Building New Panels - Sharing the Process

[rrog]

Very good! I will get some 18ga wire. Would it help if I soldered them into loops?

Should I wrap the Zeners in heat shrink?

Thanks for the eyeballs Rives!

Work Boxes are crazy expensive. Do they need to be metal for grounding or such? Could I simply ventilate a Tupperware deal or such? Maybe a Scoobie Doo lunch box?

 

[rives]

Very good! I will get some 18ga wire. Would it help if I soldered them into loops?

Should I wrap the Zeners in heat shrink?

Thanks for the eyeballs Rives!

Work Boxes are crazy expensive. Do they need to be metal for grounding or such? Could I simply ventilate a Tupperware deal or such? Maybe a Scoobie Doo lunch box?

Soldering them wouldn't be a bad idea. You could solder the two together for a bit more insurance.

Yes, use shrink on any conductor that is exposed for any distance.

Yes, boxes are higher than giraffe's balls. If you can find an appropriate size, Onlinecomponents.com has Bud Industries aluminum boxes at the best prices that I've found. Probably not as cools as a Scoobie Doo lunch box, though.

 

[rrog]

Thanks- I'll check out Bud boxes. I just like the name somehow.

Q: Can I shorten the Zener leads that I will be soldering? I have to bend them into a U to approximate what I did with the MOV. I see they're sensitive to heat.

Thanks again.

http://www.onlinecomponents.com/bud-industries-ac-427.html?p=10816564 Here's one that would work. Need the optional bottom. These don't appear to be ventilated. Should I expect much heat? Drill some holes?

Q: Do you have a toggle switch that you like? Like the ones you used on the Volks

You asked about the fuses. They are RS 2.5A 250V "Fast Acting"

MOV's are 150V 1600pF Littlefuse Varistors

Zeners are 47V 1.3W

 

[rrog]

Well my 18 ga wire came in today. Not copper... Gotta place another order tomorrow...

 

[rives]

Yes, you will need to ventilate the driver enclosure. Most of the drivers are about 85% efficient at best, so they will get pretty toasty in short order if totally enclosed. I'd put a passive vent on one end and maybe a 60mm fan on the other so that the air gets drawn through/around the drivers, and run the fan at reduced voltage to slow it down. Fan filters/guards make for a nice looking finished appearance if you use the same size holes.

That is a rocker switch on the Volks, not a toggle. Toggles are a little easier to install because you just drill a hole for them, rockers take a good bit of filing and fitting but look a little more "finished" for these applications, I think.

http://www.mouser.com/Search/Produc...TVFvirtualkey66710000virtualkey769-AJ7202BTVF

Yes, you can shorten the zener leads with no problems.

 

[rrog]

Thanks Rives-

I was thinking of the silver switches on the Hybrid... not the Volks- I mis-spoke. No big deal, though. I'll see what The Shack has.

I think I'll try a passive vent for the driver box. No idea about voltage reduction for a fan. I also may just build a box from thin plywood. I'm not seeing a reason for the use of metal. Unless the metal box is grounded? The incoming power line has a ground wire that I have nothing to connect it to.

 

[rives]

Thanks Rives-

I was thinking of the silver switches on the Hybrid... not the Volks- I mis-spoke. No big deal, though. I'll see what The Shack has.

I think I'll try a passive vent for the driver box. No idea about voltage reduction for a fan. I also may just build a box from thin plywood. I'm not seeing a reason for the use of metal. Unless the metal box is grounded? The incoming power line has a ground wire that I have nothing to connect it to.

Those switches are by Micro Switch. Most industrial suppliers will carry them, they are much higher quality than what you will usually find at RS. Holler back if you would like me to find the part number for you.

Voltage reduction can be done a number of ways, but might be as simple as simply getting a 12v fan and using a wall wart (cell phone charger, etc) of a lower voltage. You can also feed each end of a potentiometer with the (+) & (-) from a DC supply, and hook the fan up to the (-) and the wiper on the pot. It will then function as a voltage divider, and give you any output voltage between 0 and the supply maximum.

Yes, the enclosure should be metal for grounding purposes. You could possibly make the floor out of metal so that the components are mounted on a grounded substrate and then do the rest from wood. RF interference might be an issue outside of a grounded enclosure - drivers are switching power supplies, so they could potentially be pretty "noisy".

That ground wire should be connected to anything and everything metal, and needs to be carried all the way through to the heat sink.

 

[rrog]

Thanks Rives-

Since the Drivers have no ground wire, I would have to make a ground connection from the metal box to the sink.

I tell ya, this 18 AWG wire is giving me fits. It's silver, and so one would think that it's tin-coated copper. But I scrape and scrape and silver comes off and seems to reveal simply more silver. Bright silver. I look at it under a loupe and see nothing but silver. Cut the ends and nothing but silver. The supplier swears it's copper in there, but I can't find it, and clearly these are not solid tin wire strands, though it must be close...

 

[GuFF]

Rives,*
I'm planning on building some panels somewhat similar to RROGs. I ran the idea of using MOVs and Zeners past ledgroupbuy.com and rapidled.com. *Rapidled said the MOVs are a good idea but the zeners shouldn't be necessary, ledgroupbuy said neither were necessary as Meanwells are equipped from both amperage spikes and voltage spikes. *I would probably install them anyways as I'd rather have a Zener or MOV fail outside of the driver rather than one inside and have to have it shipped for repair and deal with the downtime.

I have another concern that groupled brought to my attention, inrush amperage from the drivers initial startup.

I have 4 1000 w HPS bulbs ran by 4 1000w phantoms, all plugged into a C.A.P. MLC-4X which is powered by a 230V 30 amp circuit. *

I want to replace each 1000 watt with (2) 36" Makersled heat sinks each equipped with (3) strings of 14 Cree XM-Ls, with each string powered by a Meanwell HLG-120H-48. *So thats 3 Meanwell HLG-120H-48 drivers per heatsink.

I asked ledgroupbuy about wiring 2 heatsinks (6 drivers total) into a 'work box' with terminals and power that with one cord, which would then plug into the C.A.P. MLC-4X. *That way all 8 heat sinks could be ran off the C.A.P. MLC-4X (one outlet per 2 heatsinks). *

His response was that I would not be able to start all of those drivers at once because my breaker would trip. *However my breaker does not trip when the 4 phantoms fire up, although maybe the inrush amperage of the phantom is nearly that of the drivers (70 amps)?

I would expect RROG to have a similar problem?

He also said that the HLGs must be powered by a relay, wouldn't the relay in the MLC-4X suffice?

The only solutions I can think of would be a bigger circuit or leaving the drivers constantly powered and dim them enough until they emit zero light, however I'm assuming you'd need sometime of driver controller/program to automate this to match your lighting schedule.

Thanks for the help, hopefully this can be useful to RROG as well.

 

[rrog]

You're thinking that I can't start 5 drivers at the same time? That's surprising, as you say given the smaller current

 

[GuFF]

Personally, i didn't think it was an issue, the guy at groupled had warned me about it. If you look at the spec sheet the LPF90D pulls 70amps @ 230VAC and the HLG 120H pulls 75amps at 230VAC. I'm looking to fire 8 at once which would be about 600 amps even if only for a few milliseconds I dunno if the breaker would trip or not....hopefully rives can give us some insight

 

[rives]

Interesting questions, GuFF.

Regarding the MOVs and the zeners, I would add them regardless of whether or not they are part of the driver circuitry.

MOV's have a definite lifespan - they can only absorb "x" amount of power before they are destroyed (sometimes in spectacular, explosive fashion). This power can be in the form of numerous low level hits, or far fewer high-powered spikes. If these spikes are dissipated by the driver internal circuitry, at some point the driver will be destroyed. If the spikes can be arrested before getting into the driver, you would only need to replace a component with a cost of $1 or so. Cheap insurance.

Regarding the zeners, if your driver is going to be continuously connected to the LEDs, then perhaps you don't need them. In my case, I blew mine when I took the LED bar back out to my shop to make a last minute change. Personally, I would add zeners unless the MCPCB stars had them on-board, and particularly if you plan on a remote-mounted driver. Again, cheap insurance.

As far as your breaker tripping on power-up, I would be surprised if that happened but it is possible if you have a sensitive breaker. Any capacitive load (switching power supplies, LED drivers, electronic ballasts, etc), is about the most difficult load for mechanical switching. If you look at the contact ratings for devices that are thoroughly tested, there will be ratings for various types of loads, including resistive, inductive (possibly called out as motor hp and/or ballasts), capacitive, tungsten, etc. Usually resistive loads, like heating elements, are the easiest on the contacts and capacitive loads are the hardest.

While keeping that in mind and understanding that the relay/switch contacts will be adversely affected, it isn't necessarily true of the breaker. Remember, the breaker's contacts are already closed, so what you are worrying about is how quickly the breaker will respond to the current draw. The charging of the capacitors could theoretically take place in less than a half-wave of the power cycle, but the charging is limited by the circuitry feeding the capacitor. The bad side of this is that it stretches out the duration for the breaker to respond, the good side is that it lessens the inrush current.

If you look at the time/current curve for a brand-new, in spec GE "Q-series" 20 amp breaker, it can trip in as little as 10ms to a 400 amp load, and may take as long as 16ms for a load ranging from 600 amps to 20,000 amps or more.

So, will it trip? Maybe. If the breaker is already warmed up from another load on the circuit, it increases the likelihood. One breaker may and another may not under identical conditions. I would be more concerned about what I was switching the load with, and have a back-up plan in mind if it turned out to be a problem.



GE time/current curve - http://www.geindustrial.com/publibrary/checkout/GES-9887B?TNR=Time%20Current%20Curves|GES-9887B|generic

 

[Weezard]

I just looked at the tripping curves on the circuit breakers,
You should have no problem because there is plenty of "lag"
Just google circuit breaker specs.

Aloha,
Weeze

 

[rrog]

Rives- might this panel I'm finishing benefit from a relay and timer like you walked me through a couple years ago?

I have my little board built on the AC side. Fuses and MOV. I'm stripping off that heavy external wire insulation and route just the individual wires as you did, Rives. I have a terminal block for the DC side with the Zeners on the LED side of the block. I'm trying to be diligent to the details Rives very kindly outlined. I like redundancy with safety.

 

[rives]

Rives- might this panel I'm finishing benefit from a relay and timer like you walked me through a couple years ago?

It couldn't hurt.

I just looked at the tripping curves on the circuit breakers,
You should have no problem because there is plenty of "lag"
Just google circuit breaker specs.

Aloha,
Weeze

This would be my first reaction. The only reason that I would qualify it is because of my experience with a motor a few years ago. This application had been running for years with no starting problems. The motor burned up and was replaced with a new "high efficiency" version - same frame, same everything as far as external appearances. After it was installed, it would trip the breaker instantly when it was energized. It was a very low inertia, easy starting load, but even so, the motor barely started to move before tripping the breaker. Electricity and Murphy sometimes go hand-in-hand.

I would also note that my Hybrid fixture has 6 drivers and 2 electronic ballasts, all of which start at the same time and it has never been an issue.

 

[GuFF]

Thank you very much for that response rives, was stressing I couldnt go forward with my plans. I will make my own thread when I plan to get started and I hope you can share your wisdom, until then I'm following this build very closely to learn from.....good luck rrogs!!

 

[xmobotx]

I just looked at the tripping curves on the circuit breakers,
You should have no problem because there is plenty of "lag"
Just google circuit breaker specs.

Aloha,
Weeze

yeah i like doing it 'raw dog' as well

of course i dont see anything over 100 watts in my future {so far}

 

[Weezard]

Bottom line, he could use multiple timers set a minute apart.
(Actually need less than a second, but no point in cutting it that fine.)
Timers are cheap and the plants don't mind.

A penny here, a penny there . . .

Weeze

 

[GuFF]

Rrog, are you only feeding your work box(s) with 110? Curious as to what your source is and what you plant to power the drivers up with (relay?). Some pics of your work box and how you set up your terminals would be much appreciated when you get time as I'm quite the noob when it comes to electronics.

Curious how 240 would be wired to a work box and then to the drivers?

 

[GuFF]

Sorry for the double post, not sure how to edit? I'm assuming considering 240
doesn't have a neutral the other hot lead is used in place of the neutral, therefore 2 'hot' terminal bars in the work box? And one of these hot terminals feeds the "AC/L" line for the driver, while the other hot terminal feeds the "AC/N" line? And no worries I'm not touching any wiring for a while until I have a clear understanding of it all, I also have an old friend that wired my room that could give me a hand.