Charlie - Great work! I am in the same mode here with papers in hand. I decided to just use the Sokal formulas and an old 2 dollar 2SC2078 Bipolar CB transistor. Wow is all I can say. I designed for 10W out at 15 VDC. My circuit is just as drawn in the paper with a 10 Ohm R target. I used a 1:4 UNUN that got me up towards 50 Ohms then a filtered as you did on the tail end. 91% efficiency with zero optimization, and right on the nose at 10W out. Increasing VCC to 20V gave me 18W out with only a slight impact on efficiency.
Some of the most interesting content is from people who don't present themselves as experts. I really appreciate the way you talk through your idea, theory, design, implementation and results. Most importantly, even when the results are imperfect.
I appreciate the tutorial (even though you don't think it is) and demo. I'm designing a 2nd iteration of a Class E Amplifier but with 13.56 MHz. I already have a buck converter made up to boost the sine wave signal using Vcc 1V - 24V.
@@alimaiee1647 yeah, I was able to sweep from sub 1 V all the way up to 18.5 V.. but couldn't get up to 24 V. I kept losing a lot of heat on the MOSFET (n-type) and I just couldn't figure out why.
Hello!Nice work! But your gate driver ciruit works well until 1MHz.Some weeks ago I had made simple shortwave transmitter. Output stage was push-pull.There was two IRLML0040. This transistor has a low gate charge - 3nC. IRF520 has a gate chage 16nC. Gain of BJT has been follen at high frequency.For example - 2N3904 has Beta=100 at DC.But when we have frequency 5MHz - this parameter will be goughly 10..15. We can have a fast swtching of output mosfet - to prevent heat losses (the mosfet will be work in linear mode with long rise and fall time). The best way to improve your device - using low stage gate driver IC.
Excellent description and demo, thank you. But if I heard you correctly, DC input power is about 4 W and RF output is 2 W so not as efficient as expected. More like class C.
Correct. It's definitely not working correctly Elwood. I'm going to try a totem pole drive this morning, then look to try another amp using three parallel BS170s. It would be nice to get a Class E amp working close in line with the design and the LTSpice simulation (noting that it's only a simulation).
Hi Charlie, I noticed a new sig-gen in your video, looks like a feel tech, if so what are your thoughts on quality/performance of this unit compared to the siglent ? Thanks......Peter GW4ZUA
Hi Peter. I've had the FeelElec FY6900 (they are all the same) for some time and it is a good sig gen. It actually does 99% of what the Siglent does. I decided in the end the Siglent output levels were displaying more accurately than the FeelElec. If you are on a budget then I'd certainly recommend the FeelElec. But, you do get what you pay for. I had a common problem where the sinewave output waveform was corrupted. I had to install the PC software and run the specially added 'fix sinewave' function. I hope that helps Peter.
Hi Charlie , Thanks for the feedback, here in the UK the Feel tech / elec version is about a third of the price of the Siglent (£278) , there is supposed to be a price rise coming up ! H'mm but as its close to my birthday............... :-) . Thank you for doing the videos a great source of knowledge and inspiration, between you and Pete N6QW I have learnt so much ............Thanks Peter
I’m really surprised you are able to drive the irf510 with your signal generator. Too bad you weren’t getting better results. Did you try driving it with a higher gate voltage?
I'm a little surprised too, but it looked to work. No I didn't try a higher gate voltage, but I will though. My plan is to also have the gate sitting at-say-1.2 to 2V, then have the RF drive on top of that.
I just increased the gate drive voltage up to 6.5V. There was no increase in output power. 5V provides the max output. I must be doing something wrong.
@@MegaCadr When I scope the gate, 5Vp-p on the sig gen = 8.1Vp-p on the scope. The gate does not look like 50ohm, so the sig gen display will be inaccurate. Either Way, I believe I'm getting 8Vp-p on the gate. I wonder if I should try the MC34151 again.
Charlie - Great work! I am in the same mode here with papers in hand. I decided to just use the Sokal formulas and an old 2 dollar 2SC2078 Bipolar CB transistor. Wow is all I can say. I designed for 10W out at 15 VDC. My circuit is just as drawn in the paper with a 10 Ohm R target. I used a 1:4 UNUN that got me up towards 50 Ohms then a filtered as you did on the tail end. 91% efficiency with zero optimization, and right on the nose at 10W out. Increasing VCC to 20V gave me 18W out with only a slight impact on efficiency.
Some of the most interesting content is from people who don't present themselves as experts. I really appreciate the way you talk through your idea, theory, design, implementation and results. Most importantly, even when the results are imperfect.
Thank you very much for that, I appreciate it a lot. I'm glad the videos are useful.
I appreciate the tutorial (even though you don't think it is) and demo. I'm designing a 2nd iteration of a Class E Amplifier but with 13.56 MHz. I already have a buck converter made up to boost the sine wave signal using Vcc 1V - 24V.
Have u done it?
@@alimaiee1647 yeah, I was able to sweep from sub 1 V all the way up to 18.5 V.. but couldn't get up to 24 V. I kept losing a lot of heat on the MOSFET (n-type) and I just couldn't figure out why.
Only saw this as its pretty late at night. Definitely gonna take a look at this video later
a man who knows how little he knows, knows a great deal more than one who thinks he knows how much he knows
Another great video Charlie!
Thanks Chris. I'll keep working at it as it would be nice to get it to work like the theory.
Thanks a lot Charlie. Appreciate the comment as to why you shoot in 720 as well - understood.
Cheers Graeme. We have fibre going in soon so that alleviate the bandwidth issue.
I am new to Class E, why you use sine wave as input for simulation, but the square wave as input in experiment? what is the difference between them?
Probably my mistake. It believe it should be a square wave. I will be going another Class E amp at some stage, so will look at it again then.
Hello!Nice work! But your gate driver ciruit works well until 1MHz.Some weeks ago I had made simple shortwave transmitter. Output stage was push-pull.There was two IRLML0040. This transistor has a low gate charge - 3nC. IRF520 has a gate chage 16nC. Gain of BJT has been follen at high frequency.For example - 2N3904 has Beta=100 at DC.But when we have frequency 5MHz - this parameter will be goughly 10..15. We can have a fast swtching of output mosfet - to prevent heat losses (the mosfet will be work in linear mode with long rise and fall time). The best way to improve your device - using low stage gate driver IC.
Thanks Andrew. In the end I went with a totam pole driver which worked well. The MC34151P for whatever reason did not work well for me.
Excellent description and demo, thank you. But if I heard you correctly, DC input power is about 4 W and RF output is 2 W so not as efficient as expected. More like class C.
Correct. It's definitely not working correctly Elwood. I'm going to try a totem pole drive this morning, then look to try another amp using three parallel BS170s. It would be nice to get a Class E amp working close in line with the design and the LTSpice simulation (noting that it's only a simulation).
Hi Charlie，I see the filter inductors with cors in this PCB, how do you select them under 3.525MHz operating frequency with less loss
Hi Max. I use type 43 material which is suitable for what I use them for. Check out toroids.info/FT37-43.php
Hi Charlie,
I noticed a new sig-gen in your video, looks like a feel tech, if so what are your thoughts on quality/performance of this unit compared to the siglent ?
Thanks......Peter GW4ZUA
Hi Peter. I've had the FeelElec FY6900 (they are all the same) for some time and it is a good sig gen. It actually does 99% of what the Siglent does. I decided in the end the Siglent output levels were displaying more accurately than the FeelElec. If you are on a budget then I'd certainly recommend the FeelElec. But, you do get what you pay for. I had a common problem where the sinewave output waveform was corrupted. I had to install the PC software and run the specially added 'fix sinewave' function. I hope that helps Peter.
Hi Charlie , Thanks for the feedback, here in the UK the Feel tech / elec version is about a third of the price of the Siglent (£278) , there is supposed to be a price rise coming up ! H'mm but as its close to my birthday............... :-) .
Thank you for doing the videos a great source of knowledge and inspiration, between you and Pete N6QW I have learnt so much ............Thanks Peter
@@peterwebb2224 Thanks Peter. I really appreciate the feedback.
I’m really surprised you are able to drive the irf510 with your signal generator.
Too bad you weren’t getting better results. Did you try driving it with a higher gate voltage?
I'm a little surprised too, but it looked to work. No I didn't try a higher gate voltage, but I will though. My plan is to also have the gate sitting at-say-1.2 to 2V, then have the RF drive on top of that.
I just increased the gate drive voltage up to 6.5V. There was no increase in output power. 5V provides the max output. I must be doing something wrong.
Charlie Morris does it actually look like 6.5v on the gate when you are applying it?
@@MegaCadr When I scope the gate, 5Vp-p on the sig gen = 8.1Vp-p on the scope. The gate does not look like 50ohm, so the sig gen display will be inaccurate. Either Way, I believe I'm getting 8Vp-p on the gate. I wonder if I should try the MC34151 again.
Charlie Morris according to this presentation, the irf510 needs quite a bit of drive power www.norcalqrp.org/files/Class_E_Amplifiers.pdf
Thumbs Up Charlie
We'll get there. I just need to crack the secret...
Nice video de Yi1hxh rafat 73 from baghdad
Thanks Rafat. I'm going to try BS170s today.