Qty 210 for $20 shipped, not bad... Shipping is about $5, $3 fee if under $15, so if you make it at $15, $3 fee drops.
Hhmmm. No more excuses... 
Thanx Lexel!
I keep meaning to make up a 2s driver using my resistor bank style for the 1st channel. This way you no longer have to worry about the 7135 quality. I first started using this for 2S lights with the MT09R drivers and have not looked back for 2s+ designs.
Although for these bistro drivers I just have not had any lights I am working on need one so I have not had the time or motivation to update them.
I also want to update them to use a cap + resistor combo for the OTC that should hopefully make bistro viable again, I have not even considered using it with the old OTC in years.
What resistor value are you using on the cap? Still the standard OTC?
I have not tried it yet myself, Tom Tom had the idea first IIRC. The idea is solid though, use a much larger OTC and a small resistor to bleed it at a constant rate. Basically eliminating the MCU power drain variance.
This should allow for much more consistent OTC function.
Part of why I have not tried it myself is I don’t know what values to use. If this was known you could in theory replace the OTC and stack a resistor on top to get the same end result on the current designs.
I have not touched a bistro driver in any form in over a year IIRC, so I have just not had time, motivation or opportunity to test it myself.
If someone does test it and figured out the right values please do post them here.
I think Tom tom mentioned something like a 4.7uf cap but don’t recall a resistor value to start with. The best option would be a pot for testing I would assume.
on 2S I made good results simply using the 1S HEX 1uF with 680kOhm, worked more consistent than 2S hex for some reason
had to change the voltage divider to fit 27.5k and 3k
Thanks for the info.
Interesting, why did the voltage divider need to be changed?
I finally managed to get my lights running flawlessly with TA Bistro. Many thanks to all of you for your useful tips. 
I will post again if further questions come up.
Just ordered qty 210 for ~$20 - thanx! They have Attiny85's, little more $$$ than Arrow but no Infineon FET, least not the one I've been buying: BSC009NE2LS5IATMA1
Next flashlight, next problem! 
I assembled a 22mm TA driver for the Sofirn C8F and outside of the flashlight it works as it should. If I screw it in with the retaining ring it seems to be shorting somewhere because it doesn’t work and almost melts the testing cables. Is anyone familiar with this problem or can someone explain me how to resolve this?
Possibilities I could see:
- compressed spring on the BAT side touches 7135 ground
Did you check with a DMM where you can find ground where it shouldn’t be?
Any chance you could upload pictures?
I am not a pro but have built 6 of these and they are all flawless.
I built up a couple C8F hosts - ahh, the retaining ring is pretty thick, maybe a little too thick. Think I beveled the inner bottom edge of the retaining to ring to avoid groundouts. You can remove the retaining ring and possibly test with a test battery setup or bench PS. I probably used my rotary tool to do the beveling. I know I've done this kind of thing before.
I have made for that light in 217600 a special driver with more free ground ring thickness so there is no more problem
20mm
for the 17mm drivers I can recomment after soldering the spring and check it has no short by simply potting the inner legs of the AMCs with some glue (I use temperature hardening SMD glue)
I didn’t compress the battery because I only held some cables to it. I carefully looked for unwanted ground contacts and even glued all possible contact points with silicone-like glue. Now it doesn’t produce sparks and let the cable glow like it did before but it still not works. What I can’t explain is that it works when unconnected but not in its place. Once I tried a pressfit without retaining ring, same thing. I never used the battery tube. I could try to upload pictures but it looks quite messy with the glue now.
!http://www.metronixlaser.de/bilder/flashlight/17mm_TA.png
!
red is where battery is before the 4.7Ohms resistor only there it can make bad short
black is where ground is neaR positive, so the BR or R5 is very close to ground where a short may happen
but this would mean if in body or not would always give a short
but as you say with driver not connected to the light it works
this leaves only one short option and thats the LED+ cable to the MCPCB or the MCPCB solder contacts the reflector on positive pad
I tested with my multimeter now. There is contact between the positive spring and the negative ground ring when the driver is connected to the flashlight body. The driver alone has no contact. How can that be?
Thank you, Lexel! So I will try to open the front part again, control and secure things with some kapton tape.
Thank you very much, Lexel and all others who helped me! It was as Lexel said, the wires were touching the MCPCB or the reflector. I resoldered the wires and put some kapton tape on the reflector. The C8F immidiately worked 
after screwing it together again. I’m only looking for a proper firmware now as the NarsilM.hex I flashed is going wild 
and was probably the wrong one. Would a firmware like on the Emisar D4S (RampingIOS V3) or Anduril work on a TA driver?
You need a triple build of NarsilM.hex for it to work properly, I suspect, but I din't know the exact driver and exact parts you used. I could build a HEX file form the latest, greatest NarsilM if you like, just tell me exact details on the 7135's you used - how many, hope you used the raptor claw ones... ?
Tom got a link to purchase the Raptor Claw 7135’s?
Ohh, Lexel gave me a link, I ordered like qty 210 to get slight break on shipping. Really, really hope these are the real things this time...
This is what Lexel gave me: https://lcsc.com/product-detail/LED-Drivers_MEL7135PG_C80214.html
Didn't receive them yet so don't know for sure. Tracking says they arrived at JFK NY yesterday, so close.