Hi folks,
Does anyone have any experience reducing the quiescent draw of the MicroMod ESP32 Processor? I’m new to ESP32 and when using the MicroMod Data Logging Carrier Board and the ESP32 > DeepSleep > TimerWakeUp example, I’m getting 7.2 mA when in deep sleep. Obviously, I must be missing something!
Cheers,
Adam
I tried it using the same example and it was using ~18mA in the deep sleep cycle initially, I let it run/cycle in and out of sleep for 2 or 3 minutes and now it’s reading 2-3mA while sleeping…try letting it run for a bit and see?
Hi Russel,
Thanks for the reminder.
What we’re seeing is the charging of the ML414H battery on the MicroMod Data Logging Carrier Board (MMDLCB). I’ll let it fully charge and then measure later on. The MMDLCB will be responsible for some of the quiescent draw, as it isn’t very power efficient, but I’ve characterized this draw and can simply subtract it.
Cheers,
Adam
Hi Russel,
To follow up on this issue, still no luck reducing the quiescent draw of the MicroMod ESP32 Processor.
I’ve confirmed with the MMDLCB and an Artemis Processor that I can get down to 400 uA in deep sleep. This is about as good as you can get with the Artemis + MMDLCB without removing the voltage divider resistors.
However, with the MMDLCB and ESP32 Processor, I still can’t get lower than 7.2 mA.
Are you using the same ESP32 > DeepSleep > TimerWakeUp example?
Also, very curious where the 500 uA value came from on the product page.
Cheers,
Adam
I spent the better part of the day yesterday trying to reduce the current draw of the ESP32 with no success. I tried just about every combination of GPIO pin state (e.g., INPUT_PULLUP, OUTPUT, HIGH/LOW). Unfortunately, I’m at a loss stuck at 7.2 mA in deep sleep.
I suspect the circuity onboard the ESP32 may be responsible for some major current draws. Russel, is there any chance you can inquire to Alex or the engineering team? The product page states a 500 uA hibernation current draw, which isn’t actually very good but is sure a heck of a lot better than 7.2 mA.
Cheers,
Adam
adam.g:
I suspect the circuity onboard the ESP32 may be responsible for some major current draws.
This assumption is correct. Unfortunately the CP210X USB to serial chip on the ESP32 module is connected to the 3.3V rail and not bus powered from 5V. The suspend pin is not utilized either which could reduce the quiescent current but is rather useless since the chip only needs to be active with a USB connection. A few traces need to be cut along with some jumper wires added to remove the VREGIN pin from 3.3V and add to V_USB power. It was relatively straight forward to do with a microscope and some 32 AWG wire. The bus powered configuration is shown in the typical connection diagram on page 18 of the CP210X datasheet (https://www.silabs.com/documents/public … 2102-9.pdf). Afterwards my deep sleep 3.3V current dropped from 3mA to under 10uA.
chEEm:
adam.g:
I suspect the circuity onboard the ESP32 may be responsible for some major current draws.This assumption is correct. Unfortunately the CP210X USB to serial chip on the ESP32 module is connected to the 3.3V rail and not bus powered from 5V. The suspend pin is not utilized either which could reduce the quiescent current but is rather useless since the chip only needs to be active with a USB connection. A few traces need to be cut along with some jumper wires added to remove the VREGIN pin from 3.3V and add to V_USB power. It was relatively straight forward to do with a microscope and some 32 AWG wire. The bus powered configuration is shown in the typical connection diagram on page 18 of the CP210X datasheet (https://www.silabs.com/documents/public … 2102-9.pdf). Afterwards my deep sleep 3.3V current dropped from 3mA to under 10uA.
I appreciate the explanation, @chEEm!
It’s a bit confusing the product page states “500µA sleep current under hibernation”. From what you described, this is not actually possible with the current design of the processor.
Cheers,
Adam
adam.g:
“500µA sleep current under hibernation”. From what you described, this is not actually possible with the current design of the processor.
Hibernation isn’t a sleep state of the ESP32 … Sparkfun is in the business of selling products, not correct specs or good engineering practice (sorry @sparkfun, just calling it like I see it). Here’s the mod for bus power on the USB to Serial chip:
Thanks @chEEm,
It’s unfortunate, but sometimes true. The MicroMod Artemis Processor [also has a number of issues that I’ve had to learn to live with. One of these days, I’ll get around to designing my own boards.
Thanks for the confirmation that I shouldn’t spend any more time trying to troubleshoot the ESP32 Processor.
Cheers,
Adam](Issues · sparkfun/MicroMod_Artemis_Processor · GitHub)
