Hey guys so ive taken on this project to fully convert the entire car to Led. I have been using 150w adjustable china boost dc step ups to power the 20w and 30w square smd chips I’ve integrated into the reverse and Daytime corner lamps. So far everything has been great. For my biggest challenge ive built a custom bracket and heat sinks to hold 6 cree xml in with custom made optics using 28mm convex lens and some 1” copper couplings. The problem im having is that even running the entire circuit of 6 xmls at 1.5 to 2 amps i keep blowing LEDs. Basically they will will momentarily all blink very bright then one will die and the other 5 will continue to work. Once i remove the dead LED no physical damage is present and when tested with an ohm meter it shows an open circuit yet when its installed it doesn’t interrupt the current flow.
Im stumped as to why they are going bad with no physical damage. Im by 6 dead! Its adding up.
I’m afraid I don’t have the technical depth to be too helpful, but I suggest you look into voltage spikes… apparently cars are pretty bad for that sort of thing.
Apparently, there are some drivers available that can handle car current — taskled.com and possibly our own PilotPTK (though his might be complete units… can’t remember).
That’s what I’ve been reading about spikes. I thought the Automotive battery would stabilize any spikes. Can anyone chime in if installing a capacitor in the input or output of the driver help level off voltage spikes?
Automotive power is UGLY. It can sag to 6V, during crank, and then the flyback when you let go of the key is horrendous. The battery is too far from the starter, way too high ESR, and hooked up with too small of wire to be of any use in snubbing that spike. In automotive, always assume that you could be subjected to anything from 6 to 30V with occasional transients of as much as 200V.
Since this is a boost driver, do you have all of the LED's hooked up in series? One big long chain?
What driver? Seeing it might help to define the problem.
I didnt think of the cb radio dc power supplies, thats a good idea.
The odd part is that these same power supplies 150w are powering most of my car led square chips with no issues but then those run anywhere from 32v-34v dc
Since im running 6 cree xmls with a max of 3.35v per chip that puts me at 20.1v so if you say an automotive system can peak 30v that may be might issue. The dc step up delivers those momentary 30v spikes which kill a chip in the circuit.
This is the pic of the build enjoy. Eventually it will total a series string of 10 xmls.
I bought one of these in my search for powering the mkr build I have planned. After bothering George from taskled for about a week with emails back and forth I found this and he said it should work provided it handles the voltage spikes. By no means do. I have any clue what I’m doing but I thought I’d share what I found that might work for me. http://www.prodctodc.com/dcdc-632v-to-0828v-converter-constant-current-constant-pressure-led-driver-p-76.html
Thanks the past few days I searched this site and consulted a couple engineers that are familiar with leds and their circuits. The root cause of them blowing is that they are linked up in series. Big no no its a long explanation as to why it blows them in series. just don’t do it. He rexomended feeding them their max current in parrallel with a buck converter like what you purchased. I got two 100w for about 20 bucks hopefully they come in in several weeks and I could report back. He also told me the voltage spikes deal wasn’t an issue. The buck converter most likely absorbs them and stabilizes the output.
The problem is likely to be the charging circuit, cars that are 12V do not in any way shape or form sag to 6v under cranking, most with a healthy battery don’t drop below 10v, unless in very cold climes, a starter would NOT turn the engine if fed with 6v, the std fit ecu/ecm’s will not work this low, better ones work down to around 9v.
There is no “flyback”, the regulator in the alternator is VERY good generally, pegging V at 14.4, unless fully calcium battery as std kit, ie NO TIN composition in the battery(used to strengthen weak lead), when it can be just under 16V at times.
The spikes are AC VOLTAGE, all alternators spike to some degree or other, usually exceptionally little, that’s the job of the Rectifier diodes, to stop AC and only allow DC current to pass, the diodes get leaky with age, unfortunately there are/were some especially bad USA made diodes that had more leaks than a politician.
I would put a 1N5408 diode in the positive line to try to combat this.
tabs
That sounds pretty accurate, it blew the leds because i put them in a series circuit. Now that they are in parallel they shouldn’t blow. When the buck converter arrives I will report back.
I will add that diode you mention thanks.
te=tabetha]
The problem is likely to be the charging circuit, cars that are 12V do not in any way shape or form sag to 6v under cranking, most with a healthy battery don’t drop below 10v, unless in very cold climes, a starter would NOT turn the engine if fed with 6v, the std fit ecu/ecm’s will not work this low, better ones work down to around 9v.
There is no “flyback”, the regulator in the alternator is VERY good generally, pegging V at 14.4, unless fully calcium battery as std kit, ie NO TIN composition in the battery(used to strengthen weak lead), when it can be just under 16V at times.
The spikes are AC VOLTAGE, all alternators spike to some degree or other, usually exceptionally little, that’s the job of the Rectifier diodes, to stop AC and only allow DC current to pass, the diodes get leaky with age, unfortunately there are/were some especially bad USA made diodes that had more leaks than a politician.
I would put a 1N5408 diode in the positive line to try to combat this.
tabs
[/quote]
They may be good, but surely they are slow. Load dump transients can last a couple of hundret ms and reach voltages >100V in the DC on-board supply system - this is simply the reality (there is endless material available regarding this topic in the respective journals if you doubt it.).
If one’s device cannot withstand that, it is not fit for automotive use and is endangering property and lives . . .
I do not belive this is the root of the OP’s problem of dying LEDs though.
Got it. From now on I'll design all automotive power supplies to be able to tolerate 10V to 16V and I'll put a schottky diode on the input to prevent AC from coming through. I'm sure nothing will blow up. <Eye Roll Smiley>
In case you don't get the sarcasm.. Here's a little research done by a small company (ST Microelectronics) that pretty much mirrors my comments.. 6V Crank, +150/-220V Ignition Spikes, 24V Jump start, 87V Load Dump. These are the types of events a circuit has to be able to survive to be automotive qualified.. See Page 6. http://www.st.com/st-web-ui/static/active/cn/resource/technical/document/application_note/CD00181783.pdf
