Just blew 4 diodes!!!

[whatsupadrian]

Hello.
A little frustrated today. Last night I got my quad kit, so i started populating it with diodes. Got 4 LPC-826s wired up in parallel, no lasorbs or balast resistors at the moment. Have a flexmod p3 driver, and was trying to use the PSU that came with the http://www.photonlexicon.com/forums/...le-info-thread, which is 5v 8A. So i figured to use this PSU with flexmod and 4 LOC diodes would work fine. I tried following the flexmod instructions, jumpering int to v+ and m+ to v+ also. V+ and V- to the PSU, and had my ampmeter on the LD+ and LD- leads. I could not get the red light on the driver to turn off. So then I decided to try and power a testload(burnt out 445 diode), and still got the red light. Figured the load wasn't right and so I decided to try and power the 4 LOC with the flexmod(i know stupid). Got a flash of red then the red light. I added 3 volts to the PSU thinking i may just be underpowering it and the diodes turned on for a few second then one blew then 2 second later the other 3 blew.
I was soo frustrated at that point that I didnt really care. Ill have 4 fresh diodes this afternoon and I want to give it another shot. Could anyone please offer some advice on what I may have been doing wrong. I believe the PSU I'm using may be the problem but shouldn't 4 LOC be fine since they are all wired in parallel? Doing some reading I notice some people using balast resistors when wiring in parallel. Are these required, if so what value resistors do I need?

Thanks
Adrian

[neutrino]

Did you connect the interlock connection to v+. If you don;t do this the red led will stay on and you'll get no output other than a brief flash at the beginning. You diodes may be ok as the interlock circuit was in place.
I could be wrong as I don;t have the manual in fron t of me.

CHeers
Rich

[polishedball]

Where are you getting no wait 826's from? There is no reason it wont work across your meter, you need it to to set the starting current. Connecting a LOC to a driver that comes as shipped will mean almost certain death (with 4 in parallel you could be saved if wired correct. You should use shunt resistors iirc double check this 1ohm 1/2w then you can measure the current as the voltage drop across the resistor.

Get it working across your meter, that needs to be a complete circuit prior to applying power on the flexmod.

Hello.
A little frustrated today. Last night I got my quad kit, so i started populating it with diodes. Got 4 LPC-826s wired up in parallel, no lasorbs or balast resistors at the moment. Have a flexmod p3 driver, and was trying to use the PSU that came with the http://www.photonlexicon.com/forums/...le-info-thread, which is 5v 8A. So i figured to use this PSU with flexmod and 4 LOC diodes would work fine. I tried following the flexmod instructions, jumpering int to v+ and m+ to v+ also. V+ and V- to the PSU, and had my ampmeter on the LD+ and LD- leads. I could not get the red light on the driver to turn off. So then I decided to try and power a testload(burnt out 445 diode), and still got the red light. Figured the load wasn't right and so I decided to try and power the 4 LOC with the flexmod(i know stupid). Got a flash of red then the red light. I added 3 volts to the PSU thinking i may just be underpowering it and the diodes turned on for a few second then one blew then 2 second later the other 3 blew.
I was soo frustrated at that point that I didnt really care. Ill have 4 fresh diodes this afternoon and I want to give it another shot. Could anyone please offer some advice on what I may have been doing wrong. I believe the PSU I'm using may be the problem but shouldn't 4 LOC be fine since they are all wired in parallel? Doing some reading I notice some people using balast resistors when wiring in parallel. Are these required, if so what value resistors do I need?

Thanks
Adrian

[whatsupadrian]

A member of LPF who runs group buys for diodes lives 20 mins from me. I will pick some up after work and give this another shot.

[Solonar]

Ballast resistors prevent 1 diode from hogging the current and blowing itself up.

If you have 4 precision ballast resisters, all 4 diodes should essentially split the current 25% each. If you pick the correct resistance you can also use them as current limiting resistors and maybe save the lot when 1 blows.

[whatsupadrian]

OK So I got some new diodes today and didn't hesitate to get them installed, this time i attached lasorbs.
Yes i know my solder joints suck. I had some issues cleaning the solder off when removing the blown ones and getting the new diodes in as flush as possible. I tested all the joints everything seems to be in order.

What size ballast resistors should I use?

So this time I want to get the current set before connecting to the diodes(doh). What kind/size test load can I make to emulate 4 lpc-826 in parallel and their ballast resistors?

[mccarrot]

the opperating voltage of the red diodes is aprox 3V,

the Vf of a 1N4007 diode @ 200mA = 0,8V, so if you put 4 diodes in series, and put this is series with you amp meter you can do some testing.

[FourDee]

I've said it before and I'll say it again ... Wiring diodes in parallel is a bad idea

[andy_con]

in your 6th picture it looks like your diodes arent in the mounts properly?

[SrS]

Because it seems there are quite some questions about parallel wiring diodes here are some pictures. Notice the examples are not completely accurate but it will give you an idea of what’s going on

If you put equal components in parallel then the current will be divided like you expect. In this case a driver is used to source 1A, this current is then divided over the 4 diodes so 250mW flows thru every diode



But every component is unique and as slightly different characteristics. Perfect components like above only exists in simulation. The trouble maker in this case is the different forward voltage drop in each diode (Vf or Vop) In the next picture D1 has the lowest drop D2 a bit more and D4 the highest.



Now 350mA flows thru D1 and only 150mA thru D4. As a result D1 generates far more heat compared with D4, this also influences the voltage drop of each diode! Because D1 gets more current its likely to have a shorter life.



If we put a resistor in series with every diode then yes, it will balance the current flow a bit but NOT 25% for each. Precision resistors are common and the result of the difference between them is very little. But wire wound resistors (and others) also act as an inductor and this will have impact when modulating.



Now if one of the diodes fails in such a way that no current can flow thru it anymore then the other diodes gets the extra current because your driver is configures to source 1A and will increase the voltage needed to push 1A thru whatever is connected. This would most likely result in D2 failing soon followed by the other diodes.



Another way is to put the diodes in series. This way the same current flows thru each diode but it only works when the anode(+) and cathode(-) of the diode are isolated. In this case the driver is configured to source 250mA. Because the voltage drop of each diode is added, the driver need to be able the provide 250mA @ 13V. Because there’s some voltage drop in every driver you probably need to provide the driver with a +14V power supply.

Wiring your diodes in parallel is great if using inexpensive diodes and your first & second priority is price. But make sure none of the diodes are getting close to the maximum current! Don’t rely on the current set on your driver divided by the number of diodes. If using more expensive diodes like the 637nm & 640nm then be safe and invest a little more in a quality driver with separate outputs.

in your 6th picture it looks like your diodes arent in the mounts properly?

Andy is correct and I hope you only wanted to show something on photo. If you power your diodes like this they can't loose their heat and will most likely fail. Not to mention the horrible stability.

- Dennis