1 GHz High Voltage Amplifier/Sine Generator

[vahid_ff]

Hi folks,

My professor asked me to determine if it's possible to design a high voltage (output: 30~40 V p-p) amplifier which is to operate at so high frequency range of 1 GHz or not. We want to use this dreamy device in some biomedical application, so no high output current is needed. I think available wide-band Op-amps in the market are not able to deliver such a high voltage swinging capability due to their limitation in slew rate and supply voltage. There are other problems too, for example I would need to built my own oscillator inside the circuit since no function generator is available in this frequency.

I would like to know your ideas/suggestions about this issue and challenges in front of me in this matter. Please consider my limitations as a student like lack of professional lab equipments etc.

Thanks...

 

[FvM]

I would need to built my own oscillator inside the circuit since no function generator is available in this frequency.

You mean, not available in your lab. But it's a standard RF generator.

40 Vpp correspondends to 4W into 50 ohm, well in the range of medium power wideband amplifiers.

A point that should be clarified first is the expectable load impedance range. "no high output current is needed" might apply to the test object, but a short (3") wire acts as an antenna that radiates the applied signal into free space and exposes a low input impedance. So it's inevitable to analyze the test setup in terms of electromagnetic theory and design a suitable way to control the handled RF energy. As another point, you have to observe FCC regulations and possibly need to put the test setup into a screening metallic enclosure, or restrict your generated signals to free ISM frequencies.

 

[chuckey]

As FvM said, any unshielded wire will radiate this frequency, you must use a 50 ohm load at end of your 50 ohm cable which is connected to your experimemtal point. I would suggest that you read up on RF techniques before you venture any further. The AARL or RSGB handbooks (radio amateurs organisations) are a good beginners start in this topic. At this power level you will be involved with specialized techniques, so it is unlikely that op-amps will work, though wide band RF amps are available covering an octave, say 400 MHZ -1GHZ (UHF TV) at power levels up to 200W.

 

[vahid_ff]

Thanks guys for your quick replies. So as far as I understand from your replies, implementation of this special RF amplifier is not "impossible", but it required good command of RF design. For giving more details on what I want, I would say that the load is normally capacitive (a few pf) and the amplitude could be less than what I mentioned in the previous thread, say 20 V p-p. Although I do confess that I need a lot of studying and practice in RF electronics, I would like to see some sample circuits which resembles what I'm looking for even with lower output amplitude. Furthermore, Do I have to adjust the frequency within the ISM band (902-928 MHz) or this regulation is mandatory only for commercial devices (not experimental equipments at laboratories)?

Thanks...

 

[FvM]

Limit values for radiated RF have to be kept by everyone. Unfortunately 900 MHz band limits are rather low, if you are radiating considerable part of the applied energy, you won't keep it.

"A few pF" are an impedance in the 50 ohm range at 1 GHz, but it can be compensated with an inductor if you are operating in a small frequency range. The voltage can be possibly stepped up in a resonant circuit, rqeuiring less input power. But you need to know the complex impedance of your test object, e.g. measured with a network analyzer.

It would be probably best to contact a RF engineering department of your or a nearby university and ask for some initial help with your project.

 

[jiripolivka]

At 1 GHz I do not think any suitable RF amplifier is available to generate 30-40 Vp-p except of vacuum-tube amplifiers. MOst available power signal generators are designed to have 50 OHm output impedance and not to generate such high voltage.

Refer to the ARRL Radio Amateur Handbook where high-power RF amplifiers are described; the nearest amateur band is 1296 MHz . Special tubes and special design is needed and such power amplifiers use high DC voltages, so they can be dangerous.
I have made similar amplifiers and know about the problems. Special skills are needed, not taught for >30 years.
The best advice would be to try to solve another problem. If you insist, contact your local radio-amateur club, maybe there is some expert or old-timer who can help you hands-on.

- - - Updated - - -

BTW, what you project is? The high RF power at 1 GHz can rather be used to incinerate a tissue..

- - - Updated - - -

To FvM: 40 V p-p across 50 Ohms equals 32 W of power.

 

[FvM]

To FvM: 40 V p-p across 50 Ohms equals 32 W of power.

My pocket calculator persists in 40 Vpp = 14.1 Vrms respectively 4W into 50 ohm.

But even 100 or 200W MOSFET amplifiers at 1 or 2.4 GHz are pretty standard these days.

Apart from this point, insisting in feasibility of solid state amplifiers for this power/voltage range, I share the doubts about meaningfulness of the application.

 

[vahid_ff]

What you project is? The high RF power at 1 GHz can rather be used to incinerate a tissue...

I share the doubts about meaningfulness of the application.

Thank you so much Fvm and jiripolivka.

We want to use this amplifier in some Electrophoresis applications and electro-rotation. In such a high frequency, we would expect some promising phenomenon happen for cells/bio-particles.

Could you please give me a starting point for this matter? (like a sample circuit, even in lower freq/amplitude, or mentioning somew of the high power GHz MOSFET you said).

Do you think that unwanted radiation would be dangerous for me or anyone who is working with this type of amplifier? What should we do to suppress these hazards?

Thanks.

 

[FvM]

A few watts RF are still in a range that can be handled without extensive safety precautions. Let's say a GSM phone or professional intercom range.

You didn't yet answer the question about DUT impedance, I guess you didn't yet think about it (or probably don't know how to analyze it). If you manage to keep the impedances high and don't need large signal bandwidths, you can possibly come along with < 1 W power and effective impedance transformation.

Following jiripolivka's comment, it should be clear that the integral RF energy actually dissipated in the test object must be very small, otherwise it will be simply cooked. I must confess that I'm not sure if the said 20 or even 40 Vpp numbers reflect this physical limititation correctly.

Vendors of RF MOSFET are e.g. Freescale, NXP, ST, Polyfet. Polyfet has a number of nice wide band power amplifier application notes.

 

[dick_freebird]

I don't see mention of the "load" impedance (and likely
this will vary a lot, being as it involves tissues and goop).

The voltage could be had by a transformer from many
RF PAs, the frequency puts you onto cell phone low band
or close enough so you have many inexpensive part choices.

But I think the load attributes want quantified.

The application doesn't sound like it needs a lot of
spectral purity so maybe clipping and harmonics (often
limiting factors in PAs' mismatch tolerance) are not a
big concern. And putting a box around it might suffice
for a one-off piece of lab equipment (leaving the problems
of mass market and type acceptance, for later).

I think there are "care-abouts" that have not been
expressed - like, say, "magic" frequencies to get whatever
"electro-rotation" to go on, for whatever's being rotated.
The need to vary frequency, phase, power / amplitude
would probably constrain choices.