Help with intermittent fault on a CXB3590 panel
- Thread starter kushedy
- Start date
I had a problem a while ago where I screwed the cob down to much and it cracked it worked ok for a little while then stopped working but the rest of the lights in the series still turned on
justsmokedope
Well-Known Member
it looks like you have just twisted the wires together and taped together with black tape on one of those pics ? if so you really need to use something like wago connectors or you wont get a great connection.
this does remind me a bit o something i did ages ago thinking every thing was series were i did several series loops and then used a 5pole wago connector into the dirver on possitive and the neg that basicly made it parralel .
i just though also if you only have 2 cobs on they will be super bright as the drivers power will be distributed over only 2 cobs not 4
this does remind me a bit o something i did ages ago thinking every thing was series were i did several series loops and then used a 5pole wago connector into the dirver on possitive and the neg that basicly made it parralel .
i just though also if you only have 2 cobs on they will be super bright as the drivers power will be distributed over only 2 cobs not 4
Last edited:
justsmokedope
Well-Known Member
CXB3590 Cd bin 3500k attached to a HLG-240-C1750B driver.
driver output voltage range is 71-143 volts 1750ma
but your 4 cxb3590 cd bin 3500k have a typical forward voltage of 36v this = 144v so your over max or to close to the capacity of the driver you need to have a bit of wiggle room
they prob cut out when they heat up too much due to cobs drawing more volts when they heat up slightly you could try 3 cobs or a bigger driver with more volts
ok looking at the data sheet 35.5v per cob gives you 142v but thats still a bit close to max for my liking (was from looking at graph might not be cd bin ) basically if there not enough volts you either get no cobs lighting up or just a few
driver output voltage range is 71-143 volts 1750ma
but your 4 cxb3590 cd bin 3500k have a typical forward voltage of 36v this = 144v so your over max or to close to the capacity of the driver you need to have a bit of wiggle room
they prob cut out when they heat up too much due to cobs drawing more volts when they heat up slightly you could try 3 cobs or a bigger driver with more volts
ok looking at the data sheet 35.5v per cob gives you 142v but thats still a bit close to max for my liking (was from looking at graph might not be cd bin ) basically if there not enough volts you either get no cobs lighting up or just a few
Last edited:
CannaBruh
Well-Known Member
The COBs emit small light even around 28VDC, I still think there is wiring problem or something else going on.
justsmokedope
Well-Known Member
that sounds like a short there is prob power going down the metal screws into metal case (this will make the cobs glow slightly some times) that's why i use kapton tape the attach my cobs so this cant happen ,grounding the case might help
CannaBruh
Well-Known Member
What I'm saying is when you apply a constant voltage of 28VDC across these COBs that they will light which to me is suggesting that he's likely safely within the bounds of the psu's output but maybe not?
That it worked for 1.5years without issue does not align with this theory, it wouldn't have ever worked.
All wire joins have been made by twisting, soldering & then wrapping in amalgamation tape so theoretically they should be sound.
In all honesty I could have done a better job selecting the drivers. At the time I had seen a lot of info from a certain mouse like character who was all like “max out your drivers for efficiency etc”.
I have mentally looked back over that decision a few times & wished I had chosen drivers with a little more headroom.
The panel has not been earthed. Was under the impression it was not necessary
In all honesty I could have done a better job selecting the drivers. At the time I had seen a lot of info from a certain mouse like character who was all like “max out your drivers for efficiency etc”.
I have mentally looked back over that decision a few times & wished I had chosen drivers with a little more headroom.
The panel has not been earthed. Was under the impression it was not necessary
Randomblame
Well-Known Member
I'm pretty sure the shitty Ideal holders are the culprit here.
It's a series circuit, when two COB's went off and the other two stay on, there MUST be a short within the last two holders enabling the voltage flowing thru the holders.
The reason why it does not always occur could be related to the temperature. Think of a bimetallic switch that switches when it gets warmer. The same could happen in the holders and trigger a short circuit.
I just recommend replacing the last two holders first or at least checking them for leakage currents.
You need answers for three different questions.
Sit's the copper legs proper on the COB contact surface?
Is there any voltage flowing thru the holder without a COB?
Occurs some voltage flow when they get hot?
Out of 10 ideal holders, I have to bend at least two of them to get good copper leg contacts. They bend when tightened too tightly and they tend to deform over time. I don't like them for that reason!
It's a series circuit, when two COB's went off and the other two stay on, there MUST be a short within the last two holders enabling the voltage flowing thru the holders.
The reason why it does not always occur could be related to the temperature. Think of a bimetallic switch that switches when it gets warmer. The same could happen in the holders and trigger a short circuit.
I just recommend replacing the last two holders first or at least checking them for leakage currents.
You need answers for three different questions.
Sit's the copper legs proper on the COB contact surface?
Is there any voltage flowing thru the holder without a COB?
Occurs some voltage flow when they get hot?
Out of 10 ideal holders, I have to bend at least two of them to get good copper leg contacts. They bend when tightened too tightly and they tend to deform over time. I don't like them for that reason!
Last edited:
CannaBruh
Well-Known Member
Voltage doesn't flow 
For the two COBs which were not emitting to have been shorted the remaining two lit would have to eat all the potential and current delivered by the driver, which might spell failure (can't be bothered with math) and at the least would be noticeably brighter (assumption)
When lights are all on, what is voltage across the series circuit?
What is the voltage across same points when the two lights do not fire, what is the voltage 'across' each of the non-lit COBs, what is the voltage across the lit COBs?
For the two COBs which were not emitting to have been shorted the remaining two lit would have to eat all the potential and current delivered by the driver, which might spell failure (can't be bothered with math) and at the least would be noticeably brighter (assumption)
When lights are all on, what is voltage across the series circuit?
What is the voltage across same points when the two lights do not fire, what is the voltage 'across' each of the non-lit COBs, what is the voltage across the lit COBs?
dstroy
Well-Known Member
Before I go into what I would do if it were me troubleshooting,
Sometimes, power supplies can fail but only under load. They will show nominal voltage and current when tested with no dummy load. Doesn't happen very often this way though.
First, I would look for anything obvious, such as arc burns or a funny smell or melted or deformed plastic.
Then, I would check that the voltage output of my driver was correct. If it was not I would check that the input voltage was acceptable, if the input voltage was acceptable,
I would then test the wires and connectors which are used for low (less than 4, preferably 1 or 2 ohms) resistance while moving the wire around, if the resistance reading was jumpy then I would know I had a bad connection.
I would then repeat the this test point-to-point across all wires.
Then I would test for resistance from the LED array holder connection, to the pad on the LED array PCB, this reading should also be low or 0.
I would repeat that test for all LED arrays and holders. If I found one that was high, this would cause a lot of heat and maybe cause an array to fail or cause the power supply to shut down, this goes for any high impedance connection.
Then, I would use a diode test on all LED arrays (which is much too primitive to tell you that all of the diodes in the array are functional) to see if we had a "bad diode". Very limited information from doing this test but it's better than nothing.
Hope you find what's up with it.
Randomblame
Well-Known Member
I think, you do not understand what I mean.Voltage doesn't flow
For the two COBs which were not emitting to have been shorted the remaining two lit would have to eat all the potential and current delivered by the driver, which might spell failure (can't be bothered with math) and at the least would be noticeably brighter (assumption)
When lights are all on, what is voltage across the series circuit?
What is the voltage across same points when the two lights do not fire, what is the voltage 'across' each of the non-lit COBs, what is the voltage across the lit COBs?
With a short in the holders there is no eating current. Current remains the same with a CC driver(depending on dimmer settings).
So maybe 140v(35v per COB) with all 4 lit up and ~70v with only too COB's.
There are no mystical resistors or something else pulling current, it's simply a holder short so that the COB get neither voltage nor current. Like when you use only two COB's wired in series. If there where a short to the heatsink or the case it would trigger the fuse when properly grounded.
Is my english so bad, mate? I would bet it's because of the holders!
Last edited:
Morning all (UK based). Still no change. All 4 cobs are functioning. The issue has not re-occurred for over a week now. I'm not going to rest on my laurels though. I will be taking the panel out of the grow tomorrow just before lights off. At that point I will be able to strip it down & give everything a good visual check. @Randomblame what you suggested about the cob holder carrying the current around the last 2 cobs makes sense. It's a possibility based on your logic. That said it doesn't explain why it has been happening intermittently, in fact quite rarely fortunately. There is no pressure applied to the cob holders or the wiring so I'm not sure what would be triggering it.
Again, thanks for everyone who has chimed in & suggested ways to deal with this. I haven't done a huge amount of electrical testing/trouble shooting so I am hoping whatever the cause is sticks out fairly obviously tomorrow when Ii take the panel apart.
Again, thanks for everyone who has chimed in & suggested ways to deal with this. I haven't done a huge amount of electrical testing/trouble shooting so I am hoping whatever the cause is sticks out fairly obviously tomorrow when Ii take the panel apart.
CannaBruh
Well-Known Member
Your English is excellent but convention is that potential can be had across two points and current flows but voltage does not.
Randomblame explained, "The reason why it does not always occur could be related to the temperature. Think of a bimetallic switch that switches when it gets warmer. The same could happen in the holders and trigger a short circuit."
One holder or wire is probably shorted to the frame/heat sink and one shorts intermittently. If you check, I'll bet you find voltage between the frame and the output wires from the driver.
Sorry yeah, I recall that now. My memory is not as good as it used to be & there has been quite a few different suggestions with this.
Well the panel is coming down in a little over an hour so its crunch time. Worst case scenario nothing is obviously wrong but I re-wire with solid core anyway.
Best case scenario I find the culprit re-wire. Hopefully the latter for peace of mind.
Well the panel is coming down in a little over an hour so its crunch time. Worst case scenario nothing is obviously wrong but I re-wire with solid core anyway.
Best case scenario I find the culprit re-wire. Hopefully the latter for peace of mind.
Randomblame
Well-Known Member
Totally agree! My bad...
I push the blame to SSH dabs yesterday, was probably a bit too much, LOL!
CannaBruh
Well-Known Member
Before you disassemble you might gain access to all the series wiring and give each a wiggle with an orange peeler (be safe)
See if it doesn't act up from disturbing the wire, might narrow it down.
Best of luck!
Well I took the panel into a nice bright room & gave it a good visual inspection & pulled on all the wires to look for loose connections. Nothing showed up.
I disconnected the 5 fans & still nothing obvious on the wiring under the fans. At this point I had to start cutting the wire out of the panel. All connections into the Ideal Holders appeared to be secure & nothing out of place.
I may have found the culprit (fits the bill anyway) with one of the wires I cut out. One of the solder joins on one of the cables had a large chunk of solder with a sharp point on it where the 2 cables were soldered together. This point on the solder looked like it had cut through the amalgamation tape I had wrapped around the join. I could only just about see it when the light bounced off the point.
Without cutting the wire out I would have never found it. Anyway, I have now re-wired with solid core cable & managed to wire it without having to make any cuts or joins so less margin for error. I have also replaced the potentiometer with a Vishay potentiometer.
It’s back in the grow & working as normal so its just a case of monitoring going forward.
I’m not necessarily 100% convinced this was the issue which may sound daft considering but the piece of wire I mentioned above was positioned in such a way that it did not have contact with the heatsink or anything else bar where each end connected to the Ideal Holders. I hope it was the issue but only time will tell now.
Thanks again to everyone who took the time to help with this.
I disconnected the 5 fans & still nothing obvious on the wiring under the fans. At this point I had to start cutting the wire out of the panel. All connections into the Ideal Holders appeared to be secure & nothing out of place.
I may have found the culprit (fits the bill anyway) with one of the wires I cut out. One of the solder joins on one of the cables had a large chunk of solder with a sharp point on it where the 2 cables were soldered together. This point on the solder looked like it had cut through the amalgamation tape I had wrapped around the join. I could only just about see it when the light bounced off the point.
Without cutting the wire out I would have never found it. Anyway, I have now re-wired with solid core cable & managed to wire it without having to make any cuts or joins so less margin for error. I have also replaced the potentiometer with a Vishay potentiometer.
It’s back in the grow & working as normal so its just a case of monitoring going forward.
I’m not necessarily 100% convinced this was the issue which may sound daft considering but the piece of wire I mentioned above was positioned in such a way that it did not have contact with the heatsink or anything else bar where each end connected to the Ideal Holders. I hope it was the issue but only time will tell now.
Thanks again to everyone who took the time to help with this.
