-
Any hope for this sick mouse? Razer Lachesis with nonfunctional laser sensor.
Posted by magic on 07 Dec, 2021 09:40 -
Here's the patient.
The problem: the sensor doesn't light up and the mouse doesn't react to movement. USB connection works and buttons also do.
It's a two layer board with components on top and three ICs: some MCU on the right, a 3.3V LDO in the upper left and the sensor in the center.
Sensor pinout, clockwise:
- top two pins - a 32MHz oscillator, this appears to be working
- ground
- 3.3V rail
- ground
- internal 1.8V rail apparently, bypassed by C22
- 4 signals that appear to be SPI coming from the MCU: CS, CLK, MISO, MOSI
- a mystery I/O, pulled up to 5V by R22 and going to the MCU
The voltage rails are clean.
Activity on SPI:
CS: pulses low at 1kHz
CLK: bursts of 1MHz clock, 7 × 8 bits each 1ms
MOSI: a single bit each 1ms
MISO: a repetitive bit pattern
The sensor appears to be backfeeding the 3.3V rail and drawing ~150µA through the mystery I/O and its pullup. There is 1.5V drop across the 10kΩ pullup regardless of USB voltage and 3.3V is 11mV below the mystery I/O. Therefore 3.3V rail is 3.4V on 4.9V USB and 3.6V on 5.1V.
The MCU has its internal 3.3V rail sitting at 3.2V. SPI runs at 5V and is clamped by the sensor to 4.2V (120Ω series resistors are included), MISO is 3.4V~3.6V.
I presume the sensors should be drawing several mA at least. Any idea what's going on and what could be wrong here?
-
The sensor needs the light to work. So the non working light may be the main problem. Some optical mice use a laser diode, and these are relatively susceptible to ESD damage. Even LEDs sometimes fail, though not very often.
-
Here's the patient.
don't get it, is it meant "The problem: the LED doesn't light up ..." ?
The problem: the sensor doesn't light up .. -
The sensor appears to be some Philips PLN series thing. I was able to find one summary datasheet for the PLN2020. My sensor looks about the same, has PLN written on it and even the MLCC in the center inside. The pinout matches too, but there is no explanation of the mystery pin 9 in this datasheet.
https://www.mikrocontroller.net/wikifiles/9/9b/PLN2020.pdf
The lasers are in the sensor and there is supposedly two of them. So either both diodes are blown, or the controller is blown, or maybe the SPI commands are wrong
I have no info about the protocol and they mention "upgradeable embedded firmware". Well, I suppose that could be corrupted too.
Power consumption of the sensor is surprisingly low. As I said, it appears to draw all its power through pin 9 and pull the 3.3V rail above spec. Or maybe through the SPI pins? I guess I will have to take a closer look with a scope.
-
I scoped all the I/O pins and their series resistors and found:
- pin 9 draws a constant 0.15mA
- CS draws 3mA with 80% duty cycle, more voltage is dropped by the MCU than the 120Ω resistor
- other SPI pins are low for 99% of time and don't matter
So total parasitic current consumption is about 2.5mA. An additional <1mA may be contributed by the LDO straight to the power rail because its minimum load spec for regulation is 1mA. I identified the LDO as XC6214P332 (SOT89 "DR2" marking). Loading the rail with 2.2kΩ brings it down to 3.34V which is within spec of the LDO.
Per PLN2020 datasheet, 2mA is the "idle mode" power consumption and "active mode" is 15mA. Mine is obviously a different sensor (these guys say PLN2031 but no idea where they got the info). So not sure what it means but it's almost surely too high for idle and likely not enough for active
Probably some sensor fault, then. Not sure if I even want to bother with voodoo tricks like replacing all electrolytics
-
The Philips Twin-Eye sensors PLN 2031/PLN 2032 ended up unreliable, they fail somehow and were quickly pulled from the mass market. It was supposed to be the best technology- VCSEL laser Doppler interferometer, no image sensor.
Noopy speculated there was no ESD protection in the PLN 2030 die shots: https://www.eevblog.com/forum/projects/mousesensors-lots-of-pictures/ and see his website for more pics of it.
The parts are long gone, you can't get a replacement, datasheet etc. I see no hope of repairing it. I sadly tossed mine in the pile of junk, maybe send the sensor to Noopy for a look.
Before it completely died (no USB enumeration) left-right movement would also cursor move up a bit, it was diagonal. I did change some electrolytic capacitors which actually helped a lot- but this Razer stuff is so cheap and all marketing hype.
I really liked the Lachesis ergonomics, couldn't repair it so I bought $9.99 counterfeits. Well, they are a single-sided PCB so it crashes with small static charge, no quartz crystal cheap MCU so it is a bit slow and misses mouse clicks if you happen to be moving at the same time. YUCK
Philips technology Patent US 6707027 and Patent US 7889353B2 -
Well, bummer.
But I still don't know what's wrong with it
ESD? Not sure, the main chip is clearly working and responding to SPI.
Is it possible that a power supply spike (ESD?) killed some part of the chip or the laser diodes but not the SPI circuits? Dunno, maybe.
If I were sure that the sensor is never coming back I could crack it apart and see if it's possible to test the diodes with needle probes, perhaps.I really liked the Lachesis ergonomics, couldn't repair it so I bought $9.99 counterfeits. Well, they are a single-sided PCB so it crashes with small static charge, no quartz crystal cheap MCU so it is a bit slow and misses mouse clicks if you happen to be moving at the same time. YUCK
I wonder if there is some open source mouse that could be fitted into such carcass?
Imagine a small Arduino board with a rat's nest of cables going to the switches and a sensor hot glued into a PCB hole where the original used to be.
