flexmod P3.3 catastrophic failure

[VDX]

... to avoid such 'irregular' behaviour and random killing of my diodes I've developed own simple current limiting drivers capable of high pulse rates (up to 2 MHz) and currents from 100mA to 10A per channel or "parallelizing" multiple channels for higher currents - they are based od high-speed OpAmps configured for current regulating and MOSFET's driving the current, what's producing some heat when driving in the amperes range.

Actually I'm busy to test a driver, my company developed new, where a DC-DC-converter acts as current regulator to reduce voltage drops and heating, but with specific (reproducible) conditions at higher currents than 7A it tends to oszillate with high current (PS gives max. 15A) spikes and some MHz oszillation rates, what gives funny results with the IR-diode I was testing ... a 9W-diode melted a plastic lens, which is rated for 25 Watts optical througput, before it went off !! ... but from 500mA to 6A it seems to work OK ...

Viktor

[JohnYayas]

nobody is going to put a bench/lab supply in their projector... current limiting to the max diode current seems a little silly since, you know, that's the flexmod's job...

anyway I have had several flexmods go up in flames so I gave up and sold most of mine

What do you use now? I might need some drivers in the near future. I recently found have a stack of unpopulated die4drive boards that I can build up but my eyesight isn't what it used to be and those boards are tiny. Something prebuilt is more appealing these days.

[Draco]

... to avoid such 'irregular' behaviour and random killing of my diodes I've developed own simple current limiting drivers capable of high pulse rates (up to 2 MHz) and currents from 100mA to 10A per channel or "parallelizing" multiple channels for higher currents - they are based od high-speed OpAmps configured for current regulating and MOSFET's driving the current, what's producing some heat when driving in the amperes range.

Actually I'm busy to test a driver, my company developed new, where a DC-DC-converter acts as current regulator to reduce voltage drops and heating, but with specific (reproducible) conditions at higher currents than 7A it tends to oszillate with high current (PS gives max. 15A) spikes and some MHz oszillation rates, what gives funny results with the IR-diode I was testing ... a 9W-diode melted a plastic lens, which is rated for 25 Watts optical througput, before it went off !! ... but from 500mA to 6A it seems to work OK ...

Viktor

would you be willing to share a schematic of your drivers?
I had built a driver to run a 20 watt IR bar diode a number of years back and i had oscillation when i ran it below the design current but ran well at full output, what it did is i used a comparator with a an OpAmp that drove about 7 transistors in parallel, i no longer use it as i have a different very high current power supply with a programmable current that i use for that diode bar.
But i love to see one of yours for less output current.
as for premade drivers i am going to get a sample from the guy in Russia and the guy from Australia when i need a device for 2 diodes, i wish he made single ones.

[flecom]

had i not had this connected to a current limited supply it would have caught fire and destroyed the friken power supply, so your "little silly" idea would have resulted in further damage down the line, mine did not catch fire because i limited the current to 6 amps and the supply shut down.
In any electrical drive or circuit you must have some form of current limit, at the minimum a fuse is mandatory on the power buss. you can't depend on the built in fuse or over current in some cheap switch mode power supplies.
On that note i would not build the supply into the projector anyway so it's a moot point to where the 12 volts(in my case) comes from, be it a battery, bench supply or a unicorn's ass.

what is really pissing me off is the lack of service from his company, i will never buy another one from him

it's silly because bench supplies are big, an average bench supply is not much smaller than a KVant projector

just use drivers that don't suck

I am using a modified StanHam driver... working on another revision of it which I will hopefully post soon

[Draco]

it's silly because bench supplies are big, an average bench supply is not much smaller than a KVant projector

just use drivers that don't suck

I am using a modified StanHam driver... working on another revision of it which I will hopefully post soon

your missing my point i think, i am not trying to be rude but i am at a prototype stage and doing reliability testing, before i build this thing into something portable i need to know it's reliable and i now know the flexmod p3 is indeed failed in this situation, that is why i am using a reliable power supply that has the features features that prevent a fire, you can never rely on the driver it's self for the one and only mean to prevent a failure that would result in either a fire or blowing out the power supply, and i never put the supply in the projector as it adds weight and another source for heat build up in the final device, the power supply is always external in my designs, it makes the final device smaller and i can put the power source close to an emergency shut off, one for profit company i work for does this as well, it's an added layer of redundancy, if an overload happens when the projector is mounted out of reach it is much faster to shut it down before there is a real concern for a fire or causing people to panic because of the shit load of smoke that would now be coming from the projector, with power supplies my training and experience comes from real world faults, i have seen a 120 thousand dolor heart lung machine blow the switch mode supply, shoot flames out the fan exhaust and then shut down because of a secondary layer of protection shut the system down and when i did the failure analysis found the fuses and over current protection in the supply totally fail to function, one supply even had the flame and smoke damage visible externally to the supply and find that the internal fuses where still intact and it's so called over current protection failed to function as well. That is why i find your "silly idea" to be unsafe and on a class 4 laser to be dangerous. you can not rely only on the drivers over-current "feature" to be the only level of safety, had i used a higher power diode that could result is a massive increase it the beam power and destroying the power supply as well, i always have at lest two ways to cut all power to the projector, sure it costs more at the build stage but when a fault happens it lessens the total damage to the system. In the case of this flexmod i have something that is not burned up and i may be able to repair it, since DrLava seems to not give a shit about his stuff any more i am in the process of making my own schematic, it may be in vane but it will give me a better understanding of what happened and learn from it, rather then just saying oh well and throwing it away. i still have a second working unit that i am using to better understand how these work and since mine did not catch fire i have something to work with, my current theory is it's opamp failed or the device used for current sense failed. Since my "in your view" a bench supply is not practical at this stage gives me an advantage you missed, the mosfet in the final stage is not burned up and is still working i can go farther back in the circuit and hopefully lean from the failure and fix it. right now the main reason i won't by any other thing from DrLava is the total lack of customer support. The reason i made this post was to get input from others of the possibility that i connected it wrong but now i know that is not the cause of the failure and had i used a different supply during the prototype stage i would have lost that too. so i do not consider your idea of me using a bench supply to be a valid cause of the fault, it did it's job and lessened the potential of further collateral damage.

[catalanjo]

,,,,,,,,,, i am in the process of making my own schematic, it may be in vane but it will give me a better understanding of what happened and learn from it, rather then just saying oh well and throwing it away. i still have a second working unit that i am using to better understand how these work and since mine did not catch fire i have something to work with, my current theory is it's opamp failed or the device used for current sense failed........

I totally agree with the Dragon ...hopefully he will share his valuable work with others !

@ ".....Actually I'm busy to test a driver, my company developed new, where a DC-DC-converter acts as current regulator to reduce voltage drops and heating," = I too am using this technique, ( albeit with standard convertors from China,) and the results are noticeable to say the least. However if there are significant hidden issues here, of any kind ....I am all ears !


His ideas are NOT silly!
I would be very interested in both the results of his diagnostic and any schematic he manages to create, since I have quite a few flexmods already installed "out in the field",........... as probably, do many other members here on PL.

I do give my clients proper "Service"...and finding out about causes of any possible "failure" in my own systems is a very high priority for me.

I am immensely grateful for his brave thread and shall follow his lead, and stop using them forthwith, also I will take steps protect existing set-ups adequately ,, unless he comes up with a viable solution to fix this problem .

Many thanks for sharing Draco!
Cheers

[planters]

I have used Flexmods for several years and worked with somewhere around 12 of them. Only one failed and this was a simple failure to generate output and no damage was done. These drivers were inexpensive and I never tried to operate them at their "rated maximum" of 4 A or 24 V. The maximum settings would be in the 2-3 A and 10-15V range. Also, I would not drive anything very pricey like a FC diode array with these little units. The power source was almost always a Mean Well switch mode supply that is so efficient that the heat it added to the projector was much less than the galvos, bigger diodes or god forbid, TECs!

I use bench top power supplies for powering these drivers as well as others when the lasers are still "on the bench". The convenience for setting up an experiment with these means that old saying "you can't be too rich or too thin" and which was modified by the boat building community to add "have too many clamps" can be further modified to include" too many bench top power supplies". But these are unnecessarily bulky and too expensive to pack into a mobile projection rig.

[catalanjo]

......."you can't ... have .......too many bench top power supplies".

I agree totally and have a quite a few functioning fine to date and also with Meanwells.

However new setups now include 2 x P73 per flexmod, and more importantly a NUBM07 running @ 3.8amps albeit with just a 2volt drop, by using encapsulated (in ally) chinese (Fultree) 5amp DC-DC convertors and have no idea whether they incorporate an overcurrent shutdown. I would also like to move to Chinese Switchmode PSU's in the future (price) to drive TEC controllers as well as being original source for the convertors so any info both regarding flexmod failures and/or successes I find interesting.

Seems to me that it should be the driver card being a current limiting device which should have the overcurrent shutdown not the PSU, which really can't be sure how much current it is supposed to be supplying, unless you use one PSU for each function which takes up loads of space and/or requires lots of different external supply lines.

Do you think that 3.8amps is pushing the flexmods too far?

Cheers

[Draco]

it seems that just turning a flexmod on is pushing it, lol but the stated max current for the device is 4 amps and is fused at 5 amps, what i would recommend is keeping the voltage drop the driver has to deal with as low as possible to keep the power dissipated by the FET as low as possible, he lists that value at 25 watts. this is the device IRLIB9343 and here is the data sheet http://www.irf.com/product-info/data.../irlib9343.pdf

[catalanjo]

You consider 2v drop too high?
Cheers