I have a hydroponics box, similar to this one. The problem is that the light is completely unreliable: it switches off at random times and then often refuses to be switched on again.
I would like to bypass the control unit and rather control the light using a timer on the power supply.
However, I am not sure how to do it and am afraid of burning destroying the light, so I am asking for help.
I am capable of some simple soldering, but nothing fancy.

Here is a picture of the light:


Behind the light is the control board:

which has touch-buttons on the back side (which is a top side when mounted):


On the board, the right connector is the power supply, with 24V between '+' and J3 pins. The J2-B and J1-W connectors lead to the corresponding connectors on the light panel, with 'W' provided warm light and 'B' adding a blue component.
The 3 modes selectable by the touch buttons lead to the following voltages on the light connectors:
- Normal: ca. 22.4 V on both connectors
- Enjoy: ca. 21.5 V on W and nothing on B
- Grow: ca. 23.3 V on W and nothing on B

In addition to light, the board controls a pump in the water tank. When it runs, there is 4V between J3 and C pins on the power connector (and therefore 20V between '+' and C).

The easiest fix I can think of is to connect both W and B connectors to the power supply, so they will both get 24 V - will this work? It is higher voltage in any of the modes...
This will give the strongest light, but a bit too cold. How difficult would it be to run the blue light on, say, 50% brightness?

When it comes to the pump, I guess I will try to connecting it to a separate 4V power supply, which would allow me to control it independently on the light .

Or are there some better solutions I am not seeing?

 

The Light may be over-heating.
This is the most likely reason for it turning-off and refusing to turn on again.
.
.
.

 

The Light may be over-heating.
This is the most likely reason for it turning-off and refusing to turn on again.

Sounds reasonable, except that sometimes it switches off just a few minutes after being switched on.
But if that was the case, what do I risk if I bypass the control board?

 

Smoking all the LEDs.

If You have to ask how to accomplish this, then You are not qualified to mess with it.
Even a very experienced Electronics Engineer could
easily miss just 1 important detail, and accidentally smoke the whole thing.
I would suggest calling the manufacturer and seeing what they have to say.
If it's made in China, you're on your own.
.
.
.

 

Smoking all the LEDs.

If You have to ask how to accomplish this, then You are not qualified to mess with it.
Even a very experienced Electronics Engineer could
easily miss just 1 important detail, and accidentally smoke the whole thing.
I would suggest calling the manufacturer and seeing what they have to say.
If it's made in China, you're on your own.

It is made in China, so I am willing to risk destroying it - it certainly does not work as it is...

 

The Light may be over-heating.
This is the most likely reason for it turning-off and refusing to turn on again.

I tested it again: I removed the plastic cover over the light and switched the system on (after several days off) - and it switched off after a couple of seconds.
Hence, I very much doubt it is because of overheating.

 

I would estimate that it will be quite impractical to bypass anything on your Control-Board.
But, while the Light is working, You could measure the Current flowing on the wires
connecting to the LED-Panels, ( measure the wires individually, NOT as an attached pair ).
The Voltage will also have to be measured where the wire pairs are
attached to the Board and marked plus and minus.

Then a plan could be formulated to either
purchase a factory-made Constant-Current-Power-Supply that will work,
or build your own Power-Supply.
.
.
.

 

I see there is a power indicator. What’s its status during fault operation. If the water level is low, do the lights shut off, does the the low level warning work, independently of other functions. I would check the output of the membrane switches. I’ve seen them fail.

 

I would estimate that it will be quite impractical to bypass anything on your Control-Board.
But, while the Light is working, You could measure the Current flowing on the wires
connecting to the LED-Panels, ( measure the wires individually, NOT as an attached pair ).
The Voltage will also have to be measured where the wire pairs are
attached to the Board and marked plus and minus.

Then a plan could be formulated to either
purchase a factory-made Constant-Current-Power-Supply that will work,
or build your own Power-Supply.

I have measured both voltage and current on the two light connectors (W and B), in the three modes of the lamp:
- Normal: 22.3 V * 0.26 A on W, 22.3 V * 0.19 A on B
- Enjoy: 21.4 V * 0.21 A on W and nothing on B
- Grow: 23 V * 0.34 A on W and nothing on B

Was this what you were asking for?

 

I see there is a power indicator. What’s its status during fault operation. If the water level is low, do the lights shut off, does the the low level warning work, independently of other functions. I would check the output of the membrane switches. I’ve seen them fail.

There is not power indicator on mine - as I wrote, it is a slightly different product.
However, the unit beeps before switching off.
As for the low-water level indicator, there is a little floater in the tank, but I am not sure if it does anything - but if it does, it would be detecting the presence of the floater at the bottom.

Which reminds me that I forgot to post pictures of a board at the bottom of the device:

There, DC1 is the power input (24 V), DC3 is a jack connector for the lamp (i.e., going into [+, J3, C] pins on the control board, and DC2 is a connector for the water pump.
The component on the very right (with 251 024 on it) sits just under the float element, so it could be detecting it...

Here is a different view of that component:

 

"" Was this what you were asking for? ""

I don't really need to know it personally,
but if You should decide that You want to purchase a "Constant-Current"-Power-Supply,
or build one yourself,
then You will need to know this information so that You don't inadvertently smoke your LEDs.

What You will be looking for is either .............
2 - CC-Power-Supplies that are adjustable up to at least ~350ma,
and can supply at least ~25-Volts while doing it,
Or,
a single CC-Power-Supply that can provide ~350ma at at least ~50-Volts.

This is a little less than ~10-Watts X 2, or, ~20-Watts for the single Supply.

2-separate CC-Power-Supplies would provide more versatility,
if "changing-Colors" is important to You.
.
.
.

 

I see there is a power indicator. What’s its status during fault operation. If the water level is low, do the lights shut off, does the the low level warning work, independently of other functions. I would check the output of the membrane switches. I’ve seen them fail.

I was kind of hoping the problem could be in the water-level sensor, but it is not: when I push the floater down, it sounds an alarm, but the light is not affected...

 

"" Was this what you were asking for? ""

I don't really need to know it personally,
but if You should decide that You want to purchase a "Constant-Current"-Power-Supply,
or build one yourself,
then You will need to know this information so that You don't inadvertently smoke your LEDs.

What You will be looking for is either .............
2 - CC-Power-Supplies that are adjustable up to at least ~350ma,
and can supply at least ~25-Volts while doing it,
Or,
a single CC-Power-Supply that can provide ~350ma at at least ~50-Volts.

This is a little less than ~10-Watts X 2, or, ~20-Watts for the single Supply.

2-separate CC-Power-Supplies would provide more versatility,
if "changing-Colors" is important to You.

Thanks again.
I (obviously) did not know that LEDs are current-controlled.

If I need to extra power supplies, it would probably be cheaper (and safer) to buy a new box - I need to do some searching.

You mention the possibility to use just a single power supply - how would I have to connect it, if I wanted some blue light as well as the warm one?

 

You would do what's called a "series" connection,
+LEDs- .......... +LEDs-,
"LEDs group-A" would get its negative wire connected to the positive wire of "LEDs group-B",
then the Power-supply would be connected to the positive of "group-A", and the negative of "group-B".

This would require the Power Supply -"to-be-capable-of"- producing over ~50-Volts,
but the Current would still be regulated to ~350ma, (or less).
If each set of LEDs had its own Power-Supply,
the minimum-Voltage requirement would be roughly cut in half to ~25-Volts.
( Different types/Colors of LEDs have a slightly different "Forward-Voltage-Drop",
so the Voltage-requirements of the 2 groups of LEDs might be slightly different )
If the minimum Voltage requirements can not be met by the Power-Supply(s),
it will be impossible to achieve the ~350ma maximum Current,
so the LEDs would be "dimmer", to some unknown degree, than their normal, full brilliance.
.
.
.

 

I tested it again: I removed the plastic cover over the light and switched the system on (after several days off) - and it switched off after a couple of seconds.
Hence, I very much doubt it is because of overheating.

It is possible the problem is a bad solder joint, and when it warms up, the board is warping slightly, breaking a connection. I would take a really close look at connections after it turns itself off, and see if you can find any breaks.

 

It is possible the problem is a bad solder joint, and when it warms up, the board is warping slightly, breaking a connection. I would take a really close look at connections after it turns itself off, and see if you can find any breaks.

I agree that this is a possible explanation - the sound it makes when it switches off is the same as when I switch power off.
(Actually, my very first suspicion was a bad PSU, so I bought a new one, but it did not help.)

I have had a look and I did not see any obvious problem. The most suspicious solder joint is perhaps the one on the left on the picture below, but I this it is OK anyway - it is joint for the water pump connector...

 

You would do what's called a "series" connection,
+LEDs- .......... +LEDs-,
"LEDs group-A" would get its negative wire connected to the positive wire of "LEDs group-B",
then the Power-supply would be connected to the positive of "group-A", and the negative of "group-B".

This would require the Power Supply -"to-be-capable-of"- producing over ~50-Volts,
but the Current would still be regulated to ~350ma, (or less).
If each set of LEDs had its own Power-Supply,
the minimum-Voltage requirement would be roughly cut in half to ~25-Volts.
( Different types/Colors of LEDs have a slightly different "Forward-Voltage-Drop",
so the Voltage-requirements of the 2 groups of LEDs might be slightly different )
If the minimum Voltage requirements can not be met by the Power-Supply(s),
it will be impossible to achieve the ~350ma maximum Current,
so the LEDs would be "dimmer", to some unknown degree, than their normal, full brilliance.

I did some searching and found some products that I believe could do the job and do not cost that much.
I would very much appreciate if you could check that they look correct.
First, there are AC-DC LED drivers, where some have selectable output currents, for example this one.

Then I found DC-DC step-down LED drivers, such as LDD-L and LDDS-H drivers from "Mean Well" - would these work? And do I understand the documentation correctly that they should work without dimming, i.e., only with 2 in- and 2-out wires?
I would very much prefer working with low-voltage DC, compared to AC...
The LDD-L variants go down to 300mA and the LDDS-H down to 250mA - still a bit too much, but maybe there are other variants I have not found yet...

 

"" I would very much prefer working with low-voltage DC, ""
Where will You get ~36-Volts-DC from ?

The first one accepts ~230-Volts-AC.

In either case, You will need 2,
because the Output Voltages on all of them are
too low for a series wiring scheme,
which requires ~50-Volts-DC.

Keep looking, everybody and their brother makes these things ..........
.
.
.

 

"" I would very much prefer working with low-voltage DC, ""
Where will You get ~36-Volts-DC from ?

Why 36V?
I re-read the documentation and found that the output voltage is 3V and 4V lower than input, respectively for the two types.
Should not this mean that to get, say 23V, I would need a power supply with 26V or 27V DC?

On the other hand, it must be doable with 24V input, since the box is doing it - I just don't know why.

The first one accepts ~230-Volts-AC.

What I have not found is what determines the output voltage of these things - they all have a specified range and I do not know how far from 23V I can go before doing some damage :-(
[/QUOTE]

Keep looking, everybody and their brother makes these things ..........

Possibly, but I still have not found a single one I could be confident enough about to actually try it - any suggestion?

 

I thought I will measure the voltage drop on the two light circuits (W and B) in the different modes, by measuring voltage between J3 pin (negative input) and '-' pin of the light connectors (both at the control board).
This gave me the following:
- Normal mode:
- W: 0.5V (drop from 23V on input to 22.5V in the circuit)
- B: 0.7V (drop from 23V to 22.3V)
- Enjoy mode:
- W: 1.9V (drop from 23.5V to 21.5V)
- Grow
- W: 0V .. both input and the circuit have 23.2V

I must admit the last result confused me completely. I thought I was beginning to understand LED circuits, but not I see I still do not have any idea how it works...