Skip to main content
Added related question
Source Link
Tryphon
  • 119
  • 1
  • 8

I have the following board design in mind:

enter image description here

It's a 4 layer, ~5cm diameter board where 10 slave PCB's will be soldered to it that are connected over the I2C bus. The trace line you see is roughly 8 cm in length. This bus has the clock going over the top layer and the data bus over the bottom layer, but I could always change that.

I could also route the traces over 1 layer but use via's to cross each other at the end points, if that makes sense.

As far as I can tell I'm not doing something that is against any I2C "rules" bus I'm also not that familiar with this sort of setup.

The end-point devices will add roughly 150-200 pF total load capacitance to the bus, excluding trace, solder joint and via capacitance. Which as far as I could tell is not a problem since it's far lower than 400 pF already. The bus will be running on 400 kHz.

What method of routing would be recommended here? Am I accidentally designing a radio receiver or worse, a transmitter?

edit: I forgot a case where both are routed on the top layer; Should I match the amount of vias used on both lines even if those vias aren't strictly needed?

I have the following board design in mind:

enter image description here

It's a 4 layer, ~5cm diameter board where 10 slave PCB's will be soldered to it that are connected over the I2C bus. The trace line you see is roughly 8 cm in length. This bus has the clock going over the top layer and the data bus over the bottom layer, but I could always change that.

I could also route the traces over 1 layer but use via's to cross each other at the end points, if that makes sense.

As far as I can tell I'm not doing something that is against any I2C "rules" bus I'm also not that familiar with this sort of setup.

The end-point devices will add roughly 150-200 pF total load capacitance to the bus, excluding trace, solder joint and via capacitance. Which as far as I could tell is not a problem since it's far lower than 400 pF already. The bus will be running on 400 kHz.

What method of routing would be recommended here? Am I accidentally designing a radio receiver or worse, a transmitter?

I have the following board design in mind:

enter image description here

It's a 4 layer, ~5cm diameter board where 10 slave PCB's will be soldered to it that are connected over the I2C bus. The trace line you see is roughly 8 cm in length. This bus has the clock going over the top layer and the data bus over the bottom layer, but I could always change that.

I could also route the traces over 1 layer but use via's to cross each other at the end points, if that makes sense.

As far as I can tell I'm not doing something that is against any I2C "rules" bus I'm also not that familiar with this sort of setup.

The end-point devices will add roughly 150-200 pF total load capacitance to the bus, excluding trace, solder joint and via capacitance. Which as far as I could tell is not a problem since it's far lower than 400 pF already. The bus will be running on 400 kHz.

What method of routing would be recommended here? Am I accidentally designing a radio receiver or worse, a transmitter?

edit: I forgot a case where both are routed on the top layer; Should I match the amount of vias used on both lines even if those vias aren't strictly needed?

added 1 character in body; edited title
Source Link
Michel Keijzers
  • 14.1k
  • 20
  • 80
  • 150

i2c I2C bus line over different layers in an arc

I have the following board design in mind:

enter image description here

It's a 4 layer  , ~5cm diameter board where 10 slave PCB's will be soldered to it that are connected over the i2cI2C bus. The trace line you see is roughly 8 cm in length. This bus has the clock going over the top layer and the data bus over the bottom layer, but I could always change that.

I could also route the traces over 1 layer but use via's to cross each other at the end points, if that makes sense.

As far as I can tell I'm not doing something that is against any i2cI2C "rules" bus I'm also not that familiar with this sort of setup.

The end-point devices will add roughly 150-200 pF total load capacitance to the bus, excluding trace, solder joint and via capacitance. Which as far as I could tell is not a problem since it's far lower than 400pF400 pF already. The bus will be running on 400kHz400 kHz.

What method of routing would be recommended here? Am I accidentally designing a radio receiver or worse, a transmitter?

i2c bus line over different layers in an arc

I have the following board design in mind:

enter image description here

It's a 4 layer  , ~5cm diameter board where 10 slave PCB's will be soldered to it that are connected over the i2c bus. The trace line you see is roughly 8 cm in length. This bus has the clock going over the top layer and the data bus over the bottom layer, but I could always change that.

I could also route the traces over 1 layer but use via's to cross each other at the end points, if that makes sense.

As far as I can tell I'm not doing something that is against any i2c "rules" bus I'm also not that familiar with this sort of setup.

The end-point devices will add roughly 150-200 pF total load capacitance to the bus, excluding trace, solder joint and via capacitance. Which as far as I could tell is not a problem since it's far lower than 400pF already. The bus will be running on 400kHz.

What method of routing would be recommended here? Am I accidentally designing a radio receiver or worse, a transmitter?

I2C bus line over different layers in an arc

I have the following board design in mind:

enter image description here

It's a 4 layer, ~5cm diameter board where 10 slave PCB's will be soldered to it that are connected over the I2C bus. The trace line you see is roughly 8 cm in length. This bus has the clock going over the top layer and the data bus over the bottom layer, but I could always change that.

I could also route the traces over 1 layer but use via's to cross each other at the end points, if that makes sense.

As far as I can tell I'm not doing something that is against any I2C "rules" bus I'm also not that familiar with this sort of setup.

The end-point devices will add roughly 150-200 pF total load capacitance to the bus, excluding trace, solder joint and via capacitance. Which as far as I could tell is not a problem since it's far lower than 400 pF already. The bus will be running on 400 kHz.

What method of routing would be recommended here? Am I accidentally designing a radio receiver or worse, a transmitter?

Source Link
Tryphon
  • 119
  • 1
  • 8

i2c bus line over different layers in an arc

I have the following board design in mind:

enter image description here

It's a 4 layer , ~5cm diameter board where 10 slave PCB's will be soldered to it that are connected over the i2c bus. The trace line you see is roughly 8 cm in length. This bus has the clock going over the top layer and the data bus over the bottom layer, but I could always change that.

I could also route the traces over 1 layer but use via's to cross each other at the end points, if that makes sense.

As far as I can tell I'm not doing something that is against any i2c "rules" bus I'm also not that familiar with this sort of setup.

The end-point devices will add roughly 150-200 pF total load capacitance to the bus, excluding trace, solder joint and via capacitance. Which as far as I could tell is not a problem since it's far lower than 400pF already. The bus will be running on 400kHz.

What method of routing would be recommended here? Am I accidentally designing a radio receiver or worse, a transmitter?