Re: The Transconductance of FETs

[Fred Townsend]

Declan Moriarty wrote: Recently, Somebody Somewhere wrote these words Declan Moriarty wrote: /rambling preamble My reading on transconductance goes back to Ia/Vg Curves, chasing linearity on valves (as advanced a science as surfing tidal waves) and such 'prehistoric' notions. My current requirement is in switching circuitry, and particularly, replacing fets in switching circuitry. I have designed switching circuits using FETs for many years and I couldn't tell you what the transconductance value was for any FET I used. In short I couldn't explain the relevance of gfs to switching applications because I don't believe there is any relevance. I think you are barking up the wrong tree. I wouldn't pay too much attention to gate threshold either. If I am using it for a switch I want to know at what vgs the device is guaranteed to be saturated. Then, will the FET handle the current and voltage. Finally what is Rd-on? My circuit current squared times Rd-on will give me my actual dissipation. Pretty simple. Caveat: Most power FETs have a Vgs for saturation (not threshold) of 4 to 5 volts. This is too high for interfacing with logic. There are FETs known as logic-level FETs with built-in charge pumps that saturate around 2 volts. If this is what you need, check out Harris Semiconductor. This question arose because in my particular niche, speed is everything. Speed of response to your customer or speed of the switching circuit? I fail to see the relevance to either in this discussion. I found myself trying to find a logic level fet in the stock of a supplier who didn't discriminate between logic level fets, and less sensitive ones, and wondered if could the gfs be extrapolated into a reckoning of the saturation voltage. That's worth knowing if I could do it. I sense you have mis-characterized the problem. Are you dealing with a former tube (valve) distributor that now sells a universal semiconductor line? Without naming names, these lines purport to substitute for many of the OEM lines. They give only vague, mostly useless specifications. If so would suggest you switch to an OEM distributor but you probably knew that. These guys are ones you 'know fairly well on a personal basis'. How is  the coffee or tea at their place? Sounds like a social issue, not a technical issue. Fred Townsend Secondly, I want to educate the technical support guys there who I know fairly well on a personal basis. When you are talking down to 'engineers' you have to be fairly sure of your facts. Alois gave a separate exposition on the subject, and it is a purely linear characteristic.