Desperatly looking for driver (updated kennan driver?)

[gillestugan]

I'm currently running three XR-Es at 700mA on a 14.8V pack. Haven't calculated efficiency yet, but the driver runs very cold.

I didn´t change the R1 as I didn´t have any low value resistors.

Instead I put a 100k resistor from output to R2 as shown here: https://www.candlepowerforums.com/posts/2497628&postcount=85
I kept the R2

This gives me a 700mA output.

 

[uk_caver]

I guess the value needed for that approach depends on the value of R2 (which does change between one board revision and another) as well as the number of LEDs driven, so it would either need calculation or experimentation.

 

[gillestugan]

Yes, you are right, I didn't think of that. R2 is 1k on my board.
One way is to adjust the current to whatever you want with a potentiometer, then measure the resistance in the potentiometer and switch the potentiometer for a fixed resistor.

 

[gillestugan]

I want to use two boards in parallel. Is it possible to to wire the 2:nd as slave to the first one, so that when you adjust the output from the first one, you also adjust the 2:nd?

Would it work if I remove the R2 on the 2:nd board and wire the FB pin to FB on the first one?

 

[kosPap]

hi all! In this post in another thread I have added measurements of 2 versions of this driver.

Enjoy, Kostas

 

[spun]

another way to mod R1 if you don't have these low value resistors is of course to replace R1 with 1 ohm resistors in parallel. i'm using two in parallel to get .5 ohms which runs 3 leds in series at 330ma. see parallel resistor calculators on the web or for these just simply 1/number of resistors so 4 would be .250 which i think the original kennan had. i think it's easier to do the dimmer mod and then if you want swap the pot for the resistor of your choice as gillestugan mentioned but this is another method.

 

[tobyjmack]

apologies for resurrecting an old thread but I came across this discussion when trying to find ideas for drivers and the mention of the PT4105 got me curious if dealextreme did any equivelent.

Well, after a bit of poking around I can say that that the DX SKU07425 is based on the PT4105.
I've buzzed out the board and it is identical to the data sheet example circuit except for the following-
1) It has two 0.5R resistors in parallel for the sense resistor (giving 0.25R and 800mA drive rather than the 1A quoted by DX....)
2) it has a 10k resistor from the sense to the feedback pin
3) it has a piggy back microcontroller to give the 5 modes.

As the basic circuit is as per the data sheet I can't see why it won't work to 18V rather than the 8.4V quoted. The only possible problem is the piggyback board which I haven't checked thoroughly. However, there is 5V going to the IC so I think it is regulated so the only question is whether the regulator will cope with 18V. Most will....
I've taken it up to 12V when driving a single Q5 and it still works fine

Hopefully that might be of help to someone?

Toby

 

[italianboy]

I'm new of this forum, so first, hello to everybody!
I'm posting here because I have a problem with this driver. I've bought it from KD, but I've got a new board, different from all the previous here and also from the one shown on the KD site, so I don't know where to solder the led wires.
This is the image:

I'm quite sure the "+" should go under the inductor, where there is the "+", but should I solder the "- " (looking at this image) over the R1, under the R1, or elsewhere?
Thank you all in advance!

P.S. I don't want to fry this board because I've waited over a month and I have only one (the one I have to use...).

 

[space]

The AX 2002 are simular to the PT4105 (used on previous model) in the way it operates, but the pinout is different. The AX2002 is a bit more powerfull and can handle abit higher input voltage. (Can handle abit higher swithcing current and up to 23V input voltage compared to 18V for the PT4105.) I can't see the R1-label you are reffering to, but I guess you are talking about the R220-resistor. The "-" should be conected to witch ever side of the R220-resistor that is not directly connected to "-" for the board. From your picture it seems like that is the "south-end".


space

 

[uk_caver]

duplicate post - removed

 

[gillestugan]

Data sheet can be found here.
Would it be possible to regulate output by using resistors as voltage dividers for the FB bin in the same way as on the PT4150?

Will also be interesting to see what efficiency you get and how high output it can manage. (2A in the data sheet, but that is just the AX2002, not the rest of the components.)

 

[uk_caver]

The AX2002 datasheet does mention pins 5&6 are a thermal path to the internal die, so it might be possible to arrange for some extra heat dissipation for high current usage.

 

[italianboy]

Thank you very much, I've tried to do that and, from a very quick test on a single 18650 (but I'll use a 6v lead-acid) it looks to be working correctly.
P.S. Thank you also for all the other useful information on this forum!

 

[Spykooo]

Hello from Frenchees!!!!

I have a problem with my KD régular!

For 3 regulars i have 800mA : Good!

For 6 regulars i have 400mA : No good! But why i have 400mA?

My last regular do : Psssssssssssshit!:laughing:

 

[uk_caver]

Are all the drivers the same?
Do they all have an 'R220' or 'R270'?

What input voltage are you using?
How many LEDs are you driving, and what is the total Vf for your LEDs?

 

[Spykooo]

Are all the drivers the same?
Yes.

Do they all have an 'R220' or 'R270'?
Normaly R220.

What input voltage are you using?
Alimentation made in home : 9,6 volts.

How many LEDs are you driving, and what is the total Vf for your LEDs?
One P4 rank U.

Thanks!

 

[gillestugan]

I received the AX2002 version today and will do some efficiency measurements.

Does anyone have a request for a specific condition? (Vin,Vout,Iout)

 

[gillestugan]

Have done some testing now and I like it! It is very tough. As you can see in the graph I was able to push it to 2,5A. Then it was shutdown by the thermal protection. When I turned it on again it worked like before. This happened many times during the 2A+ tests, until I put an extra heatsink on it. It wasn't really stable over 2. Had it on 2.2A for 5 min, but then it got too hot and shut down.

The test setup:
Old computer PSU
Mulimeters (Vin, Vout,Iin,Iout)
8A rectifier diodes as load. (didn't want to risk my MC-E)
Modded driver with dimming, lower R2 and a small heatsink soldered to pin 5 and 6 as suggested bu uk_caver

As you can see there are some variations, this can be because of:
I used the 20A socket on the multimeters and they are far from accurate, but both mulimeters are axectly the same and gives the same value of a current, so it shouldn't influence efficiency calculation too much.
The diodes got hot and VF changed slightly, this is of course measured and in calculation.
The dimming method used is not very temperature stable.
-A real amateur test

Here are some pictures of the board:
To do the dimming mod you have to cut one of the tracks on the board. (see picture)
Then solder A potentiometer and a resistor as shown in picture.

If you want higher output than 1A you need to change R2 to a lower value. Look in the data sheet for resistor values.

Heatsink with ugly soldering, sorry

 

[Stage X Tuning]

Well i have looked to all ends to find a good driver for the SSC P7 and cant, so i found a IC that will do a Max of 3A with 2.5V-18V in and a 89% efficiency from .25A to 3A at 5Vin and 3.3Vout. per paper calculations, still not tested. All that and you only have to use 8 to 12 or so parts if you want to use the ICs over-current protection, and short circuit protection. im hoping to have a prototype working soon

 

[uk_caver]

Nice work, gillestugan.

I may order a few more Kennans myself. I've been building a few smart chargers for 3.6V NiMH packs using the PT4015 Kennans running at 750mA, but it would be nice to be able to make some faster (~1.5A) chargers.

It looks like the soldering is to pins 7&8 (2xVss), rather than 5&6.