Thermal paste/grease preparation

[olephart]

Yep. I just did some reading about it. The AW cells use a voltage (under load) like you suggested and additionally require a relatively high, but unspecified, current at the same time before the protection circuit kicks. They are probably not gonna protect from over discharge in many typical circumstances.

Any ideas for a practical method to keep from frying the Fatman due to low voltage?

 

[Justin Case]

Measure or calculate your run time, keep a mental tally of how long you've run the light, and change out the cells when you reach some pre-determined percentage of that run time.

 

[olephart]

That'll work. I looked at some charts for the AW 2600 cell here:

http://www.lygte-info.dk/info/Battery 18650 UK.html

It appears that the battery is done for at 2.8V. Using the formula you provided, Current would would be <2.25A at all times if I reduce the max output of the Fatman to 1.2A. Am I figuring this right?

 

[Justin Case]

That'll work. I looked at some charts for the AW 2600 cell here:

http://www.lygte-info.dk/info/Battery 18650 UK.html

It appears that the battery is done for at 2.8V. Using the formula you provided, Current would would be <2.25A at all times if I reduce the max output of the Fatman to 1.2A. Am I figuring this right?

I thought the max drive for the Fatman was 1A.

I'd recalc the min voltage such that Ibatt<2A as George recommended and using a 1A drive current (not 1.2A). Then estimate total mAh capacity at 2A draw and for a terminating Vbatt equal to the min voltage determined in step 1. Finally, estimate run time.

 

[olephart]

The confusion may be that he only list the resistance values to set it for up to 1A. But, he provides a formula for setting higher values, if necessary. Here's a quote from the Fatman technical info:

"............Now, Input_current should be around 2A or less for optimal performance of Fatman. Fatman can run up to around 2.2A max. When running at high input power it is recommended to thermally epoxy the back of the Fatman PCB to a heatsink or the body of the flashlight........."

I used the formulas provided on the site to determine Input Current and got 2.28A with the max current to the LEDs set at 1.2A and battery voltage at 2.8V. Seems like this would be sorta bulletproof if I keep an eye on run time as you suggested. You think I will have problems if I run the light too long with those settings?

 

[Justin Case]

The confusion may be that he only list the resistance values to set it for up to 1A. But, he provides a formula for setting higher values, if necessary. Here's a quote from the Fatman technical info:

"............Now, Input_current should be around 2A or less for optimal performance of Fatman. Fatman can run up to around 2.2A max. When running at high input power it is recommended to thermally epoxy the back of the Fatman PCB to a heatsink or the body of the flashlight........."

I used the formulas provided on the site to determine Input Current and got 2.28A with the max current to the LEDs set at 1.2A and battery voltage at 2.8V. Seems like this would be sorta bulletproof if I keep an eye on run time as you suggested. You think I will have problems if I run the light too long with those settings?

High Ibatt probably means poor Vin:Vload ratio, which usally means lower driver efficiency. That means more heat generation, hence the advice for sinking the driver if you attempt to run at 2.2A input.

If your design can handle the thermal management, then it shouldn't be a problem running at the extremes. But that's an important "if".

I personally doubt you will notice any functional difference between 1.2A drive vs 1.0A drive. I don't think the extra power draw and heat generation will be worth it, but that's just my opinion. Are you going to change the sense resistor or does Goerge do that?

Can you go with buck instead of boost? Buck makes it a lot easier to deliver lots of power to multiple LEDs.

 

[olephart]

It will run under 2A with output set at 1.2A as long as the battery is above 3.2. Hopefully that will be all the time. In spite of that, I will give it a try at 1.0A. It will probably be better all around if the output is about the same.

The Fatman has an on board pot that you can use to select the max output. George will set it for you, but it's pretty simple to do. You can also add an external pot that allows you to vary the current from the minimum up to the maximum set with the on board pot. I will be doing that.

I suppose I could get a different driver and add another battery to do a buck mode. I already have the Fatman and the LEDs are pre-wired on a star board in series. Just trying to make something decent with some parts on hand. I'll probably stick with a boost driver on this one.

I appreciate all the help. I'm leaving in the AM on vacation. Hopefully, I'll be ready to start the build when i get back.

 

[Justin Case]

Yes, I realize the Fatman has an on-board trim pot that can be used to dial in other drive currents. But I thought the max drive current was 1A. Is that not correct and the R10 sense resistor gives a higher max drive than 1A?

 

[olephart]

I donno. I'm the one needing help. I've read the technical info on George's site and it seems to me that a fixed resistor only sets the minimum current at about 30mA. The pot sets the maximum output and an optional external pot will vary current between the min and max determined by the resistor and on board pot. I could be wrong.

As to the 1A max, it is qualified by .."depends on input/output voltage differential...". It seems that Input Current is the limiting factor and must be kept below 2-2.2A. That allows circumstances where current in excess of 1A could be output to the LED without exceeding the Input Current limit. I could be wrong.

Maybe it means both limits apply. I donno. It doesn't matter anyway as I am going to use 1A output and stay below 2A input. If you are still curious about the limits, maybe you could read the product info and the technical info and see what you think.

 

[Scattergun]

Aside from all the very advanced help above, I find it very practical to apply the thermal paste inside the flashlight body and then insert the heatsink into it, rather than putting the gooey stuff on the heatsink in the first place. Excess thermal paste tends to be pushed into the flashlight when it is applied in the body, instead of being squeezed all over the outer portion of the flashlight.
Another thing useful is having small pieces of paper and thin rubber gloves when assembling, then if I get some thermal paste on my hands, I just remove the glove and keeps on working.

 

[Justin Case]

I donno. I'm the one needing help. I've read the technical info on George's site and it seems to me that a fixed resistor only sets the minimum current at about 30mA. The pot sets the maximum output and an optional external pot will vary current between the min and max determined by the resistor and on board pot. I could be wrong.

As to the 1A max, it is qualified by .."depends on input/output voltage differential...". It seems that Input Current is the limiting factor and must be kept below 2-2.2A. That allows circumstances where current in excess of 1A could be output to the LED without exceeding the Input Current limit. I could be wrong.

Maybe it means both limits apply. I donno. It doesn't matter anyway as I am going to use 1A output and stay below 2A input. If you are still curious about the limits, maybe you could read the product info and the technical info and see what you think.

My assumption is that the board is a voltage regulated, current limited driver. The green SMD resistor at R2 on the Fatman is probably the current limiting (essentially a sense) resistor that sets the max drive current to 1A. Turning the on-board pot at R1 varies the drive current (up to the max dictated by R2) by changing the control voltage sent to the switcher IC's feedback pin.

The voltage control clearly incorporates a fixed resistor, which is what that referenced term is all about. The fixed resistor is probably in series with the pot to form a resistor divider network.

Vbatt:Vload ratio affects the max drive current through the limitation of the max switch current spec of the switcher IC, resultant driver efficiency for that Vbatt:Vload ratio, and resultant waste heat generation. You can't arbitrarily drop Vbatt to any low input voltage and expect the driver to deliver full power.

 

[olephart]

Well, I don't doubt a thing you are saying. The part about a fixed resistor setting min current and the pot setting max current is right out of the technical description. Perhaps there are some gaps and assumptions in George's description of how it works. I suppose he is writing for folks that probably don't need to read the stuff.

 

[olephart]

In case anyone is interested, I finally got around to asking the man about his driver.

Q. Can I adjust the trimpot on a fatman (V1.0) to drive the LED at 1.2A if I keep the input current below 2A?

A. yes, you can adjust the trimpot to 1.2A, just make sure you provide some thermal path to the back of Fatman to try and keep it cool.