Looking for 100 KHz - 10 MHz, 10W

What's your time worth? Most startups I know about would be better off moving fast than saving a small amount of money.

Cheers

Phil Hobbs

As Phil hinted, I'd rent it and be done with it. If you absolutely can't then look in the ham radio area. This is overkill but shows the idea, broadband amplifiers:

But you'll have to invest a lot of sweat equity because ham gear for shortwave goes from 1.8MHz or 3.5MHz to 30MHz, typically. So you'll have to redesign RF transformers and such. A 10:1 frequency range isn't exactly trivial, you need to have someone on board who knows RF very well.

You can also do it without transformers, discrete transistor-level design. But that's hard work.

I don't think you'll find anything with the required gain (about

Also try Amplifier Research: or Mini-Circuits: They're not cheap.

You also didn't say anything about linearity. If you're just pumping a sine wave into your fizzix package, almost anything will work. However, if you're AM modulating the function generator, you'll need some linearity specs. You may find yourself needing a 40-100 watt amplifier, just to get 10 watts of "clean" and distortion free RF.

As Phil suggested, startups have better things to do than save a pennies or reinvent the wheel. Find someone with a decent HF ham radio transmitter, or linear amplifier, that puts out about 100 watts, and you're done. There are some that might function down to 100 KHz if you don't care about harmonic content and linearity.

Maybe something like this home built unit: Well, maybe one of these:

Again, watch out for the low end frequency range. Going from the rated 1.5 Mhz down to your 0.1 Mhz is NOT easy.

Anything that you do yourself will take time.

If you do want to build something, take a look at the Apex opamps.

Note the "order free samples" tab.

Use two amplifiers, one for the low end and the other for the high end and a crossover filter.

The frequency drop at the extremes might be enough to have a gradual crossover response, but of course, you should not feed large power levels out of the design range, so some external crossover filtering may be needed.

Some power measurement device at the output can be used to control the amplifier input level to get a constant output, despite the filters.

"John Larkin"

So you are safe enough there.

But you are gonna HAVE to wangle a free sample to stay under your fanciful budget of $200.

.... Phil

One of the more difficult aspects of any project is getting the spec right.

I've spent a lot of time with customers who wanted "just like that, only a lot more of it!" Round numbers are a clue. After some discussion, you find you that they really needed 4 watts at 3MHz. +/- Stated another way, it's easy to add "cushion" to the spec that moves it from "simple" to nearly "impossible".

Does it have to be 50 ohms? Could you drive both ends with a bridge? Cuts your voltage requirement in half. Brute force is not always the shortest distance to success.

An ordinary solid state RF power transmitter is usually designed for

Trying to use one with some inductive load at unknown impedance will cause high SWR and a puff of smoke or the activation of any protection circuit, which drops the power.

You might have to use an "antenna tuner" (transmatch) so that the amplifier sees a 50 ohm resistive load. Depending of the impedance response of the load, you may have to switch in various inductors to cover that huge frequency range.

"mike"

One of the more difficult aspects of any project is getting the spec right.

Or... not think - as the case may be.

Might be an engineering axiom.

... Phil

snipped-for-privacy@downunder.com Inscribed thus:

My guess is that he wants to do plasma generation !

Like Joerg and Phil said: rent.

Cheap is important, but if it's a serious startup then your time is important, too. If you're a degreed engineer in the US, $1450 will pay for somewhere between one and three days of your time (remember to figure that your overhead costs as least as much as your salary when you do your computations), and there's no way that you'll put this together in three days.

There might be something available on eBay for less. But remember that if you spend three solid days getting something for free, you've just spent more than that $1450 instrument.

Wasn't that what Sinclair used to claim for an OC28 driven by a couple of 'red-spot' rejects ?

In which case he is underpowered by a couple of orders of magnitude and the frequency is on the low side for sensible inductive coupling.

A typical ICP plasma is driven by at least a few hundred watts and a peak output of 1kW or more to handle starting transients and solvents.

Also he is likely to become severely unpopular if it isn't very well screened. Spurious broadcasting into the MW and ADSL frequency band.

Depending how long it is needed for hiring the kit @ ~10% pcm may well be easier or looking for something suitable secondhand on eBay.

Maybe I am missing somthing, but 10MHz is really not that big of a deal.

A LT1210 55MHz single chip solution for 14USD:

Just add a power supply that can handle 10W (and a heat sink)

Cheers

Klaus

You might find a demo board on this page that does the trick:

"Baron"

I reckon he is having another crack at "Cold Fusion ".

Even the great Amar Bose fell for that one ......

... Phil

Ha, the 'typical app' on page 16 is almost the exact stated requirement (9W into 50 ohms over 100kHz-10MHz) using a COTS transformer and two chips in a bridged configuration.

Looks like Coiltronics have renamed their Versa Pac series at some point in the last 16 years.. hopefully the CTX-01-13033-X2 isn't discontinued. If it's off the shelf, that would be pretty fast to throw together.

Be fair. Fleischmann & Pons were respected academic electrochemists and

Sadly no-one else was able to make it work.

Martin Brown Inscribed thus:

I agree 100W is a bit low for plasma, and 13.5Mhz springs to mind.