What? You wired them in series and want to deliver them 3.5A?
Look up LED drivers or current limit them with resistors. Or wire them in series and currentlimit them.
LM338 & big honking heatsink
[editline]23rd December 2015[/editline]
[url]http://diyaudioprojects.com/Technical/Current-Regulator[/url]
[QUOTE=AGMadsAG;49376476]What? You wired them in series and want to deliver them 3.5A?
Look up LED drivers or current limit them with resistors. Or wire them in series and currentlimit them.[/QUOTE]
I am building a new light. I want to use adjustable voltage regulators instead of resistors to drop voltage and keep the LEDs alive.
The circuit operates at around 45W.
Milled the front panel for my CNC control box.
[url=http://i.imgur.com/RXjDs2M.jpg][img]http://i.imgur.com/RXjDs2Ml.jpg[/img][/url]
[url=http://i.imgur.com/1AqbtHy.jpg][img]http://i.imgur.com/1AqbtHyl.jpg[/img][/url]
[url=http://i.imgur.com/1YdOouB.jpg][img]http://i.imgur.com/1YdOouBl.jpg[/img][/url]
[QUOTE=Van-man;49376510]LM338 & big honking heatsink
[editline]23rd December 2015[/editline]
[url]http://diyaudioprojects.com/Technical/Current-Regulator[/url][/QUOTE]
How big of a heatsink are we talking about?
[QUOTE=false prophet;49376404]Yes, they're 700mA LEDs. I took the information provided above and came up with the circuit for 5 series pulling 3.5A. So I think I will need at least 2 voltage regulators, or 10 LM317 which I don't really want do to.
Can anyone suggest a voltage regulator for 2v-12v which won't blow up at 3.5A, maybe 5A? I looked at LM338K but there seems to be a big issue with them not being as advertised, usually only 1A or something.[/QUOTE]
What is the forward voltage of each LED module ?
If they're all in series you need to feed them 700mA (since the same current flows through all in series), either with a single resistor or preferably a very simple constant current source.
If they're in parallel you need to provide 700mA to each branch, again either with resistors or current sources.
The former approach takes less parts but requires a higher drive voltage (the sum of the voltage drops plus a little extra), it also draws less current from the power supply (700mA in your case).
The latter approach takes more parts and draws much more current from the power supply (the sum of each branch current) but works better at low supply voltage.
For a series combination which I recommend something like this would work very well:
[img]http://i.imgur.com/0ZaJc2t.png[/img]
The voltage applied to the non-inverting input of the op-amp sets the output current, with a 1 ohm resistor 1V gives 1A, you can use a simple zener regulator or a low power voltage regulator to make this immune to changes in load current, this may not be needed if your power supply has good load regulation to start with, in that case a simple pot will do to set the voltage.
For the transistors the first one can be any old general purpose transistor, I.E 2N3904, the second one needs to be a medium power transistor like a BD135 or you can use an all in one darlington transistor such as TIP122, you could even use a power MOSFET if you wanted to.
Op-amp can be any general purpose bipolar type that can run off a single supply rail, LM358 for instance.
I think I'm ultimately going with the 3 XFMR h-bridge design since I can implement a decoupled control loop for each one, aside from the phase synchronization. The isolation provided would also be an added bonus along with the neutral leg being artificially created.
[QUOTE=Chryseus;49377094]*snip*[/QUOTE]
It varies between 2.4 and 3.6v. I had them in series of 2.4v, 2.4v, 2.5v and 3.4v, 3.4v, 3.6v.
[QUOTE=false prophet;49378106]It varies between 2.4 and 3.6v. I had them in series of 2.4v, 2.4v, 2.5v and 3.4v, 3.4v, 3.6v.[/QUOTE]
For 5 in series then I'd recommend at least a 15V supply, 24V would be ideal (it only needs to be rated for about 1A).
If you get another LED you could have 2 branches of 3 series LEDs and that would work fine on a 12V 1.5A supply.
Or you could have one branch of 3 and one of 2, as long as you use an adjustable driver like the one above it's fine.
By the way if you use that driver a 1 Ohm resistor needs to be rated for 1W (actual dissipation is 0.5W)
remember I mentioned wanting to make a microwave telescope several pages ago and then said I wouldn't do that? Then said orbital guidance computer then dropped that when I ran out of time and realized I still wasn't qualified?
Yeah telescope project is back on my radar. I have to take an extra year in school due to my grades (saying that I'm grateful my parents are still paying is the understatement of the year), and I'm going to be an Astronomy major for a year and change due to it. And then end of next year transfer to Aero/Astro Eng. instead of go right in end of this year. Also Halofreak seems to really like RF and Signals-type stuff so I'm going to badger him, talked to me gran about the old fuckhuge dish and I might be able to use it, so now I'm just doing hardware searches. So building a radio telescope seems like the nice intersection of both majors so hey why not try it or start doing a feasibility study.
Its too hostile and blizzard-y out for me to make the trip out to the farm anytime in the next week, but if the weather lets up I'm going to go wade through deep fuckin snow and get a look at the old thing sometime.
Also still been having fun with Reaktor, made a neat sidechain compressor out of DSP logic blocks and the like and it was quite fun! And as mentioned pages ago Laplace transforms are [I]baller[/I]. Diff Eq grade came back and I got a 3.0, which is my first >2.5 grade in any math class so yay
[QUOTE=paindoc;49379509]remember I mentioned wanting to make a microwave telescope several pages ago and then said I wouldn't do that? Then said orbital guidance computer then dropped that when I ran out of time and realized I still wasn't qualified?
Yeah telescope project is back on my radar. I have to take an extra year in school due to my grades (saying that I'm grateful my parents are still paying is the understatement of the year), and I'm going to be an Astronomy major for a year and change due to it. And then end of next year transfer to Aero/Astro Eng. instead of go right in end of this year. Also Halofreak seems to really like RF and Signals-type stuff so I'm going to badger him, talked to me gran about the old fuckhuge dish and I might be able to use it, so now I'm just doing hardware searches. So building a radio telescope seems like the nice intersection of both majors so hey why not try it or start doing a feasibility study.
Its too hostile and blizzard-y out for me to make the trip out to the farm anytime in the next week, but if the weather lets up I'm going to go wade through deep fuckin snow and get a look at the old thing sometime.
Also still been having fun with Reaktor, made a neat sidechain compressor out of DSP logic blocks and the like and it was quite fun! And as mentioned pages ago Laplace transforms are [I]baller[/I]. Diff Eq grade came back and I got a 3.0, which is my first >2.5 grade in any math class so yay[/QUOTE]
The easy target is the 21cm hydrogen line which requires an antenna tuned for 1.420GHz.
A helical antenna will do nicely here, if you can find a ~1m satellite TV dish that will increase your gain and give a narrower focus.
The next easiest is Radio Jove (not microwave btw) for observing the interaction of Io as it interacts with Jupiter's magnetic field. A dual dipole sized for 20-21MHz (two 7m long dipoles spaced 7m apart in parallel) will work for this. Feed them with the same length of (RG6) coax into a power combiner at a height of about 1/4wavelength (10ft pole will do) and you'll have good vertical reception. You can even vary the lengths of the coax (typically just add extra lengths between the antenna and power combiner) to allow you to move the main lobe.
Edit: Also a farm is great for this sort of thing, good luck!
[QUOTE=false prophet;49376404]Yes, they're 700mA LEDs. I took the information provided above and came up with the circuit for 5 series pulling 3.5A. So I think I will need at least 2 voltage regulators, or 10 LM317 which I don't really want do to.
Can anyone suggest a voltage regulator for 2v-12v which won't blow up at 3.5A, maybe 5A? I looked at LM338K but there seems to be a big issue with them not being as advertised, usually only 1A or something.[/QUOTE]
I think you don't grasp just how much current 5A is. Wire the LEDs in series and apply more voltage.
If a single LED takes 700mA, 5 in series will still pull 700mA, but at a greater voltage.
[QUOTE=DrDevil;49381345]I think you don't grasp just how much current 5A is. Wire the LEDs in series and apply more voltage.
If a single LED takes 700mA, 5 in series will still pull 700mA, but at a greater voltage.[/QUOTE]
To add to this the power each LED dissipates is P = V*I
V isn't the applied voltage but rather the voltage dropped across each LED, so a single LED that drops 3V at 700mA is dissipating 2.1W.
If you have five series LEDs that drop 3V each (15V total) the total dissipation is 10.5W (2.1W * 5).
The only time it's ever really a good idea to use LEDs in parallel is when your supply voltage is very low (I.E battery powered), or you have a lot of LEDs (in that case a combination serires/parallel arrangement is used).
Putting power LEDs in parallel never is a good idea, especially with a constant current supply. When one LED for whatever reason breaks down, the current for that LED has to go through all the other LEDs, meaning that they might explode or atleast very quickly fail aswell.
Also you need very tightly matched LEDs, otherwise you can get visible difference in brightness.
Fair enough. That was an issue I ran into with my first light.
I am looking at adjustable led drivers on eBay now, as previously suggested. They also seem cheaper than the LM338's that I was looking at, and a heck of a lot simpler to use.
Anyone have any thoughts on these? [url]http://www.ebay.com/itm/DC-DC-LM2596-Step-down-Adjustable-Power-Supply-Module-CC-CV-LED-Driver-/161272395732?hash=item258c9577d4:g:oy0AAOxy4YdTRgwD[/url]
I was thinking of getting several to power LEDs with the same maximum voltages instead of having a series of for example 2.4 + 2.4 + 2.5v.
[edit] my power supply is a 12v computer psu, if it makes any difference for these drivers.
[QUOTE=false prophet;49383956]Fair enough. That was an issue I ran into with my first light.
I am looking at adjustable led drivers on eBay now, as previously suggested. They also seem cheaper than the LM338's that I was looking at, and a heck of a lot simpler to use.
Anyone have any thoughts on these? [url]http://www.ebay.com/itm/DC-DC-LM2596-Step-down-Adjustable-Power-Supply-Module-CC-CV-LED-Driver-/161272395732?hash=item258c9577d4:g:oy0AAOxy4YdTRgwD[/url]
I was thinking of getting several to power LEDs with the same maximum voltages instead of having a series of for example 2.4 + 2.4 + 2.5v.
[edit] my power supply is a 12v computer psu, if it makes any difference for these drivers.[/QUOTE]
Get some without the constant current function, as that are wonky in those switchmode power supplies anyway.
Those are also possible to get in a 10-pack for little money.
But do still have a small resistor in series with a chain of LED's, just to even out their characteristics when regulating them voltage wise.
Make something like this, I don't think you can get any simpler:
[img]http://i.imgur.com/9GSPnGZ.png[/img]
Two TIP122
Two 1k 1/4W resistors
Two 5k trimmers
Works on 9V+ (as long as it's regulated, a PC power supply is fine).
[b]Edit[/b]
Why the disagree ? I know it's not the most accurate or stable but it's fine for driving LEDs.
I will. I still have a bunch of high watt resistors from the last panel. I went with a 5w resistor for the lower voltage series and they got very, very hot. The other LEDs are working just perfectly fine, [I]for now[/I].
Also, what makes the constant current functions in those circuits wonky?
[edit]
[QUOTE=Chryseus;49384228]Make something like this, I don't think you can get any simpler:
[img]http://i.imgur.com/9GSPnGZ.png[/img]
Two TIP122
Two 1k 1/4W resistors
Two 5k trimmers
Works on 9V+ (as long as it's regulated, a PC power supply is fine).[/QUOTE]
Is there a benefit to this over one of the LED drivers available?
[QUOTE=false prophet;49384239]I will. I still have a bunch of high watt resistors from the last panel. I went with a 5w resistor for the lower voltage series and they got very, very hot. The other LEDs are working just perfectly fine, [I]for now[/I].
Also, what makes the constant current functions in those circuits wonky?[/QUOTE]
Because the switchmode IC's used in them don't have it implemented natively, and typical chinese discount designs.
Don't trust anything cheap & chinese that isn't made following the reference design.
I have had a couple from different vendors, and their current limiting were far too unpredictable for LED driving for my liking.
Their voltage regulating is almost spot-on on the other hand, even when driving a 6A load :v:
But feel free to buy one to try one out with a spare LED, just be warned that they're so-and-so when it comes to current limiting.
Ah, I see. I can't find really any adjustable drivers that don't have constant current functionality. Poop!
[QUOTE=false prophet;49386170]Ah, I see. I can't find really any adjustable drivers that don't have constant current functionality. Poop![/QUOTE]
[url]http://www.ebay.com/itm/381462079734[/url]
Get a boost converter, more suitable for your application, you can make a constant current driver like I showed [url=https://facepunch.com/showthread.php?t=1452809&p=49377094&viewfull=1#post49377094]here[/url], works perfectly from low current to many amps.
You can disable the constant current by turning the pot right up if it has CC.
Wouldn't really call it electrical engineering, but I set up these LEDS in my closet with my girlfriend. Didn't have a wall plug so I ghetto-rigged a 12v 23A battery I got from China into an adapter.
[url]http://imgur.com/a/CuLO1[/url]
I doubt that can provide 23 amps.
It's just a stack of button cells in a fancy package:
[url]https://en.wikipedia.org/wiki/A23_battery[/url]
I didn't mean to imply it was 23 Amps, I'm pretty sure it was just the name of the battery
The dish out in the country got removed :c
Guess I have to bum around near the astronomy department and try to get in with the blokes working at the observatory in the cascades or the club making a radio telescope instead
I wanted some voltage regulator modules for breadboard usage, but I wasn't sure how well I could etch a few, so I figured I'd print a bunch, and hope I get like 10-15% yield.
[t]https://i.imgur.com/1WTG9iJ.jpg[/t]
I got more than 10-15%
[editline]27th December 2015[/editline]
Assembled one, tested, works.
[t]https://i.imgur.com/uV8LneY.jpg[/t]
do you guys know of any big good resources on vacuum tubes?
[editline]27th December 2015[/editline]
so far all I've got is the wikipedia article, a couple from all about circuits, some dudes website, and a little math behind it
[QUOTE=POLOPOZOZO;49399780]do you guys know of any big good resources on vacuum tubes?
[editline]27th December 2015[/editline]
so far all I've got is the wikipedia article, a couple from all about circuits, some dudes website, and a little math behind it[/QUOTE]
You mean info on their usage or where to buy them?
Sorry, you need to Log In to post a reply to this thread.
Electrical Engineering V3
