Lighting Grow Northern Live Tech Talk

[HubbaBubba]

If that's the case i'm sure we can arrange to exchange just the driver

Thank your for the kind offer, the risk of returning my HS1 is too high. If I mail it, I need to show ID, and fill out customs forms.

We could spend months mailing parts... (I would have pick up the parts at the customs office, a nightmare) or I could just fix it.
We thought it might be the driver, right. But why did the driver forget its programming?

This was pretty easy to spot with my naked eye:

That is a NTC Thermistor in the center of the photo between the yellow LEDs. The brown smudge around it means that something bad happened. At first I thought the Thermistor failed, rare but it happens.

Anybody wanna guess what really happened? It is not directly Grownortherns fault, they use 4 parts from other manufacturers in a nice box. They used good stuff, I didn't see any damage from assembly (like pinched wires or missing parts) I think this is one of those cascade of errors things.

The box basically has 4 parts: 220 electrical, Driver, LED, and Fan/heat sink.

The circuit tracing (wires) on the copper heat sink for the LEDs failed. I reconnected the thermistor and the tracings failed again, and again, and again in different places. Pretty sure that I have isolated and duplicated the problem. Defective tracings (wires) from Luxeon (they make the LED unit) in multiple places.

 

[HubbaBubba]

Not one of the 37 LEDS failed, they all still work. The wiring on the copper heat sink is weaker than the LEDS.

So what about the "Cascade of errors"?
When a wire is too weak for the electric it carries, it gets hot. If it gets too hot it melts.
The wires to the LEDS melted on my HS1. I tested it several times with another 75 watt driver.

A related factor might be that the HS1 has Thermal Pads instead of thermal paste, big difference in thermal conductivity. Paste is messy, thermal pads are more convenient for manufacturers. But they don't conduct heat as well. They are good enough for low power stuff, even medium stuff like some CPUs, but you never see a thermal pad on a high power CPU or GPU.

That is the Thermal pad in the top left, resting directly on top of copper. On the bottom left is the mating surface of the Nuventix Synjet, is that painted? Wouldn't bare metal be better for thermal transfer? Anyhow, the weak tracings get hot and melt. The Xitanium driver senses a problem and goes into low power mode.

And that is the end of the cascade because the Xitanium driver is stuck in low power, 13 watts measured at the wall.
It is a "smart" driver but has amnesia. Apparently when it went into low power mode it forgot its programing. So I have an HS1 with defective LED wiring and a Driver that needs reprogrammed.. I need to replace the Driver AND the LED unit.

I have simplified 20 hours of troubleshooting and tried to write clearly with helpful photos.
If you don't like my posts please don't delete them, just move them somewhere more appropriate or ask me to edit my mistakes.
Surely some of this info is helpful to at least one HS1 owner.

"Many Bothans died to bring us this information."

 

[GrowNorthern]

OK, thanks for the pictures and summary. It would be super useful to get the unit back and look at the issue in more detail and send parts of to the appropriate suppliers however I understand your situation so let see if we cant arrange a solution.

Regarding your analyses
1) The xitanium driver; It is and TD+DALI control. TD meaning "touch dim" connecting the pins in the blue WAGO connector can change the power setting of the driver. (It doesn't look like it is the issue though it was a possibility given the previous info we were given)

2)The thermal pad. The MCPCB/ LED board is made of copper to spread the dense power configuration of the HS1. The material thermally bridging the heatsink and the LED board is 0.13mm graphite which has poor heat spreading in the Y axis but has much higher conductivity/lower thermal resistance than thermal grease in the X axis (the 0.13mm plane) and is used in all quality LED designs as opposed to low cost silicone grease. You can read more here http://industrial.panasonic.com/www-data/pdf/AYA0000/AYA0000CE2.pdf.

3)The heatsink is 6063 aluminium which has been anodised which is a common process for any heatsink for electrical components due to the fact that it makes the surface non electrically conductive and therefore prevents galvanic corrosion and other grounding issues. Anodising does have an effect on the radiation of IR; Black will irradiate more IR and absorb more and silver will irradiate less but also reflect more. The anodising does not effect the internal conductivity of the aluminium.

4) Rather than the track and trace not being sufficient (which it will be as the supply to the NTC is about 10ma) I think the NTC has been soldered incorrectly. We do check every board with the desktop magnifier and also read the running temps before assembly so I think the problem has become more visible during use and also the fact your unit is dimming shows it is still semi working just not correctly.

To fix
1) If you don't want to exchange any parts you could simply snip the yellow and green wires and this will remove the thermal dimming function. However If you don't assemble the unit back to the correct torque you run the risk of the LED board being damaged by heat.

2) We can exchange the LED module by post, i'm sure this would not incur any ID issues if we say there is no commercial value to the package. Again assembling at the right torque is important but atleast the LED's would dim if the torque was insufficient.

Also regardless of what we do its important you keep the silicone washers that where underneath the nut holding down the heatsink, these provide secondary insulation from ground and are an important aspect of the parts longevity

 

[HubbaBubba]

Thanks for taking the time to share your expertise in a very well written post.

And thanks for not taking offense at my layman guesses.
I don't know why there was a problem with the Thermistor, but at this point it doesn't matter, we can assume I have done more damage to the MCPCB/ LED board by NOT torquing it down properly. My damage, my responsibility, end of issue for my HS1. You have resolved this issue completely.

I do have a 2nd HS1 that I don't write about because it is working fine, no problems.

If I may ask, why does the wall plug have 2 pins instead of 3. isn't it dangerous not have a ground pin in a grow room environment.

 

[GrowNorthern]

No problem, happy to help. Are you using a nut spinner to take the nuts off and replace them at the moment? If not I would recommend getting an 5.5mm nut spinner, you can buy they for like £3-4.00 in the UK. This will allow you to apply a sufficient torque by hand but please remember the silicone washers.

There is a ground pin on the wall plug are you using the EU or UK power cable? The ground connects to the LED driver by wire and the casing on the HS1 by a grounding bar on the back of the power WAGO connector.

 

[GrowNorthern]

OK, sorry I see your picture now. The "SHUKO" EU plug has grounding clips on the top and bottom of the plug.

 

[HubbaBubba]

No problem, happy to help. Are you using a nut spinner to take the nuts off and replace them at the moment? If not I would recommend getting an 5.5mm nut spinner,.

Yes, I was using a 5.5 mm socket. Had to buy one just for the HS1, that is an unusual size. BUT, I had "help" and I wasn't watching my helper all the time or double checking that the torque was sufficient. So, really, the issue ends here, my fault, my responsibility for working on things I don't fully understand.

OK, sorry I see your picture now. The "SHUKO" EU plug has grounding clips on the top and bottom of the plug.

Now I understand. I don't have SHUKO "F" Receptacles on my power outlets so I added a ground pin to the plug to fit type "E"
You have answered all my questions, thanks again.

 

[DarkGreen]

I've had my HS1 a week now and I'm happy with it, I'm not sure I understand why it needs to be so close to the canopy (400mm) with the 60x60 diffuser, as it was my understanding that with a tighter beam than the 80x80, but I am pleased with it so far. I am currently growing alongside my ms0004, which is at 24" (is this correct? Auto day 25 from seed) so it will be interesting to note a closer comparison in wattage than most tandem grows with HS1

 

[MikeyB420]

Hi @GrowNorthern

Any idea of when you will have the 60 x 60 diffusers in stock again? I think I need to 'refocus' on certain plants!

Getting back to 'Tech Talk', regarding the control/dimming factor? I'm assuming you have the basic software sorted by now and with talking to Daniel, there will/soon be a USB-DALI program available but, would need a computer attached at all times? I'm on a waffle (forgive me?)

Would a Raspberry Pi suit the needs? Once set up.................Job Done??

Incorporated in the right sort of box/enclosure!



MikeyB420

 

[GrowNorthern]

Hi @Grow Northern

Any idea of when you will have the 60 x 60 diffusers in stock again? I think I need to 'refocus' on certain plants!

View attachment 520224

Getting back to 'Tech Talk', regarding the control/dimming factor? I'm assuming you have the basic software sorted by now and with talking to Daniel, there will/soon be a USB-DALI program available but, would need a computer attached at all times? I'm on a waffle (forgive me?)

Would a Raspberry Pi suit the needs? Once set up.................Job Done??

Incorporated in the right sort of box/enclosure!



MikeyB420

Hi MikeyB,

Sorry I missed this before (its @GrowNorthern for future ref) I will have to get back to you regarding the 60x60 diffusers I don't have an exact date on this at the moment.

Regarding our control solution we have been developing for sometime for the Holographic series; originally we developed a USB bridge which needed to be connected to a computer and allowed maximum functionality of the HS1's digital features. Along with basic dimming control it could recall temperature, permanently adjust the output of the synjet and LED module it could also program sequences such as Constant light output (this increases current of 50,000 hours to maintain the same output) and full grow cycle light on/off sequences. These would be saved to the driver memory so once disconnected from the bridge the driver would remember the settings. However this ended being a really high cost solution and did not allow you to run any sequences or turn the lights on/off in realtime without it being connected to a PC.

Our new solution is much more cost effective, it can be controlled by bluetooth and NFC from any smart phone with the app. It has reduced functionality but can remote turn on/off, dim and sequence full light schedules individually or in scenes for 32 units on one single bus in realtime. One single bus means all the units are linked up in a singular parallel DC circuit but DALI's digital addressing allows individual control of a single or group of units through one control. This is physically wired up the exact same way as the power with Wago T pieces and 2 pole link cables going from one unit to the next and the control box sits at the front of the circuit/chain.

We are going to release some more information about products we have in the pipeline for 2016 very soon.

If you wanted to have mess around with a Raspberry PI, all basic DALI protocol is available opensource. You would however need a bus supply and you may find it easier to work with the an external bridge from the PI for the DALI signal.

Here is a link to the development board we originally started with, it works well the HS1
http://www.farnell.com/datasheets/1679497.pdf
and a link to the bus supply we are currently using in the Bluetooth/App based soultion
http://www.mouser.co.uk/m_new/power/power-supplies/recom-RELV4-16-power-supply/n-axgi2Z2d178c

Hope that helps!