LND150 replacement options

[craigkimble]

Hello,

I am wanting to experiment with making an active FX loop and have found several designs leveraging the LND150 MOSFET. However, they are clearly difficult to find from the usual suspects. Everyone lists them as on backorder. I have found a few on ebay at inflated prices from China, but am afraid these are probably fake. What are others using as a suitable replacement?

Thanks!

Craig

[Ten Over]

Hello,

I am wanting to experiment with making an active FX loop and have found several designs leveraging the LND150 MOSFET. However, they are clearly difficult to find from the usual suspects. Everyone lists them as on backorder. I have found a few on ebay at inflated prices from China, but am afraid these are probably fake. What are others using as a suitable replacement?

Thanks!

Craig

I don't know what others are using, but here are four possible ones:

DN3545N3-G
VN0550N3-G
VN2460N3-G
VN2450N3-G

[Ten Over]

I have used IRF820's on occasion. To order use IRF820PBF.

[roberto]

I would not use then on the buffering send stage of the loop, instead other choices like the very old IRF820 or modern upgrades are better.
For the return stage both a DN2540 or a DN3545 can be used. Check an high voltager depletion mosfet on major sites and pick the one available, then play with it to find the sweet spot. You can then triodize their pentodeish fet-curves to get a better return stage.

[Ten Over]

I would not use then on the buffering send stage of the loop, instead other choices like the very old IRF820 or modern upgrades are better.

Why not? They work fine as source followers as do a wide array of other devices. Their small size and sufficient no-heatsink wattage make them ideal for source followers.

[Ten Over]

For the return stage both a DN2540 or a DN3545 can be used. Check an high voltager depletion mosfet on major sites and pick the one available, then play with it to find the sweet spot. You can then triodize their pentodeish fet-curves to get a better return stage.

Drain current denominations of amps instead of milliamps is problematic for single unit gain stages. They do operate very well when used in a cascode configuration, however. LND150's also work very well when used in cascodes.

[roberto]

Why not? They work fine as source followers as do a wide array of other devices. Their small size and sufficient no-heatsink wattage make them ideal for source followers.

There are other components that do source following way better than them, EG the STU9HN65M2 able to strongly drive even heavy loads in a very transparent way.

[roberto]

They do operate very well when used in a cascode configuration, however. LND150's also work very well when used in cascodes.

If two devices are needed for the return, I absolutely drive them from the source through a pmosfet and give feedback from drain to the gate to triodize the curves. Sounds way better and more natural. On top of that, I never bypass the loop, but build the amp always including it.

[Ten Over]

Why not? They work fine as source followers as do a wide array of other devices. Their small size and sufficient no-heatsink wattage make them ideal for source followers.

There are other components that do source following way better than them..

As long as you stay away from any of their limits, they all operate with 100% negative feedback. This makes it hard to believe that any of your other components are doing source following way better than any of mine.

[Ten Over]

...EG the STU9HN65M2 able to strongly drive even heavy loads in a very transparent way.

That particular device is going down the way of the dinosaur, so it may not have been the best example.

A device can either drive a given load or it can't. Let's say that we wanted to deliver 5V peak into a 10k load. Let's also say that the Source resistor is also 10k, so we need to have 5V peak into 5k. That will require 1mA peak. If we are idling at 2mA, then the device will need to handle 3mA. Being able to handle 5,500mA doesn't drive the given load with any more strength. Idling at 2mA will keep the signal well away from the zero current condition so that the load will be driven in a very transparent way.

[FUCHSAUDIO]

I like the IRF style devices as a loop follower. On the return side the LND-150 is great. Unfortunately, most suppliers are talking late 2023 for stock again.
They are available in an SMD style, and making an adaptor to put them on a through hole board isn't tough. It's a cool device and I hope it comes back into stock sooner rather than later.

[Ten Over]

If two devices are needed for the return, I absolutely drive them from the source through a pmosfet and give feedback from drain to the gate to triodize the curves. Sounds way better and more natural. On top of that, I never bypass the loop, but build the amp always including it.

I'm not sure what you are doing. Can you post a schematic?

[roberto]

That particular device is going down the way of the dinosaur, so it may not have been the best example.

So you call that prehistoric when you in 2023 propose a very dangerous schematic with an IRF820 that is more than 40 years old? Don't make your bitterness hide your brain.

A device can either drive a given load or it can't. Let's say that we wanted to deliver 5V peak into a 10k load. Let's also say that the Source resistor is also 10k, so we need to have 5V peak into 5k. That will require 1mA peak. If we are idling at 2mA, then the device will need to handle 3mA.

It is not that simple as on simulators.
Power dissipation: 3mA on a source follower supplied at a standard 450V B+ makes 750 mW dissipated in a TO92. If you supply it with 300V it will be 500 mW.
With a TO92 I wouldn't exceed 300 mW in a hot tube amp for the reliability of the system.
SOA: with 3mA you can't exceed 90V Vds, that means you will be safer around 70V Vds.
You can use it with 140V B+, working point at 70V and around 3mA with 22k on the source and around 250 mW dissipated.
I'd stick with other new brontosaurus.

Drain current denominations of amps instead of milliamps is problematic for single unit gain stages. They do operate very well when used in a cascode configuration, however. LND150's also work very well when used in cascodes.

Do you really want to propose a loop with 130V on a pot that needs to be touched by the player?
You will kill someone proposing that! Please remove that schematic!

[roberto]

I like the IRF style devices as a loop follower. On the return side the LND-150 is great. Unfortunately, most suppliers are talking late 2023 for stock again.

I agree, nowadays there are probably better options than the IRF series, but as a general concept I prefer to have a device that can handle more current on the buffer and a device easier to drive on the return.
You can even linearize it by local feedback (it will "triodize" its curves) with a resistor from drain to gate and from gate to the input. Pay attention to the input impedance that will drop.

[Ten Over]

So you call that prehistoric when you in 2023 propose a very dangerous schematic with an IRF820 that is more than 40 years old?

It is not the prehistoric nature of dinosaurs that I am referring to. It is their extinction. STU9HN65M2 will soon be extinct. I would not have recommended a device that is being discontinued, but it is fine with me if you want to do that.