Testing a driver and possible fault causes?

[norty303]

Hi All

I'm hoping the collective knowledge of the forum might help get closer to what has happened.

I have a Laserwave dual 640nm module, which the driver stopped working on (no mains power led, essentially dead). As far as I'm aware, it was working correctly and fully prior to it failing.

I contacted Laserwave, who advised that they could supply a new driver, adjusted for my head, for a price that made it worth doing.

I should describe the hardware a little first. It uses 2 single mode diodes, combined with a cube. Each diode is driven separately from the driver. The connector is an 8 pin circular plug, with LD1+, LD1-, LD2+, LD2-, pair for TEC power and pair for temp feedback.

So, I received the driver, fitted it in order to do a test run, and powered it up.
Pretty much as soon as I applied AC power, a fair amount of smoke came from the head (which is sealed, so it must've been forced out of the casing join). Note that this was before any modulation was applied, so essentially off.

I've attached a pic of what i discovered when i opened the head. I've sent this to Bill to ask what he thinks could cause a Lasorb to melt down so spectacularly.

So, what I'd like to explore is what could've happened, and what the possible cause could be.
There is the possibility they sent me a bad/misconfigured driver, but I'm also open to the fact that something was wrong with the burned diode (which could've broken the first driver too maybe?)
Note that after all this, the remaining diode and drive for it works just fine (although i believe its set to too high current, as it peaks at 2.5V mod input)

So, to my questions:

1. If a bad/shorted diode was connected to the output of the driver, could this cause what I've seen, even if the driver wasn't actually applying power to the diode?

2. I would like to test the output of the driver on those pins. How do I go about this? Do I need to connect a dummy load of some sort, and measure current and voltage in series? I can also compare the output with the known working LD pair of pins.

I have my suspicion that it might have been connected to the TEC drive somehow, which would be consistent with it coming on immediately power was applied and big current, but at the moment that is just a suspicion.

As I said, I'm open minded about this at the moment, but lacking the knowledge to fully test it and get closer to a theory.


As it stands, i think this is hosed, so probably junk or a project for someone, but unfortunately its one of a pair, so will necessitate switching out both units for new ones, which is a shame as they're awesome beams.

[aberry]

Hi Norty,

The diode failing short would not have caused the Lasorb to fail. I see two possibilities here. It could be that the Lasorb simply failed short and blew itself up, although I don't know if this is a common or expected failure mode. The other possibility is that the diode went open circuit, which would have caused the driver's output voltage to increase until the Lasorb clamped. At that point the Lasorb was continuously dissipating all of the power the driver was supplying, and since Lasorbs are meant for transient protection rather than continuous clamping the smoke came out. FWIW, we've seen exactly this happen on at least one occasion ourselves.

As far as testing the driver, best bet is to get some rectifier diodes that are rated for at least as much current as your laser diode and wire a few of them in series so that the combined forward voltage is similar to that of your laser diode. You can then hook that up to the driver with a current meter in series to measure the output current of the driver. Or, if you can identify the current sense resistor(s) on the driver, you can measure the voltage across them under load and use Ohm's law to calculate driver current, but if they are sufficiently low value the voltage may be too low to reliably measure with a basic DMM. As a third option, you can replace one or more of the dummy load rectifier diodes with a suitably rated power resistor and use Ohm's law there. That will be good enough for a basic function check, but for a thorough test of the driver's function keeping the dummy load all diodes will be a better approximation of actually driving a laser.

Hope that helps.

[norty303]

I see two possibilities here. It could be that the Lasorb simply failed short and blew itself up, although I don't know if this is a common or expected failure mode. The other possibility is that the diode went open circuit, which would have caused the driver's output voltage to increase until the Lasorb clamped. At that point the Lasorb was continuously dissipating all of the power the driver was supplying, and since Lasorbs are meant for transient protection rather than continuous clamping the smoke came out. FWIW, we've seen exactly this happen on at least one occasion ourselves.

Ok, interesting, and thank you.

So, if the diode had gone open, would I have seen this 'runaway' voltage scenario even with no modulation applied, or would that only happen in the event of actually powering the diode? I guess there may be a small offset current in effect to just below threshold.

Laserwave have supplied me some details of how I can test the current via the board, but unfortunately it was a different driver in the picture, so hopefully they can send me the correct details so i can test it.

[norty303]

Ok, so after i checked the diagram they sent, it was for a different driver, so I sent a picture back of the driver they sent me, and it turns out my suspicion was correct, and the newer driver is set up to drive 2 diodes in series, rather than 2 individually as my original one was.

So they have suggested I change out the bad diode and Lasorb and rewire the head, so I'm just waiting to confirm they'll send me the parts, and then I'll be interested to see the pinout of the connector, and see if my suspicions about it being the TEC drive that got connected to the diode are correct.

[tribble]

Since Laserwave effectively destroyed your head, I'm curious why they didn't suggest that you watch the post for a brand new one any day now. ...?...

[norty303]

The module itself is well out of warranty (3yrs+), and I don't believe they even do a 640nm any more. So yes, whilst I agree with you that it was their error, we will wait and see how this gets resolved.
Emails are still flowing back and forth.
The issue is compounded by the fact it's one of a pair, in a pair of matched projectors, so it's either bring this module back to life, or cut them both out and replace with 2 matched new ones.

I suspect it will come down to them saying "How were we to know it was an early version with different wiring to our standard now" and me countering with 'That's why i sent you the original serial number, so you could check your records for how it was configured'.

We'll see where it goes, the customer service has been pretty good to date with LW, and I've spent well over 6 thousand dollars over the last 4 or so years on their kit, so I would hope they'll see me right, however it gets resolved.

[norty303]

Ok, quick update. They're sending me the diode and Lasorb along with details of the wiring mods I'll need to make to series wire the diodes to the driver.
There's also been an enquiry whether they can 'help' with any new parts/lasers, so there will be some form of 'making it worth my while' it seems, once we can agree on something (what with me having spent all my money for the year on lasers already, at least $2k with them! )