Constant Current supply - Circuits??

[xbrite]

Hello guys,

Is there any good adjustable Constant current circuit with step up regulator for 2 NiMH application you know of??

I think the Madmax seems a simple step up voltage regulator so the current thru the LED goes up when the LED is heated up which in turn lowers the Vf. I don't know about the Badboy but it seems that the internal current limit of the output transistor determines the output current and I don't know how the current is adjusted.. Correct me if I am wrong..

I have a decent circuit with the step down regulator IC using an external voltage reference, a rail-to-rail OP amp and a current sense resistor of 0.05 ohm stepping down from 2 Li-Ion cells(7.4V nominal) for both 1W and 5W.

I tried once a Step up regulator with 2 NiMH which worked pretty good at voltage mode but once I added an OP amp, a Voltage reference and a sense resistor, it didn't work properly any more. The current didn't stay stable but varied as the battery voltage dropped. I could not make it stable thru out the battery discharging cycle. I guess the supply voltage of the OP Amp is not high enough as I only applied 2.4V.. I later modified B+ supply from battery to the output to LED which is 3.2V approx but it didn't help..

Thanks for the help in advance..

 

[tonyb]

Try a LM3574 from National Semiconductor.

 

[Entropy]

[ QUOTE ]
xbrite said:
Hello guys,

Is there any good adjustable Constant current circuit with step up regulator for 2 NiMH application you know of??

I think the Madmax seems a simple step up voltage regulator so the current thru the LED goes up when the LED is heated up which in turn lowers the Vf. I don't know about the Badboy but it seems that the internal current limit of the output transistor determines the output current and I don't know how the current is adjusted.. Correct me if I am wrong..

I have a decent circuit with the step down regulator IC using an external voltage reference, a rail-to-rail OP amp and a current sense resistor of 0.05 ohm stepping down from 2 Li-Ion cells(7.4V nominal) for both 1W and 5W.

I tried once a Step up regulator with 2 NiMH which worked pretty good at voltage mode but once I added an OP amp, a Voltage reference and a sense resistor, it didn't work properly any more. The current didn't stay stable but varied as the battery voltage dropped. I could not make it stable thru out the battery discharging cycle. I guess the supply voltage of the OP Amp is not high enough as I only applied 2.4V.. I later modified B+ supply from battery to the output to LED which is 3.2V approx but it didn't help..

Thanks for the help in advance..

[/ QUOTE ]
The output current of the BB is set via a feedback resistor on the board. It's identical to the FB settings on chips designed to be voltage regulators, except the voltage is much lower. (On the order of .1 to .2 volts, instead of 1.2 like the FB voltages used by most voltage regulators. I don't remember exactly what it is.)

It's the Zetex ZLT series that use the current of the switching transistor for regulation. In essence, the regulation of those units is of the *INPUT* current.

 

[Jonathan]

I use the LT1618. It can regulate either output voltage _or_ output current. It has an 'adjust' pin input; by changing the voltage applied to the pin, you change the thresholds for current regulation, and thus can adjust the regulated output current with a simple pot.

It will run a 1W LS from a pair of AA NiMH cells; don't know how it would do trying to run a 5W LS from 4AA NiMH cells, but it just might be able to do so.

http://www.borealis.com/~winnie/LED_BOOST/

-Jon

 

[rox]

i use the LM317. its cheap and can regulate voltage or current. you can use 2 of them to do both. it even has example circuits in the product datasheet for doing just that.

 

[xbrite]

Thanks, guys..

I am really looking for a circuit with higher than 90% efficiency.. Thus, I prefer a synchronous rectification.. LM317 is not a choice as the efficiency is not good and requires higher Battery voltage.. I am afraid if it is confidential or not but does anybody know the type number of IC used in Badboy or device equivalent to that??

 

[Doug Owen]

[ QUOTE ]
xbrite said:
Thanks, guys..

I am really looking for a circuit with higher than 90% efficiency..

[/ QUOTE ]

Nobody can blame you. I'd like a car that gets 'an honest' 60 MPG, your circuit could be in the glove compartment......

Unfortunately, real world efficiency numbers are somewhat less under the constraints. Best results will come from careful custom design the limits input voltage range and is tailored to a specific load (LED). If you look around, you'll see that typical performance probably averages something closer to 70%. A look at what folks with a lot more resources than you or I are able to deliver might shed some light on this (pun intended).

Doug Owen

 

[dat2zip]

Yup,

It's the LT1618... Ha! The cat's out of the bag now...

(And no, it doesn't get better than 90% efficiency)

Syncronous switchers are the only way to get past the 90% eifficency mark. Maxim has probably the most in this style.

The Maxim 1675 running wide open will average more than 90% most of the time. 94% with fresh set of batteries, and downto 89% when the batteries are low. It's a voltage output regulator, but, it can only deliver X Watts, so, hanging a Luxeon on it and setting it's output to 5V will run it wide open. Current and voltage regulation in this mode is quite good. The layout of this IC is aweful. The IC has a major bug in the die with hideous ground lead bounce. Maxim denies any of this.

There are several tricks to convert a voltage switching regulator to a current mode regulator.

Wayne

 

[xbrite]

Thanks, Wayne.

I do appreciate it. I will try with LT1618. Many thanks.. ^^

 

[tonyb]

Is a Dr. cuk circuit better than 90% Eff.?

 

[Doug Owen]

[ QUOTE ]
tonyb said:
Is a Dr. cuk circuit better than 90% Eff.?

[/ QUOTE ]

Do you mean 'direct drive'?

Doug Owen