DN2540/2535 fun and games

[Sheldon]

A little story: I just spent the afternoon wondering if I was finally losing it. I have an amp where I'm using a DN250 for a source follower to drive an output tube. Simple arrangement with a biasing resistor. As I'm preparing to make some other changes, I thought I'd upgrade the follower and use a current sink instead of the bias resistor. Since we might as well go for overkill, I made the current sink a cascode with a TO220 DN2540 on top and a TO92 3545 on the bottom. Then the fun started.

Assuming that the pin assignments are the same for the 2540 and 3545, I checked a data sheet I had printed out some time ago for the TO92 part. Cobbled the thing together, checked, double checked my circuits and applied power. WAY over the expected current. Dang. Checked again. Tried to think of anything I had done wrong but couldn't come up with any ideas. Well, I figured I should take a look at the 3545 data sheet at Supertex's website. WHAT? the drain and source were reversed from my 2540 sheet!!! Checked the gate resistance - cooked.

As luck would have it, I had a couple of TO92 2540 parts, so I installed those. Same outcome. OK, now I'm beginning to wonder. Computer was still on the Supertex site, so I looked up the 2540 sheet. THE PINS ARE REVERSED FROM MY VERSION!!! Foiled by a typo.

Good news is I'm not crazy (well only a little more than before). Bad news is have to wait for some new parts.

Sheldon

 

[Alexandre]

I was measuring a few DN2540 on protoboard and they fail *ridiculously* easy. Supertex should put protection diodes in it.
-Alex

 

[SY]

I've never lost one as a CCS in plate or cathode circuits. So I was tempted to try them as source followers to drive output tubes instead of the IRF820s I've used for the past 25 years. I spent an afternoon playing in Morgan Jones's lab where, after turning a few dozen of them into tiny lumps of coal, we came to the inevitable conclusion that maybe they aren't suited for that.

 

[wrenchone]

I save my money and use the smallest TO-220 mosfet with appropriate voltage rating I can find for the top mosfet of the source follower (try ST, Fairchild, Ixys for appropriate units). I use the Supertex depletion part for the bottom current sink. Always use a gate stopper resistor for both units, and gate protection with the top mosfet..

Having said that, I actually used the Supertex DN2540N5 as both driver and current sink fet for the source followers in my "Miz Piggy" screen drive SE amp. The thing is still kicking. Cascodiing the current sink for a source follower might be gilding the lily...

 

[Wavebourn]

I've never lost one as a CCS in plate or cathode circuits. So I was tempted to try them as source followers to drive output tubes instead of the IRF820s I've used for the past 25 years. I spent an afternoon playing in Morgan Jones's lab where, after turning a few dozen of them into tiny lumps of coal, we came to the inevitable conclusion that maybe they aren't suited for that.

Did you follow Morgan's advice to bake them in oven, like he did with his d3a tubes?

 

[SY]

I couldn't find the getter.

 

[dsavitsk]

I used these to make a little Gm headphone amp with some Edcor OPTs. Works pretty well for Low Z phones, actually.

 

[Wavebourn]

I couldn't find the getter.

That means, they may be counterfeit devices, they omitted getter to cut costs.

Proper gate stoppers and Zeners between gate and source would probably replace the getter, who knows...