PROTO - SkyWave DX600T Rev-C RF Linear Amplifier

A new revision of the DX600T RF linear amplifier which now incorporates an automatic over-temperature shutdown/recovery feature, and still easily puts out over 600W.

[@jasonfrye6473]

Thanks for sharing this updated version, excellent point about the loss via output transformer.
I would've thought LDMOS was gonna take over the RF amplifier world when i first seen the capabilities years ago but so far just a slow progress, maybe the higher voltage is the reason?
Anyway nice 👍 glad to see people with knowledge still experimenting

[@TelstarElectronics]

Yes, Mosfets require higher voltage... and are not very useful for mobile operation. But for home use, they certainly make more sense than bipolar devices...

[@nandrews8412]

Good to see you, been a minute.

[@TelstarElectronics]

Thanks. Was very busy moving and then retiring. But now I have time again to design some stuff...

[@freddyb2098]

The Professor is Back !! Missed You ! Looking forward to future video's and electronic offerings.

[@TelstarElectronics]

Good to hear from you. Yep, the ol' Professor is always striving for cost-effective, reliable, and high-performance amplifiers with a great feature list...

[@alsanderson4917]

It's been a while. Do you have a nano or mini VNA for fine tuning the input and output on a smith chart? VNA is also handy for dialing in pi filters, low pass, high pass, notch. Love my VNA.

[@TelstarElectronics]

Absolutely... not sure how you could do it without a VNA.

[@nonumquam1419]

Congratulations, very intersting design!
I have three questions:
1. With the new impedance matching output, what is the power amplifier bandwidth?
2. What was the reason to route back (on the PCB) the output of the 600W power amplifier, to just about 1 cm (0.4 inch) away from the input?
3. Is the long (90 deg. bent) RF output trace 50 ohms impedance?
I don't know what PCB substrate you are using and if use ground underneath, but if is about the standard FR4, thickness 1.5 mm (59 mil), with ground, the 50 ohms microstrip trace should be about 2.7mm wide. Compared to the size of the MRF300 package, this microstrip line is much wider.
tnx

[@TelstarElectronics]

The output section is much higher Q than the usual wideband transformer... so the bandwidth is much narrower. In this case, the 3dB output bandwidth is about 5MHz. The tuned circuit output has several advantages (if wideband operation is not needed)... such as very high efficiency. There is only about 5W being dissipated on the entire output section. The other advantage is that the tuned circuit acts like a built-in low pass filter... aiding in reducing harmonics... although this board also has an additional pi output filter too.
The input and output connection proximity is a result of being close to the transmit/receive relay... to keep any stray reactance to a minimum.
The 1.6mm board uses 1oz double sided copper... with a big bottom ground plane (~90% of surface). Impedance matching at lower frequencies is not critical... since the trace length is just a small fraction of a wavelength.

[@buckaroobonsi555]

Glad to see your back! I hope this finds you well and in good spirits!

[@TelstarElectronics]

Good to hear from you. Now that I'm retired, I have more time to spend on these projects...

[@k8byp]

BAD PWB design/ layout. Insufficient copper around hot components/ crowded too closely.

[@TelstarElectronics]

Huh? There are only really three higher-power resistors on the PCB, and they all have a design margin of 100% (doubled required wattage). In addition, they are not in physical contact with board itself. I can tell by your comment you haven't laid out many/any boards...