|[ Main index | Bicycle components tests | (Dynamo) bicycle lighting | Headlamps that have a cutoff, battery powered | Xanes SFL-01||Dutch: Deze pagina in het Nederlands ]|
Pics to come. Videos made and I will put these on youtube.
To judge pictures/videos of bicycle lamps there are a few problems:
- Perspective overaccentuation of the near field: The near field takes up most of the space on the screen and so being bright there gives an impression of a bright lamp... But the essence is farther away and note that what you see in reality is not a total picture, but the section of road that you focus on.
- The limited dynamic range of cameras and displays along with no accomodation of the eye to low light levels when looking at a screen in normal ligthing conditions: This means you cannot properly convey what you see at the edge of beams, and how well it lights up the road at the lower intensities.
2018-3-15: Axendo 60 USB rechargeable and Xanes SFL-01 (from Banggood.com) compared on several rides along with the Bidi dynamo headlamp in neutral white and Philips Saferide 80 neutral-white: The Xanes SFL-01 is very small and gives a lot of light but not 600lm as the site says and also not 400 lm as the manual says. I would estimate from comparing e.g. light bounce against a wall with the Saferide 80 and Axendo 60 that it produces at most what the Saferide 80 produces, and likely it is 200-250 lm. A comparison with the Silva head mounted lamp that produces about 600 lm clearly lights up far more at at least the same intensity, so that confirms the estimate, but note that it is very difficult to estimate total light output. I have disassembled the lamp and will measure the current in high mode to get an estimate of what amount of light you will really get. The light colour is blueish, not good, and it has a low lux rating which means I can only properly see the surface of the road up to about 30-35 m (the white dots on the middle of road are barely distinguishable from that point on), whereas with the Axendo 60 about 50m, with the Saferide 80 about 60-70m. The beam of the Xanes is very even and not very wide but wide enough, so riding with it is pleasant. I would recommend this lamp as a backup to take with you on long tours even if you have dynamo lighting as it's so small and quite light. I would recommend it only for low speed cycling due to the short reach. I took the Xanes apart and will do measurements of current to the LED.
The smallest T8 that I have doesn't fit, so instead of getting some more Torx keys or screwdrivers, I used a regular screwdriver. This can be easily done with Torx, as the teeth give a perfect hold in case you have a screwdriver that fits exactly in the bolthead. By the way, this is why it's silly to assume Torx to better as a theft deterrent, it's worse than a regular inner-hex bolt in this respect.
Current at low level ca. 0.322A and at high level: 0.644A (measured with an Aneng 8009, and with a Fluke 289). An XPG-R5 of ca. 6200K puts out according to the Cree XPG datasheet ca. 139 lm at 350mA. This means at 0.644A it would put out about 1.75 times that, so ca. 243 lm. As Cree datasheets have been optimistic in all cases that I know of were the XPGs were measured in light output by other people than Cree themselves, I would think 200-220 lm is more likely. And then you need to take off about 10% for the loss from the reflector and lens, which would give ca. 180-200 lm (and at best ca. 220 lm if the LED actually reaches datasheet spec, which is doubtful due to e.g. no thermal grease/glue between LED MCPCB and housing). In any case, the current measurement through the LED thus confirms that my initial estimate of the total light output was quite close to what it really is. [ Note: The resistance of the measurement leads will affect the current measurement by what I estimate about 2% in case the driver is not a true current regulator ]
On the battery is written 1200mAh, which would give at best a runtime at high of 1.2Ah x 3.7V (average voltage of the LiPo battery over the discharging range) / (ca. 2.99V (without cooling body attached), 0.644A) = 2.31h, taking away 10% assumed inefficiency of the driver gives 2.08h, which is close to the stated runtime, so it all fits. I still need to check actual runtime, and I will check power going into the device when charging.
I will probably replace the red side LEDs with yellow ones, or just remove the red ones: red to the front = stupid. Also the blue status LED for auto mode is too bright, I may just put some tape in the housing.
More to come.
|For email go to the email page|
Last modified: 2018-3-15