use datasheet of 2SK2145 to decipher approprite connection of bottom pins of test socket
then solder it to piece of veroboard, and use it as pcb for everything else
I included small relay as functional switch for JFet circuit simply because that is best option - certainly much better quality of switch than any manually operated taster or switch; there - every additional Ohm counts, and manual switches are too "dirty" for that small range of currents
then, shown "manual switch" for entire test circuit - I started with taster there, then realized that is stupid to use my digit to keep taster pressed for 15-20-30 secs, as much is needed to see mA numbers on DMM settling to final value
if you don't have handy 12-15Vdc supply, just include 7812 reg in front of everything, then you can use even your laptop supply as main PSU
then solder it to piece of veroboard, and use it as pcb for everything else
I included small relay as functional switch for JFet circuit simply because that is best option - certainly much better quality of switch than any manually operated taster or switch; there - every additional Ohm counts, and manual switches are too "dirty" for that small range of currents
then, shown "manual switch" for entire test circuit - I started with taster there, then realized that is stupid to use my digit to keep taster pressed for 15-20-30 secs, as much is needed to see mA numbers on DMM settling to final value
if you don't have handy 12-15Vdc supply, just include 7812 reg in front of everything, then you can use even your laptop supply as main PSU
Attachments
.....
if you don't have handy 12-15Vdc supply, just include 7812 reg in front of everything, then you can use even your laptop supply as main PSU
Attachments
it depends of matching purpose
say - for triplets I'm using in buffers - I'm content with having all 3 in 0.5mA ........that difference being further diminished when JFets are actually biased in circuit to significantly lower Iq than their Idss is
for 2+2 I'm using in LTP - I'm keeping them in same bracket (0.5mA), but then I'm combining them to have closest sum of twos
rough example :
10mA, 11mA, 12mA, 13mA
then 10mA and 13mA are used for one LTP side, 11mA and 12mA used for other side of LTP
trick stolen from Pa
if I tell you from which exact circ, I would need to shoot you

say - for triplets I'm using in buffers - I'm content with having all 3 in 0.5mA ........that difference being further diminished when JFets are actually biased in circuit to significantly lower Iq than their Idss is
for 2+2 I'm using in LTP - I'm keeping them in same bracket (0.5mA), but then I'm combining them to have closest sum of twos
rough example :
10mA, 11mA, 12mA, 13mA
then 10mA and 13mA are used for one LTP side, 11mA and 12mA used for other side of LTP
trick stolen from Pa
if I tell you from which exact circ, I would need to shoot you

ZM Thanks for the diagram, unfortunately I am missing some components (voltage regulators) and I need to finish the project as soon as possible, I used the dca pro 75 from atlas peak, I was able to select two quartets with intervals of 0.06 mA, the first quartet goes from 4.10 - 4.14 - 4.14 - 4.16 and the second quartet goes from 4.22 - 4.23 - 4.26 - 4.27 mA. better than this I could not do. Thank you for helping me. 

why so in hurry?
I mean - if you have other means ( Atlas) normally that you're going to use it, just asking - why so rush 🙂
I mean - if you have other means ( Atlas) normally that you're going to use it, just asking - why so rush 🙂
Some time ago, in the aleph j thread, it was mentioned that the dca pro 75 was not too suitable for jfet selection, I asked you because you are an always helpful person.
that wasn't me, that was trickster's Ggl translator
though, pity I don't know Talian .......... beautiful language
though, pity I don't know Talian .......... beautiful language
Any udates on the LuDEF P?almost forgot - and there is upcoming LuDEF P
buffer and autoformer replaced with active FE, having all bells and whistles and NFB
Lydlaug
not yet
more actual production work than I expected, so "new" ones are postponed
hope soon, after New Year
more actual production work than I expected, so "new" ones are postponed
hope soon, after New Year
Happy New Year!it depends of matching purpose
say - for triplets I'm using in buffers - I'm content with having all 3 in 0.5mA ........that difference being further diminished when JFets are actually biased in circuit to significantly lower Iq than their Idss is
for 2+2 I'm using in LTP - I'm keeping them in same bracket (0.5mA), but then I'm combining them to have closest sum of twos
rough example :
10mA, 11mA, 12mA, 13mA
then 10mA and 13mA are used for one LTP side, 11mA and 12mA used for other side of LTP
trick stolen from Pa
if I tell you from which exact circ, I would need to shoot you
![]()
Thanks for the input on matching. I got my adapter for the SOT-25(6) package and if Mouser keeps their word I should be getting a stash of 2SK2145 to start matching soon.
In the meantime I have been working on Spice sims of the FE and OS. The FE is pretty straight forward except for a model of the 2SK2145. Do you have one or know who does that is willing to share?
On the OS using 24v rails (not 22.5) it seem that I can get by with a 4u/300 heatsink from Gianluca keeping the sink in the mid to upper 50s. The big heater is the IRFP9140 with the IRFP240 just behind. LU1014 is cruising along with out much of a sweat.
I didn't have need to play with 2SK2145 in sims, but for few LTS musings I made, I was using pair of 2SK170 instead
4U/300 is OK< according to fact that many Greedy Boyz are using the same
I prefer 4U/400 from several reasons - Gianluca is selling more expensive item, I'm having better 'sinks and - of same importance - more space inside for more optimized arrangement of things inside
especially with amps having autoformer, space between signal xformer and mains Donut is critical ......... and I'm lazy, always resorting to easier solution 🙂
4U/300 is OK< according to fact that many Greedy Boyz are using the same
I prefer 4U/400 from several reasons - Gianluca is selling more expensive item, I'm having better 'sinks and - of same importance - more space inside for more optimized arrangement of things inside
especially with amps having autoformer, space between signal xformer and mains Donut is critical ......... and I'm lazy, always resorting to easier solution 🙂
I finished LuDEF last night. All voltages have been adjusted smoothly and are stable, very quiet. I really like how they sound and I would like to use them in an active 3-way open baffle speaker system. Can you tell me, please, what preamp gain it needs to get maximum power?
Thank you .
Thank you .
Attachments
gain of 4x
25W@8R means 14V14rms at output
divided with 4= 3V53rms for full blast
meaning - line level stage in front with 6db (2x) of gain is sufficient
25W@8R means 14V14rms at output
divided with 4= 3V53rms for full blast
meaning - line level stage in front with 6db (2x) of gain is sufficient
- Home
- Amplifiers
- Pass Labs
- LuDEF