Why? forgive me for my simple understanding, irrelevant of the type of device used, any modulation to the gate puts the device in source follower configuration, what am I missing?
Some people asked about the PCB in post #1286.
I have no plan to build my own any time soon.
So I shall not be offering projects based on them.
But if there is a general interest, and Nelson does not object, I am happy to post the Gerber files.
You can then organise your own GB for the boards.
Our stock of Semisouths are all accounted for.
Interest for genuine Semisouth SJEP120R100 curve tracer matched pairs & quads
So no additional Q-sets from us either.
But if people do want to use 2SK3497 instead of IRFP150 as someone else suggested, we'll make a limited amount available.
Patrick
I have no plan to build my own any time soon.
So I shall not be offering projects based on them.
But if there is a general interest, and Nelson does not object, I am happy to post the Gerber files.
You can then organise your own GB for the boards.
Our stock of Semisouths are all accounted for.
Interest for genuine Semisouth SJEP120R100 curve tracer matched pairs & quads
So no additional Q-sets from us either.
But if people do want to use 2SK3497 instead of IRFP150 as someone else suggested, we'll make a limited amount available.
Patrick
maybe I'm missing something , but I can't see in picture you posted for UJN - any sort or way of modulating it
Yes, you can use a depletion mode FET in a SRPP.
One that has a Vgs between -0.5 and -2V at bias, with a Yfs of say 3~6S ideally.
In practice, such FETs usually have a high positive tempco.
So bias will increase with temperature, whereas the Semisouth has a rather low tempco.
So you will have a DC drift issue to solve.
Can be done, of course.
Cheers,
Patrick
.
One that has a Vgs between -0.5 and -2V at bias, with a Yfs of say 3~6S ideally.
In practice, such FETs usually have a high positive tempco.
So bias will increase with temperature, whereas the Semisouth has a rather low tempco.
So you will have a DC drift issue to solve.
Can be done, of course.
Cheers,
Patrick
.
Last edited:
Attachments
An additional note to the circuit in post #1307 using a depletion MOSFET in the modulated current source (MCS).
The level of modulated is controlled by the Yfs of M1, and the ratio of R11 : R13.
The Vgs of M1, and the sum of R11 and R13, on the other hand, affect the DC bias of the MCS.
So although the circuit eventually is much simpler, it does require detailed characterisation of the depletion device upfront.
This is why the opt-coupler auto-biasing is attractive.
No free lunch.
Patrick
The level of modulated is controlled by the Yfs of M1, and the ratio of R11 : R13.
The Vgs of M1, and the sum of R11 and R13, on the other hand, affect the DC bias of the MCS.
So although the circuit eventually is much simpler, it does require detailed characterisation of the depletion device upfront.
This is why the opt-coupler auto-biasing is attractive.
No free lunch.
Patrick
The "J1014"
And this is just me having fun.
I am sure people remember the Zen V9 with a LU1014 Triode Cell.
So why not use that here instead of the Semisouths ?
This is how it could work.
A bit more complicated, but one can now change the open loop gain easily.
This allows you to adapt for different speaker loads, or the level of NFB.
Nelson still have 4k of those to give away ?
Cheers,
Patrick
.
And this is just me having fun.
I am sure people remember the Zen V9 with a LU1014 Triode Cell.
So why not use that here instead of the Semisouths ?
This is how it could work.
A bit more complicated, but one can now change the open loop gain easily.
This allows you to adapt for different speaker loads, or the level of NFB.
Nelson still have 4k of those to give away ?
Cheers,
Patrick
.
Thank you, now you have me thinking along the lines of a SIT 2 and others
And this is just me having fun.
I am sure people remember the Zen V9 with a LU1014 Triode Cell.
Or maybe a combination
While we're having fun, here's a draft of a FirstWatt form-factor board using Ohmite TO-126 package power resistors in place of the parallel Panasonics.
Pardon my ignorance, but how hot are the axial resistor when in use?
Aren't you heating your power resistor while thermally coupling them to the JFET?
Any benefit doing so (resistor value drifting with temp, etc)?
Ohmite TNP series can only take 1W in free air, and 10W on heat sink.
It is thin film and low tempco, but has limted choice of low resitance values.
For the same foot print as 2x Panasonic 2W, one can also use 10x Beyschlag MMB0207 in series parallel.
More rated dissipation, thin film, low tempco, low tolerance, low inductance, .....
Patrick
It is thin film and low tempco, but has limted choice of low resitance values.
For the same foot print as 2x Panasonic 2W, one can also use 10x Beyschlag MMB0207 in series parallel.
More rated dissipation, thin film, low tempco, low tolerance, low inductance, .....
Patrick
Given their respective power dissipations I suspect the heatsink will slightly heat the source resistors and slightly cool the drain resistors. Not that I can imagine either of those making any difference.
The primary benefits (such as they are) are that it lowers the parts count just that little bit more, and I think they look cool.
The primary benefits (such as they are) are that it lowers the parts count just that little bit more, and I think they look cool.
I resolved this issue by starting out with high resistor values initially,
trimming them down for each individual Fet by paralleling additional
values. Makes for additional holes in the board, but works like glue...
