These parts are being operated in their linear region, so Rds(on) isn't relevant. As long as they can take the heat, what we need is agility, that is, minimal Ciss/Coss/Crss.
________________________________ From: [email protected] <[email protected]> on behalf of Chris Elmquist via groups.io <chrise@...> Sent: Wednesday, April 2, 2025 11:18 AM To: [email protected] <[email protected]> Subject: Re: [TekScopes] horz output FETs in 2245? On Wednesday (04/02/2025 at 05:47PM +0000), Dave Wise via groups.io wrote: If I had a 2245 with your problem I'd be looking for... Hmm. Your parts are running naked, right - no heat sink? That's maybe 1W.Yes-- they have no heatsinks and are spec'd at 3W max in the TO-39 can. The left plate pair gets so hot I can't touch it but I realize that's not a calibrated measurement. An eBay seller has TO-92 VP0116 $15 for 10, glue on a copper fin. But nobody is selling VN1316 AFAICT.There are also VP0120 and VN1320 which are similar devices with a higher Vdgs (200V). These would likely work just fine too if they could be found. Look at Mouser.com . They don't have any 160V P-channel RF MOSFETs so how about the regular ones? In 150V-300V select the lowest Ciss proxied as gate charge, that's Diodes Inc ZVP2545 in the SOT26-6 package. Put a copper fin on the drain pins and you should be okay.Right. OK. Thanks. Ciss is the critical parameter then for performance here? Was also wondering how critical the Rds (on resistance) would be in this type of circuit? Modern FETs measure that in tenths of ohms while these golden oldies measure it in 10s of ohms... eg, VN1316 is at 40 ohm and VP0116 is at 25 ohm. There's not a lot of clearance between the tops of these TO-39 and the scope's enclosure though so some challenges with fitting new/different stuff. Could envision a little PCB that carries the SMT parts but has "leads" coming out the bottom to mate up to the scope's PCB. Chris ________________________________-- Chris Elmquist |