Max current #2
Comments
|
It needs 5v supply to have the +/-5A range with 400mv/A sensitivity but in
this design we run it on 3.3V only. But anyway, +/-2.5A peak is more than
needed for most 3D printers.
BTW, there is a better current sensor with the same pinout but it's very
hard to find since the pandemic, especially non fake ones (I did get fakes
once from Aliexpress, it's easy to identify since pins 1,2 and 3,4
respectively should be shorted internally). The ACS70331EESATR-2P5B3 has
higher bandwidth, lower noise, and lower output impedance. The output
impedance is important because the ESP32 ADC has relatively low output
impedance and loads the CC6920BSO-5A such that we don't really get 400mv/A.
This is why we calibrated it using actual measurements with ESP32 and not
based on the datasheet.
https://github.com/zapta/ble_stepper_motor_analyzer/blob/main/platformio/src/acquisition/acq_consts.h#L11
The two configuration resistors R17, R18 were designed to support 4
different kinds of sensors, to let the firmware know which one is
installed, but CC6920BSO-5A is the only one we used in this design because
it's available on JLCPCB.
Zapta
…On Sat, Jun 10, 2023 at 5:03 AM dzid26 ***@***.***> wrote:
If the design uses CC6920BSO-5A, shouldn't it be -/+5A range? The readme
says +/-2.5A per phase
—
Reply to this email directly, view it on GitHub
<#2>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAQVMQLEMB7F2EG64LQDRPLXKRPAJANCNFSM6AAAAAAZBUJ2TU>
.
You are receiving this because you are subscribed to this thread.Message
ID: ***@***.***>
|
Are you saying that the real sensitivity of CC6920BSO-5A is 400mV/A ? |
|
Yes, you are right, I may mixed it with another sensor. BTW, I see that digikey carries now some Allegro GMR sensors, they have higher bandwidth and significantly lower noise. Ideally the +/-2.5A with 400mv/A, but the +/-5A with 200mv/A can still be more quieter than the one from JLCPCB. file:///C:/Users/user/Downloads/ACS70331-Datasheet.pdf (Having all the components available from JLCPCB was a key design goal here, hence the selection of the inferior current sensor. In previous designs which I soldered myself I used the Alegro ones, e.g. https://github.com/zapta/simple_stepper_motor_analyzer/blob/main/kicad/stepper_analyzer.pdf ) |
|
Today I designed a voltage-measuring circuit that filters scales down, and offsets the phase voltage. It is using a differential opamp configuration - so the ground doesn't have to be connected. I don't know how noisy it will be and what can be done with it, but power calculation should be possible. But now I am thinking - it would be interesting to use the same structure (but with different gain and less filtering) to also measure current across 0.1ohm shunt resistors. I think I will put it in parallel with CC6920BSO to compare the noise of both solutions. I wonder if the noise you mention is also related to flawed ADC from ESP32 and additionally lack of filtering of analog VDD3p3 in Wroom module. Would be cool to replace it with a custom design, but I have enough of Kicad. |
|
I believe that the noise is dominated by the JLCPCB current sensor because when I hooked the ESP32 to a ACS70331 +/-2.5A the noise reduction was very visible. As for your circuit, one thing that the magnetic current sensors provide is galvanic isolation which the opamp circuit doesn't provide, so make sure to get the connections and voltages right so you don't damage anything or experiment with an ESP32 that runs on a battery. There are also isolated amplifiers if you want to get fancy ;-). Never used them but they seem interesting. E.g. https://www.ti.com/product/AMC3311-Q1 |
If the design uses CC6920BSO-5A, shouldn't the range be -/+5A?
The readme says +/-2.5A per phase
The text was updated successfully, but these errors were encountered: