P82B715: Pull-up voltage for I2C- P82B715

Hey Arunkumar,

"In datasheet it is mentioned that Voltage logic levels are independent of VCC and also the IC has to sink 30mA."

-Looking at the datasheet, it implies that the Lx/Sx pins can be Vcc+0.7V or lower and work.

-The IC does not HAVE to sink 30mA, what the datasheet is saying here is it is capable of sinking 30mA on the "L" side of the device.

"So can I use pull-up voltage of 1.8V for Sx/Lx signals with VCC of 5V and calculate pull-up resitance values to provide sink current value of maaximum 30mA. Also I calculate RC value to provide 1uS rise and fall time."

-You do not actually have to sink 30mA. A 1uS fall time violates I2C standard as a maximum time allowed is 300ns, I assume you are actually just talking about rise time in this case.

A. What pull up values and capacitance are you expecting?

B. How far are you planning on communicating? (distance of cables)

C. What is your expected max frequency?

"If I do so, there will be still any problem in sinking current or timings? or else any other problem will arise??"

Can you show me a schematic/diagram with the pull up resistors for both boards and expected capacitance? To be honest, I have no had much time with this device in the lab so I may need to put it on a bench and do some testing if you provide the details.

Thanks,

-Bobby

Hi Bobby,

Assume I need to have bus capacitance of 3000nF( maximum capacitance that can be driven by the IC) and assume distance to be 20meters(single mater and single slave).

Let Speed be 400KHZ as per I2C standard.

I need to put pull the bus lines at 1.8V.

Following flow diagram will shows outline of my circuit.

Please let me know that this circuit works and what pull-up value to be calculated.

Thanks & Regards,

M.Arun kumar

Hey Arun,

The length of the wire was very short in this test (I wanted to confirm that the device would work at Vcc=5V while I2C @ 1.8V). I did try to put a 3000pF capacitor on the L line but I had a difficult time finding a capacitor that would fit on my set up. I ended up trying to solder 3x 1000pF 1210 caps together in parallel and tying wires to it but I'm not sure if the wire's made a good enough connection to my set up. When I added and removed the 3000pF load from the set up, I did not see much variation in the L line (I did see the L line change slightly with the cap added but not by much). This is likely due to the VERY strong pull ups on the L line.

-Bobby