Where have the microwave transistors gone?

I have need of designing a discrete microwave transistor amplifier. A few years ago there were wide selections with detailed data sheets. Now the links to transistor companies in my browser give me web sites of famous transistor companies with just a few transistors that are scantily documented. For instance they do not always show S parameters or IP3 values.

Are these famous companies being put out of business by new companies? Are they trying to save money with limited web sites?

With the exception of power transistors I'd say everything has been replaced with MMICs and RFICs. Maybe NEC / CEL still makes some discretes and the stuff from Agilent is probably leftover stock (ATxxxx)

As I know , Philips,NEC-CEL,Ericsson,Motorola and a little bit ST are going on to produce MW transistors. But as mentioned before, transistors are replaced by RFIC and MMIC's.

Unfortunately those are "Old Skool Days..."

To my knowledge, only Agil.... and Philips ( also NEC ) are continuing to produce some RF transistor.
Probably the new RF engineers don't like the Smith chart and they prefere to find everything 50 Ohm matched without any tuning component.!

Mandi

This is all very bad. I liked to custom set the gains of each stage (with feedback) to optimize the system performance plus get the added benefit of increased IP values. I notice that these RFICs have miserable noise figures, frequently above 4 dB.

I have used the Agilent MGA-6xxxx recently. You can set the bias with a resistor and it still requires mathcing.

I also look for dual-gate FET. it seems that it is no longer exist since they move to IC for mixer. But I need it for my experiment in active antenna. too bad.

Dual gate types were very good for their low S12 values. You could do more with them without worrying about stability problems.

These links may help you, at least to remember some forgotten names. http://www.interfacebus.com/FETs.html

Hi,

Yes, I was very desperate too when searching for a device for my designs was without any result. Then I approach some manufacturers in person on some exibitions.

It was then clear to me that there is something else going on with the discrete transistors, they didn't just disappear. For some reason that I can not understand, the companies are hiding the existance of packaged transistors from wide public! I have beed working since then close with come manufacturers, they gave us samples for their devices two years ago, and they are constantly improving them (and selling to us negotiating price of milions in quantities!), but you cannot find any informations about the product anywhere. It is not the case for one company only, it is the case for the least 3 of them.

I am sorry I cannot disclose any name here, not they were willing to tell me the reason for this secret developments.

Packaged MMICs and RFICs are great if you are building exactly what they are inteded for, if you want to use them in a slightly changed environment or conditions you might get very disapointed. Not to mention the work with bare dies that can be good for experiments or space research, but when you start to add costs of technology needed to use them reliably, you can never get in line with the cost of packaged discretes and pick-and-place assembly.

So, I think (Old School!?) that discretes are not easily replaceable with integrated, off the shelf, one size fits all MMICs or RFICs.

flyhigh

The problem is pretty obvious. Why did Agilent just sell off its semiconductor group? They are not making much money there. Build a high volume product with a transistor in it and what do you do, you select the transistor that is 0.005 cents cheaper than the next one available. So, is it any wonder that the semi companies do not market to low volume engineers anymore? They are selling these applications specific devices that have large volume and some brand loyalty still.


That said, I still get good responses from NEC. The future of these devices will probably reside in much smaller companies, such as Hittite, who do not mind servicing niche markets.

At one time the oceans were so full of Cod fish that you could "walk from America to England by stepping on their heads", then overfishing brought the fishermen woes. We engineers apparently do not learn from history! If we want to have a source of supply for discrete semiconductors, we had better start designing in these smaller niche providers, and be sure to let them have the larger production wins when they happen too, or there won't be any fish left in the ocean in another few years!!!!!