I've always noticed that when you get to that magical-20W that you can almost guarantee return contacts from anyone you can hear! When I design for 'survival' radios, I always make sure the PA is designed to acheive 20W; past that and you gain little and should just then use your 'appliance' rig! Would really like to see and update video whereby you add a buffer-amp and see how that affects the chirp. Anyway Peter, excellent video with great explanations--thanks for posting! 73...
20W wow... that's really nice from such a simple schematic. I only got to built the original version of the transmitter... and I just received the power transistors a few days ago thinking to build the other version you made for 5W. Not sure if I will make this one too... but it's good to know I can get more out of it if I need it )) Another great video Peter. I must watch closely the transformer made on the FT50-43 as I don't really understand how it's made ))) 73 and thank you for sharing this... as always.
Nice job Peter. QRP is good for portable and 5 Watts allows more operating time than 20 Watts so I'd go for 5. Xtal transmitters are so easy to make. I use G QRP Club xtals and QRP-QRP contacts are great. 7.030 is good in Europe on this QRP frequency. Bill, G4GHB.
Great fun! A MOPA Out of interest, on 30v, chirp aside, how much power could you get on 20m? My guess is around 5 to 6 watts, due to the rapid drop off in power of the IRF FETs. WA3EBY runs two of them at around 35v (?), wear safety glasses!
Paul - It was more when I tried it - around 12-15w from memory though that was just into the power meter with no low pass filter. Yes, you should mount the 510 in a way it can easily be replaced!
I've got some IRF531 transistors. IIRC they are the same as the IRF530's except for the D-S voltage rating. I might want to put a tuned circuit on the output of the oscillator so I can use some war surplus 5.x mhz rocks I have in the junkbox on 30 meters. Also could use an Si5351 "DDS" vfo instead of a crystal with an arduino to drive that.
You will lose oscillator drive if you use that stage as a doubler. Look for cheap 10.111 MHz crystals on ebay. These will work in this circuit. A DDS VFO could also work but you'll need more stages to amplify its output.
Hi Peter, what great little idea, Just as a heads up, if you are a member of the G-QRP club you can get a whole host of xtals on qrp frequencies as well as some ceramic resonators. 72 Mark G0NMY
Nice construction! A preferred power supply, I think, would be 8 Li-Ion 18650 batteries in series (~30 V) with an LM7812 linear regulator for an oscillator. Having a receiver in the same enclosure would be nice too. 73s from Russia!
You can fix the sidetone delay but having a local receiver (no antenna connected). Tune it to a pitch different (preferably higher) than what comes out of the SDR and concentrate on that when sending.
Well, if the BD139 can be used as an oscillator, then if you get a buffer and an audio modulator circuit, how far is it possible to transmit with 3 BD139 transistors?
Nice video and great fun. I have built a 4-transistor XCO circuit (2n3094-2n3094-2n3053-IRF510) that can do 20W on about 18V power. I am trying to build another transmitter (si5351-BD139-IRF510). On LT Spice it seems to be able to push >20W at 18V power, but after building it I am only getting 5W... hihi Some more playing to do. A quick question, how did you come with 10T on FT50-43 transformer. I only have 5T now and it gets a bit hot. Also what's the effect of stacking two cores? Equivalent to a larger core??
👍very good, thanks Peter. Your video prompted me to order some more BD139's 😃 (the last couple I blew up trying to get a Pixie kit to work). I like the plug in LPF idea. Calling CQ on CW QRP, in the UK, I have found the same 20 min. "rule" applies. Although I sometimes copy a quick reply, it can often take 15 to 20 minutes of calling to get a QSO. As I think you mentioned in a previous video, sending QRZ? at end of a QSO may help - chances are some operators will have been listening and may reply if they hear QRZ?
I put my Pixie on 60m and got a SWL report from 220 miles away and worked a couple of stations 180 miles and some nearer stations. Not bad for 350 mW out. Using a low inverted vee. Less QRM from other stations on 60m but not always a busy band. Unfortunately the Pixie blew up and the copper track lifted when I attempted a repair. Bill, G4GHB.
Yes the IRF 510. my nemesis...waiting on a delivery of 2 of them.....blew mine in the kit I'm building :( It had a gate drain resistance of 2 ohms :((((( I killed it good!!!! Mosfets are amazing powerful but so fragile?? I am a static machine.....might need to buy a wrist strap?
I've not worried about static with them. But depending on the circuit they are super fragile if you adjust the DC bias to the gate too high. With enough drive you can get away with no DC bias if you're just building a CW transmitter.
@@vk3ye yes …. That’s what I did…. Cranked the pot too far and let the smoke out! Good to know static is less of an issue (pretty sure they have protection diodes in them?) But I ordered 2, so one will go into my little transmitter which now has agility at bottom of 80 m band. Was chirping but lowered coupling from oscillator to second “pa” stage and it is ok now . But even lower power abt 300 mW. So will beef it up with an IR510. Also added brass contacts to Morse key. So much better now. Breaking stuff teaches you heaps! 🤡…. Do u ever sleep??? 😁
Nice project Peter! I'd like to build this as an amplifier for a 20m QRP rig. Will 250 mW of drive into the IRF510 be enough to get 20W output? Any idea how much power your are getting out of the BD139 oscillator circuit?
I don't know if this question is still relevant, but I'll answer nevertheless. 250mw will get you quite far. He's putting in 200mw. BD139 is quite beasty, it can do 4w. Single IRF510 can do 50w+ (class E), just don't mistreat it. Add a reversed diode right onto the gate of the IRF510(parallel to resistor), will prevent blow up, I speak out of experience. Also if it is oscillating add a couple of ohms worth of resistance between the gate and the cap (diode will then sit between cap and res to ground, reversed so it doesn't open up with positive V). If you are not using a crystal, reverse the resistor at bd139 to ground, will directly work with input from si5351 to 20w. if you want more than irf510 go for irf530, works up to 75w-ish. T50-2 or T68-2 can also be used instead of ft50-43, amount of windings not so important. 10 bifilar windings around one of those work fine.
@@pa0fxx509 Thank you for the comments! After I posted the question, I saw that Peter answers this question in the video, duh. I was inspired enough to do some experiments on my own and can fully believe all your comments regarding power output. I used an IRF520. It I saw 55W at 32V and that was class-c. There appear to be never-ending opportunities to tinker with these devices. I think that the key to success with the FETs is low drive and 'high' voltage on the drain. Once the drive gets up to a certain level, the gate waveform really deteriorates and the output signal became rich in harmonics. But low drive and high drain voltage...that gets it done :) Thank you, Brad (AD0RX)
@@paulvandergroen9569 Cheers, I was thinking that the oscillator would take time to stabilise once power is applied, so I can see how your idea would work
Whats next? Class 'E' Amp? A 75w valve Novice transmitter? How about powering the transmitter from a 19v/20v laptop PSU (a good one, like Lenovo/IBM )? regards G7VFY
@@Steve-GM0HUU Lenovo 80watt, 20v psu's a really high quality and are rarely noisy. I picked up a few of these at a car boot sale and they are very beefy. I've been meaning to crack one open to see if I can turn the voltage down to 13.8volts. www.ebay.co.uk/itm/274656000818
@@Steve-GM0HUU Slim tips: From ThinkWiki: power rating resistance. ua-cam.com/video/XO3D0d8a_fc/v-deo.html 36 W 7.3 kΩ 45 W 120 Ω 65 W 280 Ω 90 W 550 Ω 135 W 1 kΩ 170 W 1.9 kΩ 230 W 4.6 kΩ
Another couple of questions. Is the transformer 4:1 (Z) because it is wound as an auto transformer with a 2:1 turns ratio? Also, is that ferrite bead on the transistor necessary (VHF parasitics prevention)? Thank you.
"Is the transformer 4:1 (Z) because it is wound as an auto transformer with a 2:1 turns ratio?" Yes. Look at it this way: the input voltage Vin is halved and the input current Iin is doubled at the output. As resistance/impedance is defined as V/I, at the output of the transformer, Zout = 0.5*Vin/(2*Iin) = (1/4)*Zin.
@@vk3ye Peter, I built the first stage, but it does not seem to be oscillating. Does it need the load of the gate of the following stage to get it to oscillate? The BD139 is getting warm at key down, but no tone in a nearby receiver.
@@vk3ye It is working now. I noticed the source of the FET was not grounded. Xtal gets HOT. It is the smaller type HC-49. Probably the old FT-243 would work better, or maybe a buffer stage to run the oscillator a little cooler.
@@MegaCadr Steal from the best and take a lucky guess. If it gets hot, use more stuff. Ferromagnetic Core Design & Application Handbook by W1FB I downloaded it from archive dot org. PM if you want a copy
I've always noticed that when you get to that magical-20W that you can almost guarantee return contacts from anyone you can hear!
When I design for 'survival' radios, I always make sure the PA is designed to acheive 20W; past that and you gain little and should just then use your 'appliance' rig!
Would really like to see and update video whereby you add a buffer-amp and see how that affects the chirp.
Anyway Peter, excellent video with great explanations--thanks for posting! 73...
20W wow... that's really nice from such a simple schematic. I only got to built the original version of the transmitter... and I just received the power transistors a few days ago thinking to build the other version you made for 5W. Not sure if I will make this one too... but it's good to know I can get more out of it if I need it )) Another great video Peter. I must watch closely the transformer made on the FT50-43 as I don't really understand how it's made ))) 73 and thank you for sharing this... as always.
Well done Peter love your work
Really interesting, 20w is a very workable wattage. Thanks Peter.
Nice job Peter. QRP is good for portable and 5 Watts allows more operating time than 20 Watts so I'd go for 5. Xtal transmitters are so easy to make. I use G QRP Club xtals and QRP-QRP contacts are great. 7.030 is good in Europe on this QRP frequency.
Bill, G4GHB.
20 watts is amazing, Good job
Great fun! A MOPA Out of interest, on 30v, chirp aside, how much power could you get on 20m? My guess is around 5 to 6 watts, due to the rapid drop off in power of the IRF FETs. WA3EBY runs two of them at around 35v (?), wear safety glasses!
Paul - It was more when I tried it - around 12-15w from memory though that was just into the power meter with no low pass filter. Yes, you should mount the 510 in a way it can easily be replaced!
The frequencyswitch is f*ckin kewl.
Wow that 1dB makes a big difference. (Your 5 watts version was completely readable on the kiwiSDR though)
I've got some IRF531 transistors. IIRC they are the same as the IRF530's except for the D-S voltage rating.
I might want to put a tuned circuit on the output of the oscillator so I can use some war surplus 5.x mhz rocks I have in the junkbox on 30 meters.
Also could use an Si5351 "DDS" vfo instead of a crystal with an arduino to drive that.
You will lose oscillator drive if you use that stage as a doubler. Look for cheap 10.111 MHz crystals on ebay. These will work in this circuit. A DDS VFO could also work but you'll need more stages to amplify its output.
Hi Peter, what great little idea, Just as a heads up, if you are a member of the G-QRP club you can get a whole host of xtals on qrp frequencies as well as some ceramic resonators.
72 Mark G0NMY
Good work
Nice construction! A preferred power supply, I think, would be 8 Li-Ion 18650 batteries in series (~30 V) with an LM7812 linear regulator for an oscillator. Having a receiver in the same enclosure would be nice too. 73s from Russia!
Very nice design. And impressive focus. The delay in the the sidetone would wreck me!
You can fix the sidetone delay but having a local receiver (no antenna connected). Tune it to a pitch different (preferably higher) than what comes out of the SDR and concentrate on that when sending.
In over 20 years I never got round to putting sidetone on my h/b rig.
G4GHB.
Well, if the BD139 can be used as an oscillator, then if you get a buffer and an audio modulator circuit, how far is it possible to transmit with 3 BD139 transistors?
BTW, just re-watched this. Was that your first harmonic getting spotted on RBN?
Nice video and great fun. I have built a 4-transistor XCO circuit (2n3094-2n3094-2n3053-IRF510) that can do 20W on about 18V power. I am trying to build another transmitter (si5351-BD139-IRF510). On LT Spice it seems to be able to push >20W at 18V power, but after building it I am only getting 5W... hihi Some more playing to do. A quick question, how did you come with 10T on FT50-43 transformer. I only have 5T now and it gets a bit hot. Also what's the effect of stacking two cores? Equivalent to a larger core??
Very cool
👍very good, thanks Peter. Your video prompted me to order some more BD139's 😃 (the last couple I blew up trying to get a Pixie kit to work). I like the plug in LPF idea. Calling CQ on CW QRP, in the UK, I have found the same 20 min. "rule" applies. Although I sometimes copy a quick reply, it can often take 15 to 20 minutes of calling to get a QSO. As I think you mentioned in a previous video, sending QRZ? at end of a QSO may help - chances are some operators will have been listening and may reply if they hear QRZ?
I put my Pixie on 60m and got a SWL report from 220 miles away and worked a couple of stations 180 miles and some nearer stations. Not bad for 350 mW out. Using a low inverted vee. Less QRM from other stations on 60m but not always a busy band.
Unfortunately the Pixie blew up and the copper track lifted when I attempted a repair.
Bill, G4GHB.
Yes the IRF 510. my nemesis...waiting on a delivery of 2 of them.....blew mine in the kit I'm building :( It had a gate drain resistance of 2 ohms :((((( I killed it good!!!! Mosfets are amazing powerful but so fragile?? I am a static machine.....might need to buy a wrist strap?
I've not worried about static with them. But depending on the circuit they are super fragile if you adjust the DC bias to the gate too high. With enough drive you can get away with no DC bias if you're just building a CW transmitter.
@@vk3ye yes …. That’s what I did…. Cranked the pot too far and let the smoke out! Good to know static is less of an issue (pretty sure they have protection diodes in them?) But I ordered 2, so one will go into my little transmitter which now has agility at bottom of 80 m band. Was chirping but lowered coupling from oscillator to second “pa” stage and it is ok now . But even lower power abt 300 mW. So will beef it up with an IR510. Also added brass contacts to Morse key. So much better now. Breaking stuff teaches you heaps! 🤡…. Do u ever sleep??? 😁
Is there a nice Bill of Materials list of all the parts used in this circuit?
So few parts that it's not needed.
Nice project Peter!
I'd like to build this as an amplifier for a 20m QRP rig.
Will 250 mW of drive into the IRF510 be enough to get 20W output?
Any idea how much power your are getting out of the BD139 oscillator circuit?
I don't know if this question is still relevant, but I'll answer nevertheless.
250mw will get you quite far. He's putting in 200mw. BD139 is quite beasty, it can do 4w. Single IRF510 can do 50w+ (class E), just don't mistreat it.
Add a reversed diode right onto the gate of the IRF510(parallel to resistor), will prevent blow up, I speak out of experience. Also if it is oscillating add a couple of ohms worth of resistance between the gate and the cap (diode will then sit between cap and res to ground, reversed so it doesn't open up with positive V).
If you are not using a crystal, reverse the resistor at bd139 to ground, will directly work with input from si5351 to 20w.
if you want more than irf510 go for irf530, works up to 75w-ish.
T50-2 or T68-2 can also be used instead of ft50-43, amount of windings not so important. 10 bifilar windings around one of those work fine.
@@pa0fxx509 Thank you for the comments!
After I posted the question, I saw that Peter answers this question in the video, duh.
I was inspired enough to do some experiments on my own and can fully believe all your comments regarding power output. I used an IRF520. It I saw 55W at 32V and that was class-c. There appear to be never-ending opportunities to tinker with these devices.
I think that the key to success with the FETs is low drive and 'high' voltage on the drain. Once the drive gets up to a certain level, the gate waveform really deteriorates and the output signal became rich in harmonics. But low drive and high drain voltage...that gets it done :)
Thank you,
Brad (AD0RX)
What gauge of magnet wire was used on the toroids?
About 0.5mm
There is a drift on the tone.. just wondering how you would go stabilising it. Cheers VK5LEX
Don't key the oscillator, use a J310 to drive the BD139.
@@paulvandergroen9569 Cheers,
I was thinking that the oscillator would take time to stabilise once power is applied, so I can see how your idea would work
Whats next? Class 'E' Amp? A 75w valve Novice transmitter? How about powering the transmitter from a 19v/20v laptop PSU (a good one, like Lenovo/IBM )? regards G7VFY
Laptop power supplies are really under represented in ham radio builds, imo. Decent quality, very available, decent power for medium power and CHEAP.
That's a good idea. I have a few of these but never really thought of using them for radio.
@@Steve-GM0HUU Lenovo 80watt, 20v psu's a really high quality and are rarely noisy. I picked up a few of these at a car boot sale and they are very beefy. I've been meaning to crack one open to see if I can turn the voltage down to 13.8volts.
www.ebay.co.uk/itm/274656000818
@@Steve-GM0HUU Lenovo power supply plug configuration
ua-cam.com/video/XO3D0d8a_fc/v-deo.html
@@Steve-GM0HUU Slim tips: From ThinkWiki:
power rating resistance. ua-cam.com/video/XO3D0d8a_fc/v-deo.html
36 W 7.3 kΩ
45 W 120 Ω
65 W 280 Ω
90 W 550 Ω
135 W 1 kΩ
170 W 1.9 kΩ
230 W 4.6 kΩ
Where did you get the knowledge to do the 4:1 transformer for the fet?
Details of good books, including some for transmitter construction, are on my website at: vk3ye.com/equipment/revbooks.htm
Another couple of questions. Is the transformer 4:1 (Z) because it is wound as an auto transformer with a 2:1 turns ratio? Also, is that ferrite bead on the transistor necessary (VHF parasitics prevention)? Thank you.
"Is the transformer 4:1 (Z) because it is wound as an auto transformer with a 2:1 turns ratio?" Yes. Look at it this way: the input voltage Vin is halved and the input current Iin is doubled at the output. As resistance/impedance is defined as V/I, at the output of the transformer, Zout = 0.5*Vin/(2*Iin) = (1/4)*Zin.
If I just want to try the 5W version, which FET do I use? I already have the BD139. TNX es 73 de KQ6UP
IRF510 would be fine.
@@vk3ye TU es 73 de KQ6UP SK :D
@@vk3ye Peter, I built the first stage, but it does not seem to be oscillating. Does it need the load of the gate of the following stage to get it to oscillate? The BD139 is getting warm at key down, but no tone in a nearby receiver.
@@kq6upA load may help. I'd also try adding a capacitor from base to ground of the BD139 in case that heps.
@@vk3ye It is working now. I noticed the source of the FET was not grounded. Xtal gets HOT. It is the smaller type HC-49. Probably the old FT-243 would work better, or maybe a buffer stage to run the oscillator a little cooler.
I can't tell which channel is you on the water fall??
It's in the 7 MHz amateur band, as described.
What I meant was that i couldn't tell which trace was your transmitter on the waterfall.
WOOW!!
i get 100watts from a single irf510 at 24 volts
❤❤❤
Hey class C is great if you are just plugging it into AC
class- C amplifier. wow.
Why two toroids?
More power handling, less turns.
@@paulvandergroen9569 thanks! Is there a guide to how much power a given core size/mix can handle?
@@MegaCadr Steal from the best and take a lucky guess.
If it gets hot, use more stuff.
Ferromagnetic Core Design & Application Handbook by W1FB
I downloaded it from archive dot org. PM if you want a copy
@@paulvandergroen9569 haha
I might have some type 61 in that size.
Google dye granberg motorola