Depletion Mode MOSFETs in UK?

[merlinb]

Does anyone from the UK know of a convenient supplier of depletion mode MOSFETs? I can't find any on Farnell etc.

[R.G.]

Amplifier triode replacement time, eh? 

Look for whomever carries the Supertex line of high voltage depletion mode stuff. The DN2530, 2535, 2540, and 3545 work well that way, although they sound more pentode-like than triode-like. The fun idea here is that they can be direct replacements in the high voltage milieu of a preamp setup, as they have the 300+ volt rating to take it and the -1 to -3V cutoff on the gate-source to bias right(ish) in normal 12AX7 circuits.

Anglia, in Wisbech, Cambs is listed on the Supertex site as a UK, distributor. Proper chatted up, they might part with samples.

[tca]

Does anyone from the UK know of a convenient supplier of depletion mode MOSFETs? I can't find any on Farnell etc.

ah, nice question! I've been looking for them to!

The DN2530, 2535, 2540, and 3545 work well that way, although they sound more pentode-like than triode-like.

I guess if you could bias them the "right" way you could get a triode responce.

[R.G.]

Spain/Portugal
Matrix Electronica, SL
Location: Madrid
Phone: +34 91 560 2737
Fax: +34 91 565 2865
supertex@matrix.es
www.matrix.es   

For reference:
http://www.supertex.com/contact_intl_distis.html

[tca]

Spain/Portugal
Matrix Electronica, SL
Location: Madrid
Phone: +34 91 560 2737
Fax: +34 91 565 2865
supertex@matrix.es
www.matrix.es  

For reference:
http://www.supertex.com/contact_intl_distis.html

Thanks, R.G.!!

I guess I'm going to build an amplifier. So much to do, so little time!

[R.G.]

I guess if you could bias them the "right" way you could get a triode responce.

That's the supposition, anyway. I tend to look at the equivalent of the plate characteristic curves for devices to compare to triodes. Pentodes have a certain plate current/plate voltage setup with varying grid voltages. This is remarkably like the one for bipolars, JFETs and MOSFETs. None of them have the characteristics of a triode, excepting for very small regions near the origin for JFETs - only a volt or two for most JFETs.

My tinkering with depletion mode MOSFETs a  few years ago when I noticed them in the Mouser catalog did indeed sound a lot like the pentode first tube of some AC30s in isolation. But that's not what most people thing of as the tube sound. I still have most of the first batch I ordered somewhere in an antistatic bag.

[tca]

That's the supposition, anyway. I tend to look at the equivalent of the plate characteristic curves for devices to compare to triodes. Pentodes have a certain plate current/plate voltage setup with varying grid voltages. This is remarkably like the one for bipolars, JFETs and MOSFETs. None of them have the characteristics of a triode, excepting for very small regions near the origin for JFETs - only a volt or two for most JFETs.

I was thinking in bias them as the Fetzer valve or using the Schade mechanism to transform a pentode tube into a triode, by connecting the plate to the grid using a resistor network. Have to put my hands on to them and play.

[defaced]

There's been quite a bit of work done with the LND150 that's out in the wild.  KMG (who posts here) has some excellent stuff on his site.  There are also some good threads at SS Guitar.  I've been working through them as I tune this project: http://www.diystompboxes.com/smfforum/index.php?topic=99142.msg879595#msg879595
Right now I'm working through how to configure the source components to get an approximate of the gain, frequency response and clipping characteristic of each stage. Unfortunately I don't have a tube circuit for comparison that's finished, so I'm doing this by ear.  Going by what I'm hearing, there's too much gain which is resulting in fuzzy distortion.  

As for sourcing, I buy mine from Mouser, but I'm also in the states.  Not sure if shipping cost would be prohibitive for ordering from across the pond. 

[R.G.]

Right now I'm working through how to configure the source components to get an approximate of the gain, frequency response and clipping characteristic of each stage. Unfortunately I don't have a tube circuit for comparison that's finished, so I'm doing this by ear.  Going by what I'm hearing, there's too much gain which is resulting in fuzzy distortion.  

That's the problem I ran into, if what you want is the standard triode gain stages of most guitar amps.

MOSFETs in general and the modern MOSFETs in particular have much more transconductance (change in drain current per volt of change in gate-source) than triodes (change in plate current per volt of change in grid-cathode). If you try to use them with the same resistance values, the voltage gain is much, much higher. Then there is the plate/drain curves thing I mentioned. MOSFET curves are much more like pentode plate curves, which also matches the higher transconductance. And frequency response, come to think of it.

The best I could do was to redesign the whole circuit down in impedance, scaling the plate resistor down to a much smaller resistance to give a similar voltage gain without inserting either source or drain-gate feedback. Feedback acts to idealize the gain and sharpen the transition into clipping, which is what most guitarists don't want. Scaling the impedance levels down lets you run them raw, without the high feedback covering things up.

This means that the power supply has to be scaled back down, too, to keep things from burning up at the higher currents. Capacitors have to be scaled up to lower their impedance and keep the same frequency response. It just keeps rippling.

I played with it long enough to decide that there was probably something there that if I kept hacking long enough could be tweaked out, but then I'd wind up with another semiconductor simulation of tubes, and the world is full enough of them already. I decided it's not as simple as subbing in a MOSFET for a triode and changing the bias for really good stuff, and for so-so stuff, why bother?

But Merlin may have some clever insights beyond that.

[defaced]

Good insights RG.  I'll keep those in mind.  My next step is to look at increased source resistance while using the -V bias point to keep the Vgs at a similar point as what one would get with conventional source resistances.  Or put a simpler way, make the stage more like a cathodyne phase splitter.  Basically I want to see if I can dump some of the voltage gain to the source.  I've also considered a split drain resistor, that's the next path I plan to try.  Then the last one, which I'd like to avoid, is re-designing all of the interstage voltage dividers and filters to work with the higher voltage gain. 

That said, (3) cascaded LND150 stages sound pretty good.  (4) takes it over the edge, it may just be the cold clipper I'm bypassing with the (3) stage configuration that is really causing things to go overboard. 

[R.G.]

Increased source resistance is negative feedback, too, unfortunately.

I tried a split drain resistor with a BFC bypassing most of it to force the AC current load to much smaller than the DC current load. That helped.

I also used some interstage dividers to simply dump the gain. That ran into problems where the signal, even divided, was enough to make the drain clip.

Pentodes can sound good, as witness the popularity of the pentode input stage on some AC30s among Vox afficionados. It's just not the standard cascaded-triodes sound that guitarists expect.

Then too, there is a fair amount of history behind *any* clipper being usable as long as the EQ before it and after it work well. After all, a clipper really only makes harmonics for the EQs to voice. I suspect that this is the reason that JFETs dropped into triode circuits make people happy. It's not the same, but hey, sounds good, and I made it myself! 

I've done a fair amount of tinkering with high voltage MOSFETs in amplifier circuits, a part of which I posted on Geofex as the MOSFET Follies, at least the more accessible bits.

[tca]

I played with it long enough to decide that there was probably something there that if I kept hacking long enough could be tweaked out, but then I'd wind up with another semiconductor simulation of tubes, and the world is full enough of them already. I decided it's not as simple as subbing in a MOSFET for a triode and changing the bias for really good stuff, and for so-so stuff, why bother?

My state of mind exactly!

[AdamM]

http://uk.mouser.com/Semiconductors/Discrete-Semiconductors/_/N-awhng?Keyword=depletion+mode&FS=True

[merlinb]

http://uk.mouser.com/Semiconductors/Discrete-Semiconductors/_/N-awhng?Keyword=depletion+mode&FS=True

£12 postage!? Think I'll pass...

[craigmillard]

Did you ever find a source Merlin? Im looking for a few of these in the uk too!!

[merlinb]

No, nowhere reasonable.

[R O Tiree]

BF994 available at Farnell, along with a few other dual-gate depletion mode MOSFETs, about 30p each, but SOT143 package only. Max Vds 20V on those.

[smallbearelec]

Is there one particular number that would be of best/greatest use? I can't put major $ into these right now, could buy 100 or so from Mouser and offer small quantities @ maybe $1.25 until I can invest in a couple of thousand. Let me know.

SD

[craigmillard]

Hey Guys,

I have found a potential source!:)

http://www.anglia-live.com/

Although they have a min order quantity of 50 and a handling charge of 5 quid! Free shipping though, would come in at £30 for 50, that 60p each...

Anyone wanna go in on an order with me??

Sorry forgot the most critical part:) they are Supertex LND150N3-G

[MetalGuy]

Another source for LND150 in Europe:

http://www.tube-town.net/ttstore/product_info.php/info/p5342_MOSFET-LND-150.html