So why CAN'T the driver go in the tail ?

27 posts / 0 new
Last post
wle
wle's picture
Offline
Last seen: 13 hours 57 sec ago
Joined: 01/07/2015 - 13:49
Posts: 497
Location: atlanta ga
So why CAN'T the driver go in the tail ?

So why CAN’T the driver go in the tail ?

Someone on this thread: “The FW3A, a TLF BLF special, elegant, triple, powerful flashlight” ( http://budgetlightforum.com/node/54239 ) asked this question.

It seems like that would help with the heat-dissipation problem: the LED gets hot, and the driver gets hot, so why not separate them?

The same current has to travel from battery to driver, as from driver to LED, so that can’t be it…

It would make ‘active switches’ or ‘soft switches’ and lighted tail caps easier to implement..

What would make that not work?

wle

L4M4
L4M4's picture
Offline
Last seen: 1 hour 11 min ago
Joined: 12/08/2013 - 13:43
Posts: 1041
Location: Germany

Well, you would need very long cables between driver and the LED – plus a thicker tube with place for those cables – or if you make it flatflex, you still need that.
It makes it just more complex – for example, how do you change the battery? You will have some cables in the way, no matter where you open the case

The Miller
The Miller's picture
Online
Last seen: 8 min 8 sec ago
Joined: 12/14/2015 - 12:08
Posts: 9210
Location: Charente France

Yes it can be done
But wires need to be guided to the LEDs 2 drawbacks here, 1 the round rube nicely fits the round cells and resistance of longer wires

In case of drivers with some sort of thermal stepdown one want it close to the LEDs.

wle
wle's picture
Offline
Last seen: 13 hours 57 sec ago
Joined: 01/07/2015 - 13:49
Posts: 497
Location: atlanta ga

The Miller wrote:
Yes it can be done
But wires need to be guided to the LEDs 2 drawbacks here, 1 the round rube nicely fits the round cells and resistance of longer wires

In case of drivers with some sort of thermal stepdown one want it close to the LEDs.

OK, so one wire, or two (with thermal sensing).

Enderman
Enderman's picture
Offline
Last seen: 5 hours 6 min ago
Joined: 11/03/2016 - 22:42
Posts: 1805
Location: Vancouver, Canada

Not needing any wires going up the tube is the main reason, it simplifies manufacturing a ton.
Drivers typically produce very little heat, a few watts or tenths of a watt, so it’s not an issue to have it near the LED.

Even flashlights with charging in the tail cap choose to use a double ended battery (with both + and – on each end) to avoid any wires or other complications in the tube.

eebowler
eebowler's picture
Offline
Last seen: 1 day 17 hours ago
Joined: 08/13/2012 - 11:34
Posts: 1801
Location: Trinidad and Tobago GMT -4

The Imalent DM 21 has driver and USB charger in the tailcap without wires running through the body. It is possible. Driver, battery and switch are all in series, so it shouldn’t matt where these component are in the string.

My gratitude to those who are willing and able to help others (in whatever way you can)! Being human is more than just existing for yourself. Smile

emarkd
emarkd's picture
Offline
Last seen: 1 hour 3 min ago
Joined: 04/14/2015 - 22:04
Posts: 1371
Location: Georgia, USA
eebowler wrote:
Driver, battery and switch are all in series, so it shouldn’t matt where these component are in the string.

I’m not sure that’s true. They’re all in series along the negative current path, but the driver has to have access to Vcc too, just like the emitter. So that side of the circuit is parallel.

Enderman
Enderman's picture
Offline
Last seen: 5 hours 6 min ago
Joined: 11/03/2016 - 22:42
Posts: 1805
Location: Vancouver, Canada

eebowler wrote:
Driver, battery and switch are all in series, so it shouldn’t matt where these component are in the string.

Nope.
Most drivers have + and – contacts for the battery, then + and – wires for the LED.
Lexel
Lexel's picture
Offline
Last seen: 11 hours 49 min ago
Joined: 11/01/2016 - 08:00
Posts: 2407
Location: Germany

It is easily doable with 2 tubes for the driver positive and LED-, like Klarus build their lights

there is no need for the LED positive cable to go to the driver, as its connected with the battery +

It can be done without any cables or double tube, with a battery that provides positive and ground to the driver in the tail and use a p channel fet for the LED+ and battery ground for LED-

Ozythemandias
Offline
Last seen: 1 hour 49 min ago
Joined: 09/21/2016 - 08:33
Posts: 215

Or the HDS method with a wire in the body.

jimahi
Offline
Last seen: 8 hours 1 min ago
Joined: 10/17/2015 - 08:25
Posts: 51
Location: The Tropics

The Imalent DM21T does not have a wire going to the head. The driver (with brightness control knob) and usb charging are in the tail. It can be done without a wire running from driver to battery +.

The one thing the DM21T is missing is LVP. The driver does not have direct access to the battery voltage. LVP in their design would require a more complicated driver that approximates battery voltage by measuring current and knowing the Vf of the LED at that current.

Enderman
Enderman's picture
Offline
Last seen: 5 hours 6 min ago
Joined: 11/03/2016 - 22:42
Posts: 1805
Location: Vancouver, Canada

Lexel wrote:
there is no need for the LED positive cable to go to the driver, as its connected with the battery +

This is not true.
Maybe on some very basic drivers, but not the large majority.
Just take a look at the PCB, you can see it does not just go straight through.
The Miller
The Miller's picture
Online
Last seen: 8 min 8 sec ago
Joined: 12/14/2015 - 12:08
Posts: 9210
Location: Charente France

So you see it can be done, but all that do charge a hefty price for their lights

For Narsil and Bistro based drivers that allow temp stepdown using the MCU it still is best to have led and driver close together or a temp sensor needs to be installed, adding complexity in parts and assembly AND sacrificing a MCU pin for the read out.

kiriba-ru
kiriba-ru's picture
Offline
Last seen: 1 hour 1 min ago
Joined: 01/17/2016 - 02:34
Posts: 1094
Location: Russia

The Miller wrote:

For Narsil and Bistro based drivers

FET is not driver Grad

Meanwhile in Russia: https://twitter.com/CrazyinRussia

The Miller
The Miller's picture
Online
Last seen: 8 min 8 sec ago
Joined: 12/14/2015 - 12:08
Posts: 9210
Location: Charente France

kiriba-ru wrote:
The Miller wrote:

For Narsil and Bistro based drivers

FET is not driver Grad

???
I am talking about BLF developed firmware that has the option to use the temp sensor build into the MCU for user calibrated thermal stepdown vs using extra parts to measure temperature.
It is quite useful to have the MCU located in the area where it can actually sense temp Wink
kiriba-ru
kiriba-ru's picture
Offline
Last seen: 1 hour 1 min ago
Joined: 01/17/2016 - 02:34
Posts: 1094
Location: Russia

The Miller wrote:

I am talking about BLF developed firmware that has the option to use the temp sensor build into the MCU for user calibrated thermal stepdown vs using extra parts to measure temperature.
It is quite useful to have the MCU located in the area where it can actually sense temp Wink

You are talking about pretty names that some useless parts of firmware used to call here. Simple thermal resistor on pcb can do this job (thermal stepdown) better (much better) than mcu-integrated sensor (regardless what firmware do you use). There is no sense in applying so much efforts in firmaware which should manage so simple hardware.
FET is not driver. You can not garantee any value. Each of them will be always different. Simple things that are usual printed on brand light boxes – smth like runtime@100 lumens – cant be calculated, because fixed mods do not exist.

Meanwhile in Russia: https://twitter.com/CrazyinRussia

wle
wle's picture
Offline
Last seen: 13 hours 57 sec ago
Joined: 01/07/2015 - 13:49
Posts: 497
Location: atlanta ga

Good Discussion! Smile

About the “modern/fancy drivers need direct access to + voltage” argument, though …

Couldn’t there be a “leakage path” up at the LED, so the driver is never really without current?

Then the driver, or micro, or whatever you call it (the active electronics) will have some juice to run on, without a direct connect to the + terminal.

The Miller
The Miller's picture
Online
Last seen: 8 min 8 sec ago
Joined: 12/14/2015 - 12:08
Posts: 9210
Location: Charente France
kiriba-ru wrote:
The Miller wrote:
I am talking about BLF developed firmware that has the option to use the temp sensor build into the MCU for user calibrated thermal stepdown vs using extra parts to measure temperature. It is quite useful to have the MCU located in the area where it can actually sense temp Wink
You are talking about pretty names that some useless parts of firmware used to call here. Simple thermal resistor on pcb can do this job (thermal stepdown) better (much better) than mcu-integrated sensor (regardless what firmware do you use). There is no sense in applying so much efforts in firmaware which should manage so simple hardware. FET is not driver. You can not garantee any value. Each of them will be always different. Simple things that are usual printed on brand light boxes – smth like runtime@100 lumens – cant be calculated, because fixed mods do not exist.

Hm. Sounds a bit negative.
I am not talking about FETs just stating hoe Narsil and Bistro achieve thermal stepdown without the need for a thermistor.
Yes as I wrote a dedicated part for measuring can be done.
Yes as you write, there is a big difference in how each individual driver without it measures temp.
(I have read it is the MCU doing the measuring not the FET)
Hence the clever workaround, one selects thermal stepdown instead of the default timed stepdown and the light starts on turbo, one selects the desired stepdown temp and it doesn’t really matter that other same hardware drivers measure temp different for it is setup based on that individual unit.

So i don’t disagree with what you say, I merely explain why IMHO things are done how they are done in the lights I know a little bit more about Wink

kotobuki
Offline
Last seen: 6 hours 43 min ago
Joined: 10/17/2016 - 09:41
Posts: 144
Location: US

This Costco Duracell 700 lumens flashlight has driver circuit in a tail:

Batteries, LED and driver circuit are in serial, but it needs a 24 Ohm resistor parallels with LED to pre-feeds power for driver circuit. Circuit likes that will “eat” some more energy (~ 200mA) from batteries when turns ON (because of current goes through LED and resistor at same time).

pilotdog68
pilotdog68's picture
Offline
Last seen: 1 week 2 days ago
Joined: 05/30/2013 - 23:31
Posts: 6346
Location: Held against my will in IOWA, USA

Yes it can be done (at least with a linear driver). It is a DC circuit, current is flowing in one direction so Pos+ Voltage is available anywhere in the circuit without any extra wires. It is this principle that allows the illuminated tailcap to work.

So why isn’t it popular here?
1. It makes a bigger tailcap necessary, which can be really ugly and is not accommodated by many hosts.
2. Driver modding started with Qlite drivers (which have a Pos+ battery connector), and evolved from there.
3. See below:

jimahi wrote:
The driver does not have direct access to the battery voltage. LVP in their design would require a more complicated driver that approximates battery voltage by measuring current and knowing the Vf of the LED at that current.

My Favorite Modded Lights: X6R, S8 , X2R , M6, SP03

Major Projects:  Illuminated Tailcap, TripleDown/TripleStack Driver

warzi1975
Offline
Last seen: 6 hours 8 min ago
Joined: 08/16/2015 - 05:30
Posts: 53
Location: Munich, Bavaria

Ultrafire UF10 and UF9 have Drivers in the tail and they are my coolest 16340 lights. You can drill the tube to fit a 18350, I made it on three lights.

Robert

Enderman
Enderman's picture
Offline
Last seen: 5 hours 6 min ago
Joined: 11/03/2016 - 22:42
Posts: 1805
Location: Vancouver, Canada

pilotdog68 wrote:
Yes it can be done (at least with a linear driver). It is a DC circuit, current is flowing in one direction so Pos+ Voltage is available anywhere in the circuit without any extra wires. It is this principle that allows the illuminated tailcap to work.

An illuminated tail cap is literally just another LED in the circuit, that’s why it only needs to be in line to work.
For any kind of modern not-super-basic driver there needs to be separate positive and negative to both the driver and LED for it to work properly.
Stuff like LVP, battery indication, cc regulation, etc…

That stuff is not possible if you simply have battery->driver->led->battery.
In this case the driver would literally just be an in-line resistor, it would not have any other functionality.

EasyB
Offline
Last seen: 8 hours 49 min ago
Joined: 03/09/2016 - 15:24
Posts: 1167
Location: Ohio

With our drivers the battery positive does “pass through” to the LED, but it’s still connected to the driver. The MCU needs the connection to the positive to work.

Jerommel
Jerommel's picture
Online
Last seen: 13 min 1 sec ago
Joined: 01/04/2014 - 13:18
Posts: 4044
Location: the Hague, Netherlands

So it’s impossible to have a driver without +

Looking for:

5” parabolic reflector (for recoil light)

Enderman
Enderman's picture
Offline
Last seen: 5 hours 6 min ago
Joined: 11/03/2016 - 22:42
Posts: 1805
Location: Vancouver, Canada

Jerommel wrote:
So it’s impossible to have a driver without +

Technically you can, it would basically be the same thing as a resistor.
Maybe it would be possible to have it switch between different resistances too to change the brightness.

But for most drivers, the advanced kind we use in flashlights, it would not work (not properly, at least) if you connected + directly to the LED bypassing the driver.

pilotdog68
pilotdog68's picture
Offline
Last seen: 1 week 2 days ago
Joined: 05/30/2013 - 23:31
Posts: 6346
Location: Held against my will in IOWA, USA

Enderman wrote:
pilotdog68 wrote:
Yes it can be done (at least with a linear driver). It is a DC circuit, current is flowing in one direction so Pos+ Voltage is available anywhere in the circuit without any extra wires. It is this principle that allows the illuminated tailcap to work.

An illuminated tail cap is literally just another LED in the circuit, that’s why it only needs to be in line to work.
For any kind of modern not-super-basic driver there needs to be separate positive and negative to both the driver and LED for it to work properly.
Stuff like LVP, battery indication, cc regulation, etc…

That stuff is not possible if you simply have battery->driver->led->battery.
In this case the driver would literally just be an in-line resistor, it would not have any other functionality.


“just another LED in the circuit” still requires Pos+ and Neg- just like a full driver does.

As I quoted in my post LVP would be difficult to get accurate, and I agree with all of your Pos+ running throught LED the LED would have to be on for the driver to be powered, but this is currently true of clicky lights anyways. This configuration would not work for a an e-switch driver, granted.

I mentioned the illuminated tailcap because I was met with the same resistance with it “It’s not possible, it won’t work” even after it had been proven to work.

We already have existing lights with the driver in the tail and no additional wires, so it’s ignorant to still say “It’s impossible!”. It’s just a matter of “how much do you want it?”

My Favorite Modded Lights: X6R, S8 , X2R , M6, SP03

Major Projects:  Illuminated Tailcap, TripleDown/TripleStack Driver

Lexel
Lexel's picture
Offline
Last seen: 11 hours 49 min ago
Joined: 11/01/2016 - 08:00
Posts: 2407
Location: Germany

drivers in tail without double tube or wires
or special battery with both battery contacts to tail

the lights from Imalent have big drawbacks

first 33 Ohms bleeder resistor, jes 33 no 0 missing, so in lowest mode the bleeding current is on a normal Vf of the LED a bit over 100mA

2. slow PWM at 400Hz even at full brightness to power the driver in the off time
this pwm can be noticed by eye in multiple situations

3. LVP is almost impossible to do in the Off time, because it may be tricky to sense this while the inductor is draining power through the bleeder, to drive the MCU during on time