100W Class E 160M AM Transmitter build (FAT5)

[VK3HN]

Hi all

This is my first post to the Class E Forum. I recently got my first Class E Tx working on 160m AM.  The Tx is a reproduction of the FAT5 RF PA and linear/series modulator from Dave GW4GTE and Eric GW8LLJ.

I initially started with a crystal oscillator and 74HC04 inverters to get the two anti-phase drive signals, with fairly poor results. I replaced this with an Arduino Nano/si5351 sand LT1016 comparator which produces near perfect drive waveforms.  The script also keys the drive and does T/R sequencing.   

The LM3886 modulator drives three 2N3055s as emitter followers, with the RF PA acting as the load.  No issues. Preamp is a $3 ebay SSM2167 module followed by an LM358 op amp for extra gain. 

The 100W carrier RF PA is TCA4422 drivers and four IRF630s on a 40V rail, at around 3 amps.  I have not been able to accurately measure efficiency, not having a digital 0-10A ammeter, but the analogue meter makes it appear to be around 90%.

the power supply is a 300W toroidal transformer with 40+40V, 15V and 15V secondaries.  Linear LM317HV regulator with four emitter follower 2N3055s on heatsinks.  I know, a switching PSU would have saved a lot of work, but I did it the old fashioned way.  It supplies around 40V (variable) at up to 10A. 

I have a more comprehensive blog post here, complete with Youtube video:
https://vk3hn.wordpress.com/2019/04/25/160m-100w-class-e-am-transmitter-fat5/ 

I am now a big fan of Class E.  Eventually, I'd like to build a bigger transmitter, for which this forum is a hugely valuable resource.  For the moment, I'm enjoying rag chewing across Melbourne on 160m. 

73, Paul VK3HN.   

   

[VE3ELQ]

Paul,
Welcome to the forum.  Great to see another experimenter/builder with an AM TX project up and running. Nothing like your first QSO on a HB rig.  There is a wealth of more modern high performance components now available to experiment with as you gain experience.  Happy building.

73s  Nigel

[VK3HN]

Thanks Nigel VE3ELQ.  I am discovering some of the 'more modern' design options and components, starting with classeradio.com, and the rather impressive set of circuits and design notes in this forum.   I think the best way forward is 'up', meaning the next transmitter will be bigger (higher powered), so that it can loaf along at the allowed VK power levels.  The drivers, FETs and cores are available via international vendors, some with local offices, so parts are not a huge problem.  It's mainly just time.  No lack of enthusiasm!

73 Paul VK3HN. 

[M0VRF]

Same way I started.

I wasn't happy with the quality of the FAT5 PCB's so made my own and started to redesign the circuitry.

I'd forget linear mod over 50W as it's rather wasteful on power being class A, you'll be better off with PWM.

The PUWMA is a decent PWM design, just use different FETs (Cree SiC) and tidy up the PSU.

RF O/P wise stick to the latest GaN devices or SiC and use a 'modern' Fet driver, NCP81074A.

Plenty of design work on here, mainly by Nigel and lots of help and advice.

Good Luck!

JB.

[VK3HN]

Thanks JB (M0VRF),

The FAT5 project is an excellent introduction to Class E AM, and Eric GW8LLJ actively responded to all my questions with some very timely and sound advice.  The documents are well written, accurate and make construction look doable. 

Since posting on my blog, I have created a kicad circuit A3 for the transmitter, which is fairly much complete. It's now embedded in the post, or at:

https://vk3hn.files.wordpress.com/2019/04/fat5_160_am_tx-1.pdf

To a prospective builder, without theoretical knowledge (that's moi), a design needs to be fully resolved.  That means every component and particularly the high voltage capacitors and inductors fully specified, and also available. Even buying something like a TransZorb (TVS diode) can be challenging if you have never dealt with these components before.  This is one reason why the FAT5 project appealed.  I could source and work with every single component. 

In my case, I collected the components over several months, including buying from three online retailers, (Minikits, RS, el14), and two physical stores (Jaycar, Altronics).  Researching, sourcing and collecting the components is all part of the fun.   

Having built a first transmitter and debugged it, I now feel capable of delving in further.  But I really need to follow an established proven design, I am a maker, not an RF engineer.   

I'll keep nosing around here for ideas.  Something around the 200 to 250 watt carrier level.  With PWM.  and preferably something that will run on a server power supply (50V). 

73 Paul VK3HN.   

[M0VRF]

ALL Good, however you'll need more volts as 50V is fine for say, 100W as the currents are 'reasonable'

At 250W (1KW peak) you'll be needing 25A+ (accounting for inefficiencies etc).

You'll be much better off with 150V (say) and <10A.

ALL the 'modern' FETs, both GaN and SiC are VERY high voltage and will have no problem at (class E peaks at 3.5 X Vcc) 600V or so.

I'd go for the CREE SiC devices and as your on 160m, check out the C3M0065090D.

Again lots of advice on here from folk who've actually built these monsters!

I'm happy with 100W and have used 50V (server PSU).

Regards

JohnB.