apparent AMC7135 voltage drop problem

[jesusthepirate]

I've used the wiring diagram that netkidz posted a while back in order to use a pair of standard 1400mA AMC7135 boards from DX in parallel, and I'm having a little trouble. I've been trying to run a Cree XR-E P4, a Cree XR-E Q5, and an SSC P7 (all separately of course) at ~3 amps. The XR-E's are TEC cooled. I haven't measured vF on the SSC, but for the Cree's it's about 4.5vF @ 3.2A, and I just can't do it on 4x NIMH AA (Energizer 2500mAh) or generic 3000mAh Sub-C... I can get about 2.4A going through the P7 and around 1.5A through the Cree LEDs with that setup. Thing is, if I up the cell count to 5 I have no problem, 3.0-3.3Amps works fine through any of the mentioned LEDs.

Voltage sag with -4- cells is not bad, under load they are maintaining 5.5v pretty easily, so there's a whole 1.0 volt between vBattery and vF... I just don't see what's going wrong. In my frustration I've wired up a second pair of those AMC7135 boards, removed the reverse polarity protection diodes (to no real effect of course), etc...

Sorry for the wordy post and thanks for any help!

 

[ahorton]

How are you measuring current? With a digital multimeter in series?
If so, work out what the resistance of the multimeter is and take that into consideration.

Mine has a resistance of 1 Ohm. At 1.5A, that's a voltage drop of 1.5V which you don't notice when you have 5 cells.

 

[jesusthepirate]

How are you measuring current? With a digital multimeter in series?
If so, work out what the resistance of the multimeter is and take that into consideration.

Mine has a resistance of 1 Ohm. At 1.5A, that's a voltage drop of 1.5V which you don't notice when you have 5 cells.

Yes, I'm measuring current with a DMM in series with the LED. Do you have a link or some quick instructions on how I can figure out my DMM's resistance?

Unfortunately I've only got one multimeter, I know that puts a cramp in my style... I'll have a second (cheap) one from DX in a couple of weeks, but in the mean time are there any tricky ways I can set things up to get a better idea what's going on? I didn't try measuring the amperage the 7135's are pulling from the battery pack, but that seems like a pointless measurement if the DMM's resistance is my problem here.

Maybe tonight I'll try measuring the voltage at the LED when using a 4-cell and a 5-cell pack... I'll let voltage settle and if it's the same... does that tell anything with much certainty?

 

[ahorton]

When I set my DMM to find resistance and touch the probes together I get (in the 200 Ohm range) I get a reading of about 1 Ohm. Of course this may not work because you can't be sure the DMM doesn't take into account its own resistance and the circuit is different for testing current.

So just measure the voltage at the LED since you know what the Vf is at a given current.

Another option is to find the smallest resistor you can and put it in the circuit. Then measure the voltage drop across it and calculate the current that way.

Why are you pushing so much current through these little LEDs? Do you notice any real difference in light output?

 

[jesusthepirate]

Why are you pushing so much current through these little LEDs? Do you notice any real difference in light output?

I can't see the difference... to my eyes anything above 700mA really just registers as "very bright". I'm thinking of building a toy for throw, and the Cree XR-E die is the brightest light source I knew of that is appropriate for use with an aspheric.

Thanks for the suggestion on measuring the resistance of my multimeter, I get similar results... I forgot that not only is this a crappy multimeter, I also have extra-crappy leads hooked up to it.

I've never been 100% positive about the way an LED's vF works, if someone would write this in stone for me I'd appreciate it: There's no way you could measure the same voltage across an LED at two different current drains, right? So if I determine that vF is 4.5 at 3 amps (using 5 cells), then measure voltage across the LED with only 4 cells and it is 4.5, then I am certainly operating at 3 amps? Like I said, I was never 100% clear on this elementary issue...

 

[Mr Happy]

I can't see the difference... to my eyes anything above 700mA really just registers as "very bright". I'm thinking of building a toy for throw, and the Cree XR-E die is the brightest light source I knew of that is appropriate for use with an aspheric.

There may in fact be no difference. If you drive an LED too far beyond the manufacturer's design current it may even get dimmer.

 

[LED_astray]

... I've never been 100% positive about the way an LED's vF works, if someone would write this in stone for me I'd appreciate it: There's no way you could measure the same voltage across an LED at two different current drains, right? So if I determine that vF is 4.5 at 3 amps (using 5 cells), then measure voltage across the LED with only 4 cells and it is 4.5, then I am certainly operating at 3 amps? Like I said, I was never 100% clear on this elementary issue...

Yes, no, maybe.

There is a knee in any diode's I-V curve. (IIRC- LEDs aren't as sharp as signal or rectifier diodes. But it's still there.)

First, the very simple view. For voltages below Vf very little current flows. Around Vf current starts to flow freely and will increase with very little voltage change (until you kill the diode by being out of spec.)

I suppose if you have accurate and precise enough equipment you can measure diode V and predict the current. (For that diode, once you've characterized it. There is process variation and the measurement won't hold for another instance of the same diode type.) I've never tried it for an LED, but I bet it is not possible with the kind of equipment we'd have at home.

The current is (mostly) set by the rest of the circuit. That is why we can size resistors for simple circuits well enough with a simple calculation. (Based on Ohm's Law: V=IR. We control the current in the resistor and, since they are in series, the same current flows through the LED. Voltage across the resistor is V = Vbatt - Vf, and so R = V / I. E.g. Vf = 3.7v, Vbatt = 4.5v, & desired I = 350mA. Then V = 4.5-3.7 = 0.8v and R = 0.8/0.350 = 2.3ohms.)

 

[jesusthepirate]

@ Mr Happy: I do/did realize that, I just sort of assumed that the limit should be (somewhere) around 2-4A, and since I had everything soldered up to supply about 3A already, that's where I started. Seeing a reputable member (who I've seen post his own findings...) bring it up made me dig a little, and here we go, sure enough 2.5A w/ sub-ambient cooling is my limit (for recent bins, which is what the eventual plan involves).. good job making me get off my butt..

@LED_astray: hmm. unfortunate.