10-Ohm Resistors in Current-Sense Circuit

Hi mjsadaway ,

As the R1 and R2 are quite far from the MAX20050, the R3, R4 together with C18 add some low-pass filtering to the current sense signal.

Cheers, heke

Thanks, Heke.

That sounds reasonable, but a low-pass filter with R3, R4, and C18 would have a cutoff frequency of 79.6 MHz, which doesn't sound reasonable.

Mike

Hi Mike,

Yeah, that's above the operational frequency range of the current sense amp, but it does not mean that higher frequencies cannot enter to the chip and by-pass the amp. For some reason the eval board designer has decided to add the filter. Maybe he/she lives near an FM station and saw the FLT pin toggling. It is perhaps safe to leave them out if you can place the sense resistor(s) closer to the chip. If need to use the filter, I'd put the R3, R4 close to the chip instead of being close to the sense resistor(s).

Cheers, heke

Your FM radio theory somehow shook loose in my cluttered brain the fact that 79.6 MHz is within the FM broadcast band in Japan. Don't think Maxim has a facility there. Silly thoughts.

It wouldn't cost much to retain the filter. Board space is limited, but the chip doesn't need a huge number of other external components.

It's always worrisome when an eval board includes components not shown in the suggested application circuits in the chip's datasheet.

Although looking just now, Figure 10. "Section from MAX20051 EV Kit PCB Layout" in the datasheet shows specific locations for the filter components and recommends routing the current-sense lines on an inner layer.

I might not have noticed this without having this discussion.

Thanks yet again for your help.

Mike

Hi Mike,

mjsadaway said:

It's always worrisome when an eval board includes components not shown in the suggested application circuits in the chip's datasheet.

That is a good point, and would say that the discrepancy is quite common among all chip vendors. Product managers should pay attention to that.

The shown filtering may be somewhat related to a thing Chris Mangelsdorf calls "Bandini Mountain"*, i.e. all chips have parasitic resonances making them sensitive to certain dV/dt's, mystically usually around 100MHz.

Cheers, heke

* IEEE Xplore: Shop talk: What you didn't learn in school / VDD is NOT your BFF: