Anduril ... 2?

AVRDD support is now good enough for production I think.

that’s the usual way, or you can use the DAC of the T1616 and AVRDD, 8 and 10 bits respectively, but we use different VREFs, allowing to get 1:1159 dimming ratio on the T1616 (0.55V and 2.5Vref), and 1:2500 on AVRDD (1.024 and 2.5Vref).
with 2S you can use 5Vref (T1616) and 4.096Vref (DD) for higher dimming ratio, DD : 1023*4.096/1.024 = 1:4092 dimming ratio
I recommend going with DD if you want to use the DAC, the T1616’s DAC pulses on the level 1-2.

4 Thanks

Ah, so analog output of the DAC is supported as well - even better. Thanks for all the info, now I only need time… Lots of time :smiley:

Yeah, thermal stepdown doesn’t affect memorized_level (and in some cases you have target_level which is used where the intended level needs to be used and not the current level after thermal regulation.

I have a potential project in the works, can you go into more detail on how vref is used? That’s a side of it I don’t really understand yet.

@ToyKeeper posted a graph somewhere illustrating it, maybe she can post it again ?

But basically the DAC outputs the VREF chosen, for example on T1616, : 2.5V, down to 2.5V/255 at level 1/255, but if you use 0.55VREF, you can go lower down to 0.55V/255.
So you generate a ramp table that uses 0.55V first, and then switch to 2.5V for the top end.

// 4 ramp segments:
//   - low 0.55V
//   - low 2.5V
//   - high 0.55V
//   - high 2.5V
// PWM1: DAC Data
#define PWM1_LEVELS \
           2,  3,  4,  5,  7,  9, 11, 13, 16, 19, 23, 28, 33, 39, 45, 53, 61, 71, 81, 93,106,121,137,155,175,196,220,246, \
          60, 67, 74, 82, 91,100,110,121,133,146,159,174,190,207,224,244, \
          8,  9,  9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 20, 21, 23, 24, 26, 27, 29, 31, 33, 35, 37, 40, 42, 45, 47, 50, 53, 56, 59, 62, 66, 69, 73, 77, 81, 85, 90, 94, 99,104,109,114,120,126,132,138,144,151,158,165,173,180,188,196,205,214,223,232,242,252, \
         57, 60, 62, 65, 67, 70, 73, 76, 78, 82, 85, 88, 91, 95, 98,102,105,109,113,117,121,126,130,135,139,144,149,154,159,164,170,175,181,187,193,199,206,212,219,225,232,240,247,255
// PWM2: VREF selector
#define PWM2_LEVELS \
        V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05, \
        V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25, \
        V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05,V05, \
        V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25,V25

That’s one of my T1616 driver, additionallly there are 2 sense resistors (”low” and ”high”), so there are 4 segments in total.
First it goes up to 246/255 at 0.55Vref, then continues at 60/255 at 2.5Vref.
Then it chages sense resistor and do that again.
The resulting ramp is a smooth curve.

1 Thank

Is that specific to linear, or similar principle for buck drivers? I still don’t 100% understand why some build targets have an external voltage divider and some just use VCC.

This is for drivers that needs a voltage to set the current (it’s proportional to this voltage). That can be linear or buck/boost drivers, or any driver that uses analog dimming.

If you’re talking about the voltage divider for measuring the battery voltage that’s unrelated. Regarding that, the MCU can measure the VCC voltage with it’s internal voltage divider. But that’s only possible if VCC is the same as the battery voltage ( or offset, with a diode for example, VCC =Vbatt-0.3V). If VCC is regulated, for example from a LDO, or a boost converter (not main led boost converter), then an external divider must be used to bring the battery voltage down below the ADC VREF used (usually 1.1V for the T1616 builds).

For example, @ebastler driver project is 2S input, that’s too high for the MCU so he’s going to power it with a LDO, probably a 5V one so that he can use 4.096VREF for the DAC.
For reading the battery voltage the default ADC VREF is 1.024V with DD builds, he could use something like 300k and 39k divider so that 8.4V is divided down to 966mV, just below 1.024V ADC VREF.

3 Thanks

Pretty much what I’ve been thinking about, yes.

To me that’s very confusing, the fact that the 2C behavior is depending on invisible variables (temperature/voltage and thus a variable time window). Especially when the stepdown is small and you didn’t notice you are no longer at turbo. But yeah, if others don’t feel that way I can just modify it for myself. That target_level you mentioned does seem like it is exactly what I was looking for.

Also it’s not like re-turboing gets you much turbo time, on my light it starts ramping back down almost immediately. If i want turbo without regulation, it’s easy enough to use momentary turbo.

Indeed, that’s why I mentioned using 1C to turn off at the end, to go back to the memorized level. Can definitely do what I want, it’s just non-fixed time-sensitive states bug me.

1 Thank

Yeah, I completely understand where you’re coming from on this, it’s probably worth at least a discussion with TK. I’ll work on a more elegant solution when I find the time to too.

I made a merge request for your previous change, added it as the 3rd turbo style (3 clicks), also posted a build for if you want to use it with the current version because there’s a big backlog of requests right now: Anduril 2 feature change suggestions - #588 by wolfgirl42

2 Thanks

Hello.

I’ve been reading thru the threads related to anduril and I saw somewhere where Toykeeper mentioned the candle mode to have a max of about 4.5hrs timer. I’ve been enjoying candle mode since I’ve gotten a few TS10s and they’ve been relegated to full duty as candle lights in the bedrooms.

Any chance someone has a TS10 firmware with candlemode defaulted to have the max timer? Starting to get tiring setting timers on these TS10s everynight. Appreciate it. :slight_smile:

Also, since I updated my other lights to the latest build (Dec), I noticed some bugs in the TS11 and TS25 firmwares when the aux channels are enabled. Goes something like this:

  1. From On->9H and Enable (red) aux
  2. 3C to switch to red channel
  3. Switch Off. Then On->9H again.
  4. No buzz light follows to disable it back. And when going back to ramp mode, the main channel no longer works with a 3C. (bug?)

I updated all my lights to the new version 03/12/2023 and saw something weird but didn’t know what and yesterday I received a new one with the 2022 version.

I compared and noticed the following:

Factory version 2022:
Turns on instantly and turns off instantly.

New version 2023:
Turns on with almost half a second delay in progressive and turns off in progressive.

It is really annoying.

I press and it turns on quickly that is ideal.

But with the new version that does not happen, it is slow to turn on, it happens to me in all those updated to 2023.

Any way or setting to turn on instantly?

Thanks.

I believe this is in the Ramp Extras Config.
10H from ON
Release after 5th blink
Configure SMOOTH STEPS
0C to disable
1C to inable

2 Thanks

Ok, I’ll give it a try.

I think it should be the other way around if you want this mode set it but default to fast mode.

Because sometimes I want to turn on an instant to see and I want quick light and quick off.

Now it is slow on and slow off.

I’ll look for a version that doesn’t do that so I don’t have to configure anything.

It is very annoying this mode.

Thank you.

But this changes the ramp mode not just the ignition, it doesn’t work for me.

Because I like the smooth ramp mode so it goes up and down smoothly.

I’m just looking for instant on and off as usual.

Thanks but that setting doesn’t work for me.

It doesn’t affect smooth ramp at all. It just controls the fading for on/off and between steps.

1 Thank

Perfect! I will try it.

I think it is an additional option and he who needs it is the one who should configure it.

And by default fast power on.

I like to have one turned off and if I need it for a moment to have all the light when I press and when I turn it off it disappears, I use it a lot because I have it turned off and I turn it on and off to see something.

As it is now, personally for me, it’s annoying.

Although I can configure it, I prefer to use a previous version that does not have it, in case there is a blockage and I have to reset the flashlight.

Thank you.

Hi.

I just got my KR4s and now very confused with the info I got from Hank. So, just wanted to check here and tap on the knowledgebase.

Do these 2 firmwares have the option to enable the FETs in the KR4 during turbo?
anduril.2023-12-03.hank-noctigon-kr4.hex
anduril.2023-12-03.hank-noctigon-kr4-2ch.hex

anduril.2023-12-03.hank-noctigon-kr4.hex does (-nofet is the non-FET build)

anduril.2023-12-03.hank-noctigon-kr4-2ch.hex doesn’t (emisar-2ch-fet to enable FET on 2ch)

1 Thank

Thank you for the info, @wolfgirl42. Appreciate it.

The dual channel KR4 I ordered was suppose to have 519a in Ch1 and W1 in Ch2. Instead, I got the reverse of the setup. When I checked with Hank, he said he doesn’t distinguish channels in the orders unless specifically requested (although his site already specifies it in the drop list…ugh!) I should have checked with him.

  1. For this firmware, does this mean the upper 150 value of the ramp is still regulated? - anduril.2023-12-03.hank-noctigon-kr4-2ch.hex

  2. I have a DT8K dual channel incoming but, unfortunately, not FET-enabled as per Hank. This one has W2’s in Ch1 and SST20s in Ch2 (and I’m really hoping it’s not swapped again like my KR4.). Would you know if this can be reflashed with the FET version to enable FET in Ch1?

Thanks.

I see there’s a new “joined” 2 channel FET firmware for the d4s lighted switch? Neat! Thanks!
Would (should?) this be able to be flashed on existing d4s’? Or is it more like, going forward?

Sorry if this has been answered already somewhere and I missed it.