comfychair-inspired quick+cheap+lazy single-sided 17DD FET-driver (poor man's nanjg92)

305 posts / 0 new
Last post
luminarium iaculator
luminarium iaculator's picture
Offline
Last seen: 1 hour 35 min ago
Joined: 07/07/2013 - 09:09
Posts: 2195
Location: X

New potting method:

This can survive almost anything.

djozz
djozz's picture
Online
Last seen: 13 sec ago
Joined: 09/07/2012 - 17:04
Posts: 18090
Location: Amsterdam

Looks neat.

Just wondering: if that stuff is kind of stiff and has some weight and it does not completely fit the pill-cavity from side to side, is it not doing the opposite what is the intention?: applying extra stress on the components on impact?

luminarium iaculator
luminarium iaculator's picture
Offline
Last seen: 1 hour 35 min ago
Joined: 07/07/2013 - 09:09
Posts: 2195
Location: X

That stuff is very solid(plastic solid) but very light(lighter than fujik) but it completely fills pill cavity from side to side.

That stuff melts together with all components for permanent bond and it is non conductive.

So it is permanent stress relief for all components on the driver. Components and that stuff are now one chunk Smile

Agro
Agro's picture
Online
Last seen: 2 min 21 sec ago
Joined: 05/14/2017 - 11:16
Posts: 6868
Location: Ślōnsk

Will you tell us what is this stuff?

LightRider
LightRider's picture
Offline
Last seen: 3 years 7 months ago
Joined: 08/05/2015 - 09:52
Posts: 2007
Location: U.P. MI, USA

My results look similar and are not permanent. Though it’s not necessarily fun to have to dig back into the stuff. The texture of mine is like a hard rubber with a fine grit. I guess that should be obvious though. I use silicon carbide powder 400grt, 2000grt alumina oxide, sensors safe high temp copper permatex, a drop of silicone oil and a squirt of thermal paste just for kicks. Smile

There are a number of tricks you learn to get it to pack in just right but it’s been working well for me. Not sure how it would hold up to li’s method though…

luminarium iaculator
luminarium iaculator's picture
Offline
Last seen: 1 hour 35 min ago
Joined: 07/07/2013 - 09:09
Posts: 2195
Location: X
Agro wrote:
Will you tell us what is this stuff?

It is 5 minutes epoxy similar to JB weld but it hardens in 5 minutes. So yes this stuff in picture is done in 5 minutes Smile
Abro steel epoxy. Permanently. No repair here like with fujik or above mentioned stuff. looks similar to this one(athough my package bought locally looks bit different):

I used all kind of epoxies and probably 2 most famous ones(uhu endfest 300 and JB Weld), but this Abro one bought me Smile Done in 5 minutes!

Edit:

How can I rework this driver to single mode only(high or turbo only)? But I want of course that reverse polarity protection works after mod. Pretty please guys…

Hellie112
Hellie112's picture
Offline
Last seen: 1 day 9 hours ago
Joined: 05/18/2013 - 08:48
Posts: 110
Location: The Netherlands

Does anybody have working firmware?

The OCT firmware uses a different schematic for the NANJG 105C

Like this:

pb

All mine NANJG 105C drivers have the pb1 connected to the 7135 chips.
And the OCT firmware uses pb0 for the 7135 chips..

Cereal_killer
Cereal_killer's picture
Offline
Last seen: 1 year 11 months ago
Joined: 07/22/2013 - 13:10
Posts: 4005
Location: Ohio

You can compile the code for whatever pinout you need however the easiest thing to do would just be to physically jumper(short with solder) pins 5 and 6 on the attiny.

 RIP  SPC Joey Riley, KIA 11/24/14. Now I am become death, the destroyer of worlds.

Hellie112
Hellie112's picture
Offline
Last seen: 1 day 9 hours ago
Joined: 05/18/2013 - 08:48
Posts: 110
Location: The Netherlands

But when I short the two pins, I can’t use the 7135 for a moonlight/ low mode. I really looking for someone who have modified the firmware to work. I’m still trying to get it to work by myself by changing the firmware (and testing with trial and error because I don’t have any programming skills :p)

Also the screenshot for the minimalist driver posted by Major (in beginning of this thread) shows only half of the required code. So I can’t use that either.

Cereal_killer
Cereal_killer's picture
Offline
Last seen: 1 year 11 months ago
Joined: 07/22/2013 - 13:10
Posts: 4005
Location: Ohio

So you’re trying to run it as FET+1?
Is the pinout you posted above what you’re actually using? If not can you post that?

ATTiny13A I assume? BLF A6 firmware will work for that configuration.

 RIP  SPC Joey Riley, KIA 11/24/14. Now I am become death, the destroyer of worlds.

Hellie112
Hellie112's picture
Offline
Last seen: 1 day 9 hours ago
Joined: 05/18/2013 - 08:48
Posts: 110
Location: The Netherlands

Yes FET+1 (with ATTiny13A Smile )
Someone else in this thread posted this drawing and it’s actually the same as the drivers I use. So I presume I need to program that the FET uses pb0 as output.

Cereal_killer
Cereal_killer's picture
Offline
Last seen: 1 year 11 months ago
Joined: 07/22/2013 - 13:10
Posts: 4005
Location: Ohio

 RIP  SPC Joey Riley, KIA 11/24/14. Now I am become death, the destroyer of worlds.

Hellie112
Hellie112's picture
Offline
Last seen: 1 day 9 hours ago
Joined: 05/18/2013 - 08:48
Posts: 110
Location: The Netherlands

I could only flash blf-a6-13-rmm.hex (but it also uses the wrong chip layout). The other hex files are also too big to flash for an atiny13a. When I try to compile it, I have to delete a lot of code and then it’s still 3% too big..

Maybe I can edit pilotdog68’s xintd-x3 firmware, it’s within the attiny13a size range and I don’t need to include other files..

Cereal_killer
Cereal_killer's picture
Offline
Last seen: 1 year 11 months ago
Joined: 07/22/2013 - 13:10
Posts: 4005
Location: Ohio

Do you have optimization set for size (-0’s)?

 RIP  SPC Joey Riley, KIA 11/24/14. Now I am become death, the destroyer of worlds.

Hellie112
Hellie112's picture
Offline
Last seen: 1 day 9 hours ago
Joined: 05/18/2013 - 08:48
Posts: 110
Location: The Netherlands

Nope, don’t know anything about that..

But I experimented a bit further with the xintd-x3 firmware and I think I have solved my problem.
By changing the input at the following fields: define PWM_PIN, define ALT_PWM_PIN, define PWM_LVL, ALT_PWM_LVL. By swapping the pb0 & pb1 and again swapping OCR0A & OCR0B

Cereal_killer
Cereal_killer's picture
Offline
Last seen: 1 year 11 months ago
Joined: 07/22/2013 - 13:10
Posts: 4005
Location: Ohio

Most all of our BLF firmwares require optimization settings for size.

Its a preference setting in atmel studio that effects how the code compiles. I don’t know enough to know what it actually does or explain it, just enough to know what to set it to…

 RIP  SPC Joey Riley, KIA 11/24/14. Now I am become death, the destroyer of worlds.

Hellie112
Hellie112's picture
Offline
Last seen: 1 day 9 hours ago
Joined: 05/18/2013 - 08:48
Posts: 110
Location: The Netherlands

Oke I will look into it. But for my regular Nanjg 105C drivers I still use the luxdrv (really easy to personalize and has everything I need).

Is it normal I only get 3,4 amps on the max setting? I used the PH4030AL FET in combination with a Sony / Murata Konion US18650VTC6 3120mAh – 30A cell. With cheap cells I only get 2,2 amp..
Also replaced the leads with some ticker ones, without any effect. Wil try a spring bypass just to be sure.

luminarium iaculator
luminarium iaculator's picture
Offline
Last seen: 1 hour 35 min ago
Joined: 07/07/2013 - 09:09
Posts: 2195
Location: X

Thread resurrection needed…

Djozz and other fellow members who may know this…

Is there any way to restrict current draw to not more than 5A in mentioned diy FET driver?

djozz
djozz's picture
Online
Last seen: 13 sec ago
Joined: 09/07/2012 - 17:04
Posts: 18090
Location: Amsterdam

luminarium iaculator wrote:
Thread resurrection needed…

Djozz and other fellow members who may know this…

Is there any way to restrict current draw to not more than 5A in mentioned diy FET driver?


No, not if you want a 5A current limit regardless of led voltage or battery type/charge state. You could add a resistor in the circuit (maybe in the form of thinner longer ledwires) but that would limit current at all modes, not just over 5A, it does not regulate.

What you hope for is current regulation and that is what the original 7135 chips on the original AK-47 driver did (3 chips for 1.05A), and for 5A you need 14 of them Shocked . Similar current regulator chips that regulate more than 350mA do not exist (maybe because of overheat problems in such a small sot23 package) or are not suitable in this application.

The way to go for you may be using a different type of FET’s that can regulate the current instead of just switch on and off. Led4power’s drivers are like that but are not as cheap as these NANJG drivers. A relatively new solution that you may check out are the 5A Convoy drivers, https://nl.aliexpress.com/item/33007449093.html?spm=a2g0z.12010612.81483... , they are 17mm, regulated at 5A, have 4 simple modes and memory, and have lightweight low profile components that just may survive rifle recoil pretty well. And they are cheap, 5 drivers for 16 dollar.

luminarium iaculator
luminarium iaculator's picture
Offline
Last seen: 1 hour 35 min ago
Joined: 07/07/2013 - 09:09
Posts: 2195
Location: X

Thanks Djozz,

I really hoped there is a way to restrict current on this driver for certain applications like OSRAM leds cause I have hundreds of them(ak47c1) and since they have proper mode order for mine application(c1 firmware, it always starts on high mode).

While thinner wires are not suitable for mine build Resistor in a circuit sounds good but I really don’t know where to place them and what resistors should I buy? I hoped you could show that? Wink
With FET drivers we use high current cells like INR30Q so this Djozz FET driver should be modded so it can’t pull more than 5A of current in a setup with INR30Q and Osram WF emitter. Probably with proper resistor this is doable?

Convoy drivers unfortunately does not have suitable mode order for hunting purpose low>high. And I really don’t need regulation for 1×18650 cell setup. Short burst of turbo is all I need.

Jerommel
Jerommel's picture
Offline
Last seen: 1 year 6 months ago
Joined: 01/04/2014 - 13:18
Posts: 6380
Location: the Hague, Netherlands

For guns / rifles you need a fat rubber or even plastic ring around the battery spring, as a bump stop, like in car suspension (for the same reason)
It will prevent spring crush and PCB impact.

Yes, luminarium iaculator, get that driver from Simon, for Pete’s sake. Wink
Doesn’t it have mode memory?

luminarium iaculator
luminarium iaculator's picture
Offline
Last seen: 1 hour 35 min ago
Joined: 07/07/2013 - 09:09
Posts: 2195
Location: X

Simple spring bypass with 18AWG wire acts like a bump stop and perfectly presses flat top battery to Djozz FET driver spring(which is also bypassed with 22awg wire). Holds heavy centerfire magnum recoil without any issues.

I would order from Simon but I don’t like mode order.

And I still hope that there is some kind of resistor that can be added to this Djozz FET driver which would restrict current draw to not more than 5A with INR 30Q and OSram WF

Jerommel
Jerommel's picture
Offline
Last seen: 1 year 6 months ago
Joined: 01/04/2014 - 13:18
Posts: 6380
Location: the Hague, Netherlands

luminarium iaculator wrote:
Simple spring bypass with 18AWG wire acts like a bump stop and perfectly presses flat top battery to Djozz FET driver spring(which is also bypassed with 22awg wire). Holds heavy centerfire magnum recoil without any issues.

I would order from Simon but I don’t like mode order.


Doesn’t it have mode memory?

A tough rubber ring around the spring moves the force from the center of the driver PCB to the rim, so it won’t bend or break.
It’s a simple mod.
You can cut a ring from rubber hose, or even garden hose or something similar.
It will also reduce the impact on the battery itself.

luminarium iaculator
luminarium iaculator's picture
Offline
Last seen: 1 hour 35 min ago
Joined: 07/07/2013 - 09:09
Posts: 2195
Location: X

Thanks for rubber ring idea but in mine mode there is no impact on battery at all… Once when you put the bat it stays firmly in place.
During few years I done hundreds of them without any issue. No need for rubber :), although we really do know that proper rubber is really needed for certain applications Beer

Jerommel
Jerommel's picture
Offline
Last seen: 1 year 6 months ago
Joined: 01/04/2014 - 13:18
Posts: 6380
Location: the Hague, Netherlands

luminarium iaculator wrote:
Thanks for rubber ring idea but in mine mode there is no impact on battery at all… Once when you put the bat it stays firmly in place.
During few years I done hundreds of them without any issue. No need for rubber :), although we really do know that proper rubber is really needed for certain applications Beer
I saw Djozz you mention recoil hunting, so i thought you used lights on rifles.

(edited)

luminarium iaculator
luminarium iaculator's picture
Offline
Last seen: 1 hour 35 min ago
Joined: 07/07/2013 - 09:09
Posts: 2195
Location: X

luminarium iaculator wrote:
Simple spring bypass with 18AWG wire acts like a bump stop and perfectly presses flat top battery to Djozz FET driver spring(which is also bypassed with 22awg wire). Holds heavy centerfire magnum recoil without any issues.

I would order from Simon but I don’t like mode order.

And I still hope that there is some kind of resistor that can be added to this Djozz FET driver which would restrict current draw to not more than 5A with INR 30Q and OSram WF

Jerommel,
Magnum centerfires are rifles. Magnum means badass rifles LOL
Lets stick to my primary task on this thread. Maybe you could help?

Do you know what resistor should I buy and where should I place it in a circuit of this Djozz FET driver? Pretty please?

Edit: Simon drivers has mode memory and is OFF time based which I don’t prefer. Old grandpa hunters likes simplicity and any kind of complicated UI is out of a question.

Jerommel
Jerommel's picture
Offline
Last seen: 1 year 6 months ago
Joined: 01/04/2014 - 13:18
Posts: 6380
Location: the Hague, Netherlands

Oh, i see.
We here in EU only have cutlery for weapons, so i’m not familiar with gun / rifle lingo.

You mean it will change modes when you use it again within seconds, i guess.
No, i wouldn’t know what value resistor you would need.
Also, i assume you want to use the 1mm² white flat?
You should know it’s not worth driving it on 5 Amperes.
It maxes out on 4.5 Amperes and the difference between 4.0 and 4.5 Amperes is quite negligible.

I see 2 possibilities to best solve your question:

1] Don’t use fully charged cells, and forget about resistors and thin wires and crappy springs and what not.
You’ll have much more maximum output runtime when you start with 3.8 Volts cells, the discharge curve of the battery is much less steep under 3.8 Volts.

2] Get a Nanjg driver with 8× 7135 and the four stars to choose groups, and stack 3 or 4 more 7135 on top of the original 7135’s.
It’s not that hard if you have a fine soldering tip and some flux grease.
The trick is to
a) remove the middle leg
b) bend down the 2 legs
c) stick them on top of the original 7135’s with a tiny drop of super glue
(the bent down legs will help positioning them, but you can also first glue them in place and then bend down the legs, but then the glue may break loose))
d) solder bridge them to the 2 outer legs of the originals
e) bend down the thermal tab (which is also the middle leg)
f) solder thermal tab to the original’s thermal tab

The second possibility gives you safely regulated current for a considerable time.
Use fairly high drain cells (30T or VTC6 or similar) and optimize the electrical path to prolong regulated current drive.
You eliminate many unknown / inconsistent variables with 11 or 12× 7135.
You will never over drive the LED.

(I assume you don’t use a remote switch, with long thin wires.)

luminarium iaculator
luminarium iaculator's picture
Offline
Last seen: 1 hour 35 min ago
Joined: 07/07/2013 - 09:09
Posts: 2195
Location: X

Jerommel,

Thanks on a help but please tell me something I don’t know! Facepalm Smile I have asked you simple question: “Do you know what resistor should I buy and where should I place it in a circuit of this Djozz FET driver?” If you know ok if you don’t know that is also fine… That was the question that demanded answer. It is ok if you don’t know that Innocent

I sold over 1000 modded lights in mine life. I have a master degree on soldering AMC 7135 and 7138 chips to linear drivers and you really don’t need to cut middle leg. I solder all 3 of them even without bending. I know what are my potential solutions but I refuse them since I have hundreds of AK 47 C1 drivers and if there is possibility for current restriction than I will try to mod them.

It is the most stupid thing to stack so many amc chips on small linear drivers cause it heats a lot and failure will obviously happen sooner or later. FET driver is much much better than any high current 17mm linear driver with hundreds of stacked 7135 chips cause it does not heat that much at all…

I also know how FET drivers work since I made hundreds of them this diy Djozz type and Texas ace one (without stencil; just soldering iron and calm hand).

If you want to run FET with Osram WF just use low current 18650 cell at any charge state and don’t bypass springs and that is it. Or add thick steel contact springs for example.

But this is not solution for the problem of how to restrict current in FET driver.

Potential solution in this below mention style but I would like to take that to higher level and make restriction on driver only… EasyB!

http://budgetlightforum.com/comment/1458705#comment-1458705

http://budgetlightforum.com/comment/1443460#comment-1443460

So any advice regarding my question should led to solution(or possible solution) how we can simply restrict current draw in our FET drivers? Simple resistor mod or something else? It is probably doable but we need electronic engineer for this.

Beer
Jerommel
Jerommel's picture
Offline
Last seen: 1 year 6 months ago
Joined: 01/04/2014 - 13:18
Posts: 6380
Location: the Hague, Netherlands

luminarium iaculator wrote:
Jerommel,

Thanks on a help but please tell me something I don’t know! Facepalm Smile

Maybe youŕe not the only reader here.
Quote:
I have asked you simple question: “Do you know what resistor should I buy and where should I place it in a circuit of this Djozz FET driver?” If you know ok if you don’t know that is also fine… That was the question that demanded answer. It is ok if you don’t know that Innocent
Demand..
I gave you the answer to that particular question:
I don’t know what value you need (but i could calculate it if i look up the specs / test results), and it also depends on the battery you use, the health of the battery, how well the contact between battery and springs is, the condition of the switch you use, the ambient temperature, differences between individual LEDs, LED wires.
All uncertain variables.
Oh and of course the charge of the battery..
Quote:
I sold over 1000 modded lights in mine life. I have a master degree on soldering AMC 7135 and 7138 chips to linear drivers and you really don’t need to cut middle leg. I solder all 3 of them even without bending.
So they stand upright then, and have no decent thermal path to the driver.
Quote:
I know what are my potential solutions but I refuse them since I have hundreds of AK 47 C1 drivers and if there is possibility for current restriction than I will try to mod them.
Okay, i didn’t realize you’ve got hundreds of those 3× 7135 drivers. I can understand you’re not planning to stack those up to 12× 7135.
Quote:

It is the most stupid thing to stack so many amc chips on small linear drivers cause it heats a lot and failure will obviously happen sooner or later.

It depends on how much work you put into it.
at 4.1 Volts (freshly charged with 0.1 Volts battery sag) it’s less than 1 Volt x 4.5 Amperes, is a maximum of around 4 Watts of heat the driver has to dissipate. With the thermal tabs soldered together and the driver sits tight with a alu / brass retainer ring in the pill or head, it will survive.
Quote:
FET driver is much much better than any high current 17mm linear driver with hundreds of stacked 7135 chips cause it does not heat that much at all…
That’s true.
The maximum 4 Watts will be dissipated elsewhere in the electrical path, including inside the battery itself.
Quote:

I also know how FET drivers work since I made hundreds of them this diy Djozz type and Texas ace one (without stencil; just soldering iron and calm hand).

If you want to run FET with Osram WF just use low current 18650 cell at any charge state and don’t bypass springs and that is it. Or add thick steel contact springs for example.

But this is not solution for the problem of how to restrict current in FET driver.
But it’s a series chain, so it’s the same as adding a resistor to the driver, but the resistor is approximately the same whether it’s hot or cold.
Quote:

Potential solution in this below mention style but I would like to take that to higher level and make restriction on driver only… EasyB!

http://budgetlightforum.com/comment/1458705#comment-1458705

http://budgetlightforum.com/comment/1443460#comment-1443460

So any advice regarding my question should led to solution(or possible solution) how we can simply restrict current draw in our FET drivers? Simple resistor mod or something else? It is probably doable but we need electronic engineer for this.

ok.
luminarium iaculator
luminarium iaculator's picture
Offline
Last seen: 1 hour 35 min ago
Joined: 07/07/2013 - 09:09
Posts: 2195
Location: X

“So they stand upright then, and have no decent thermal path to the driver.” Hmmm Question Smile

Yes they stand above or “upright” soldered directly on their twin AMC 7135 brother soldered at 4 points: 3 legs and one back soldered so it is the most superior way of stacking 7135 chips with the best thermal path unlike your 2 legged one LOL

But lets stay on topic…

Edit:
“But it’s a series chain, so it’s the same as adding a resistor to the driver, but the resistor is approximately the same whether it’s hot or cold.”

When you sell thing to potential buyer you really don’t know which cell he plans to use so you must protect yourself.

Pages