TPS2044A: TUSB2046_TPS2044A : abnormal operation

Hi Sun Young,

/ENX is a input pin. I don't see the cause of the behavior.

Can you please share more details about your test condition and your schematic?

I already attached  our schematic and figures.  Refer to our schematic.  

/ENx is a input pin  of TPS2044. And  /ENx is a out pin of TUSB2046. 

Dear Bing,

Thank you for your support.

(Sun Young is my customer.)

the biggest problem is that if they make overcurrent situation over about 3 seconds, TPS2044A cannot recover to normal status.

(You can see the waveform on 2nd slide of pptx file.)

Furthermore If it is connected to TPS2046B, we even cannot see the change of \OCx pin. it is just high status although overcurrent situation happens.

(but we can see the change of \OCx pin if we test TPS2044A alone, so we suspect that there are some problems with the combination of TPS2044A and TPS2046B.)

anyway we have to solve the issue about cannot recover situation. do you have any idea to solve this problem?

and please let me know if you need more information about it. Thanks.

Best regards,

Chase

Hi Sun Yeong, Chase,

TUSB2046 is not the product of our line, /EN signal is controlled by it. I assume the control logic is that when /OC of TPS2044 is asserted, TUSB2046 will set /EN to disable the switch, is that right? Can you zoom in the waveform when the /OC pulse occurs? For the second figure, I see pulse on /EN pin.

Hello Bing Hwang, 

We have attached  the zoom figures of /OC. Pls check the waveform. 

Hi Sun Young,

Thank you for your update. I think the waveform you attached is also with /EN device not EN device right? But the waveform does not keep consistent with the switch you use, seems that when /OC occurs, /EN also asserts low (enable).

By the way, can you please share more details about control logic of TUSB2046? Cause the enable signal is from TUSB2046, I want to know more about it.

Dear Bing,

Thank you for your support.

first we want to know that the schematic has a problem or not. it is very important to my customer. please review the schematic below.

3757.TUSB2046_TPS2044A_OCS ISSUE_schematic.pdf

second, Sun Young tested TPS2044A alone, using transistor IC between \OCx and \ENx. (if \OCx is low, \ENx would be high.)

the result was the same just like the waveform right above, we think that it is not a problem due to TUSB2046.

so could you check the reference schematic including any pull-ups, pull-downs or any other additional circuit if we need?

and please provide the data which you tested TPS2044A yourself.

please let me know if you need more information about it. Thanks.

Best regards,

Chase

Hi Chase, Sun Young,

The schematic looks good, but we may need a bulk capacitor on output pins. I will do the test after I get TPS2044A.

But before that I want to based on your test results do the debug, can you please help clarify below questions?

1. For the previous waveform, CH1 refers to which /OC and CH4 refers to which /EN? TPS2044A has 4 /OC, 4 /EN pins. On which power path did you test?

2. 

Chase Jeong said:

Sun Young tested TPS2044A alone, using transistor IC between \OCx and \ENx. (if \OCx is low, \ENx would be high.)

For "tested TPS2044A alone", it means that we tested it on board but cut off all connection to TUSB2046?

3. How did you do the test? By using a 2Ohms resistor to connect between TPS2044A's output and GND directly? And before connect 2Ohms load, has there existed a load such as 0.5A , or we do the 2Ohms load test with no load before (except for the 10k Ohms and Diode shown on schematic)?

4. Did we monitor the status of VIN? Did it keep at about 5V? -> I would suspect the sudden increment of the output current will pull low the voltage level on VIN in short time, and then the voltage level for TPS2044A INx cannot have it work normally.

Thanks.

Bing

Hello Bing Hwang, 

We have tested  all /OC and /EN.  I think  certain path  don't produce this problem. 

But we have tested  many times 3/OC, 3/EN. 

Thanks. 

BR,

HAM

Dear Bing,

I add more answers for your question.

2. Yes, they cut off all connections to TUSB2046.

3. they used 2-ohm load and also used electric loader. and Yes, they connected 2-ohm load between output and GND. current consumption is small in normal operation.

4. we have to check it, but I think it is not such a large current that causes VIN voltage drop.

Best regards,

Chase

Hi Sun Young,

Thanks. So you mean each path may have the behavior that you observed? And what is the fail rate, how many fails in all tests you have? (Fail cases refer to those that /EN cannot auto-recover).

Thanks.

Bing

Hi Bing Hwang, 

The fail rate of this issue is 100%. 

Thanks. 

BR,

HAM

Dear Bing Hwang,

We  have already answered  to your question.  Pls refer to answer of Chase Jung. 

Hi Ham,

I will get samples in this week, will feedback to you..

Thanks.

Bing