WorldwideDX Amateur Radio Forums

When one tells another to 'get a clue', one must ensure they have a clue as well.

It was stated 20W RMS was the level to keep the single device at. A pair would then be 40 watts.

YOU stated you get 200 watts PEAK.

I'm going to make the assumption he meant carrier power at 20 watts (which generally would mean 20 watts of AVERAGE Pdiss on the device under AM 100 percent modulation conditions, give or take.. You COULD squeeze 30 out MAYBE).... BUT, that's 160 to 180 watts PEAK.

So, there's your clue for the evening.

--Toll_Free

http://www.rfparts.com/pdf_docs/ERF7...20Rev2%200.pdf

The datasheet quotes 75W pep at 12.5V so I'd expect around 100-125 at a solid 14V with everything optimised.

Shame these things have such low gain.

Phil
Unit 148

You will be lucky to see much more than the stated 75w output Phil. And as they heat up they back off the power in a hurry too.

I imagine with such a small die it's imperative to have perfect finished heatsink plus fan cooling to keep things under control.

You can push RF fet's as I'm sure you know.

We have seen a genuine 1Kw out of a pair of these at work:
http://www.st.com/stonline/products/...e/ds/15501.pdf

However for reliability and the general public we use four in our 1Kw FM PA.

Is this new ERF fet the absolute highest output 12V fet out there at the moment?

Phil
Unit 148

It's the class of service people operate them at.

As you increase efficiency, gain goes down.....

--Toll_Free

The real issue with the gain is that these are cheap MOSFET's. As such they have higher internal capacitance then you would find on a more expensive RF MOSFET. When you drive a cheap MOSFET at it's extreme frequency limitations, you have to overcome this gate capacitance that appears to shunt some of the drive power. This lowers the gain of the device in use. Quality RF power MOSFET's can have gains around 22 db at frequencies into UHF.

The IRF-520 series transistors have been around longer then a decade. They were designed for the "new" high frequency switching power supplies of the day. Some of the spin-offs like the ERF7530 actually claim to be inexpensively designed for HF RF amplification to 30 MHz. There is the possibility they are nothing more then higher power versions of transistors designed for switching supplies that have been marketed for RF because they happen to function up to 30 MHz. with lower then typical gain.

There is more connection between class of service and efficiency then there is with either of the two effecting the gain of the active device itself. As you approach 360 degrees of conduction, the efficiency will drop due to the power dissipated as DC bias. Interestingly, this does not have a major impact on the overall gain of an RF power stage unless we are talking about very low signal levels that would have to overcome biasing the transistor on. As drive levels go from milliwatts to watts, the difference in gain becomes negligible with respect to bias class. It is true that input or output circuits having a poor match will waste some of the gain the active device can provide due to losses in the mismatch.

You speak the truth.

But there is more: its also about the drain-gate capacitance, also known as the miller effect. It has been said this capacitance varies over the RF cycle because it depends on drain-source voltage.

This probably has something also to do with packaging. You really cant expect good results when the RF input pin is right next to the RF output pin, which is the case with TO-220 type of package. The gate needs to be on the opposite end of the device from the drain.

All this results in poor efficiency and poor IMD.
Would you trust a vendor without a street address?

I was getting my hopes up on these plastic fet's but you're painting a poor picture re their capabilities.

I'm very familiar with the even cheaper IRF's as we used to sell a 40w FM PA with a pair of 510's in a broadband design covering the FM broadcast band 88-108MHz.

I am no designer and gave up diagnosing faults at component level many years ago moving into marketing of RF and DSP based broadcast products (we exhibit at NAB each year).

I think I may give up on these fet's and go back to my 16 x 2879's or the Metron in my avatar

Btw, are the IMD figures any better with these plastics that the bi-polars?

Phil
Unit 148