On Mon, Mar 10, 2014 at 01:50:01PM +0000, Laszlo Papp wrote: > On Mon, Mar 10, 2014 at 1:26 PM, Guenter Roeck <linux@xxxxxxxxxxxx> wrote: > > On 03/10/2014 02:59 AM, Laszlo Papp wrote: > >>> > >>> The reason is (most likely) that your fan input does not have a pull-up > >>> resistor. Per datasheet, the fan inputs need a 10kOhm pull-up resistor. I > >>> confirmed this with my test board - with the pull-up resistor, inputs read > >>> 0, Without pull-up, the reported value is 1, which translates to 30 RPM. You > >>> might also need the 10 uF capacitor on the FB pin. > >> > >> > >> We have both of them in place as per requirement for the chip as far > >> as I understand the datasheet. The circuit would not work properly > >> without it. > >>> > >>> sudo i2cdump -y -f 1 0x48 > >>> No size specified (using byte-data access) > >>> 0 1 2 3 4 5 6 7 8 9 a b c d e f 0123456789abcdef > >>> 00: 00 00 0a 0a ff ff 00 00 00 00 00 00 01 01 00 00 ..??........??.. > >>> 10: 01 01 01 01 1f 1f 02 02 91 91 91 91 91 91 91 91 ???????????????? > >>> 20: 00 00 0a 0a ff ff 00 00 00 00 00 00 01 01 00 00 ..??........??.. > >> > >> > >> /usr/sbin/i2cdump -y -f 1 0x48 > >> No size specified (using byte-data access) > >> 0 1 2 3 4 5 6 7 8 9 a b c d e f 0123456789abcdef > >> 00: 00 00 3a 3a fc fc b4 b4 00 00 00 00 01 01 01 01 ..::????....???? > >> 10: 09 09 01 01 1c 1c 02 02 91 91 91 91 91 91 91 91 ???????????????? > >> 20: 00 00 3a 3a fc fc b4 b4 00 00 00 00 01 01 01 01 ..::????....???? > >> 30: 09 09 01 01 1c 1c 02 02 91 91 91 91 91 91 91 91 ???????????????? > >> 40: 00 00 3a 3a fc fc b4 b4 00 00 00 00 01 01 01 01 ..::????....???? > >> 50: 08 08 01 01 1c 1c 02 02 91 91 91 91 91 91 91 91 ???????????????? > >> 60: 00 00 3a 3a fc fc b4 b4 00 00 00 00 01 01 01 01 ..::????....???? > >> 70: 08 08 01 01 1c 1c 02 02 91 91 91 91 91 91 91 91 ???????????????? > >> 80: 00 00 3a 3a fc fc b4 b4 00 00 00 00 01 01 01 01 ..::????....???? > >> 90: 08 08 01 01 1c 1c 02 02 91 91 91 91 91 91 91 91 ???????????????? > >> a0: 00 00 3a 3a fc fc b4 b4 00 00 00 00 01 01 01 01 ..::????....???? > >> b0: 08 08 01 01 1c 1c 02 02 91 91 91 91 91 91 91 91 ???????????????? > >> c0: 00 00 3a 3a fc fc b4 b4 00 00 00 00 01 01 01 01 ..::????....???? > >> d0: 08 08 01 01 1c 1c 02 02 91 91 91 91 91 91 91 91 ???????????????? > >> e0: 00 00 3a 3a fc fc b4 b4 00 00 00 00 01 01 01 01 ..::????....???? > >> f0: 08 08 01 01 1c 1c 02 02 91 91 91 91 91 91 91 91 ???????????????? > >> > >>> groeck@desktop:~$ sensors max6651-i2c-1-48 > >>> max6651-i2c-1-48 > >>> Adapter: i2c-diolan-u2c at bus 003 device 002 > >>> fan1: 30 RPM (div = 4) > >>> fan2: 0 RPM > >>> fan3: 30 RPM > >>> fan4: 30 RPM > >> > >> > >> sensors max6651-i2c-1-48 > >> max6651-i2c-1-48 > >> Adapter: DaVinci I2C adapter > >> fan1: 30 RPM (div = 4) > >> fan2: 30 RPM > >> fan3: 420 RPM > >> fan4: 30 RPM > >> > >>> Only fan2 (TACH1 chip input) has the resistor installed. > >> > >> > >> We have it both on fan1 and fan4. > >> > >> If I use our scope for checking the tachometer, I see level zero on > >> the scope, but 1 when checking the open collector level, since that is > >> the purpose of the pull-up resistor. > >> > > With the pull-up installed, you should always see a high voltage on the > > scope or with a volt meter. The scope would not see or be able to know > > that the pin is OC if the pull-up resistor is there. > > > > If it is truly OC, it would be floating and might show low. If you can, > > I would suggest to manually add a 10k resistor against VCC to the open > > connector and see what happens. > > > > Another way to find out for sure might be to disconnect power and > > measure resistance between the tachometer pin and VCC. > > Perhaps I was not clear. So, we do have both the resistors and > capacitor as required. I double checked the schematics as well as the > pins coming from the fan and that was zero. The tacho0 input, however, > was 1. > How can the pin coming from the fan be zero, and the tach0 input be 1 at the same time ? Those should be the same ? Anyway, I don't think I can help you further. All I can say is that the chip operates as expected on my test board. Guenter _______________________________________________ lm-sensors mailing list lm-sensors@xxxxxxxxxxxxxx http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
