HF linear amplifier advice / Electromagnetic Radiation / Forums

Forums

4hv.org :: Forums :: Electromagnetic Radiation

« Previous topic | Next topic »

HF linear amplifier advice

Move Thread

LAN_403

GeordieBoy

Sun Oct 12 2014, 05:54PM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

Hi RF Gurus

For some work i'm doing with telemetry from rotating machines, i've been experimenting with near-field wireless power / data transfer in the HF range. Specifically the 13.56MHz ISM band commonly used for RFID tagging, mobile NFC, etc.

I'm in need of a reliable and robust source of up to 15W of RF power over the range 5MHz to 30MHz. I'd ideally like work to shell out for one of those nice Amplifier Research broadband Class-A amplifiers. They are designed to withstand load mismatch and lots of reflected power for EMC RI testing so they would be ideal for what I'm doing. Except for one thing. The COST !

How easy is it to build a fairly decent Class-A RF power amp to give me up to 15W over the 5 to 30 MHz range? (I'm more interested in flexibility and reliability than low harmonic output as the application is not intended to radiate a significant far-field signal.) I've seen several linear amp circuits online intended for the HF band and they don't seem that complicated but i've never tried to build a linear RF power amp before so would be interested to hear other's experiences or advice. I'd probably start with something like a pair of IRF510's in push-pull and go from there...

I have extensive experience in the design, manufacture and testing of high-power switching RF amplifiers like Class-D ,E, etc, although I really want a broadband amplifier that is tolerant of high SWR. That rules out any of the switching amplifiers with their high-Q matching networks, at least for the experimentation stage of the project development.

Any thoughts/advice would be appreciated,

Best regards,

-Richie Burnett,

Dr. Slack

Sun Oct 12 2014, 06:45PM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

How much was the AR amplifier that you didn't want to afford? You could try going down market a bit with Mini Circuits. Their LZY-22+ is 100k-200MHz, 30 watts, safe into short/open at full power, $1500 (but probably not much less than that in GBP).

Sulaiman

Sun Oct 12 2014, 11:19PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

I'm no guru but I have a little experience with HF power amplifiers;
your quickest cheap route
would be to take the PA section from a second hand amateur radio transceiver (eBay etc.)
The PAs are wideband, with input and output filters added to match the selected band.
Many have output power limiters.
Some transcivers can be easily modified for continuous coverage.
If you want to go this route I'll ask Aard-Wolf
.. he knows a lot as he used to repair ham equipment, and works at the same place as me.

If you take for example a simple 120W PA
and add a resistive 6dB attenuator for 30W to load
you will not damage the PA with any load mismatch.

DIY is fairly straightforward if you don't use feedback, a little more tricky with feedback but do-able
power supply bypassing is more challenging than for an audio PA but not too difficult up to 30 MHz

If going the DIY route you should buy proper RF transistors,
I've used the famous IRF510 at 14 MHz but it is useless at 30 MHz
Don't use the 10's GHz transistors for 10's MHz ... sure to make an oscillator
Class-A push-pull is probably the easiest PA to successfully design and build
something like the old MRF150 has loads of data, application notes, complete PA schematics etc.
e.g. eBay

I'm currently playing with LDMOS transistors - easy to use, stable, need a good bias circuit.
(there are two photo's of my LDMOS PA in Ash's electronic ignition thread

)
(un-necesary overkill in the ferrites and bypassing but good for hobby testing/torturing etc.)

EDIT: Aard-Wolf recommends a Yaesu FT897 or 857 which have the same circuitry, different casing.
They can be set to continuous coverage by changing a few jumpers near the processor
and have enough power and protection for your needs.
They are multi-mode so in FM mode or CW mode have continuous carrier/power.
PLUS you'll have a nice radio to play with !

For 13.5 MHz only

Wideband Class-A but 15W

1.8 - 30 MHz all mode 300W PA

Steve Conner

Mon Oct 13 2014, 12:59PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

I like the idea of buying a radio. I have an old FT-757GXII modified for general coverage transmit, that I've used in all sorts of cruel and unusual RF experiments. It's good for about 100W from 1.8 to 30MHz. The FT857 or 897 are more modern and better suited to the power level you want. I have a ham license, but I find lighting up plasma tubes a lot more interesting than talking to people on 80m.

The main drawback with using a ham radio is that the modulation bandwidth is limited to about 3kHz, so you won't be able to achieve a very high data rate. If this is a problem, I'd buy an amp module similar to the ones suggested by Sulaiman. When you blow the transistor you can reuse all the other parts.

GeordieBoy

Mon Oct 13 2014, 02:08PM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

Hi guys,

Thanks for your help with this. I think the AR linear amp was going to be a little over Â£5k. I was also looking at this one from E&I (or ENI) but waiting for a quote back:

Thanks for the suggestions about using a ham radio in transmit. I had wondered about this but thought that it would either shutdown or blow-up if the load wasn't a nice well-behaved 50 ohms. How low can you typically turn down the output power? I was budgeting for about 10 or 15W maximum power needed at the source, because I don't really know the efficiency of the inductive power transfer system as a whole, and can't test it without a source

I only need about 1.5W or power at the actual destination end of the link, so would be a little bit worried about an amplifier capable of 100W doing some damage if it's output power was to momentarily overshoot. (I know I can use attenuators (and that will harden the amplifier to SWR) but that's additional cost and complication.)

The links to RF amplifier modules are interesting too. Obviously i've got power supplies and heatsinks and fans around so the idea of buying a power amp module (where someone has already done all the technical RF magic) and just supplying the power and thermal managment sounds tempting to me.

Although the intended application is to operate in the ISM band at 13.56MHz I need the flexibility to tune the RF source significantly either side of this frequency during testing of antennas etc. For the final application I will design a little Class-E power amp optimised to provide whatever power is needed efficiently at 13.56 MHz into whatever the load impedance turns out to be.

Steve, The modulation bandwidth of the RF source isn't too important to me at this stage. The part that goes on the rotating machine (and is wirelessly powered from the RF source) actually transmits the data back by load absorption modulation. So I don't think the modulation bandwidth of the RF source is as important as how well it behaves when the load impedance is varied. I will need to send some messages in the other direction, but they will be simple low bandwidth ones like "start measuring", "stop measuring", "change gain" etc... Any Tx that is fast enough for morse should be okay for what I need to send in that direction

Cheers,

-Richie,

Steve Conner

Mon Oct 13 2014, 03:16PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

The ham radio transmitters typically turn down their output power when they detect excessive SWR. Should be quite resistant to blowing up. A high power model might overshoot quite a bit when dialled down to low power though.

The amplifier modules typically don't have variable power. You are expected to manage it yourself by varying the drive.

I would use the attenuator method to "harden" it. At 13.56MHz a couple of non-inductive power resistors should do it. The attenuation doesn't need to be huge to help with the SWR. Even with 6dB, an open or short circuit load will give a SWR of 2:1 instead of infinity.

GeordieBoy

Mon Oct 27 2014, 10:03PM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

Hi Guys,

Thanks for all your tips and advice on this. I ended up making a lash-up Class-E power amplifier capable of producing 20W at 13.56MHz into 50 ohms in the end. I padded the output with an attenuator to make it more tolerant of mismatch like Steve C said.

I just bit the bullet and designed it to drive into 50 ohms as I had a decent 50 ohm dummy load, and it turned out that the actual load was very close to that, so the matching network didn't need much tweaking! It also turned out that the application only needed a couple of watts of power, so even with the attenuator the Class-E PA had enough power to get things up and running.

Now that i've got enough power to make things tick, I can take some measurements and optimise the PA design and the inductive power/data transmission parts. It was like a "chicken and egg" scenario before, i couldn't design the RF power amplifier without knowing the impedance of the load, but I couldn't determine the impedance of the load without a decent amount of RF power first!

As a bonus a 20W 13.56MHz RF source would also make a nice driver for a 1kW HF tesla coil PA

Now to build a directional coupler to allow me to measure the refelected power. I was thinking of one using a ferrite VT and CT at this frequency given that the power level is only a few watts. I've found a few designs online, but am not sure whether to use high-permeability (but more lossy) MnZn ferrite, or lower-permeability (but much lower loss) NiZn ferrites? I'm only interested in killing off most of the forward power so that I get a better measurement of the variation in reflected power due to the load modulation, so I will "calibrate" the directional coupler to give me minimum reflected power into my 50R dummy load.

-Richie,

johnf

Tue Oct 28 2014, 07:46AM

Registered Member #230 Joined: Tue Feb 21 2006, 08:01PM
Location: Gracefield lower Hutt
Posts: 284

It is very easy
use the ham rig and buy a 3dB attenuator chip this will give 6dB of return loss to the amplifier so that is the worst mismatch the amp will see
from memory that will be an VSWR of 3: 1 which all ham rigs will handle

look at my web site IEL-RF.com there is an explanation there

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob