Emisar D4S review

well, the protected 26650 in my Rofis MR70 will not fit in my D4S (too long)
butt I can charge the Shockli IMR26650 5500mAh from RMM in the MR70 just fine
and I can use either in the MR70 to charge anything using USB for recharge
very handy and useful

I can’t speak for vendors in other countries, but after doing some Googling it looks like all the U.S.-based vendors are out of stock, and most list a 6 week lead time.

This might be a good thing for intl-outdoor or Mountain Electronics to stock…? I’d almost certainly buy one from them (assuming the cost of shipping weren’t $8 like Mouser. :weary: )

this D4S throws just as far as my Klarus XT11GTXHP 35 Hi! I’m new here so forgive my ignorance. I showed a buddy at work my newly acquired D4S and he instantly bought one yesterday from mtn! Toykeeper, I could see the smooth ramp function on this light alone selling Big to the general public! My wife’s mom thought it was.. and I quote, “a really nice mood light” lol.

Now the momentary tactical mode In my opinion is perfect! I love that it cannot exit this mode without untwisting the tailcap for a set amount of time. This is truly a re-invention of the modern day LED flashlight. This light is actually fun to operate and learn all the modes too with the flow chart!

It’s a better mood light with Anduril installed. That adds, among other things, candle mode. Works well with a traffic wand or something over the bezel.

Click the Link in my signature for more information.

Also, for the pinouts, this might help. I refer to it while figuring out where to poke wires through my D4S driver.

http://flashlightwiki.com/AVR_Drivers#Set_up_your_clip

@ToyKeeper
How do you feel about these arguments for an estimate of levels?

./level_calc.py 2 150 7135 1 0.5 400 FET 1 10 4300

Yes.


It’s a bit primitive, but it works and it uses parts which are easy to get.

You should be able to check the accuracy by comparing known levels to what the computed ramp says. For example, level 85 is the 3x7135 mode and should be about 400 lm with XP-L HI emitters.

I’d suggest checking a few points along the curve, like 1, 40, 85, 120, and 150. Then adjust the parameters until all of those line up with the expected values.

Because the calculator and the actual light won’t necessarily agree, it may be necessary to split the ramp into two parts and calculate them separately. Probably levels 1 to 85 in one segment, and 86 to 150 in another segment. It’s a little tricky getting the calculated shape to match, which is why I haven’t already made a reference chart.

Additionally, the 219c curve will be different because all its FET values are multiplied by 0.8. This is because Hank was getting complaints about 219c emitters burning themselves out on turbo, so the 219c “turbo” is 80% power.

I’ll try this.

How are the discrete modes in the actual firmware calculated? Sorry if I missed that, if it’s already written somewhere.

They’re evenly-spaced along the ramp. Floor at one end, ceiling at the other end, and “num steps - 2” levels between.

Because a few people have asked, I added a stepped ramp calculator called “steps.py” in the anduril directory. Here is an example of its usage:

> ./steps.py 1 150 10
1: 1
2: 17
3: 34
4: 50
5: 67
6: 83
7: 100
8: 116
9: 133
10: 150

Or, using the D4S default values, here is the result:

> ./steps.py 20 120 7
1: 20
2: 36
3: 53
4: 70
5: 86
6: 103
7: 120

Another solution until the pogo pins arrive: just use a piece of breadboard with some crooked pin headers as the programming vias are not aligned on a 100mil grid.
It may not be the best solution, but it’s faster than connecting all 6 pins seperate each time.

Adapter for the D4S is on the left, Clip for D4 / Q8 is on the right.

Also, if you don’t have an AVR ISP but own an arduino you can program the arduino to act as an ISP.

I love the timer feature on the Olight S1R. If Hank could give us a diffused side light, motion sensor and timer in the D4GT I’d have his baby.

Received a grey Nichia D4S today from Intl-outdoor - ordered August 3 (20 days ago).

Anyone got a pic of the magnet tailcap?

I couldn’t wait either, my XP-L 5000k came today and I got 4650 lumens at 4.10v with a Shockli on the KG-Lightbox, the shockli are taking hours to fully charge.

My ROT66 came too, faulty.

Bummer! What’s wrong with it?

One of the LED’s isn’t working. The Nichia ROT66 is fine though.

Ok. Starting with the basics, assuming everything has to be bought.

1. Buying the hardware

First you need the programmer itself, an USB ASP Programmer.
The original can be found here USBasp - USB programmer for Atmel AVR controllers - fischl.de while you can find it cheaper on Aliexpress or on Amazon
For the D4 or other boards without the nice connectors, you will need an SOIC8 programming clip to connect to the chip

Also, to get nearly any kind of connection, I suggest a set of Dupoint connectors (male-male, male-female, female-female)
You will use this cable to make connections between the Pins of the programmer and either the clip or the D4S programming pads.

2. Connections
Now lets look at the available connections, with an example for the D4S.
You have the programmer

And you have the D4S

Now you take the female-male dupont cables and connect it the following way:

USBasp -> D4S
GND (any) to V- (top left)
MOSI to MO
RST to RS
SCK to SC
MISO to MI
VCC to V+ (bottom right)

The result may look like this

(for the BLF Q8 ,D4 or so, you do the same thing but with the connector of the clip. But that’s out of scope for now)

3. Flashing itself

Get the hex-file from here: Index of /torches/fsm - take the latest date of anduril for your light
Tutorial for avrdude is here: AVR Tutorial - AVRDUDE

Short version of the tutorial:

  • Download hex from toykeeper.net
  • Download avrdude
  • Run this command: avrdude.exe -p attiny85 -c usbasp -Uflash:w:anduril.YYYY-MM-DD.EMISAR_D4S.hex:a

Cells were fully charged and I have 2 of each cell.
Tests done in the KG-Lightbox.

Shockli 5500 mah: 4600 lumens
AWT 5200 mah: 4650 lumens
(D4 XP-L with VTC5A: 4300)

Very tough to call as it’s around 4600 for a split second. The Shockli both came on at 4400 at 4.20v though which was weird, they perform a bit better around 4.10-4.15v

A quick ceiling bounce with the ceiling bounce app calibrated to read 4300 off an Emisar D4 XP-L yeilded the same results:
D4 XP-L 4300
D4S XP-L 4600

Can anyone commend on the Neutral White - XP-G2 S4 5D, 4000K tint and the current draw on full power of the XP-G2 in general? How about tint shift and amount of yellow/green/red? Thinking about getting it as a warmer addition to the XP-L 5000K.