Elly low voltage start up problems

LEDBE
C

Corona

Enlightened
Joined
Aug 1, 2005
Messages
246
Location
Chicago Area
I've noticed the Elly (MXDL 1W) that I've used for a violet laser project has trouble starting with a partially discharged alkaline. It won't start at all with a NiMH cell.

When it is in this mode, loosening the tailcap a bit and tightening it again seems to help get it to fire off. Working the clicky doesn't have the same effect.

Is this a common issue, or is the one I have dodgy?

I suspect this is either a connectivity problem via the anodizing on the 2 aluminum mating pieces (tailcap & body) or some inherent startup voltage issue with the wee little boost controller IC.

Perhaps the act of making an intermittent connection introduces small voltage spikes on the controller and allows it to start. I'm going to try adding a small tantalum cap right on the PCB and see if it makes this better or worse...

It's easy to see that the boost controller's 3 pins have these functions (left to right looking at it with the heatsink tab at the top)

Pin 1 - Negative supply rail
Pin 2 - Boost voltage input (positive supply for IC and sense input for boost voltage regulation)
Pin 3 - External FET gate drive output

EDIT - found the datasheet, it sure looks like it's a Rohm RH5RH!! Supposed to operate down to 0.9V at light load

Hmmm...
 
Last edited:
Yup, I've found the same thing with the driver in the Elly. If a NiMH is slightly used, the circuit refuses to start, didn't try any used Alkalines. I thought it was just me. If you find an answer, please post it as I'm using one to drive a DVD LD.
Here's where I'm at: first the setup
setup.jpg

Then the filtering / protection network:
ldfilter.jpg


Rob
 
Knowing Robk wel, I can almost guarrantee it is ! Now- can you identify where the circle of metal is from ??


Muahahahaha ! Just kidding - I have no idea !
 
You guys are pretty good! Yes, a fuse puller with a 1/4-20 nut epoxied to it for tripod mounting, stuck to a hard drive magnet on a circular metal scrap. Now - where did the circle come from? I save everything, I make things from garbage. I have a Maha 801 charger and some cells heat up a bit when quick charged, so I needed a fan to keep things cool. I took an old ATX power supply, took the fan out, used half the enclosure as a stand to sit over the battery charger, mounted the fan in the center and you can guess where I got the circle of metal using a nibble tool!
cooler.jpg


Rob
 
LOL entirely too funny. Great idea for the cooler though. and for the mount. I work at an automotive parts retailer (they just notified us that we are not supposed to mention ^&^&^#% in forums or chat) and I recognized the fuze puller. that and I have about 7,000 magnets all pulled from HDD. What is it with geeks and thier pull or attraction to lights (ooooh shiney bright thing....Pretty!) and magnets that are strong (must seperate to remove broken finger!!! rrrrraaaahhh....)
 
HaHa I am a geek! Only here and a few other places on earth I find people who understand my mental condition. My wife thinks I'm crazy - I save weird things like product packages (especially plastic cylindrical), the little triangle platforms that come on delivery pizzas ( I swear I will find a use for them), all kinds of stuff. I enjoy making useful things from what others consider junk! Last week I cut down a hairspray aerosol cap to about 1/2" deep to cover my can of Miller Lite when I leave it in the garage to go back inside to take care of business - keeps the bugs out (this is Florida!). It's amazing what you can build with stuff that other people consider junk. You should see my garage - it's like a museum - I can't throw anything away!

About the magnets - they are extremely useful in a thousand ways. Every dead hard drive that passes by me gets torn down for magnets and some surface mount parts like caps, power sockets, whatever.

Rob
 
Thread-jackers!! :D

lol - the startup problem I'm experiencing was due to my modding the circuit to add a silicon diode in series with the power/sensing pin on the boost controller - while it did bump the boost voltage up by 300mV, it also increased the startup voltage by 300mV. Anything but a freshly unwrapped alkaline would NOT start it, and worse, I could run the thing for 10 minutes, no problems with the +5V output at all once it was started - but click the switch and it would refuse to come back on.

Carry On
 
Ok, well... After some looking at this thing, I know how the circuitry operates and how the chip works now, but I can't distinguish the pinouts. I have a couple possibilities on paper, nothing solid though. I have no idea how to test this, and I blew my DMM a few days ago. :(

If anyone can help, please... any input is very much appreciated;
DDL
 
Further testing has uncovered something very odd. The LED is not running when 1.5 volts are applied, but does turn on when 3.0V are applied. The circuitry does not get a 3V laser to lase, But powers the LED. Testing voltage while across the LED while it is running still shows 1.5V output. Obviously... there is something I am doing wrong. I really have run out of ideas on this one... :(

--DDL
 
Well, while the part number DOES say "2106", that can't be a dual diode or it would not work as a boost circuit. That p/n is either an abbreviation or a "house number" - large OEMs can order parts from the semiconductor maker by their own in-house part number; this also makes reverse engineering the circuit tougher as finding datasheets is near impossible.

Or it could be a Chinese semiconductor fab house disguising counterfeited parts with another part number, to avoid royalty payments (to Rohm, for example).

Looking at it I can see it's still a classic boost circuit, just simplified by using a boost controller with an internal FET. The older Elly circuit used that obsolete Rohm part (I posted a datasheet link somewheres on CPF), with an external MOSFET.

The boost diode, being a glass part, may not be a Schottky (more cost savings). If true, this will certainly increase the startup voltage. In this case changing it out for a Schottky part will at least improve that. But, it could be OK as-is, just free associating...

Here's what I think is wrong; you don't have a good connection to negative. Remember that the flashlight body is the negative connection for the circuit, and that the negative (blue wire on this one) must be touching the Elly case in order for the battery power to get to the circuit.

So your voltmeter is measuring the battery voltage through the components of the circuit, but the circuit is not operating. It should make 5V, just like the original circuit.

Also I see you have chipped a bit off the inductor; not a big problem though it will lower the inductance a bit and result in higher peak currents in the internal FET. This new circuit is probably not as robust as the old one with external FET, but probably works just as well for violet and DVD red laser diodes.
 
So Did you get it to work?
I want to know if it works before I risk my cash on it:cool:
Thanks
Jonno
 
I'm really sorry. My internet took a dump and I finally got it up today.
I sent the picture over to Conrona and once I hear anything I'll let you all know.
I also bought another one and got it in yesterday, as well as the other 1W LED flashlight. I'll post both of them as soon as I get the time.

--DDL
 
Had a look and I don't like what I see. This new circuit is a cost-reduced version of the original Elly boost design and has some significant shortcomings.

First is the fact that the load (LED or laser) is connected to the "flying" end of the inductor, not the rectified & filtered output of the boost circuit. This means that the load will see some very high voltage spikes. I've seen this done in other LED drive circuits and while it works, it's not a very good idea.

Also, because the load "rectifier" has a different voltage drop than the wee little rectifying diode on the PCB, the regulation is not going to be very predictable.

And, since there is no filtering of the pulses, an LED or laser diode connected will not be truly CW; it will be pulsed. At least, it will as long as it withstands those spikes. It will be 100kHz or higher, so you wouldn't notice it visually - but it will be pulsed - and in order to get the average current up to where a DVD burner diode will actually make enough RMS power to do cool stuff, the PEAK current will be nasty indeed; upwards of an amp or more.

Did I mention the high voltage spikes? LEDs are fairly robust and a white LED might be OK with this, but I would NEVER connect a sensitive and costly laser diode to it as-is.

The "filter" capacitor in this design is nothing more than a tiny ceramic capacitor, meant to filter the supply to the IC, and nothing more.

I would not use this one for laser duty without modifying it to:

1) Replace the little MELF (round glass SMT) diode with a beefier Schottky SMT part. Remove the old diode, can't piggyback MELFs very well and you don't want them both in there at the same time anyway.

2) Add a significantly larger tantalum filter capacitor where that little ceramic part is (10uF 16V or so). Piggybacking it would be fine.

3) Most critical - move the (+) load connection (the red wire) from the flying end of the inductor, to the (+) side of the added filter capacitor.

And of course, provide some method of limiting the load current, after evaluating the regulation capability of this new circuit at higher load currents.

If I come across one of these, I'll do the mod and testing and report back, but I'm not going to consciously seek one out. :D
 
You must log in or register to reply here.

Similar threads

Accutronitis
I Updated My NUBM44-V2/81 7W+ Blue Laser!
Replies
2
Views
7K
Accutronitis
Accutronitis
karlthev
Low power, long running laser suggestion
Replies
2
Views
4K
karlthev
karlthev
D
Stickies updates?
Replies
1
Views
21K
nbp
nbp
D
Sell 5mW 650nm Red Laser Diode Module As Low As $45
Replies
3
Views
2K
FRITZHID
FRITZHID
C
40W Laser Tube No Idea How To Wire Up
Replies
5
Views
7K
TEEJ
T

Latest posts

Top