search

Severson-Group / AMDC-Hardware

Circuit board designs for controlling advanced motor systems
http://docs.amdc.dev/hardware
34 stars 3 forks source link

uInverter revB Schematic Review #198

Closed purohitparth25 closed 2 years ago

purohitparth25 commented 2 years ago

This PR outlines changes in schematic for Rev B uInverter board. Following are the changes incorporated in this revision:

  1. The polarity of sensed phase current is reverse because of the inverted connection of differential amplifier as pointed out in issue #193 . In this revision the differential amplifier's input connection is swapped to get the correct polarity of the sensed current.

  2. This revision also addresses issue #192 by adding a 0.1 uF filtering capacitor to filter out the high-frequency noise in DC link voltage measurements. An important thing to consider while placing this capacitor is its sizing. This additional capacitor along with the potential divider network of DC bus voltage sensing will form a low pass filter (see image) and this should not affect the nominal sampling operation of the DC bus voltage.

image

Bandwidth of the low pass filter in this particular configuration is given by:

image

Here, a bandwith of 1 kHz is desired from this low pass filter and for this case, the capacitor value comes out to be around 100 nF

DC-link voltage waveform (300 Hz output fundamental frequency, 3.4 A load current):

LeCroy38

As it can be seen from the above waveform, the DC-link voltage has two components. The high-frequency component is well beyond the 1 kHz cutoff frequency, however, the low-frequency ripple is around 900 Hz, which will be captured at this 1 kHz cutoff frequency.

  1. Revision A uInverter has two screw connectors, however, the user would anyways keep those shorted while using the board as mentioned in issue #197 . In this revision, the screw connectors and a cutout is made in the PCB so that the user can measure currents using a clamp (LeCroy CP030A Oscilloscope Probe).

  2. An additional 10 uF ceramic capacitor is added for high frequency noise decoupling at DC link issue #199

  3. 0603 sized packages for Capacitors: C3, C4, C5, C6 and C7 and Resistors: R1, R2, R3, R4 and R5 (In attached images) used in revision A are now replaced by 0805 sized packages for ease of board 'Bring-up'

Rev-A Schematic

Sch_Before

Rev-B Schematic (Changes highlighted in green boxes):

revB (1)

Board 3D view:

image

Other changes
elsevers commented 2 years ago

Hello Parth, it looks like the GitHub issues are not yet linked in here. Can you link these?

Screen Shot 2022-02-21 at 8 37 06 AM
elsevers commented 2 years ago

Please revise your PR description of item 2 to explain how you have selected this capacitor value (i.e., discuss the filter bandwidth).

I seem to recall you planning to do some additional experimentation regarding DC link noise. Is this still forthcoming?

purohitparth25 commented 2 years ago

I could not get the DC link current waveforms last week. I am doing that today and will update this PR once collect the data.

purohitparth25 commented 2 years ago

I have made the requested changes. I skipped the 16V Zener part since we are no longer having screw terminals. Regarding the DC link voltage ripple frequency, as expected there are two components in it. The high-frequency component is well beyond the 1 kHz BW set for the filter, however, the low-frequency component would change with the output fundamental frequency (Can be thought of as a rectified version of the AC output voltage) and may go above 1 kHz at higher output frequency (eg. 700 Hz).

purohitparth25 commented 2 years ago

I have made requested changes in the schematic and updated the image in this PR. I have also created a CHANGELOG.md in the uInverter folder. The PR is ready for review.