, 10-27-2017, 09:23 AM

Hi all,

as you know, to calculate the width of track, there is 2 rules:

the first one is based on current and the temperature like here

and the second one is based on the impedance controller for high speed signal to prevent reflection.

now the question is how should we calculate the track's width

based on frequency (skin-effect)? for example for the track carried a current with frequency of 10MHz, what is track width?

thanx

as you know, to calculate the width of track, there is 2 rules:

the first one is based on current and the temperature like here

and the second one is based on the impedance controller for high speed signal to prevent reflection.

now the question is how should we calculate the track's width

based on frequency (skin-effect)? for example for the track carried a current with frequency of 10MHz, what is track width?

thanx

robertferanec , 10-27-2017, 05:05 PM

- Impedance is usually important for signals (low currents).

- High currents are usually important for power supplies (usually you only would like to make the connections low impedance = wide tracks or using planes)

I have not really needed to design a board where high currents and high frequencies are running. What kind of circuit it is?

- High currents are usually important for power supplies (usually you only would like to make the connections low impedance = wide tracks or using planes)

I have not really needed to design a board where high currents and high frequencies are running. What kind of circuit it is?

Luca , 11-13-2017, 03:34 PM

Im not sure what Nakh0d4 are looking for, you need to calculate the minimum track width for certain amoung of current, at fixed frequency? I mean for example in CW signal?

Or you need to calculate the width for the better impedance matching for no reflection?

By the way, in both cases i know exactly all the parameters of signal and PCB material and geometry, then i use ADS from Agilent, now becomes Keysight.

I hope this can help you

Or you need to calculate the width for the better impedance matching for no reflection?

By the way, in both cases i know exactly all the parameters of signal and PCB material and geometry, then i use ADS from Agilent, now becomes Keysight.

I hope this can help you

, 11-13-2017, 03:59 PM

Thank you all,

As I know, the width of tracks are determined by two ways. First one is based on maximum DC current and second one is based on the impedance matching.

But my colleague asked me a question if we have high AC current at high frequency( the current's output of an inverter) how could we calculate the width of track. Because, we have the skin effect, so we need to consider the skind effect also.

As I know, the width of tracks are determined by two ways. First one is based on maximum DC current and second one is based on the impedance matching.

But my colleague asked me a question if we have high AC current at high frequency( the current's output of an inverter) how could we calculate the width of track. Because, we have the skin effect, so we need to consider the skind effect also.

robertferanec , 11-13-2017, 06:58 PM

Hmm, I am still not sure what is the application, but it looks like something out of my focus range I hope someone else can help.

Luca , 11-14-2017, 11:25 AM

Im pretty sure that the application is about SMPS in middle or high power application.

Tipically in this case you have to know exactly the amount of DC current and AC current that tracks have to carries.

For example 5A for DC current + 1A for AC triangular current ripple at 500kHz, and dont forget that 500kHz is the main frequency of that current but there are also the armonics to take in account...

If this is you scenario the only way to make somenthing with sense is to calculate the width for the maximum DC current and then add some percentage of safety fot both, DC and AC current.

Remember that is more important to keep a safe "back way" for the AC current then skin effect especially in high power SMPS that tipically have frequency around 50kHz to 200kHz, depends on topology and application.

Only if you have very hard design with a lot of amps at MHz the skin effect becomes predominant, and the solution is not on PCB track width

Tipically in this case you have to know exactly the amount of DC current and AC current that tracks have to carries.

For example 5A for DC current + 1A for AC triangular current ripple at 500kHz, and dont forget that 500kHz is the main frequency of that current but there are also the armonics to take in account...

If this is you scenario the only way to make somenthing with sense is to calculate the width for the maximum DC current and then add some percentage of safety fot both, DC and AC current.

Remember that is more important to keep a safe "back way" for the AC current then skin effect especially in high power SMPS that tipically have frequency around 50kHz to 200kHz, depends on topology and application.

Only if you have very hard design with a lot of amps at MHz the skin effect becomes predominant, and the solution is not on PCB track width

Use our interactive **Discord forum** to reply or ask new questions.