LND150 Hybrid Cascode

[stratomaster]

The new AC15 has a boost function that adds an LND150 between the plate resistor and plate of the first stage.  I wanted to experiment with this circuit so I cribbed it almost directly and put to into the first stage of a Princeton non-reverb.

The circuit as I installed it is attached.  It works and gives about 3.5dB of boost. 

The tweaks I made are keeping the 100k plate resistor instead of the 220k used on the AC15, and I used a 0.001μF cap to ground off the gate to give a bit of low end roll off. I also left the cathode network stock as I wanted to have the stock sound available by shorting across Drain and Source on the FET.

I understand there can be a very large gain boost in this configuration--more than just 3.5dB, and if there's some juice to squeeze and something to learn here I'd really like to.  My familiarity with solid state is slim to none.  Does anyone here have any guidance and/or resources to better understand the LND150 in this application? 

Thanks.

[Merlin]

The gain of a triode stage is: A = gm (Ra || ra)
The gain of a cascode is approx: A = gm Ra (Where gm is the transconductance of the lower device)
Cascoding only takes ra out of the equation, so unless ra was much smaller than Ra to start with, the increase in gain is not necessarily very large. When you switch in the LND150 you are probably reducing the idle current in the tube, giving you less gm. You also get less headroom from a cascode. I'm not sure there actually is much more juice to squeeze out of this, do you have a link to the Vox schem?

[stratomaster]

Thanks, Merlin.  Your page on the cascode was very helpful in wrapping my head around what was happening.

Unfortunately there are no schematics available that I know of for the new Vox design.  I took the circuit from a screenshot of a video of the turret board and used your page to fill in the rest.

I do have a free set of contacts on the push/pull switch, so paralleling another resistor across the cathode network to boost current/shift the bias in cascode mode isn't out of the question.

[Merlin]

You did well to decipher that layout! If you want more gain I guess you could increase the anode resistor back to 220k, and maybe switch the cathode bias resistor to a smaller value when the FET is in circuit. Incidentally, there is no need to use a depletion FET in this configuration, a reguler enhancement FET would also work fine.
However, I'm concerned by your switch. When the FET is shorted out, your anode can swing up and down a lot, relative to the gate, possibly destroying the gate-source insulation.

[stratomaster]

Yeah, switching like that has me concerned as well.  I can't think of a good way to use protection diodes that doesn't affect the stock configuration. Perhaps tweaking the fixed bias at the gate to minimize the voltage difference? But again, very little knowledge of solid state.

[kagliostro]

Will it be possible to have a link to the video from which you sourced the image ?

Many Thanks

Franco

[stratomaster]

Will it be possible to have a link to the video from which you sourced the image ?

Many Thanks

Franco

Here's the video:

&t=50

[kagliostro]

Many Thanks

One thing I'm not able to understand, is that amp build by VOX with the LND150 or it is a custom mod ?

Franco

[stratomaster]

It's stock from Vox on their newest AC15 variant.  It's meant to substitute for the EF86 channel.

[stratomaster]

Yeah, switching like that has me concerned as well.  I can't think of a good way to use protection diodes that doesn't affect the stock configuration. Perhaps tweaking the fixed bias at the gate to minimize the voltage difference? But again, very little knowledge of solid state.

Perhaps use the spare set of contacts to disconnect the source/short or to the gate? 
Or Parallel in a resistor that sets gate voltage close to the triode plate voltage. This would limit the variation to just the AC signal swing .. which on the first stage should be in the single digit voltage. This would otherwise disturb the bias, but since the FET is "out" of the circuit at that point, the gate bias is immaterial.

[kagliostro]

Thanks for the replay

The circuit you draw is what is on the original VOX design or you added the switch ?

Franco

[stratomaster]

Thanks for the replay

The circuit you draw is what is on the original VOX design or you added the switch ?

Franco

The switch is in the Vox design. The differences between what I installed and the Vox design are in the first post.

[kagliostro]

Thanks

Yes, I was asking because of the discussion about the problem that the switch introduce and so I was interested to know if VOX introduced it or it was a mod

About this cascode circuit ask to ChatGPT, he do some consideration that aren't about boost as improvement in the circuit, he report other difference/benefits

Franco

[passaloutre]

I’d be very curious to hear more about this and see how you like it. I’m starting to experiment with LND150s, as I have an amp that needs a preamp and has no more juice for heaters. I’ve also always been very interested in the cascode, but never tried one. How about a double mosfet cascode?

[Merlin]

How about a double mosfet cascode?

It wouldn't make sense to build a double MOSFET cascode, because a MOSFET already has characteristics that resemble a pentode/cascode. The LND150 is an unusual part, I kinda think it is worth reserving for situations where you really need a depletion FET, whereas in this case the upper device can be a regular MOSFET (IRF820 etc). Up to you though.