SM72442: SM72442

Hello Pietro,
Thanks for reaching out to us via e2e.

You wrote:
> By default, via HW the current limit is set up to 4A
Can you please explain what exactly you mean?
I cannot find this information anywhere in the datasheet.

Are you sure that the I2C communication works properly?

Can you execute the following sequence:

- Read register 4 and write down the original content
- Modify a bit of the register content that you have read and write this modified content to register 4
- Read register 4 again and verify if your modified content can be read from the register
- Write back the original content that you had noted in the beginning
- Read the register again and verify if the content has changed again

Best regards,
Harry

Hello Harry ,

thank you for the replay, here you can see the answers to your question:

1) from the datasheet at page 8 there is the section

"CURRENT LIMIT SETTING Maximum output current can be set by changing the resistor divider on A4 (pin 18). Refer to Figure 2. Overcurrent at the output is detected when the voltage on AIOUT (pin 21) equals the voltage on A4 (pin 18). The voltage on A4 can be set by a resistor divider connected to 5V whereas the voltage on AIOUT can be set by a current sense amplifier. "

My design has on A4 pin a voltage defined via HW by a voltage divider to 2.67V , that it is equivalent to a setting of 4A of max current out, cause the Iout sense network is made by :

- Rsense (30mohm) ,  INA193  (gain 20V/V, Vcc= 5.1V) .

2) Via I2C registers 3 and 4 at power-on are in the following configuration:

- SACU 0: MPPT 0 configuration:
Reg 0: 0x000072A19E1B16
Reg 3: 0x0008FFFFF6DFE0
Reg 4: 0x00000000000000
Reg 5: 0x00000A0180A018

after sending the following command eps sacu wr_conf 0.0 3 0x00 0x48 0x06 0x2C 0x96 0xDF 0xE0 ( overwrite of register 3 for current setting):

- SACU 0: MPPT 0 configuration:
Reg 0: 0x000072A19E1B16
Reg 3: 0x0048062C96DFE0
Reg 4: 0x00000000000000
Reg 5: 0x00000A0180A018

-  after send the following command eps sacu wr_conf 0.0 4 0x00 0x00 0x00 0x10 0x00 0x00 0x00 ( overwrite of register 4 for offset setting):

SACU 0: MPPT 0 configuration:
Reg 0: 0x000072A19E1B16
Reg 3: 0x0048062C96DFE0
Reg 4: 0x00000010000000
Reg 5: 0x00000A0180A018

- restoring the original value of register 3:

SACU 0: MPPT 0 CFG reg 3 updated
eps_pt_X # eps sacu rd_conf 0.0
SACU 0: MPPT 0 configuration:
Reg 0: 0x000072A19E1B16
Reg 3: 0x0048FFFFF6DFE0
Reg 4: 0x00000010000000
Reg 5: 0x00000A0180A018

 eps sacu wr_conf 0.0 3 0x00 0x08 0xFF 0xFF 0xF6 0xDF 0xE0  ( command with the overwrite bit at 0 )

SACU 0: MPPT 0 CFG reg 3 updated
 eps sacu rd_conf 0.0
SACU 0: MPPT 0 configuration:
Reg 0: 0x000072A19E1B16
Reg 3: 0x0008FFFFF6DFE0
Reg 4: 0x00000010000000
Reg 5: 0x00000A0180A018

- restoring the original value of register 4:

eps sacu wr_conf 0.0 4 0x00 0x00 0x00 0x00 0x00 0x00 0x00

SACU 0: MPPT 0 CFG reg 4 updated

SACU 0: MPPT 0 configuration:
Reg 0: 0x000072A19E1B16
Reg 3: 0x0008FFFFF6DFE0
Reg 4: 0x00000000000000
Reg 5: 0x00000A0180A018

as you can see the I2C communication seems to be correct.

In both cases, the current limit function doesn't work.

Thank you for the support

Pietro

Hi Pietro,

can you measure the voltage at A4 an AIOUT at a current level of 4A and also above e.g. 4.5A.

Best regards,

Stefan

HI Stefan,

here you che find the measurement of A4 (yellow trace) and AIOUT ( green trace) at :

- 1 A Load

- 4 A load

- 4.5A load

- 4.6 A load

thank you for the support,

Pietro 

Hi Pietro,

It looks strange for me that the signal at A4 which should come from a resistor divider looks so noise.

If this signal is really so, it might interfere with the logic.

Can you check that again?

Why is the signal so noise?

What happens if you add a capacitor to filter?

Best regards,

 Stefan

Hi Stefan,

the resistor divider on A4 is filtered with 100nF.  I repeated the test changing the filter with 1uF here the trace that I take: (A4 yellow trace, AIOUT Green trace) I take the trace at 10us time division.

-  Load 2A 

-  Load 3A

-Load 4A

-Load 4.5A

The noise on the trace I think is due the PWM switching and it reduces increasing the current load.  ( the reduce of the noise will be an improvement on the next PCB release).

My question is: What should I expect when the current reaches the threshold?  Switch off or change regulation mode, or limitation and decrease of the output voltage?

Thank you 

Pietro 

Hi Pietro,

if the current limit will be reached the output voltage will be reduced to not hit the current limit level.

Best regards,

 Stefan

Hi Stefan,

thank you, If I understood well when the limit is reached the behaviour of the dcdc is CC mode where the max current is the current limit, is right? This behaviour is different from what is written in the datasheet. 

Can you suggest me to how to trigger this limit, to understand if is a design problem or not? From my images, I see that the board is not limiting.

I'm testing the board with the following setup:

- DC power supply 60V , 5A

- Electronic load in CC mode.  

- MPPT design 50.4V vout, 4A Iout max.

Can you suggest me a correct setup for the current limit test?

Best regards

Pietro

Hello Pietro,

Thanks for your feedback.

Stefan is already in the weekend.
He will look into it on Monday.

Best regards
Harry

Hi Pietro,

Just checked the datasheet again. AIIN is over current and not current limit.
So the controller should stop and not go to CC mode.

I would need to double check this on an EVM but will need sometime.

Best regards,

 Stefan

Hi Pietro,

have you noticed that this device has been set to last time buy.

Best regards,

Stefan

Hi Stefan,

yes, I noticed that, and my protoboard has 2 different MPPTs SM72442 and SM72445. the last waveform was taken on the section with the SM72445. 

Hoping that SM72445 will be available.

Pietro

Hi Pietro

unfortunately I do not have an EVM available to check in on my side.

I also tried to get more information on how the function is implemented internally and what are the behavior in this condition but could not get this so far.

Best regards,

 Stefan

Hi Stefan

thank you for the support, I have an SM3320-BATT-EV RevB and reading the documentation I understand that the current limit has been done by HW with a dedicated circuit, and the A4 pin has been set at max current.

Pietro

Hi Haroon ,

honestly I think the tread has not been resolved because I think a solution has not been found, as the component is not behaving as per the datasheet.

BR

Pietro

Hi Pietro,

In that case could you tell me what is the latest issue you are facing?

BR,

Haroon

Hi Haroon,

this is the lastest issue:

"Hi Pietro

unfortunately I do not have an EVM available to check in on my side.

I also tried to get more information on how the function is implemented internally and what are the behavior in this condition but could not get this so far.

Best regards,

 Stefan"

BR 

Pietro 

Hi Piero,

The issue is that we do not have EVM for this part, I will try to get some information internally as well like Stefan but hopefully I will be able to get it. I also looped in one of my more experienced colleagues to help me out on this.

Thank you for your patience.

BR,

Haroon

Hi Haroon,

thank you,  I'm waiting for your feedback, If you want you can close the topic.

BR

Pietro

Hi Pietro,

I will update you sometime next week regarding this, I am sorry my bandwidth at the moment is very short at the moment.

Thank you for your patience.

BR,

Haroon

Hi Pietro,

I think that board does use this device, so what was the conclusion after testing and working with this board?

BR,

Haroon

Hi Haroon,

 from the tests I have done on my prototype boards with SM72442 and SM72445 the current protection configured on pin A2 and via I2C is not working. From the SM3320-BATT-EV evaluation board using the SM72442 , the current limiting or protection is done with an external comparator. To find a solution to my design do you suggest using the same philosophy? Or are there other tests or configurations I can perform on my prototype?
Thank you

 Pietro

Hi Pietro,

I cannot suggest a solution as I do not have the evaluation board you mentioned, you could try and see if the same "philosophy" works for your design, but I do not have enough data to suggest whether that method would help your cause or not.

BR,

Haroon