Ok, adding some LDO versions, The 30mm version should be a drop in replacement for the L6 driver.
Now sure how small I can make then, possibly 22mm with no edge clearance at all, not sure if that would even be useful.
TA-LDO series - Here are some versions of the existing drivers with an added LDO option
These LDO versions of the drivers are mostly for e-switch lights that will be used in 2s or more configuration. Zeners cause a lot of excess battery drain when the light is “off” with e-switch drivers.
You can bypass the LDO by putting a normal SOD-323 diode across the furthest right LDO pins, with the line pointed towards the MCU.
The added R6 and R7 resistors are either 0 ohm jumpers or R7 can have a small resistor installed to correct for a C2 cap with the wrong ESR.
C2 also needs to be increased to 10uF, you can use the same one as you do on C1.
C1 must also have a cap rated for the full voltage you will be using!
Besides that they are exactly the same as the rest of the Texas Avengers drivers, just with less edge clearance for a given size due to the LDO.
Ok, managed to get the LDO on the 21mm version but I think that is as small as she is going. Even with this one you will have VERY little edge clearance.
Ok, I kinda had that impression. But I'm a bit confused. If the MCU can measure its own Vcc, without needing a reduced voltage and a sense pin, then your comment suddenly makes perfect sense. Measuring Vcc won't help if Vcc is from a regulator.. so with LDO, still need a divider and sense pin, not with the battery trick, so that would all add up.
So for powering these drivers with 1S (17mm version in particular), can R1 and R2 be left out and still get LVP? If not, I assume any ratio close to 22/4.7 (ex: 33/6.8) will get LVP within that close of right no? Also why not use 220K with clicky? I understand why use it with e-switch but does hit hurt anything to use it with clicky?
Correct, for multi-cell drivers the internal voltage reference won’t work.
Yes, R1 and R2 can be left out of any of the drivers if used in 1S. I am hoping that someone that knows how can inject that code as a definable option into bistro, narsil already has this option I think.
22/4.7 have proven to be a good combo and work well. The issue with 220k with clicky drivers is it will extend the OCT time to about 6 seconds. You can recalibrate but it doesn’t work as reliably from what I understand.
Ah, ok, but then wouldn't leaving them off entirely extend OTC even more? (or at least as much?)
So this isn't hypothetical. I'm considering making a couple of these up and don't mind making a fairly universal shopping list. Sounds like maybe I better leave the resistors in for now. Speaking of narsil your instructions say use attiny25 on 17mm and also use attiny25 for bistro. Can bistro work on attiny85?
And attiny85 can't fit on 17mm? I assumed the size was different but somewhere around page 204 of here:
It has not been tested but I assume this to be the case. The thing about leaving them off is that what you gain is worth the hassle of recalibrating ect since you free up a pin for other uses and/or are able to build the driver smaller (aka, the 15mm version).
If you will be using them anyways it is simply easier to use the proven values.
The 30mm looks ideal for the L6!! The LDO doesn't need any other components, resistor or cap? I'm not sure - thought there was in some designs before, maybe not though. Looks like the original wight LDO and MtnE LDO doesn't use any other components.
Note: There's no reason for Bistro not to work on a 85 - just compile it for the 85. The 25 is available in the same size package as a 13A, while the 85 is not. A few of us use the "bent pin" technique for mounting 85's on 13A pads - I've done lots of drivers successfully this way.
It does need an output and input cap, the input is combined with C1 (obviously untested if the 10uF we normally use is ideal or not but should work)
The output cap is handled by the larger C2.
R7 can be used for a small resistor to increase the ESR of C2 as specified in the spec sheet if you do not have the correct type of cap to start with.
None of these are tested and will most likely need a bit of tinkering to make perfect but based on what the datasheet and DEL said the designs above should work.
Wut? I got 10 or more 85 based 17 mm drivers in operation??
What is ESR? The C2 cap is the smaller one, not larger? I'm confused... The C2 is not DEL's C2, and not the OTC cap? Dunno - can't see the traces so can't see how it's wired up. So the C2 is dedicated just for the LDO?
According to TK bistro is actually compatible with a tiny85, she just hasn’t worked out the fuses for it, she said the ones Tom is using in narsil should work in theory though.
Far as mounting the 85 on the 17mm, it will indeed mount if you bend the pins, there is just not room for full sized 85 pads, it takes up the entire board if I put them on there.
I have no idea what ESR is, I have not taken the time to look it up yet. It has something to do with the specs on the capacitor and someone mentioned that you can use a normal capacitor with a resistor to get the same results. I am guessing it is some kind of internal resistance. Either way, the other pads were needed to jump traces so they are there regardless.
The picture with the traces should be live, I will have to check.
C2 is the same as normal from it’s position and purpose. It is just now pulling double duty as the output cap for the LDO as well. As such the size needs to be increased to ~10uf and since we use those as the C1 anyways it makes sense to use the same 10uF in both C1 and C2. This was suggested by DEL earlier in this thread IIRC.
The OTC is still in the same place but the name was switched to “switch” (forgot to note that) as these will mostly be used as e-switch drivers.
Ah, ok. Was confused because in the OSHPark listing is where the parts list is for these, and it says "C2: 0.1 uF", but your posts above says to increase it to 10 uF.
Silly me - ok, of course I can follow the tracing in post #381 -- ?? got confused
Sorry, I forgot to update that part of the oshpark listing, too hard to keep track of that and the thread here so I put the first focus here and update oshpark when I have time and think about it.
Yes, the hex file I currently have uploaded is for a 19.1k R1. I forgot to note that.
I stole TomE’s 22k calibration file but I have not actually moved it over to my Linux machine to use it yet.
I am waiting to update the firmware section of the OP till I finish the latest version of bistro, updates are being made in the TA firmware thread. Right now I am waiting on TK to have time to finish implementing the updates to compact the firmware and allow more space for mode groups ect.
As soon as I get that back I plan to put together the latest version with ~30 mode groups and update the OP.
I would like to create a “standard” firmware to go with these “standard” drivers so that they are interchangeable and easily worked with for many years.
TA, for some time I’m using BISTRO as my standard firmware in almost all my lights, and really enjoyed it. Can’t wait it’s new revised version to use with TA triple channel drivers.
I have high hopes for the combo of firmware and drivers, if for no other reason then they make my own life way easier and do what I personally need. Hopefully it fits the bill for some others as well.
Let me be clear though, I am simply building on what others have done before me and combining lots of work from different places into one single project.
ESR is equivalent series resistance, sot TA's description was exactly correct. Normally low is considered better, but I can see where it could be convenient for it to have some resistance built in, in some cases.
I see now what I missed in the spec sheet, the SSUR version of attiny25 comes in the S8S1 package which is not available for attiny85. Got it. Bending pins sounds like a bad idea for a first try at this.
