[dfarr67]

Two issues here:
1/ The system I purchased about two years ago is been plagued by reliability issues with LEDs going out- I have been replacing them almost as a maintenance item, one or two a month- this is an issue in itself, never mind I have been ordering replacements from the manufacturer in Australia which shipping cost and time are a factor. See pics to give me a North American source I can order 3W LEDs from, would be appreciated. While troubleshooting I didn't argue against the suggestion it was water damage being in the environment it functions in- saying that I see no evidence of water damage, corrosion, etc. The only thing I might comment on is the thermal compound is dried out and hard. Just seems to me for the premium price I paid it should be lasting much longer trouble free. As it stands with shipping costs/time/price....look elswhere if you're in the market.

3W Power Led - MakeMyLed

makemyled.com.au

2/ I'm about to throw in the towel on my fourth tc420, this is a freshwater planted tank and I didn't need anything fancy- just a ramp up and down with the ability to adjust light intensity and duration to account for algae issues I was having. The first unit purchased as an option starting flickering on the dimming stages and has since been replaced three times- I don't know whether I got into a bad run on units or I am fundamentally missing something as when you put in the time to learn the machine- it seems fairly simple. If someone with Pled could look at the program and see a flaw- it would be welcome. I can program the lighting events, clear the unit, upload, sync and run accelerated in a loop, looks good but when the laptop is unhooked the unit seems to freeze and even tho it reads the correct time, the timing is not followed and when manually switched on in the morning it will randomly switch off again. Currently am looking for a higher quality controller (Storm?) or set the controller at 75% light 100% of the time and use a basic receptacle

timer.

 

[MockoB]

On 1st point, the reason for leds burning is the leds itself, no water damage I believe. If you don’t insist buying those from North American supplier you could check AliExpress, first result I found is 50 leds for less than $5 shipping included. Of course you must be familiar with the parameters of your leds, but I suppose the manufacturer of your light will give you the specifications.
On the 2nd point. I’m not familiar with the driver you are using but I believe it’s not the correct way you could dim your light. I see on your pictures tiny smd chip PT4115, which is basically led driver. And the correct way to dim your light is to supply pwm signal on its’ PIN number 3. Which should be in the range from -0.3 to 6V. Currently, I suppose you are dimming the power supply of your light, which messes with the led driver on the board. I believe there is nothing wrong with your led drivers at all, so stop investing in those. Unfortunately I can’t help you further without having the light with me and have some measurements.

I’m not 100% sure with the above and I know there are much more knowledgeable ppl around here, so someone else could help you further.

Good luck!

I just saw that there’s not soldered jumper between the pcb boards named dim. If you could measure, if there is link between that jumper and PIN number 3 on the chip, on all boards. I suppose if you solder the jumpers, you could provide dimming signal at the one end which will be enough to drive all the chips and dim the light. But that needs additional measurements to not burn your light.

 

[jeffkrol]

Have no idea why your tc-420's should die so easily.
Certainly goes against " common experience"

As to the diodes, either bad diodes, poor thermal transfer to the "heatsink", or too much voltage.

Edit . See they may have a "driver" so change it to too much current.

Tc-420 can be used to dim 5v pwm but needs a modification.

The company rebrands tc' s but no idea if it is a normal or modded tc.

Upon further thought
IF you have one string that's problematic it could be a driver issue.

I "believe" that the power supply is 24volts sooo each on board driver is limited to that.
1 driver per about 6 diodes (like 3.4x 6).
Standard design
If the driver chips current control resisors are wrong or " open" you could have more current going to the string than it is designed for ..burning out your leds.

There are 2 resistors which I believe are the Current control signals.
Not knowing if in series or parallel ( if both are used as I believe they are) I can't determine the setpoint.

Got a good meter?
A few ways of checking driver current w ampmeter (2A or better range)
He as to be " in line" of course.
Solder 2 leads where a diode goes and measure there. There are other methods using shunt resistance but since you have or will have to remove a diode (running less than 6 diodes is no problem) you have an easy contact point.
You will need to turn the light on which is why I don' suggest " touch" measuring.
Driver at max can output 1.2 amps.
See if the driver is behaving.

Also when attaching the new diodes make sure ALL the old compound is off the circuit board and use a teeny bit of new paste
Suggest heatsink plaster as a glue type thermal compound.

Next.
Can you open up and show me the inside of one of the tc's?

 

[dfarr67]

Thanks for the reply. I have given up on the TC-420 from MMLED, I have purchased the last two off of Fleabay- except for the sticker- seems to operate identical at half the cost.
This was a package- not DIY, so any deficiencies of construction is a design flaw on their end.
Looks like specs have changed since I purchased mine.
I typically never use the LEDs to 100%, normally peak at 75% for a few hours a day.

MML 120cm Flora Sun VividMax Smart Panel- MakeMyLed

makemyled.com.au

In service power supply (original), spare purchased because of shipping times.

 

[jeffkrol]

Thanks for the reply. I have given up on the TC-420 from MMLED, I have purchased the last two off of Fleabay- except for the sticker- seems to operate identical at half the cost.
This was a package- not DIY, so any deficiencies of construction is a design flaw on their end.
Looks like specs have changed since I purchased mine.
I typically never use the LEDs to 100%, normally peak at 75% for a few hours a day.

MML 120cm Flora Sun VividMax Smart Panel- MakeMyLed

makemyled.com.au

Yea I know, just trying to figure out HOW they use a TC-420(421) with constant current drivers.
Out of the box they are designed for constant voltage.


"Dimming" the power supply would be ridiculous.
Something is missing here..

The driver chips can dim with voltage (like 0-2.5V DC) or pwm
again though if it is ONE particular string failing it is a driver (and builders) issue most likely.
Currents too high.
Check the ohms on the 2 resistors 0.43 and 0.40 Ohms each.

Dying for more info..

 

[dfarr67]

The LEDs that I have replaced so far seem to the 'lead' LED in the strip that opens the circuit for the rest. Once repaired it is another strip that goes down. So far the faults are same for each strip. Four so far. The original two MMLED controllers were deep sixed.

 

[dfarr67]

Top image was the base manual controller.

 

[jeffkrol]

So not one particular group of 6..
Hmm.. that doesn't make sense.

I've built constant voltage and constant current arrays.. what you are showing me makes no sense "TO ME"
I know there is ONE other fixture that DIY-ers have hacked in this manner. Believe it was an ada unit or something that ran constant current but people added "strip" dimmers to it.
Not a manuf. thing though.

It's a goofy way for a production unit unless I'm (and not saying I'm not) missing something.
As the TC is wired a simple resistor works but all 6 diodes need to be electrically similar.
No need for drivers.

Sorry tagging out till I can find some info on why someone designs it that way.

IF "I" used those parts you would need 2 extra wires for 24V direct to the circuit board bus.
1 wire per to each "dim" port

 

[dfarr67]

Hacked? just trying to keep what was supplied going. The original 420 controller started to flicker on the dimming end of things- I took it apart when I got the new one and the only thing I saw was a capacitor falling apart.
I don't know what to add, this is a three channel light with the colour LEDs mixed from factory in the strips- there is no separate reds, blue, etc channels.
I'm thinking the single LEDs are overheating due to thermal compound degradation.

XB2 7W 300ma PCB - MakeMyLed

makemyled.com.au

Is there a more suitable controller out there?

 

[jeffkrol]

Falling apart capacitors is never a good sign.
Nothing is jumping out..


How about a pic of the end where all the wires come in.

Really looks like double pwm.
Getting it to work " normally" could be an issue.
Getting it working without a controller may be simple.

If the orig controller is the same as your new ones and if wired the same way there "should" be no problem wiring the plus/ minus directly to the power supply .
IF'S I can' t answer.
Exactly how is it functioning now?

Safest way is testing each channel on a bench power supply set for really low current.
See if one if the cicuit boards is messed up

 

[dfarr67]

 

[dfarr67]

Not ideal- but right now have timer at the receptacle.

 

[jeffkrol]

Ok soooo basically direct 24 v dc to the board...tc-420 at 100% is straight 24v dc.

Now any "strip light" controller should work .
It may be possible to fix it correctly, changing it to 5v pwm.
More work, wire and expenses (controllers are more expensive).

First thing would be to find if the "dim" contact is continuous to the dim pin on the driver ic.

Needs to be on the same board.
Pin closest to the 0 resistor on the choke side.

https://www.electroschematics.com/wp-content/uploads/2014/07/PT4115E-datasheet.pdf

Plan would be to use a hurricane, storm or blufish dimmer.
You can modify a manual dimmer as well.
Just like the tc 420 I posted.

How it should be done.. There are things I'm not sure of.. Dim for each driver is in parallel. "Think" that is what the "0" resistor is for.

 

[dfarr67]

These boards is are firmly attached to the light housing/heat sink- not coming off unless to replace. I'm weak on the electrical side and instead of building/modifying something would prefer to just buy a well regarded controller (within reasonable $$).
Took another look at the Pled program- very corrupted- I straightened things out, but have no confidence in it now.

 

[jeffkrol]

No need to remove the boards.
Anyways I def. understand.

Bluetooth RGB LED Controller - Smartphone Compatible - 10 Amps/Channel | Super Bright LEDs

Control RGB LED strip lights with your smartphone! This Bluetooth LED controller allows for color-changing or single-color modes. Easily adjust speed, brightness, and timer settings, create custom colors, sync lights to your phone's music, or match colors in photos.

www.superbrightleds.com

Pick one....

I'm going to explain why "IN MY MIND" that is a bad design.

Starting with the TC-420 on the power supply.
It functions by pulsing the 24v output..which is fine if that was it.
So it goes from 24 to "zero" volts in a certain span of time.
Usually based on 500Hz.
So at 50% the power coming out is 24V and at 50% 0 volts

Now you are feeding this pulsed voltage to a pwm-ed buck voltage regulator.
That regulator's job is to set the output voltage to supply a set amount of current (believe in this case it is like 500mA)
It does this just like a switch mode power supply.

Now the issue..
That regulator gets 24V then needs to drop it to whatever voltage generates a 500mA draw on the led string.

So every pulse of the power supply generates a pulse on the driver that it needs to pulse to regulate the current STARTING at 24 volts and dropping the voltage.
So the voltage is see sawing constantly.. twice.

so your strings probably experience 24V (overdriven) way more times than necessary.
May explain why the lead LED burns soo easily. First to feel the "wave" so to speak.

Doesn't explain bricking the tc's though.
Back emf from the chokes dumping is a possible thought.

I also am not an electrical guru, just a hack w/ enough knowledge to fry things..

 

[dfarr67]

I'm not sure whether I'm bricking the 420's- other than the first one, but they sure aren't operating to my expectations- maybe me not understanding something, but I thought they were straight forward.
I've been looking at a few options, the Hurricane seems to have a problem with customer service, the Storm seems to fit the bill. Your opinion? Same boat? I was also thinking of just running a rotary timer- does the controller need to be in the string here- or can I just hook the power supply straight to the LEDs?

 

[jeffkrol]

AFAICT ..you can bypass the controller for 100% light

Since it isn' t my money nor can I control any variables I do need to add "at your own risk"

But it's s relatively simple system.
That ( +) part on the tc. The 2 are internally bridged.
Each negative is a strip.
Do one at a time to the power supply.

The hurricane/ storm are assuming " we" can get the driver chips to work ad expected.
They are not tc direct replacements.

Matter of fact sort of the opposite.

Manual dimmers to replace tc:

Hiletgo 3pcs DC12-24V 8Amp 0%-100% PWM Dimming Controller for LED Lights, Ribbon Lights,Tape Lights,Dimmer is compatible with Hilight, LEDwholesaler, fillite, and others' strips

Buy Hiletgo 3pcs DC12-24V 8Amp 0%-100% PWM Dimming Controller for LED Lights, Ribbon Lights,Tape Lights,Dimmer is compatible with Hilight, LEDwholesaler, fillite, and others' strips at Amazon. Customer reviews and photos may be available to help you make the right purchase decision!

www.amazon.com

 

[dfarr67]

I forgot about that pesky 100%. My no cost options are to set the 420 at 75% and run everything through a timer or use the base supplied manual controller through a timer. If the Storm is the 'opposite' isn't that what I'm looking for?
It gets even better- I just having break fast, the 420 randomly changes mode...wtf is with that. Things were looking a little promising as the 420 turned off in auto last night and turned on this morning. Putting the base controller on timer an will see what happens.

 

[jeffkrol]

No the dimming is electronically different.


Tc, manual strip dimmers use mosfet " switches" to pulse the power supply.
Storm uses logic level 5v.
You don't "pass" 24 v nor amps of current.

Storm would go to the " dim" circuit (if configured and when its daisy chained)
ASSUMING the circuit board isn't crippled this is what you need to add to use a Storm or any 5V PWM controller
You end up with 4 drivers on one dim channel. 3 dim channels

STORM doesn't pass power.. seperate circuit.

 

[dfarr67]

An update, have been using the manual dimmer switch w remote that initially was the base equipment for the LEDs. As of when I switched over- no LEDs have failed yet- it is an inconvenience to manually dim the lights before turned off by regular receptacle timer. I must had had a decent light program as algae is starting to be a concern again. I am disappointed at the availability of simple reliable controllers.