Parts organic 5W transceiver to make it frequency mobile would be a lot more rewarding, esp if it included either:
audio annunciation of frequency or
drove a two line LCD display.
👍🏻👋