vk3alk
Member
Offline
Posts: 265
|
|
« on: March 20, 2024, 01:17:39 AM » |
|
Have been playing around with this project using Nigels circuit as a guide and it really has been quite successful...... Have uploaded 2 photos..... The efficiency is mid 90s .... a little unbelievable to say the least... Have made the load around 8R and using my test Power Supply that can only supply about 5 amps the output is 173 watts going by my CRO at the dummy load which is getting very warm.....
Are going to order some more FETs and makeup a smaller module using CMCD..... My aim is to hopefully have 2 small modules with good efficiency into a combiner producing a total of around 150 watts....
Wayne
|
|
|
Logged
|
|
|
|
KD6VXI
Contributing
Member
Offline
Posts: 2692
Making AM GREAT Again!
|
|
« Reply #1 on: March 20, 2024, 06:17:37 AM » |
|
Awesome news!
Good to see the higher frequencies being explored.
I miss building. Hoping to buy a house in the next couple months and get started again. Idle hands and all.....
--Shane WP2ASS / ex KD6VXI
|
|
|
Logged
|
|
|
|
vk3alk
Member
Offline
Posts: 265
|
|
« Reply #2 on: March 20, 2024, 06:12:27 PM » |
|
Hi Shane... Thanks for your remarks.... Not that I'm spying on you but you live in an interesting place Right next to Cuba and the Bermuda Triangle Do you go swimming down at the beach ?
|
|
|
Logged
|
|
|
|
KD6VXI
Contributing
Member
Offline
Posts: 2692
Making AM GREAT Again!
|
|
« Reply #3 on: March 20, 2024, 06:49:34 PM » |
|
Swimming, diving (snorkel and scuba).
Love the water, it was a major factor in moving here.
Hurricanes are a downside but having moved from California where you never know if you where going to have an earthquake and fall intk the ocean or a wildfire burn your house down, this is manageable!
--Shane WP2ASS / ex KD6VXI
|
|
|
Logged
|
|
|
|
vk3alk
Member
Offline
Posts: 265
|
|
« Reply #4 on: March 20, 2024, 09:27:11 PM » |
|
Hmmm yeah.... California does have shocking fires ..... We get quite a bit of news of those fires and makes me wonder if there is any bush left.... Fires over here are definitely ferocious too but those around Californa seems to top the list which is not something you want.... I dont really want to think about that super valcano as I suppose it is what it is ..... so to speak.... Have fun swimming Wayne
|
|
|
Logged
|
|
|
|
M0VRF
Member
Offline
Posts: 177
|
|
« Reply #5 on: May 31, 2024, 08:38:07 AM » |
|
Hi Wayne, what drivers are you using I wonder? Still the NCP's?
I'm trying to find faster devices as need to get to over 15MHz now.
I doubt you're getting the efficiencies you quote and will be more high 80's.
I've used SMT GaN devices with 2.2nC Qg (GS-065-011-2-L) and can get around 89%. Measured using a fully calibrated Agilent system with a Python efficiency meter script.
Sure I'd like to get to mid 90's but I don't think it possible especially above 5MHz.
I hope to be proved wrong!
Regards
JohnB.
|
|
|
Logged
|
|
|
|
vk3alk
Member
Offline
Posts: 265
|
|
« Reply #6 on: May 31, 2024, 11:19:42 PM » |
|
Hi John
Sorry to disappoint you but I'm rather boring and decided Sinewave drive was a better option.... I only build for myself so picked the easier path that offered success for me anyway..... Have attached a photo of my Sinewave Driver and if you look closer .... yes an analog VFO...!!!! All that work for 5 watts of drive... Was very surprised with the efficiency so thought this TX deserved to be on air.... The FETs are the key here in my opinion. They work well on 20M and going by the specs are fully on at 8 volts which was another reason why I used Sinewave Drive.... 50% DC would be obtainable. The PWM is already built... Whats left is the front panel assembly and a Bridge for SWR and Power display etc: It will have sequencing too as well as StevesQIX " no drive no transmit " circuit....
My Chrome display has gone all weird again and cannot add Smiley Faces .....
Wayne
|
|
|
Logged
|
|
|
|
M0VRF
Member
Offline
Posts: 177
|
|
« Reply #7 on: June 02, 2024, 01:40:15 PM » |
|
Hi Wayne, maybe it's the 'Fashion' on here not to use a PCB!
So cheap these days so I wouldn't consider all that effort to use any other method.
That said your construction is a work of art!
So what are you using for drivers and O/P devices, I wonder?
Greetings from the UK.
|
|
|
Logged
|
|
|
|
vk3alk
Member
Offline
Posts: 265
|
|
« Reply #8 on: June 03, 2024, 01:36:56 AM » |
|
Ohhh some people use PCBs etc: John JSW does.... This build is all experiential really.... The Sinewave VFO thing was made up on the go from stage to stage and for me cutting little bits of PCB and glueing as I went was much easier... If I was to make another the outlay would be different and better... There is a MMIC there somewhere too thats modern isnt it .... Just after the mixer and before the BPF ......
The FETs are TPH3206PS that Nigel discovered.....
The Sinewave drive goes into those ferrite cores for the phasing and into the gates...
When testing the drive level was only 2.5 Watts of carrier so have plenty up my sleeve....
Thats about it I think..... Been a fun build really....
Actually I bought a Sysjoint SV4401 VNA a little while ago and was interesting setting up all the impedances and things.. a great unit....
Wayne
|
|
|
Logged
|
|
|
|
M0VRF
Member
Offline
Posts: 177
|
|
« Reply #9 on: June 04, 2024, 03:49:35 AM » |
|
Hi Wayne, all good stuff altho' you didn't mention the drivers?
I guess still the NCP81074's?
2W5; then that's why it's so efficient!
4 GaN's will happily give 100W carrier @ 24-30V
I've made a few of these and are in commercial service 24/7.
I use one driver per fet driven direct thru' small R.
However I'm wanting to get to 250W carrier using Nexperia GaN's and also (maybe a different design) increasing the frequency to 20MHz so may (as been suggested elsewhere) have to use 'sinewave' drive but am going to try some different drivers.
Lots of experimenting to do and am determined to get the designs CE certified.
Good purchase on the VNA, an invaluable instrument!
Great value at the mo' are MXA's (N9020A) as CH are selling them off cheap and relatively simple to enable ALL the options for a fully configured Spec Ani.
Keep up the good work, I'll post up any findings re driver improvements etc.
Best Regards from the UK.
JohnB.
|
|
|
Logged
|
|
|
|
vk3alk
Member
Offline
Posts: 265
|
|
« Reply #10 on: June 04, 2024, 11:21:52 PM » |
|
Sounds like your full on trying to get higher in frequency....
Not using drivers like the NCP81074 just sinewave from that module just built into the ferrites for phasing and into the gates of the FETs... Thats all....!!!
Will post more pictures as I proceed along....
Wayne
|
|
|
Logged
|
|
|
|
vk3alk
Member
Offline
Posts: 265
|
|
« Reply #11 on: August 13, 2024, 08:48:30 AM » |
|
Just a bit of an update...
The TX is nearly finished although had a few little problems.... The Analog RF driver was effected by the Class E stage and caused feedback issues so unfortunately looks like it will have to be in its own box outside the main case.... Are currently driving the Transmitter with my IC706 thing... The electronic boards at the front left hand side are a protection circuit as the IC706 puts out full power for a few mS and would destroy the FETs... All that will be replaced by the overload shutdown circuitry....
The output transformer had to be rewound a few times and also are using two T200-2 cores... The windings are 5 turns each side with a 1 to 2 turns ratio.... The power supply under load is 88 volts so the carrier voltage is around 38 volts... Are settling on 150 watts carrier output but under testing pushed to 250 watts quite easily and the FETs remained cool....
The overall efficiency is 85% including the PWM so the TX is probably around 89-90%... Are happy with that....
Wayne
|
|
|
Logged
|
|
|
|
Opcom
Patrick J. / KD5OEI
Contributing
Member
Offline
Posts: 8301
|
|
« Reply #12 on: August 15, 2024, 12:47:39 AM » |
|
Like that VSWR/power meter in there. Really like that these class E transmitters have tuning and loading controls. I know they have to, but its much more dignified than broadband amplifier output.
|
|
|
Logged
|
Radio Candelstein
|
|
|
vk3alk
Member
Offline
Posts: 265
|
|
« Reply #13 on: August 15, 2024, 07:11:10 PM » |
|
Yes those dual meters are great although are a little expensive...
I get sick of flicking buttons going to forward and then reverse etc:
|
|
|
Logged
|
|
|
|
M0VRF
Member
Offline
Posts: 177
|
|
« Reply #14 on: October 17, 2024, 11:06:38 AM » |
|
Hope everyone out there is fine.
OK, I'm still experimenting as new (GaN) devices are coming out all the time.
Nexperia has some 4 Qg TO247 devices (GAN 111-XXX)I couldn't wait to try and just got delivery yesterday.
I'm using the UCC 27614 drivers which seem to be an improvement on the NCP81074's
14MHz is no problem and will probably go higher but drivers would need heatsinking or forced air!
Regards to All.
J.
|
|
|
Logged
|
|
|
|
vk3alk
Member
Offline
Posts: 265
|
|
« Reply #15 on: October 17, 2024, 06:41:28 PM » |
|
Hi John.. Can feel your excitement tremors down here at the bottom of Australia... Are you using CMCD or have you just applied a signal to the gates.... The TPH3206PS FETs look like they are going obsolete ... sort of typical really ... How much are these Nexperia FETs...? Anyway great news... Wayne
|
|
|
Logged
|
|
|
|
M0VRF
Member
Offline
Posts: 177
|
|
« Reply #16 on: October 20, 2024, 12:25:31 PM » |
|
Hi Wayne, thanks for the reply.
Funny as I've just been speaking to a Ham in Perth via the 10m repeater in Moscow, crazy!!
The newer Gan systems devices have a higher Qg now but useable.
The Nexperia devices are around £10 ea.
I think Mouser have some but got mine direct from manufacturer..
My designs are push pull class e with a pair a side giving 125W carrier at 48V.
Regards.
JB.
|
|
|
Logged
|
|
|
|
|