There're no data in the data sheet that would give any clue as to the linearity of the device.
Tom
Tom
Good old Sanyo's transistors. Unfortunately you won't find something similar in the SMD package. You might try with TSC497. Does it have to be an SMD transistor in your project btw ?
Nexperia and Infineon both make or made SOT223 versions of their old TV deflection amps such as BF469,BF470 which were TO126 and very popular for VAS use. BF721-3 are made by both and are in stock at Digi-Key, for example. Nexperia has more in another BF4XX number series as well but you'll have to search that as my notes don't cover them. Cut tape unit price? - about US 50c ea.
Most VAS transistors will dissipate some heat so SOT223 or SOT89 forms and copper foil or flag style heatsinks are a wise choice if you're thinking of any medium to large size power amp.
Most VAS transistors will dissipate some heat so SOT223 or SOT89 forms and copper foil or flag style heatsinks are a wise choice if you're thinking of any medium to large size power amp.
Following on from Ian's comments, what are the rail voltages and the VAS current in mA for this application? This is a good start.......
HD
HD
Rails are +- 75v and VAS current is currently 2ma ( exicon lateral mosfet outputs ) This is a design of mine that I am changing everything except the outputs to SMD.
Using LND150 for ~2ma current source for input diff pair and for the vas stage
Toshiba HN4A06 is diff pair
Toshiba HN4C51 is the current mirror
I would also think carefully about the drive current and safety margin required for just a single pair of lateral mosfets and how many pairs will be needed to make use of those 75V rails. This short thread talks about the issues: How much VAS current to drive a pair of lateral FETs?
PCP1208 is 120MHz, 200V and 9pF Cob in package with a heat slug - its still a small package though, but rated for 1.3W onto ceramic substrate.
Would 2 mA of current be enough to drive the input capacitance of even *one* pair, let alone multiples? Boot strapping Cgs as a follower only goes so far. The local feedback divides the capacitance down, it does not make it zero. By the time I get rail voltages that high I usually want at least 10 mA, as that’s driving EF3’s. When you start approaching a watt, SMDs start getting your PCB *hot*.
2mA is not sufficient to clear charges to and from the gates.
With 75V rails you either need the double die Profusion (ECW20N20) or up to four pairs of the single dies (EC10N20). You will be dealing with a gate capacitance of many just over 1nF, typically N500pF and P700pF for the smaller, and 2mA won't cut it. Try increasing it to 8mA, then you will be dissipating 600mW from one VAS device, so you will be needing at least a large smd, a SOT89, or even a TO126 thru hole.
You can use lower VAS current in a Class A because the outputs never turn off. But with AB you have to deal with the transitions from off to full on. Too little VAS current will lead to slew distortion at high frequencies.
HD
With 75V rails you either need the double die Profusion (ECW20N20) or up to four pairs of the single dies (EC10N20). You will be dealing with a gate capacitance of many just over 1nF, typically N500pF and P700pF for the smaller, and 2mA won't cut it. Try increasing it to 8mA, then you will be dissipating 600mW from one VAS device, so you will be needing at least a large smd, a SOT89, or even a TO126 thru hole.
You can use lower VAS current in a Class A because the outputs never turn off. But with AB you have to deal with the transitions from off to full on. Too little VAS current will lead to slew distortion at high frequencies.
HD
Look at some of the Diodes Inc/Zetex devices. They have some good SOT-223 devices for example DZT5401/DZT5551 - similar to 2n5401/2n5551 but in a large package capable of dissipating enough heat to be suitable for VAS use
A good trick is to heatsink the copper plane nearby. I've found a Faston blade terminal soldered in actually works quite well for that 🙂
