JFET = fire and pitchforks !!😱
Many through-hole are going end of life on mouser. I have a mouser account ,
so they are my main source - 2 days from Texas ....
As they go "end of life' many unscrupulous (even some here at DIYA) are buying them up.
I see ksa1381 drop from thousands to 0 in one day.
So , expect these "horders" to fleece you with .80c Mouser parts at a big markup.
$$$ $crew them , will make easy HQ projects with bigger SOT and avoid
"unobtanium" B$.
OS
Don't give it up, there is 7000pcs of KSA1381, on order...
Personally I still trust them, as Fairchild still keep them manufaturing.
Sajti
But I keep getting PM'ed like I can pull 1381's out of my ar$e.
So, I gave the best alternatives I could.
I wash my hands of them for new designs.
Edit - greenamp will allow me to offset many >200V devices with <100V ones.
EF3 and low current will allow for SOT.
I AM ready for generation 3 designs.
OS
Last edited:
I don't really understand why. Frankly speaking, I hate to solder the smd parts, maybe I'm too old, and my eyes are tired....
Sajti
But, SOT223 is bigger than a TO-92.
I think it is easier.
SOT223 , you can cool the device die with PCB copper.
OS
Attachments
Last edited:
Not to try to convert you , but this is
how you will will get your through - hole
replacements in the (very near) future. (below).
REALLY !!
Edit - a ZTX957 could be on that adaptor.
OS
how you will will get your through - hole
replacements in the (very near) future. (below).
REALLY !!
Edit - a ZTX957 could be on that adaptor.
OS
Attachments
Last edited:
OK, it looks funny 🙂 Yesterday I soldered some smd,
than I put order for 200-200pcs of 3503/1381...
Sajti
Not want to criticize you - but 'm sure they will not. The KSA1381E and the KSC1305D are still available and working perfect in a pnp based VAS design.
BR, Toni
and hard to match
But expensive, hard to find, hand matched parts make the perfect cult audio product for the hi-end market😉
The K170 is only ~$3, not too bad, but I didn't realize the J74 was so rare.
I knew LS had made some but apparently they are once more out of stock.
That does make it unreasonable for a DIYaudio project.
I still have a pair, I could make a very exclusive amp for the hi end, and like OS I could call it a cool name to increase the appeal - the "Unobtainitron" perhaps?
Best wishes
David
Problem is for a symmetric design. There is no any more KSC1305E type or KSA1381D type and not possible to do matching.
Oh, David, nice to see you!
The main advantage of the self-biased IPS - is a high CMRR, this is intrinsic oftenly unknown property.
So, why not to increase typical up to “set and forget” level?
I don't mind . Criticism is the mother of invention for me 😀 .
But , you fail to consider my use of the 1381/3503 = cascode.
Lin/'blameless" ... you are spot on. My cascode shows almost no THD penalty
with the MJE340/350 (below).
I do see a performance "hit" in the CFA's slewrate , no perfect 100k squarewaves today.
>200V/uS - 1381/3503 150V/uS with the mje340/350.
No , I would not use the 340/350. But I HAVE simulated a mid performance
device as my VAS cascode. 13pf/85mhz has little effect on MY design.
OS
Attachments
Last edited:
That is funny , I could become yet another "snakeoil" salesmen here at DIYA.
Actually , a good idea. Make my IPS's SOT23/223 on a 6 x 4 cm PCB ,
thermal epoxy a copper plate on top .... sell as a proprietary "module".
OS
Hello OS, when you refer to the ZTX957/857 SOT223, I only see it in TO92, it will be the same as FZT857/957
Quite right - FZT is the SMD variant.
I have the models and data sheets for the ZTX.
OS
The circuit in post-122 is NOT class-H or class-G; rather, just a fancy cascode. I don't see how it is any more efficient than any other amp with the same rails? Waste heat is just shared over devices in series rather than in parallel.
I was using my reference EF3 slewmaster for testing (below). Your Sanken models gave me the same 9ppm/104v p-p/4R/20khz output. What I find interesting is that I had to (optimally) bias the EF3 to 50ma -
JUST LIKE the real physical devices. ON/Fairchild "like" 80-100ma biases. The Sanken devices are different in the real world. Your models VERY accurately reflect this !! Thanks.
OS
Attachments
If you read post #128 you could see this was already pointed out to me by Dave Zan. It is a "stacker" , no more efficiency but still splits the SOA between the 4 devices. By this "deviation", my last posted update "slides" from a true class H at the 30V level and approaches almost full series operation "stacker" as it nears the high rail. EVEN in full switched OEM class H's , full power use of the inner rails is insignificant to the point of just adding the dreaded switching distortion + NO added efficiency.
(BELOW) is D107 and R116 sharing current at the onset of class H.
But the outer devices are still class AB (always on) not B , like other OEM designs. R123 (the trimmer) can both adjust the outer device idle and , if you go further , make the outer devices transition to "stacker" mode. Soon after post #128 , I pulled the patent for the "stacker" in my reply. OS
Attachments
The ZTX957 is 85MHz at 100mA, the KSC3503 is 150MHz at 10mA, so the 3503 will be faster in application. At low frequencies the ZTX957 might perform better than the KSC3503, but HF performance will depend on how well you can drive the base and how well you can manage it's lower GBW contribution.
Early voltage is 270 here:
Measuring the Early voltage "VA" of some PNPs at 8mA
- Status
- Not open for further replies.