Desperatly looking for driver (updated kennan driver?)

[ambientmind]

I've been looking through this thread, and I'm a little lost with my quest for an answer. I'm just looking to mod this board so it can output about 350-450mA. What size and where do I need to solder the resistor? Thanks!

 

[uk_caver]

For the newer (AX2002) version of the board, for 400mA, you'd want a current-sense resistor of 0R625, replacing the 0R22 resistor (marked 'R220')

Looking at my electronics supplier, resistors in 0805 size are available in values of 0R56, 0R62 and 0R68, which would give theoretical output currents of 446mA, 403mA, and 367mA respectively.

 

[bshanahan14rulz]

I don't understand how you search for these boards, or why they are called kennan. I have a 17mm driver board that needs replacing soon (it smells like burning electronics) and am thinking that this board, with all of your research, seems to be the most fun. So, from what I understand, it is a generic board from KD or DX that provides CC (as opposed to a controller board or a buck/dd board)

 

[TorchBoy]

Kennan was a constant current driver board from KD (back in the days when their site worked) that scored three out of three: It was inexpensive, efficient, and versatile - it accepted 5-18 V in, and output a constant current for 1 to 3 LEDs. It claimed a lot, did what it claimed, and developed a fan following because of it. It got the name Kennan because KD and DX used to make up pseudo-random names for all their products back then.

 

[bshanahan14rulz]

hah, they just made up names? So this was an old board that has since been replaced by less efficient but higher current boards? Like, is it not available anywhere anymore? seems like a backwards step for batteries, though I wouldn't care so much if it were for a wired application...

 

[Justin Case]

I don't know about the efficiency of the original Kennan, but this new KD board based on the AX2002 seems quite efficient.

I am using a modified version to drive a 2S2P MC-E at ~700mA per core. The board uses a sense resistor setup to deliver 1445mA (nominal, based on the output current equation in the AX2002 datasheet) out.

I didn't measure actual Vf or Iout, but on my bench power supply, the driver requires about 10W of power in when running in regulation. If we assume Vf~6.5V and Iout~1.4A, then Pout~9.1W, or about 90% efficiency.

The driver runs in full regulation starting at about 7.0V, which seems quite good in terms of voltage overhead.

One caution. It appears that the inductor may be undersized for this higher current output application. Above 12V for Vin, the driver becomes unstable and Pin starts to behave oddly (it isn't 10W anymore -- it fluctuates from about 4W to 8W). To run above 12V Vin, the inductor should be 33 uH or so. If the board I have uses the same 22 uH inductor that is shown on the KD web site, then that could explain the instability. Unfortunately, the inductor on my board is scrubbed of all markings.

 

[Capo_au]

Just received two of the AX2002's in the mail today. Having not had a chance to play with them yet I can only comment on the shipping and appearance.

It took exactly 2 weeks from time of order with 8 of those being business days. Packaging was also more than adequate. The inductor also appears to have had no markings or the markings have been removed the same as the post above.

I will be sure to post some pic's when i get the dimmer mod up and running.

-Capo

 

[Capo_au]

Well I finally got the time to have a play with my new AX2002's. Did the dimmer switch mod to a pair of the drivers and connected them to 4 ER-E's, 2 a pc.

Works like a charm, the only problem I have encounter is a slight fluctuation in the current in the first 30 seconds or so of switch on. My guess is the battery voltage is dropping to quickly in that first 30 seconds or so for the drivers to be able to keep up.

Now some photos for good measure.

I used a 10kOhm dual rotary pot to control the current to both drivers at once. I only wanted a 1A max current so I have put a 1kOhm resistor (not shown) in series with the pot. With out the 1K resistor it it would dial all the way up to about 1.45A where everything got very hot very fast even with large heat sinks on the LED's.

A close up of the mod'ed AX2002. Green wires are for the pot, red is + led black is negative. The red and black wires both came with the driver. Note:Its hard to tell in the photo but none of the soldered connections are contacting the outer ring.

Simultaneous readings of the current output from both drivers. Close enough that i cant tell the difference by looking at them.

Many thanks to all the previous posters, all the info came in very handy.

-Capo

 

[Capo_au]

In the post above I dimmed the two drivers at once using a "dual rotary" pot. This means that each drive is essential dimmed by a different pot, so to speak.

Does anyone know if it would be damaging if i was to use a single rotary pot to dim both drivers at once? They would be powered off the same power source but driving separate led's.

Thanks. lovecpf

 

[gillestugan]

Hi, it wont be damaging, but unfortunately it doesn't work. You will never get the same level on both drivers. When one of the drivers goes high, it will cause the other one to go low, so they will never be in balance. (as long as the pot is connected to output. Maximum level is still restricted by the sense resistor, so it wont damage the driver or led.
It may be possible to make it work if you use a diodes to block current going both ways between the circuits, but I haven't tried this.

You may be able to calibrate the output by changing the value of the fixed resistor of one of the circuits, but this is probably not worth it. As you say, it is not possible to perceive the difference in brightness.

 

[Capo_au]

Maximum level is still restricted by the sense resistor, so it wont damage the driver or led.

Good to hear, that was my major concern.

It may be possible to make it work if you use a diodes to block current going both ways between the circuits, but I haven't tried this.

I will give this a try one day when i get the time and post my results.

Thanks for the help :twothumbs

 

[tobyjmack]

has anyone used the enable pin on the AX2002

Hopefully not deviating too far from the original thread but has anyone used the enable pin on the Kennan board?

According to the AX2002 datasheet it should turn on with anything over 2V. I am finding that 3.3V (output of a PIC) only turns it partially on. I get only about 100 to 150mV on the sense pin and a dim light. Reconnecting the on board 100K pullup and it works fine with 250mV on the sense pin.

Has anyone else done anything similar, I've testing it on two boards so I am pretty sure I am right but it would be good to get confirmation I am not doing something silly!

Toby

 

[uk_caver]

I've driven the AX2002 Kennans from PICs, leaving the pullup resistor in place, and pulling the enable pin to ground via a MOSFET when I want to disable the chip.

However, I have been using them as cheap buck drivers in smart battery chargers, so I haven't needed rapid on/off performance (as would be needed for PWM control for a lighting application).

When you're trying to drive with a PIC, how are you making a connection - through some resistor, or a direct connection?
The voltage readings you were getting do seem a little odd at first sight.

 

[tobyjmack]

I've used it with both direct connection and with a 47k series resistor, both with the 100k onboard pullup removed.

I've also tested it with the 100k pullup in place and just probed the enable pin with a high output from the PIC (3.3V). When the pic line is disconnected then the driver is on full, when connected it goes to half brightness. So it seems pretty conclusive that it needs greater than 3.3V to turn it on.

Using a fet to pull the enable low to turn it off does look the simplest option. The downside is an extra 100uA current draw when off but that will not be a major problem as long as I don't leave it connected for months.

A shame if this is the case but it is not the end of the world.

Toby

 

[tobyjmack]

a quick update in case if affects anyone else.
I have put the on board pullup back on the enable and used a transistor from the pic to pull that low (as suggested by uk caver).
So far at least it is behaving itself properly. :thumbsup:

It is an extra component and 100uA on the off current but I can live with that.

Toby

 

[Capo_au]

Thanks for the info tobyjmack.

Any chance you could post a simple schematic, or even just a few photos?

I'm considering doing something similar but experience in electronic components is limited. So any extra info would be great.

Thanks!

 

[tobyjmack]

Pics and schematic are posted here.... http://forums.mtbr.com/showthread.php?t=554555

Note that the picture of the board does not have the transistor in the enable line.

It was more hacked together rather than properly designed so it could be improved (eg filtering the battery monitor, reducing standby current) but it works ok.

Toby

 

[Capo_au]

Thanks for the link, thats just the info i have been looking for.

-Capo

 

[Tobias Bossert]

Does somebody still have old Kennan2 lying around?

I have an application using analogue dimmed driver boards in the range 30mA to 700mA for a single LED load and supply voltage starting at about 3.5V.

Kennan2 was optimized for that! But it is sold out at Kaidomain and replaced by Kennan3, which is much more powerfull for currents above 1000mA and increased input voltage range. But it is less efficient with dimmed output, especially when dimmed down below 300mA.

I'm interested in getting this old stuff. To prevent from misunderstanding, I would like to get those blue Kennans on which all components are arranged orthogonally:

In case you have such drivers and like to switch them to actual Kennan3 with AX2002, please send me a message on how many of them you can give to me. It doesnt matter if the board has other or more components in place, it is sufficient that it has the same layout only.

I propose that I order the same amount of Kennan3 at KD in my account but under your shipping address. When you got them, you send your Kennan2 to me. Also direct payment would be possible if you want to get rid of the old stuff without replacing it. There is also possibility that I may 'pay' with rotary switches 1 by 1. I have left about 25 rotary switches with 5 positions (30°) and 4 poles, open before close, upt to 1,5A, axis 6mm, max diameter 18mm including contacts, ITT SB16. This switches are still on the market, now from ELMA

Reason is, that my actual project (testdevice_3) no more requires high current switching and 2 poles but only low current switching of 1 pole. For this I use a smaller switch from Marquardt (axis 3mm, diameter 12mm)

 

[spun]

rf shielding tips?

i've got a couple of gen2 kennans running 3 leds in series each at different ma outputs. both are powered by older style wallwarts, not the switching type. running either of them cuts out a few over-the-air dtv stations. i've enclosed one circuit in an altoid box and it's not much help. the wires coming out of the altoid box are very short and the leds are maybe an inch apart on a heatsink. i am not using the altoid box as a 'ground'.