What? double blink every 8sec?
So I should change
to
Edit: or blinks the main LED?
You could. I'm incorporating all the LVP details into a "Advanced Notes and Usage" manual. Working on it now. Basically I want something available in greater detail, we can refer people to when they have some in depth tech questions.
Tom, is it okay to dump a problem here?
I’ve got a TA FET driver built by Lexel with NarsilM that has some problems only in the mode sets area and Lexel thinks it might be in the programming. This is a 2S driver.
Dang, lost my typed response...
There's no 2X, 3X, 4X fast clicks in mode set operation. It's described in the manual, in the OP link. Fast clicks would cause countless # of inadvertent actions in mode set operation -- the two methods don't get along.
Mode memory in mode set operations is "Not Recommended". I used to say that explicitly in the manual. I should have taken out that feature, really. It doesn't work well.
What you are seeing, I think, is the lock-in time. You must be in a mode a min length of time in order for it to be saved. If you don't hold in a mode long enough, it won't be remembered.
Hi Tom,
I'm about to try to stuff a filed down TA board into an olight S15. It fits but only single sided. So with the 3 channel TA board running as a 2 channel FET+1 will the 3channel setup.h work or do I need to eliminate the bank 7135s?
Could I just reassign the pwm pins in the 2channel setup to work with the TA layout?
Way back I asked TA if he could change the 3 channel mapping to I/O pins to match up to a 2 standard channel FET+1, but it didn't happen. If he would have put the FET on pin #6 and the 7135 bank on pin #3, it would have been no problem.
So for what you got, pin #5 is the single 7135, pin #3 is the FET, I believe... I think this is the same thing Lexel asked just up above?
Even in my small OTF M3 16340 light, I was able to use a TA 15 mm board, using 3 7135's in the bank, but kept the original driver as a contact board, and mounted a filed down TA 15 mm slanted, at a angle.
I haven't tried doing what you are asking, and it's certainly possible with a little tweaking. It would be much easier now in NarsilM, since NarsilM has much more flexibility from the prior Narsil versions. I don't think Lexel knows code/programming enough to make and test the changes, though he knows a lot about the MCU and hardware designs, etc. I didn't hear back from him bout what I posted above, so I guess it didn't work out.
Boy, I could get this working, and I'm sure it will help out more than just you, and I can even find usage for it, I'm sure. Problem is time, with so much going on right now with the Q8 and GT projects, plus other stuff on very limited time availability.
How about if my next weekend, I can get to this, maybe have a tested solution. Do you have the capability to re-program the board, download a HEX file?
Hey Tom I just realized I never checked back in after this bit of advice. Just wanted to say that changing R1/R2 to these values (or very close, my set didn’t have exact matches) seems to have fixed mine right up. Works great now. Thanks again for all your help.
Cool! Glad to hear!! 
Okay, so when in mode sets you loose instant turbo (2x), batt check (3x), lock out (4x) and all strobes/beacons etc…(2×2)?
That is a big loss, I didn’t know that.
And you say it’s normal for mode set memory to be glitchy as well? Momma mia! Lol
Do you have any idea why in mode 8 I only have 2 of the 3 levels or why I have to press the side button several times just to get the light to turn on?
No, did you read the manual? Tells you how to access all that stuff in mode set operation - I ain't repeating it here 
Ohh - the mode set #8 problem... I can't see any problem there, least in the code. I'll test it on a Q8 proto I have here w/NarsilM later, though it's not the same options you have, Q8 is 1S, not 2S. I don't have a NarsilM v1.0 2S light with me at work.
For the issue of several clicks to turn the light on, I dunno. I have to re-read your description.
Are you certain it's NarsilM v1.0 you are running, that Lexel setup for you? I'm confused about all this...
Edit: Ooops!!
Ok, I see the problem with mode set #8: a 3 channel configuration uses a different set of modes. Only 8 sets of modes are defined, not 12, and mode set #8 should be 2 modes only:
- max of all 7135's (single and bank channels)
- Max FET
So what you are seeing is correct for what it's defined as. I never got around to filling up all compatible 12 sets of modes for triple channel configurations. I'll need to document it this way some where. I'll update the OP for now to mention it.
Also, to explain the modes:
- mode percentages are roughly based on 1,500 lumens being 100%, typical XM-L2 type of LED, near full battery, moderate resistance reductions
- The single 7135 is assumed to be 150 lumens which is 10%
- if I rated a mode at like 40%, I was assuming 30% of the FET is used, and the 7135 at max (30% + 10% is 40%)
- 2S setups are roughly double that. For example on my MT03 running 2S and triple XHP70.2's, one 7135 at max is about 320 lumens. A single XHP70 should be roughly 300 lumens.
- of course the modes will work with any LED compatible to the driver, so an XP-E2 for example might max out at 600 lumens on the full FET, while a 4X XP-L2 SRK is ~6,000 lumens max, so the percentages kind of lose their meaning. This is the problem of rating output percentages based on 1 or 2 channels that are fixed in amps, and another channel that is basically wide open, where output depends on LED's, batteries, driver, and resistance in the full circuit.
Manual? I don’t need no stinkin manual. Lol jk
Okay, so I see that in mode sets you can still access batt check and lock out, but the procedure is different. (I can’t get strobes, but I may have strobes turned off)
Wow, it’s like 2 completely different firmwares loaded onto the same chip.
How are things looking for the Q8? Are the mode sets working well and spaced evenly? They are not spaced evenly on my 7k+ lumen light. That’s okay though. I plan to make a user video then go back to ramping mode. I like it a lot more.
OK, all good. Actually just played around in the 4 modes set (plus moon), 5 total, and with mode memory turned on and it seemed to be working pretty good, consistent functionality.
The Q8 was locked down to NarsilM v1.0 a while back. Now I can't recall, but I did bring it up that the mode sets could be tailored/tweaked in both settings and documentation to match up with a 5,000 lumens light, as the Q8 is, roughly. I don't think that ever happened though, so unfortunately it is what it is now. Triple channel mode sets are different from 2 channel mode sets, at least in the firmware, so I'm not sure how it would compare with your set up.
One nice feature added in to NarsilM was a "Reset to factory defaults" ability, so if you think the settings are all messed up, you can return back to a known good configuration: get to displaying the revision # of the firmware (in ramping, 3X click to battery status, 2X click to temperature reading, 2X click to firmware version), then press& hold til the light blinks -- this comes in handy 
Please go to http://bit.ly/narsil-cs
load the newest PDF. In the first rightmost leftmost Table you see clearly the different klick comands from RAMPING and MODES. ONLY ON and OFF is the same.
What are you showing me here? These look like the the copies I already have. Did you misread that I was only joking about not having a manual? I have manuals for NarsilM and V1.1 which I think covers V1.2. I couldn’t find one that said V1.2.
Time is always a bottleneck, but it sounds like you’ve been particularly overloaded lately. How have you been doing? Is there anything you’d like help with? I’ve been trying to get in touch, but I don’t want to take away time you don’t have.
Wait, is the Q8 a 2 channel driver? I was thinking it was 3 channel for some reason.
EDIT: I scanned post 1 on the Q8 thread, but didn’t see it mentioned. Looking at the driver board I only see one 7135. I guess that means it’s 2 channel.
Yes, the Q8 uses two channels. They came last year with a way to free a pin for a third channel.
Even the Emisar D4 is a dual channel.
The FW3A will be a tripple
JasonWW
Yes I missed you were joking.
If you need the cheat sheet for Narsil v1.2 please PM me your email address.
Dang, yes, saw your pm, but didn't get a chance to read thru it. Still have over 45 unread pm's.
We do need to collaborate. Man, if you got the time, it would be great. Thanks for bailing me out on the D4, btw - I got overwhelmed, reached a point were I couldn't get anything done because I had so much to do, and other issues, limiting my workshop access. I guess I had to turn off BLF for a while to keep my sanity, I dunno, but also my regular consulting work, single client, has gone ballistic with pressure on projects over the last couple of months and will continue for a while.
I also know you had some really good suggestions for Narsil, think maybe early in this thread.
The Q8 is pretty much a done deal at this point - ThorFire is making 2,000 pieces, so it's the full batch, all with NarsilM v1.0, but trying to write up an advanced guide now, plus on-going support issues.
At this point, I forget or dunno where we are with the GT, it may be locked down now, not sure, Lumintop isn't communicating much so we don't know exactly where they are, but they are in a rush apparently. I thought one more single updated GT proto would be sent out, but not sure. We are at least asking for pics and clarifications over the battery carriers. Might be able to sneak in an update to NarsilM, but dunno.
The GT buck driver is kind of a whole new world - true regulating buck, no PWM output on the LED, but using PWM's to control amps, and can do it quickly and smoothly, so full smooth ramping works well -- it's bout as ideal as you can get. Our main problem is the voltage supplied by the cells is not higher enough than the draw of the LED, so, it will fall out of regulation if driven at high amps. At 2 amps, it can run long in regulation, but at 2.5A, it's getting dependent on voltage drops from the springs, circuit, carriers, etc. The XHP35's Vf is way high -- it's not based on the new generation LED's that have a lower Vf, but rather uses high Vf XP-E2's. DEL did a fantastic job though updating Narsil in implement the ramping and mode sets, and other changes for supporting the two PWM outputs. All I did was merge it into the NarsilM header structures so we, again, have one code base supporting 1, 2, and 3 channel 7135/FET outputs, as well as the GT buck driver as a "special" 2 channel option.
So, in the BETA NarsilM, I added support for a max ramping level, separate from the max output. so we have the GT configured for 2 amps max in ramping, and ~2.5A turbo (2X click), and 6 mode sets set for 2.0 amps max, and 6 mode sets set for 2.5 amps max.
Oh boy, sorry we didn't make that clear. Not sure how this all originated, but think the Q8 project was started just before, or about the time the triple channels started catching on. At the time, I probably didn't have the support for 3 channels and a switch LED, but now I do, least for a 1S battery configuration.
The good thing about all this is that DEL really engineered a much improved design to support the ATtiny 25/45/85's, which TA and HQ incorporated into their driver designs, so something really useful came out of this effort, mainly for the Q8, early on. Though TF dragged their feet along the way, we always though we were just couple weeks away, and didn't think an upgrade of the driver was possible without delaying the project. Little did we even think it was possible it would be 18 months in the making... 
