Inside B+M IQ Fly Senso Plus

[syc]

I had been kind of curious if anyone had opened up one of these, and it looks like someone did, and documented it as well.

Here's a picture of the most interesting part, the emitter/reflector assembly:



Looks pretty straightforward to either upgrade the emitter, or use the reflector assembly for other projects.

 

[ihab]

That's my blog. Glad I could be of help!

 

[syc]

Cool! Did you ever consider upgrading the LED?

That's my blog. Glad I could be of help!

 

[ihab]

Cool! Did you ever consider upgrading the LED?

Good question. So if it has an XR-E Cool White R5 emitter right now then, from the data sheet, at 700mA I would get:

170% * 107lm = 182lm

If I upgraded to an XP-G Cool White R5 then, at the same current, I would get:

185% * 139lm = 257lm

This is a nontrivial increase for sure, but does not seem worth risking downtime of my bike light that I rely on every day.

(=> Is there some other emitter that I'm missing?)

Plus note from the picture that the design requires a somewhat specially shaped MCPCB to fit into the reflector assembly properly and have the wires coming out of the back of the MCPCB.

Ihab

 

[syc]

Ihab,
Are you sure about the LED's bin and current? I'm curious what bin XRE they put in there - I came across an earlier thread (which incidentally also had a disassembly of the IQ Fly - I missed that one) that claims the 60 lux of the Supernova E3 is about 220 lumens. I believe the IQ Fly uses the same reflector, so the 40 lux IQ Fly might be more around 140 lumens. Normalizing for the current, the output for binning purposes is around 82, for a P4 bin.
Going to an R5 might come close to doubling the output. But I agree, if its your only light, its not really a good candidate for tweaking.

Good question. So if it has an XR-E Cool White R5 emitter right now then, from the data sheet, at 700mA I would get:

170% * 107lm = 182lm

If I upgraded to an XP-G Cool White R5 then, at the same current, I would get:

185% * 139lm = 257lm

This is a nontrivial increase for sure, but does not seem worth risking downtime of my bike light that I rely on every day.

(=> Is there some other emitter that I'm missing?)

Plus note from the picture that the design requires a somewhat specially shaped MCPCB to fit into the reflector assembly properly and have the wires coming out of the back of the MCPCB.

Ihab

 

[markus_i]

There's a similar thread currently running on a german bike light mailing list. According to the experts in the field (i.e. the guys who already tried this), replacing the XR-E with an XP-G won't help that much since the (apparent?) emitter size in the XP-G is larger than in the XR-E, making it harder to impossible to get a similar good focus for the XP-G, resulting in less light on the road.

Get an Edelux instead ;-)

Bye
Markus

 

[syc]

I saw photos of a Cyo mod to an MCE on a German site and the beam pattern was very dispersed, so I'm not surprised.

Its sounds like the only reasonable way to mod these lights would be to get it to the R2 bin if it isn't already, and maybe use the boost caps inline with the power from the dynamo so that you get the current peak.

There's a similar thread currently running on a german bike light mailing list. According to the experts in the field (i.e. the guys who already tried this), replacing the XR-E with an XP-G won't help that much since the (apparent?) emitter size in the XP-G is larger than in the XR-E, making it harder to impossible to get a similar good focus for the XP-G, resulting in less light on the road.

Get an Edelux instead ;-)

Bye
Markus

 

[markus_i]

Forget about the boost cap. The electronics inside the light already does a (roughly - minus losses) unregulated 1:2 current transformation for the LED. Being unregulated, this becomes dangerous if you exceed the usual 500 mA generator current (or connect a DC source > 2* LED forward voltage). You might get a little bit of extra light if you can find a way of improving the cooling, especially the IQ Fly with its heat sink _inside_ the plastic case is a somewhat unlucky design.

Bye
Markus

 

[deNaranja]

Hi, i'm from the german bike forum and also i modded my cyo, which uses nearly the same reflector as the IQ Fly.

i drive my lamps with a shimano hub.

First i exchanged the xr-e with a mc-e (wired to a mosfet-rectifier) all 4 DIES's in series. the lightbeam-pattern was okay, a little bit softer on the edges but even better than my doppelE6halbe (beam-shot).

But in fact of the high foward voltage of the mc-e the light is at normal speeds 10-30 km/h not bright enough respectively at lower speeds it flickers too much.

thats why i exchanged the mc-e with a xp-g (R4 bin) and used the original cyo-cuircit (with "Standlicht" and currrent-doubler)

the lightbeampattern is also okay and a little bit brighter.

the xp-g runs at 0,8A and 3,1V, this must be around 295lumens (a bit lower), without optic losses

But, it is not bright enough :( (especially on rainy streets)

Here are some pictures of my cyo-mod:

1
2
3
4

When i drive my xp-g direclty with the mosfet-rectifier, i get a current of 0,5A, so why does the current-doubler of the cyo-circuit does not double it?
can anybody make some mesurement with an xp-g and the circuits of the cyo or iq fly?

thanks so far, tino

 

[markus_i]

...
When i drive my xp-g direclty with the mosfet-rectifier, i get a current of 0,5A, so why does the current-doubler of the cyo-circuit does not double it?
...

Two possiblities:
a) (most likely): losses. The doubler uses a simple (read: cheap) circuit that has some potential for optimization. Since we're running only at 3.5 volts, even the losses on the (Schottky?) rectifier diode become significant. Other potential losses: inductor resistance, inductor saturation, too slow driving (i.e. not enough driver current) of the FET
b) (possibly): intentionally running the converter not at 50% duty cycle (I'm too lazy at the moment to work out whether you'd need to run at lower or higher duty cycle - ask me again in half a year or so...) in order to keep safely below the LED limit of 1000 mA peak current.

BTW, what kind of instrument do you use to measure the LED current? The associated wave forms have a very ugly shape and unless you use something that can measure TRMS at high crest factors, all you're getting will be house numbers that may be correct within an order of magnitude.

Bye
Markus

 

[yellow]

how about squeezing two XP-G inside?
two - better three - led is simply the best number when using a dynamo
(but three wont fit, two should),

or - to get light at lower speeds - connect just three of the MC-E plates, ;)

or put an Elko into the circuit, leaving one floating freely, so the Mosfet still works.
Might fight flickering a bit

 

[deNaranja]

@markus_i
i think i measured the peak-values, my multimeter is not true RMS. I used the Voltcraft VC220.

@yellow
yes i thought the same, to put two xp-g on one pcb, but therefore i need to practice soldering them, it would be the first time.

i tested the cyo with the mc-e with only 2 dies connected and i get light earlier (this was okay, like my doppel E6 with two xr-e) but the enlightend area was smaller, because the near-field was missing.

i think with the mc-e i should use the strassikowski-circuit.

i have now a testversion of lucidshape for designing a reflector i will try if i get some results.