MTE Rebel 100 Converter data

[chimo]

I picked up a few of the MTE Rebel 100s from DX (reviewed in this thread). I cracked one open to conduct some tests on it.

(mods: feel free to merge this with the linked thread if you would like them combined)

Light Data:
1xAA Primary
Multimode boost driver: High-Med-Low-Strobe-SOS (the strobe is very fast)
Very lightweight (almost flimsy)
Smooth reflector (IMS17-like)
Reverse tail clicky

My power supply would only source about 3.4A of current so the High mode test does not go to 1.4V

Measurement data

Current in/out (it's hungry on high!)

Coverter efficiency

 

[koala]

Hello Chimo,

Been busy with circuits lately?

Is this toroidal based voltage step-up?

 

[chimo]

Is this toroidal based voltage step-up?

That's the one! I had a bunch of stuff arrive around the same time - the problem is finding the time to play...

 

[yekim]

I opened mine up and the board is identical to the sku7880 that everyone likes. I cannot see any physical differences between the MTE board and the 7880's that I got in at the same time anyhow.

 

[Nake]

I opened mine up and the board is identical to the sku7880 that everyone likes. I cannot see any physical differences between the MTE board and the 7880's that I got in at the same time anyhow.

I've noticed that with a few of the MTE AA(1.5V) only lights. Some are 20 modes, some are 5.

 

[MorePower]

Were you measuring Vout and Iout at the emitter? Your Vout measurements look a bit off to me, because a white LED won't light at 2.2 volts. That's at least part of the reason the efficiency for the low level is only ~30-40%.

I'd expect a Rebel 100 to light at around 2.5 volts, minimum, and at 100mA, Vf would be near 2.85 volts.

 

[chimo]

Measurements were taken off low value sense resistors (0.05/0.1 ohm) inserted in the cct. Recall that there can still be current flowing through a LED without noticeable light. Cheers,

Were you measuring Vout and Iout at the emitter? Your Vout measurements look a bit off to me, because a white LED won't light at 2.2 volts. That's at least part of the reason the efficiency for the low level is only ~30-40%.

I'd expect a Rebel 100 to light at around 2.5 volts, minimum, and at 100mA, Vf would be near 2.85 volts.

 

[koala]

chimo, by (0.05/0.1 ohm) do you mean 0.05% tolerance?

I just tried powering a NWOD with my linear power supply. There's about 4.5uA flowing through the LED at 2.2000V. Not trying to prove anything here but perhaps you got a C bin vf Rebel!!!

Can you try connecting the Rebel directly to the power supply without the boost circuit and check if it matches your chart? It could very much be an unobtainium bin Rebel!

 

[chimo]

I forgot to mention that I was using an old Lux or LuxIII and not the rebel. Sorry for the short responses - I am still on the road and am using my iPod Touch - not fun for a lot of typing.

 

[MorePower]

Measurements were taken off low value sense resistors (0.05/0.1 ohm) inserted in the cct. Recall that there can still be current flowing through a LED without noticeable light. Cheers,

Thanks for the responses. No rush on replies, but when you get back to an actual computer, could you explain where / how you measured V out and I out?

I assume you had the sense resistor in series with the Luxeon and measured V drop across it, then used that to calculate current through the Lux via I=V/R, right? My question is, where were you measuring V out?

 

[chimo]

Thanks for the responses. No rush on replies, but when you get back to an actual computer, could you explain where / how you measured V out and I out?

I assume you had the sense resistor in series with the Luxeon and measured V drop across it, then used that to calculate current through the Lux via I=V/R, right? My question is, where were you measuring V out?

Correct on the method. Vout was measured across both Rsense and the LED (of course). Vin and Iin were measured in a similar manner, however, Vin was the voltage to the driver only (not the driver and sense resistor).

As mentioned, the LED was an old Lux or LuxIII. It had a fairly high Vf (perhaps a K or L bin)

 

[VegasF6]

Chimo, how long ago did you order this light? I waited 3 1/2 months for them to ship it, before cancelling about 2 weeks ago! I figured it was because of the rebel recall and we would never see it again. Let's see what you think of it, and if you get tired of them, maybe I could take one off your hands.

I don't really see that the light would serve any purpose for me that my others don't, but I don't own any Rebel lights, and that is just a shame

 

[chimo]

Chimo, how long ago did you order this light? I waited 3 1/2 months for them to ship it, before cancelling about 2 weeks ago! I figured it was because of the rebel recall and we would never see it again. Let's see what you think of it, and if you get tired of them, maybe I could take one off your hands.

I don't really see that the light would serve any purpose for me that my others don't, but I don't own any Rebel lights, and that is just a shame

They were ordered around 29 Mar. Looks like it took about 3 weeks to arrive.

Koala: The resistors were 0.05 ohms and 0.1 ohms - both were 1% tolerence. I used small resistors as to not affect the circuit operating parameters too much.

 

[jtr1962]

I bought some of these boards from DX recently. To add another data point on high with Vin at 1.25 V (measured right at the board), I measured Iin=4.29 A (!), with Vout of 3.70 V and Iout of 979 mA. That's a conversion efficiency of 67.5%, which isn't horrible considering the size of the board and the currents involved. I didn't want to take things further as the board was getting seriously hot. Also, even from a freshly charged AA NiMH in actual use the board won't pull much over 3 amps anyway due to cell/switch resistances.

My figures for 1V were Iin = 2.08 A, Vout=3.35 V, and Iout=477 mA. Conversion efficiency was 76.8%, not bad at all. Another data point was Vin=0.865 V, Iin=1.00 A, Vout=3.13 V, Iout=225 mA (efficiency = 81.4%). The board appears to use a transformer-based step-up circuit rather than the usual MOSFET/inductor/schottky diode.