COB Ringlight

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

Miljenko
Posts: 171
Joined: Sat Jun 01, 2013 11:53 pm
Location: Zagreb, Croatia
Contact:

Post by Miljenko »

Thank you, Mike for your kind offer. While having M.Sc.degree in electronics and designing (tube) amps whenever I don't measure lenses, I believe I can manage :D
IMHO, keeping cob leds at very constant current makes no much sense since those are mostly internal series/paralel combos where you don't have total control over individual led currents anyway. Besides, led light output is not proportional to current so +/- 10% current drift is actually no big deal. And it actually gets pretty stable just after couple of minutes. After all, we are talking here about very cheap solution so making it much more sofisticated is a no go for me.
Just my 2 cents....
All things are number - Pythagoras

Saul
Posts: 1781
Joined: Mon Jan 31, 2011 11:59 am
Location: Naperville, IL USA
Contact:

Post by Saul »

mawyatt wrote:...I just got a cheap eBay supply for $35 to supplement my older lab supply...
Hi Mike,
Be careful - I have a 30v 10A " cheap eBay supply" (similar to yours)- this thing went mad and made lots of damage to my electronics ....

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Saul wrote:
mawyatt wrote:...I just got a cheap eBay supply for $35 to supplement my older lab supply...
Hi Mike,
Be careful - I have a 30v 10A " cheap eBay supply" (similar to yours)- this thing went mad and made lots of damage to my electronics ....
Saul,

Thanks for the warning!! I'll use my old lab supply for critical stuff and this will be used for less demanding use like maybe running these LEDs.

I did open it up (I aways investigate new equipment), the build quality is just so so. That alone alerted me to possible weak design, so I studied the design.

The design is based upon a switch-mode PS that produces a few volts above the present demanded voltage output. A linear regulator then regulates this voltage to the final output voltage, which reduces the emphasis on the switch-mode PS as the linear PS can improve the final regulation and noise properties. This isn't new, has been employed for decades, we've used this concept for a very fast envelope tracking modulator for an RF power transmitter.

I may build a "crow bar" circuit for the supply in case I use it for more sensitive and expensive stuff, the idea is the "crow bar" with short the supply and burn it up before the expensive electronics. If Murphy is around the expensive semiconductors will burn up first and protect the cheap PS :shock:

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

ray_parkhurst
Posts: 3416
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

I'm lucky to live in a high tech area. New companies are formed and older companies go out of business constantly, and their surplus equipment ends up in various surplus dealer stock, garage sales, estate sales, or the electronics swap. I make a point to pick up HP power supplies when they are available cheap, and have several different models laying about. I picked up a beautiful Lambda 40V 500mA supply with current and voltage adjustments at an estate sale a week ago, and a dual 40V HP/Harrison supply a few weeks earlier. They don't have digital readouts, and can't be programmed, but for bench use these things are great.

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Those old power supplies are great, I grew up using them. We had a philosophy of only using Tektronix scopes, Fluke Meters and everything else was HP...although we did have a few of the Lambda power supplies. One Lambda I recall had a separate scale that was in 10uV increments!

Wish we had those old electronic estate sales here :(

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

ray_parkhurst
Posts: 3416
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

mawyatt wrote:Those old power supplies are great, I grew up using them. We had a philosophy of only using Tektronix scopes, Fluke Meters and everything else was HP...although we did have a few of the Lambda power supplies. One Lambda I recall had a separate scale that was in 10uV increments!

Wish we had those old electronic estate sales here :(

Best,
I worked for HP, and then Agilent, so we had a tendency to buy our own stuff. It wasn't mandated, but of course you support your own company. When the wireless division became Avago, we still had Agilent discounts for a while. At some point the discounts ended and we looked more to open market, but pretty much the only thing we bought from outside were a few R&S analyzers so that we had at least one bench setup that was similar to some of our customers who preferred R&S over Agilent.

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

ray_parkhurst wrote:
mawyatt wrote:Those old power supplies are great, I grew up using them. We had a philosophy of only using Tektronix scopes, Fluke Meters and everything else was HP...although we did have a few of the Lambda power supplies. One Lambda I recall had a separate scale that was in 10uV increments!

Wish we had those old electronic estate sales here :(

Best,
I worked for HP, and then Agilent, so we had a tendency to buy our own stuff. It wasn't mandated, but of course you support your own company. When the wireless division became Avago, we still had Agilent discounts for a while. At some point the discounts ended and we looked more to open market, but pretty much the only thing we bought from outside were a few R&S analyzers so that we had at least one bench setup that was similar to some of our customers who preferred R&S over Agilent.
It's a shame what happened to HP, with Agilent spin off. Agilent should have been the name of the computer company, not HP. HP was founded on instruments, not computers :shock:

Keysight spun off from Agilent because Agilent became more interested in medical equipment than instruments. They are trying to recreate the old HP and I wish them success. We need the old HP back!!

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Miljenko wrote:Thank you, Mike for your kind offer. While having M.Sc.degree in electronics and designing (tube) amps whenever I don't measure lenses, I believe I can manage :D
IMHO, keeping cob leds at very constant current makes no much sense since those are mostly internal series/paralel combos where you don't have total control over individual led currents anyway. Besides, led light output is not proportional to current so +/- 10% current drift is actually no big deal. And it actually gets pretty stable just after couple of minutes. After all, we are talking here about very cheap solution so making it much more sofisticated is a no go for me.
Just my 2 cents....
While I agree that LED output isn't perfectly linear with current, it's much more linear than with voltage. Semiconductor physics shows that a photon is released when charges recombine in the LED junction and since charge movement per unit time is current, then the more current the more photons released. Semiconductor junctions are highly non-linear with voltage and have a large temperature dependance, so fixed voltage operation with forward biased junctions is usually avoided unless some means of resistive ballast is employed.

Paralleling junctions of any sort is usually done with ballast, think emitter resistors in BJT audio amplifiers for example to prevent "thermal runaway" and "current hogging". I wouldn't be surprised if the LED chip has a ballast resistor built-in to allow parallel operation.

Anyway, understand the reluctance to create a complex, expensive LED driver and completely agree. However, this doesn't need to be the case, a simple power transistor or MOSFET and a TL431 reference diode is all that's required to create a simple, cheap ($1~2), adjustable, current source for LED control. You can find reference designs on-line for this. If you require a large range of current then a Widlar current source could be employed, this requires only 2 transistors, but is somewhat temperature dependent.

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

Miljenko
Posts: 171
Joined: Sat Jun 01, 2013 11:53 pm
Location: Zagreb, Croatia
Contact:

Post by Miljenko »

mawyatt wrote:I wouldn't be surprised if the LED chip has a ballast resistor built-in to allow parallel operation.
Believe it or not, there are no balast resistors for parallel leds in COB arrays. They just count on tight manufacturing tolerances. But the life isn't that simple, of course. That's why you'll see burned out leds on led strips made with old 5050 led chips (not used any more by serious manufacturers). 5050 have 3 chips in paralel without any balancing resistors inside. And of course, triplet segmet which draws more current gets hotter which reduces resistance even further and leads to self destruction.
mawyatt wrote: ... a simple power transistor or MOSFET and a TL431 reference diode is all that's required to create a simple, cheap ($1~2), adjustable, current source for LED control.
Since I have a box full of TL317/337 regulators, I usually make simple enough current source which is totaly adequate for driving leds up to 1 amp. The one below is tailored for COB ring lights we are talking about.

Image
All things are number - Pythagoras

ray_parkhurst
Posts: 3416
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

Miljenko wrote:
mawyatt wrote:I wouldn't be surprised if the LED chip has a ballast resistor built-in to allow parallel operation.
Believe it or not, there are no balast resistors for parallel leds in COB arrays. They just count on tight manufacturing tolerances. But the life isn't that simple, of course. That's why you'll see burned out leds on led strips made with old 5050 led chips (not used any more by serious manufacturers). 5050 have 3 chips in paralel without any balancing resistors inside. And of course, triplet segmet which draws more current gets hotter which reduces resistance even further and leads to self destruction.
mawyatt wrote: ... a simple power transistor or MOSFET and a TL431 reference diode is all that's required to create a simple, cheap ($1~2), adjustable, current source for LED control.
Since I have a box full of TL317/337 regulators, I usually make simple enough current source which is totaly adequate for driving leds up to 1 amp. The one below is tailored for COB ring lights we are talking about.
My preferred way to drive the COBs, and how I've been doing it for years now since @rocky introduced us all to the Angel Eyes COB ringlights, is to use a voltage source with a limiting resistor. This allows me to use general-purpose supplies, and tailor the currents and thus LED intensity to the application. I have around 10 different COB ringlights of different diameters and masking configurations that I use for different purposes, and have settled on the connector configuration from the Angel Eyes as my preferred interface to allow quick changes.

Post Reply Previous topicNext topic