Misterbenn
Active Member
Hi All,
I'm using this thread as a way to document my repair efforts for my LED bike light. This is mainly for my benefit so I can have all the information I need in one place however I hope that it will be of use to other people with similar problems and others may find it interesting. I've no particular need of assistance or answers but as always the advice and experience is very welcome.
I will be updating this first post as I make progress, I know this brakes the usual thread structure but this way all the info will (eventually) be in one post.
The Light
The LED light in question is a 4 Cree LED lamp with a 4 cell Lithium-ion battery. It comes with 3 modes; mode1 On; mode2 dimmed; mode3 flashing, I also consider that it has a forth mode - mode0 Off.
It is an import from china and a very common listing on eBay, The following link is to an identical looking light
https://www.ebay.co.uk/itm/like/400554413825?limghlpsr=true&hlpv=2&ops=true&viphx=1&hlpht=true&lpid=108&chn=ps&device=m&adtype=pla&crdt=0&ff3=1&ff11=ICEP3.0.0-L&ff12=67&ff13=80&ff14=108
The Problem
While messing about with this light I accidentally reversed the polarity. On realising my mistake and correcting the polarity the light initially seemed fine in its off state (which has a small green power LED) however upon switching it to mode1 the main LED came alive for a short amount of time (<1s) and then died.
Subsequent monitoring of the battery voltage showed that when the LED was turned on the battery voltage dropped significantly, initial speculation is of some kind of short in the LED drive circuit.
Drive Circuit Photos
Underside
Topside
The topside contains the LED drive circuitry while the underside is the mode select and undervolts logic
Drive Components
SS54 Diode
https://www.farnell.com/datasheets/1441695.pdf
The diode was checked with a DVM and looks to be working correctly
D38NH02L Fet
https://www.farnell.com/datasheets/1441695.pdf
Checked with DVM to confirm no hard S/C across the FET and also checked for P/N junction voltage drops.
5241 Driver IC
https://monitor.espec.ws/files/led_driver_sot-23-6_marking_5241b_745.pdf
... and the google translate link
https://translate.google.com/translate?hl=en&sl=zh-CN&u=https://monitor.espec.ws/download.php%3Fid%3D96287&prev=/search%3Fq%3Dqx5241%26biw%3D1366%26bih%3D683
Checked some of the functions with a DVM, the Vref output has been measured to be 5V and the Fet drive output is also working.
Circuit topology
The datasheet for the 5241 driver tells us that it is a buck type constant current driver and gives the following type of circuit configuration
But this circuit topology doesn't stack up with test results.
For example I removed the main LED and the circuit still pulled a large current when switched to mode1 (LED on), based on the circuit above that shouldn't / couldn't happen.
Further investigation and probing of the PCB revealed the actual circuit topology used, a boost circuit as shown below.
With a boost circuit such as this you can never turn off the LED while power is available, so that raised the question of what was happening in mode0 (LED off).
It turns out that the battery voltage is 8.4Vmax and the minimum Vf of the LED string is 9.3V at the point they visibly emit light. So in mode0 this circuit is just using the Vf of the LEDs to turn off the circuit while in mode1 the voltage is boosted. This isn't what the driver is designed to do but I guess it works.
This also explains why the circuit was drawing high current even with the LED disconnected, as the current feedback in the driver was adjusting the fet duty cycle to compensate.
I also discovered the undervoltage shutdown for the IC is set at 7V, so I've only got 1.4V of battery discharge! Not sure what's safe for a 8.4V Lithium-ion battery but 1.4V seems small to me.
Testing Update
With my new understanding of the circuit I did some additional testing yesterday using a bench top supply set to 8.4V and 10A.
With the LED full on current draw was 7.4A so that's 62W, although that's probably a peak value? Anyway the 'power' LED started to flash red after a short time - I'm not sure what that signals overcurrent? overtemp? undervolts?
With the LED dimmed the current draw is 2.4A which is a more reasonable 20W and the light seems ok I think, if a little nosier than I remember.
Testing with a new bike light and the current draw is 1.5A full power, 0.35A dimmed. So there is definitely something wrong with the driver, although it is working (otherwise the LED would not be lit!) So perhaps the current feedback is fried??
Battery Tests
at 2.5A the battery voltage dropped 0.6V from 7.95 to 7.35V. so that's 0.24Ohms of internal resistance which looking on the internet seems about nominal. And the voltage went below 7V at 4A.
Further work
Take another look at boost controller? To confirm the current feedback on/off voltage by removing the sense resistors and stimulating with a sig gen.
I'm using this thread as a way to document my repair efforts for my LED bike light. This is mainly for my benefit so I can have all the information I need in one place however I hope that it will be of use to other people with similar problems and others may find it interesting. I've no particular need of assistance or answers but as always the advice and experience is very welcome.
I will be updating this first post as I make progress, I know this brakes the usual thread structure but this way all the info will (eventually) be in one post.
The Light
The LED light in question is a 4 Cree LED lamp with a 4 cell Lithium-ion battery. It comes with 3 modes; mode1 On; mode2 dimmed; mode3 flashing, I also consider that it has a forth mode - mode0 Off.
It is an import from china and a very common listing on eBay, The following link is to an identical looking light
https://www.ebay.co.uk/itm/like/400554413825?limghlpsr=true&hlpv=2&ops=true&viphx=1&hlpht=true&lpid=108&chn=ps&device=m&adtype=pla&crdt=0&ff3=1&ff11=ICEP3.0.0-L&ff12=67&ff13=80&ff14=108
The Problem
While messing about with this light I accidentally reversed the polarity. On realising my mistake and correcting the polarity the light initially seemed fine in its off state (which has a small green power LED) however upon switching it to mode1 the main LED came alive for a short amount of time (<1s) and then died.
Subsequent monitoring of the battery voltage showed that when the LED was turned on the battery voltage dropped significantly, initial speculation is of some kind of short in the LED drive circuit.
Drive Circuit Photos
Underside
Topside
The topside contains the LED drive circuitry while the underside is the mode select and undervolts logic
Drive Components
SS54 Diode
https://www.farnell.com/datasheets/1441695.pdf
The diode was checked with a DVM and looks to be working correctly
D38NH02L Fet
https://www.farnell.com/datasheets/1441695.pdf
Checked with DVM to confirm no hard S/C across the FET and also checked for P/N junction voltage drops.
5241 Driver IC
https://monitor.espec.ws/files/led_driver_sot-23-6_marking_5241b_745.pdf
... and the google translate link
https://translate.google.com/translate?hl=en&sl=zh-CN&u=https://monitor.espec.ws/download.php%3Fid%3D96287&prev=/search%3Fq%3Dqx5241%26biw%3D1366%26bih%3D683
Checked some of the functions with a DVM, the Vref output has been measured to be 5V and the Fet drive output is also working.
Circuit topology
The datasheet for the 5241 driver tells us that it is a buck type constant current driver and gives the following type of circuit configuration
But this circuit topology doesn't stack up with test results.
For example I removed the main LED and the circuit still pulled a large current when switched to mode1 (LED on), based on the circuit above that shouldn't / couldn't happen.
Further investigation and probing of the PCB revealed the actual circuit topology used, a boost circuit as shown below.
With a boost circuit such as this you can never turn off the LED while power is available, so that raised the question of what was happening in mode0 (LED off).
It turns out that the battery voltage is 8.4Vmax and the minimum Vf of the LED string is 9.3V at the point they visibly emit light. So in mode0 this circuit is just using the Vf of the LEDs to turn off the circuit while in mode1 the voltage is boosted. This isn't what the driver is designed to do but I guess it works.
This also explains why the circuit was drawing high current even with the LED disconnected, as the current feedback in the driver was adjusting the fet duty cycle to compensate.
I also discovered the undervoltage shutdown for the IC is set at 7V, so I've only got 1.4V of battery discharge! Not sure what's safe for a 8.4V Lithium-ion battery but 1.4V seems small to me.
Testing Update
With my new understanding of the circuit I did some additional testing yesterday using a bench top supply set to 8.4V and 10A.
With the LED full on current draw was 7.4A so that's 62W, although that's probably a peak value? Anyway the 'power' LED started to flash red after a short time - I'm not sure what that signals overcurrent? overtemp? undervolts?
With the LED dimmed the current draw is 2.4A which is a more reasonable 20W and the light seems ok I think, if a little nosier than I remember.
Testing with a new bike light and the current draw is 1.5A full power, 0.35A dimmed. So there is definitely something wrong with the driver, although it is working (otherwise the LED would not be lit!) So perhaps the current feedback is fried??
Battery Tests
at 2.5A the battery voltage dropped 0.6V from 7.95 to 7.35V. so that's 0.24Ohms of internal resistance which looking on the internet seems about nominal. And the voltage went below 7V at 4A.
Further work
Take another look at boost controller? To confirm the current feedback on/off voltage by removing the sense resistors and stimulating with a sig gen.
Last edited:
