Depletion mode FET needed, which type?

[neazoi]

Hello, my application design suggests a J110 (10mA Idss) mosfet in depletion mode.
The J108 is similar but has 80mA Idss.

Can you propose a similar fet with more Idss (interested only in depletion mode, normally on at Vgate=0V)

 

[FvM]

Both are P-JFET (and depletion mode as all JFET).

What are you looking for? Depletion mode P-JFET with higher IDSS, or depletion mode MOSFET?

 

[neazoi]

Both are P-JFET (and depletion mode as all JFET).

What are you looking for? Depletion mode P-JFET with higher IDSS, or depletion mode MOSFET?

I apologize, but I have really no idea,
The circuit is from the https://dev.patents.com/pdf-4734658.html‎
He uses a J110, but I would like to get more current out of this, so I thought the J108 will be ideal.
Any proposition about other more powerful devices?

 

[FvM]

Sorry, I confused the FET types. J110 and J108 are in fact N-JFET. You can use any low ohmic switch JFET. But the transistor also in differ in terms of Vgs,off. E.g. J108 has considearble higher pinch-off oltage than J110. So it might be insuitable for the application.

E.g. U291

 

[neazoi]

Sorry, I confused the FET types. J110 and J108 are in fact N-JFET. You can use any low ohmic switch JFET. But the transistor also in differ in terms of Vgs,off. E.g. J108 has considearble higher pinch-off oltage than J110. So it might be insuitable for the application.

E.g. U291

Hm.. this is quite rare, but thank you

 

[FvM]

U291 is in production at Interfet. Possibly less rare are depletion mode MOSFET, e.g. BSS159. Don't know if they can work in the circuit.

 

[erikl]

U291 this is quite rare

BF246C, BF247C are available from Philips (NXP), Siemens (Infineon), and (formerly) Fairchild. Their Idss = 110 .. 250mA.

BTW: you can always operate two or more in parallel to increase the current.

 

[Kral]

Try the 2N4391 https://alltransistors.com/pdfview....st4392_2n4393_pn4393_sst4393.pdf&dire=_Vishay. Its Idss range is 5 to 30 mA. It is being discontinued by many manufacturers, but is still available.

 

[schmitt trigger]

Interesting circuit...perhaps if one swaps the (+) and (-) of the thermopile supply, one could possibly use a N-chan JFET, which may have a broader availability.

 

[FvM]

It is in fact a N-JFET (diode arrow towards the gate bar), but the circuit can work with JFETs of both polarities.

Depletion mode FETs are an essential means for designing low voltage DC/DC converters, also used in "energy harvesting chips like LTC3108 (in this case a depletion mode MOSFET).

 

[schmitt trigger]

"It is in fact a N-JFET (diode arrow towards the gate bar)"

Correct!
Having seen so much of the most common Mosfet, I got my arrows confused!

 

[neazoi]

I have made a prototype of the patent. I used a j310 which was available. The toroid was a high permeability FT50-75-J. It worked at once!
The primary was 1, 2, 3 and 4 turns. I found a good point at 2T, empirically.
I will try some of the other FETs you mention.

Having a white LED connected, I noticed a low distortion 122KHz sine on the output, about 5Vpp, when the circuit operated at 86-150mV input. At higher voltages the top of the waveform clipped seriously. The frequency was depended on the input voltage.
Oscillation started at 86mV and the output was a pure sine at this voltage!!!!
Hey this gives an idea about an ultra low voltage oscillator

Although the result was a lower brightness LED than the joule thief even at higher voltages, I believe this has to do with the current capability of the j310, so I will try the other fets you mention.

 

[neazoi]

Here is the tests I have made **broken link removed** (end of the page)

 

[schmitt trigger]

I have visited your website, very interesting.
I was particularly intrigued by the "Diamp".
Nice that you still use an analog scope, although with digital readout. Is that an Iwatsu?


Now a comment for your "JFET joule thief project" for lack of a better word. At input voltages above 0.9V, the reverse bias appears to go higher than 9V...that is not good for leds.
Perhaps if the main voltage source will be a single alkaline battery, you could reduce the primary/secondary turns ratio a little.

 

[neazoi]

I have visited your website, very interesting.
I was particularly intrigued by the "Diamp".
Nice that you still use an analog scope, although with digital readout. Is that an Iwatsu?


Now a comment for your "JFET joule thief project" for lack of a better word. At input voltages above 0.9V, the reverse bias appears to go higher than 9V...that is not good for leds.
Perhaps if the main voltage source will be a single alkaline battery, you could reduce the primary/secondary turns ratio a little.

It is an analogue scope and a frequency counter in the same instrument, it does not read out the waveform digitally in any way (out of budget)

I was thinking it the other way around. Instead of trying to alter turns, I could use a low drop, low consumption regulator at the output, to regulate it to the desired voltage (typically about 2.5v for charging ultra capacitors). Compensating for different input voltages might be more difficult than using an output regulator. I am thinking of better "juice out" as far energy as I can at the input, and manipulate the output later on (do not forget the rectifying diodes drop). Also, that way, more output sources with different voltage requirements, could be fed by the same input power source, by using different regulators.