A question concerning 7135 chips

[tx101]

The ATMEL chip used for multimode on the 7135 boards get less than a few mA (much less if I remember correctly). Well within specs. The ATMEL on my two level 7135 boards survive in my 3xP7 Mags powered by 3 x D-size LiION. I've also been overdriving a 4x CREE R2 setup running it at 5A for a long time without problems.

Now that is a handy piece of information, Thanks Stefan :thumbsup:

Is 5A the highest you have pushed the ATMEL chip ?

 

[Al Combs]

The ATMEL chip used for multimode on the 7135 boards get less than a few mA (much less if I remember correctly). Well within specs. The ATMEL on my two level 7135 boards survive in my 3xP7 Mags powered by 3 x D-size LiION. I've also been overdriving a 4x CREE R2 setup running it at 5A for a long time without problems.

Thanks for the info Stefan.:wave:

 

[jar3ds]

I used your idea DZ of drilling through the board from the sales thread I first saw that picture in. I used a #60 drill hole to run a 20 gauge resistor lead to make the parallel connection. It looks so much neater than a wire looping around the edge of the board. Thanks for the idea!:wave:

I wondered if I could do that the other night, glad to see the drill idea works well.

can't find this sales thread... can someone link it? :candle:

 

[Al Combs]

can't find this sales thread... can someone link it? :candle:

There really wasn't anything else in that thread besides the same picture he posted here. The picture caught my eye though. I only mentioned it to thank DZ for the idea.

 

[Justin Case]

KD ProductID=1770 also uses a hole drilled through one of the boards to make a 2-board, 8xAMC7135 sandwich (the photos on their web site, which show an external wire connecting LED- to LED-, don't match the actual product). The boards are oriented with the non-component sides touching, so KD1770 is probably better suited for something like a Mag mod than a P60 pill.

 

[Der Wichtel]

Now that is a handy piece of information, Thanks Stefan :thumbsup:

Is 5A the highest you have pushed the ATMEL chip ?

The 5A is not going through the Atmel chip. The AMC chips have an enable pin which can be seen on this pic ( from the link above):

The red connection is the enable pin of the AMC chip. If that pin senses a high level then the chip will "turn on" otherwise it disconnects the circuit.

So what the Atmel does is just turning the chips on and off very quickly, not noticable for usual human eyes 🙂.
if the on-time is equal off-time then you'll see 50% brightness. That's called pwm 😉. Therefore the brightness is controlled by the ratio of on-time and off-time.

 

[Mettee]

great post yitao

 

[Der Wichtel]

great post yitao

thanks, thought I could clear a few things up.

From the TaskLED site, "Maximum drive current 3.4A (70C ambient) or 4.5A (25C ambient)". Here is a post from George himself in a cmacclel thread that there is apparently a new version D2Flex that can handle 10 amps. If you want to get a D2Flex, maybe you should email TaskLED first to make sure you get the right one. If you look at the top of this thread, cmacclel says he used a D2Flex with no regulator on an SST-50 with a BatterSpace 26650. I'm not sure if these are IMR's as they have a max discharge of 10 amps. It might behave differently with the AW IMR26500's. Here's a post from moviles using a pwm (unregulated) dimmer from an MTE SF-15. He says the FET handles 71 amp. I guess that should be enough.

If the Attiny13V or whatever µC is used on the amc is desoldered and the pwm output of the D2Flex is soldered on the red connection then you won't have the maximum drive current problem of the D2flex.
A low esr capacitor paralleled with the batteries will reduce input ripples --> more runtime, since batteries are "more efficient" with lower currents

 

[tx101]

The 5A is not going through the Atmel chip. The AMC chips have an enable pin which can be seen on this pic ( from the link above):



The red connection is the enable pin of the AMC chip. If that pin senses a high level then the chip will "turn on" otherwise it disconnects the circuit.

So what the Atmel does is just turning the chips on and off very quickly, not noticable for usual human eyes 🙂.
if the on-time is equal off-time then you'll see 50% brightness. That's called pwm 😉. Therefore the brightness is controlled by the ratio of on-time and off-time.

Thanks for the explanation Der Wichtel

 

[Al Combs]

The 5A is not going through the Atmel chip. The AMC chips have an enable pin which can be seen on this pic ( from the link above):

The red connection is the enable pin of the AMC chip. If that pin senses a high level then the chip will "turn on" otherwise it disconnects the circuit.

So what the Atmel does is just turning the chips on and off very quickly, not noticable for usual human eyes 🙂.
if the on-time is equal off-time then you'll see 50% brightness. That's called pwm 😉. Therefore the brightness is controlled by the ratio of on-time and off-time.

Yes, thanks for the explanation. It also helps explains the debate about removing the protection diodes from the 7135 boards. The pdf on the AMC7135 says the minimum Vdd voltage is 2.7 volts. But the output can still be more than that. There was a good post about the 7135 from Drewfus2101. In post 30 of this thread he gave the results in a graph after removing the reverse polarity diodes. There was no difference. But he only went down to 3.4 volts supply. Which is still 0.7 volts above the trigger threshold. I get it now, thanks.:thumbsup:

 

[Justin Case]

Yes, thanks for the explanation. It also helps explains the debate about removing the protection diodes from the 7135 boards. The pdf on the AMC7135 says the minimum Vdd voltage is 2.7 volts. But the output can still be more than that. There was a good post about the 7135 from Drewfus2101. In post 30 of this thread he gave the results in a graph after removing the reverse polarity diodes. There was no difference. But he only went down to 3.4 volts supply. Which is still 0.7 volts above the trigger threshold. I get it now, thanks.:thumbsup:

Yes, Drewfus2101's measurements didn't show any difference even after removing the diodes. But #1, his measurements didn't go low enough in input voltage as you say, and #2, it isn't a "trigger threshold" issue. It is being able to still run in full regulation even with a low Vf LED and relatively low Vin. See my explanation in Post #44 of that thread.

 

[Packhorse]

As Der Wichtel points out the control chip does not see the total output current. BUT it does have to control all AMC 7135 chips and there will be a point where it will not be able to control any more chips. Where this point is I have no idea.

Anyway who has hooked up the most AMC's? What battery pack did you drive it off? What problems did you have?
I am wanting to run 20 of them off a 20ah Li Ion 3.7v pack.

 

[darkzero]

As Der Wichtel points out the control chip does not see the total output current. BUT it does have to control all AMC 7135 chips and there will be a point where it will not be able to control any more chips. Where this point is I have no idea.

Anyway who has hooked up the most AMC's? What battery pack did you drive it off? What problems did you have?
I am wanting to run 20 of them off a 20ah Li Ion 3.7v pack.

Most I have ever done is 10 but only because there was no sense in going higher for the P7. There were no SSTs around back then but I'd be willing to try.

Like the "AMC" boards that uses the Atmel MCU to control the ADDtek regulators, the FLuPIC is similar which uses a Microchip MCU to control Texas Instruments regulators. I built a Tri-FLuPIC running 15 regulators (5 per channel wired in parallel) to power a SST-50 & the PIC did not like that. Aside from the UI acting funny, at times it would trip out & push 9.5 - 10A to the poor SST-50.

 

[Linger]

your work is art

 

[tx101]

:bow: DZ :bow:, if only my soldering was half as neat as yours :sigh:


=========================================


I still have not gotten around to "slapping" a couple of 7135 boards together 😱
I keep getting distracted by another project 😀

 

[StefanFS]

As Der Wichtel points out the control chip does not see the total output current. BUT it does have to control all AMC 7135 chips and there will be a point where it will not be able to control any more chips. Where this point is I have no idea.

Anyway who has hooked up the most AMC's? What battery pack did you drive it off? What problems did you have?
I am wanting to run 20 of them off a 20ah Li Ion 3.7v pack.

I did run a total of nineteen AMC7135 chips controlled by a Atmel Tiny chip (the "new?" two level 1A board from KD). Three on the 1A controller board and four 1.4A boards with four schips on each. Three IMR about 18650? size/three D LiION, both setups in parallell. One SSTx0. No problems that I could observe. The low level is still ~10%. 10% of a lot more 100% level, but that is also true of my P7 and MC E lights with AMC7135 based drivers. It's a simple design, as Yitao explained the Atmel chip just switches the string of AMC chips, it doesn't seem to matter if it's one or fifteen. I stepped it down to ~5A to control heat from the emitter in a Mag D size package before I took it apart permanently.

 

[Der Wichtel]

Well the Atmel µC can only deliver 20mA max per pin. The more AMC boards are added to the setup the higher the capacity on that pin will be.

If the capacity is too high on that pin the edge between on and off state of the AMC boards will be too long which will cause some problems.

Furthermore the switching current will rise up and can destroy the output of the Atmel µC if there is no internal protection.

So we need to find out the enable pin's capacity of the amc board and the pwm frequency of the µC in order to estimate how many boards can be connected in parallel.

 

[Packhorse]

For my application 10% is too low for low mode. I really want about 30-50%.

The way I figure it is get the control chip to control 70% of the AMC's and run the remaining 30% as single mode.
This way hi is 100% and low is 30% + (10% of 70%) or 37% of high.
I can then also run a thermal switch connected up to the dual mode driven AMC's and if /when it over heats it will shut down to 30%.
It also means the Atmel chip is driving less AMC's.

 

[saabluster]

As Der Wichtel points out the control chip does not see the total output current. BUT it does have to control all AMC 7135 chips and there will be a point where it will not be able to control any more chips. Where this point is I have no idea.

Anyway who has hooked up the most AMC's? What battery pack did you drive it off? What problems did you have?
I am wanting to run 20 of them off a 20ah Li Ion 3.7v pack.

Wouldn't it be possible to have the Atmel run one AMC which then in turn switches the remaining AMCs? That way the current load on the Atmel pin is reduced. I'm no electronics expert so don't roast me if I'm way off.😱

 

[Packhorse]

No I dont think that would work. You could use a pair of transistors though. ( I guess). I not think it is required though.