How to dim a Meanwell ELN-60-48D...

[der_wille_zur_macht]

Originally Posted by Skeptic_07

after doing some more reading about this electronic type of stuffs.. even if it is not true analog output, couldn't I do a filter between the pot and the amp? like this? yes?

Yep.

edit: wait a minute.. Couldn't I just filter the PWM signal from the arduino and then run it through an op amp to double the voltage? no need for the digital pot.

Yep.

FWIW you might also find that the driver dims fine even with a lumpy signal.

[Chemistry Guy]

Just to let everyone know, I too have used an op amp to double (well almost, I could only find an 8V transformer, still looking for true 10V) the arduino PWM. I'm not filtering because I have the PWM drivers, and they work fine. They shut off somewhere above 0 signal from the arduino, something like 50-70 out of 255, I'm guessing because of the minimum signal needed from the drivers is significantly more than 0% duty cycle (~11% according to data sheet).

[Skeptic_07]

which op-amp did you use? LM358?

[der_wille_zur_macht]

Got the ELN shield PCBs yesterday, soldered one up, and tested this morning. It was kind of a comedy of errors; I got the transistors and the voltage regulator in the wrong packages. The transistors were supposed to be a small surface mount transistor package but apparently I ordered the infinitely microscopic package by accident. At any rate I got it soldered up and began testing.

Things looked good - I began testing by reading the voltage across the output pins, and I could turn the pin on and off and pick an intermediate level and the output pins would show an appropriate voltage. I set my multimeter to duty cycle to see how it looked, and the transistor popped.

Calling on the more experienced EE's - is this due to some fundamental flaw in my approach, or did I just pick an inappropriate transistor (i.e. I need one with a higher power rating)? I can't find meaningful specs in my multimeter's manual to know what it pulls when it tries to take duty cycle measurements. This is the transistor I used:

http://www.semiconductors.com/acroba...C546_SER_6.pdf

It chugged along great until I tried to measure frequency and duty cycle. . .

[Chemistry Guy]

Originally Posted by Skeptic_07

which op-amp did you use? LM358?

I used LM741.

V+: 8V
V-: Ground
+ input: Arduino PWM
- input: At half 8V rail voltage via two equal resistors (8V - Resistor - (- input) - Resistor - Ground)
output: to meanwell

When the +input is at least half the rail voltage, the op amp outputs 8V. So in theory this should remain PWM, right?

[akindbro4u]

Originally Posted by der_wille_zur_macht

Got the ELN shield PCBs yesterday, soldered one up, and tested this morning. It was kind of a comedy of errors; I got the transistors and the voltage regulator in the wrong packages. The transistors were supposed to be a small surface mount transistor package but apparently I ordered the infinitely microscopic package by accident. At any rate I got it soldered up and began testing.

Things looked good - I began testing by reading the voltage across the output pins, and I could turn the pin on and off and pick an intermediate level and the output pins would show an appropriate voltage. I set my multimeter to duty cycle to see how it looked, and the transistor popped.

Calling on the more experienced EE's - is this due to some fundamental flaw in my approach, or did I just pick an inappropriate transistor (i.e. I need one with a higher power rating)? I can't find meaningful specs in my multimeter's manual to know what it pulls when it tries to take duty cycle measurements. This is the transistor I used:

http://www.semiconductors.com/acroba...C546_SER_6.pdf

It chugged along great until I tried to measure frequency and duty cycle. . .

Have you got an answer on your dilemma yet sir?

[der_wille_zur_macht]

I soldered new transistors on the board and sent it off to someone with a meanwell to test, so results are pending. I think the fried part was a mistake on my part because it looked fine otherwise on my bench.

[OSUav8er]

Thanks for all of the great info guys. I'm an avid reader but I have a question I can't find the answer to so I'll try a post.

My Meanwell's are on their way and I'd like to devote 3 drivers (36 LED's) to my build and I'd like to have them all on one potentiometer. I'm sure it's been addressed but I'm having trouble locating the thread.

Thanks in advance for a little help.

[akindbro4u]

Originally Posted by der_wille_zur_macht

I soldered new transistors on the board and sent it off to someone with a meanwell to test, so results are pending. I think the fried part was a mistake on my part because it looked fine otherwise on my bench.

That's good news. Hope the test goes well.

[PhilipOSU]

I know this thread is primarliy geared towards a DIY solution to the dimming, but does anyone know how many Meanwell drivers the Apex can control the dimming on at one time? Thanks. Phil

[HeneryH]

xxx

[hllywd]

Originally Posted by PhilipOSU

I know this thread is primarliy geared towards a DIY solution to the dimming, but does anyone know how many Meanwell drivers the Apex can control the dimming on at one time? Thanks. Phil

I hope at least 2 channels of 4...
Seriously I've been looking for the same info, I asked some similar questions on the Neptune forum and didn't get a lot of useful info back. Neptune was non-committal due to not knowing specifics for the Meanwells...
Right now I have an ACIII with plans to move up to an Apex for this purpose as well as the other advanced features it offers.
Tim

[der_wille_zur_macht]

Originally Posted by der_wille_zur_macht

I soldered new transistors on the board and sent it off to someone with a meanwell to test, so results are pending. I think the fried part was a mistake on my part because it looked fine otherwise on my bench.

An update. the "beta tester" I sent the board to (RC user SWINGRRRR) just emailed me a video of the thing in action, so at least we know it works. It's running off a 9v battery (until he can find a match for the whacky plug on the Seeeduino Mega he's using) and controlling one ELN using the example "fade" sketch that comes with the Arduino IDE.

More robust testing to follow, but wanted to share the news with everyone. Assuming we can verify functionality a little more thoroughly I'll post up more details and the EAGLE design files.

[stugray]

"so at least we know it works."

Good to hear!

Now, can he measure the current that the Meanwell pulls from the PWM, so we can tell how many drivers you can "drive"?
Just set to max brightness & put a DMM inline.

Stu

[der_wille_zur_macht]

That's one of the things I'm hoping we can do in the next few days. I also want to send one of these boards to someone with the "analog" D-type ELN just to see how it likes a ~500hZ 10v PWM signal.

Any volunteers? You'll need an Arduino (bonus points if it's shield-compatible) and a D-type ELN.

[SWINGRRRR]

Originally Posted by stugray

"so at least we know it works."

Good to hear!

Now, can he measure the current that the Meanwell pulls from the PWM, so we can tell how many drivers you can "drive"?
Just set to max brightness & put a DMM inline.

Stu

Will do. Hopefully I get some more trigger time with it tomorrow. Also, I got to find a 10V WW. I am running it all off a 9V WW.

Originally Posted by der_wille_zur_macht

That's one of the things I'm hoping we can do in the next few days. I also want to send one of these boards to someone with the "analog" D-type ELN just to see how it likes a ~500hZ 10v PWM signal.

Any volunteers? You'll need an Arduino (bonus points if it's shield-compatible) and a D-type ELN.

I'm going to put a shout out locally to see if anyone has a "D" I can borrow as well. There are several of us doing builds.

[der_wille_zur_macht]

Cool. Shoot me a PM or post here when you get some more results.

FWIW to expand a bit on Stu's comment - once we know the current sucked by the meanwell for sure, we'd need to determine the max current any given dimming signal generator (i.e. this shield, or a commercial controller) can handle, then we'd be able to determine the max.

Just looked at the Seeed website and the "recommended" max input voltage for the Mega is 12v but they claim the actual limit is 20v, so your 12v wall wart is probably fine. You're only drawing a few mA on one of the PWM pins to trigger the transistor on the shield so the regulator won't have much of a load to worry about.

[Skeptic_07]

I've done some more prototyping on this (with help) and have figured out that you don't need to supply voltage to the driver at all. There is 10V coming from the + line on the driver. I used an NPN transistor and a 1K resistor and got the dimming working beautifully on the P type drivers. (still no luck on the D type drivers for me) You hookup the PWM pin you're using on the the arduino to the 1K resistor to the middle pin on the transistor, hookup the - from the meanwell and the arduino ground to the ground pin on the transistor and then hookup the + from the meanwell to the other pin. Sorry I can't be more specific then that, I don't know what letters (C, B, E or whatever) to connect to. On my transistors, if the flat side is facing you, the left pin is ground, middle is pwm, and right is +, if that makes any sense.

I think we can get the D models to work this way too, we just need to add a 10k resistor to the + from the meanwell, but i haven't tried it.

FWIW I've also found that tunze pumps work the same way. I hooked them up to the same exact circuit and have been able to control them via Arduino! The only difference is that it was inverted so 1 was full blast and 255 was off. I spent all night making a standing wave in my tank so now I'm completely shot, but it was worth it. I wonder what my boss thinks when I show up looking like hell and hes like what were you doign last night and i'm like messing with my fish tank

[der_wille_zur_macht]

Someone needs to sit down for a few hours with both meanwells, an oscilloscope, and a few different signal sources and figure out what's really going on here. Every time this thread pops back up, there's wild new information coming out of the woodwork. I'm glad you and I both seem to be getting some success but with vastly different approaches and the datasheet seems pretty meaningless given the info that's come out!

The Tunze at least is a "problem" that's been solved before - there's +8v on one of the pins in that DIN connector so it's really easy to do.

[nanojg]

For non circuitry experts, can I just plug the meanwell into this http://papatek.com/AllOtherFeatured/...er-Supply.html and be done with it? I know it only has the option of switching 1.5v at a time, but will it work?

[der_wille_zur_macht]

That would work for a D-type meanwell but wouldn't be ideal, since it has no equivalent "max" setting (10v) and the 12v setting is out of the safe range.

[Skeptic_07]

Originally Posted by der_wille_zur_macht

Someone needs to sit down for a few hours with both meanwells, an oscilloscope, and a few different signal sources and figure out what's really going on here. Every time this thread pops back up, there's wild new information coming out of the woodwork. I'm glad you and I both seem to be getting some success but with vastly different approaches and the datasheet seems pretty meaningless given the info that's come out!

The Tunze at least is a "problem" that's been solved before - there's +8v on one of the pins in that DIN connector so it's really easy to do.

Agreed the datasheet is really cryptic. All it shows is that dim - goes right to ground. as for where dim + goes.. it shows it going into the over current protection which goes into the detection circuit but thats all meaningless to us without knowing what those circuits are.

I know a guy with a scope and he knows his stuff. Maybe i can get him to put in some work for us with his scope, no promises on that though! He's deep into lots of projects as it is like using a micro-controller to change his engine timing tuning. Its tough getting him to help me because I'm a complete newb doing things he doesn't care about. I have to like, lure him over to my house with promises of sandwiches and beer.

[der_wille_zur_macht]

I just bought a DIY scope kit so I'll (finally) have on within the next few weeks. Of course, I don't have an ELN to test.

Can you at least probe the two DIM connections with a multimeter and tell us what they show? i.e. when the unit is plugged in to the wall but nothing is connected to the DIM circuit.

[nanojg]

thanks dwzm, are there any better plug and play options? I can't seem to find a 9v variable

[Skeptic_07]

Originally Posted by der_wille_zur_macht

Can you at least probe the two DIM connections with a multimeter and tell us what they show? i.e. when the unit is plugged in to the wall but nothing is connected to the DIM circuit.

will do. I'm at work now though. Can't do it until i get home. hmmmmnnnn should i fake illness so i can go home now??????