Page 1 of 1
RP2350 devices: 3V3 max current?
Posted: Thu Sep 19, 2024 10:49 am
by cp_addict
Hi,
what is the current limit that the 3V3 output on the RP2350-xxx boards supply? I can see from the schematic that the regulator itself is rated up to 1.3A, but I assume that there are limitations since this has to power the RP2350 (and C6) as well.
Bernhard
Re: RP2350 devices: 3V3 max current?
Posted: Fri Sep 20, 2024 4:12 pm
by Pontus
Hello Bernhard,
The power supply have been designed to be able to source up to 1A with reasonable ripple. This was estimated from the requirements from Raspberry Pi for the RP2350 (650mA) and the ESP32-C6 (350mA) combined. In theory that would leave you without any budget for any external peripherals. But while the Raspberry Pi recommended a very powerful power supply it generally does not consume that much power at all. We have been able to run LTE modems and NFC controllers on the same supply and I would estimate that you have a budget of around 400mA for external devices.
We'll experiment a bit and update the datasheet with our findings.
/Pontus
Re: RP2350 devices: 3V3 max current?
Posted: Fri Sep 20, 2024 7:34 pm
by cp_addict
Hi Pontus,
thanks! I will do some measurements too with RP2350 and C6 active. These 650mA seem a bit high, I have never seen something higher than 150mA on a Pico-W with active WLAN. With one exception: I saw a very short spike of 1A when turning the RP2040 on, but I assume that is only because the PPK2 picks up extremely short (rush-in) spikes with it's 100K samples/s.
Bernhard
Re: RP2350 devices: 3V3 max current?
Posted: Mon Sep 23, 2024 4:01 pm
by cp_addict
Hi Pontus,
Here is a measurement with UDP-transfers:
- challenger_udp_transfer.png (96.82 KiB) Viewed 504 times
Start and setup is t=0 to t=10. Then there are 15 seconds sleep, with the LED on in the middle of that period. The current of the LED is so low that you cannot spot that period. At about t=25 the loop-processing starts: 10s of UDP transfers (max speed), followed by 10s sleep.
So we can see about 28mA@5V during sleep (which is implemented as a busy loop in the RPxxxx-SDK). During active transfers, we have a current draw of about 90-95mA@5V. Voltage is fine with at most 1.5% voltage drop under load.
The same program running with a Pico-W (RP2040):
- picow_udp_transfer.png (75.31 KiB) Viewed 504 times
Observations: setup time is 5s faster (expected). There is much less noise. Idle current is 60mA@5V, current during transfers is 75mA@5V.
So nothing to worry about. I will do some additional measurements with the RP2350 BConnect, just to see if the noise is due to the C6 or not.
Bernhard