TLC555-Q1: output abnormal for pin2, 3, 6

Hi Carols,

From my understanding, the quality team looked at the device on the bench and found no issues. However, the customer is still seeing an abnormal waveform in their application circuit. Is this correct?

I am trying to understand the waveforms, but it seems that there is some inconsistency. I see in the schematic that pin 2 is directly connected to pin 6. As these pins are directly connected, I would expect the voltage at these pins to be equal, however this is not the case in the waveforms shown. I believe this is because you probed these pins at different times and the voltage is continuing to rise over time Can you increase the time scale and monitor this voltage for a longer period of time?

Can you also update the schematic to show the values of all R and C components? I will need to know all component values in order to properly review the schematic.

Thanks,

Zach

Dear Zach

Thanks for your support! I answered you as below, if any other question, please let us know.

From my understanding, the quality team looked at the device on the bench and found no issues. However, the customer is still seeing an abnormal waveform in their application circuit.

---Yes, It was correct.

I ask our customer to provide a bigger scale for their pin 2,6 waveform, did it match your hope?

The schematic with value on it was as follow:

Hi Carols,

Thank you for the updated schematic and scope capture, this gives me a much better understanding of the fault condition.

I have reviewed the quality report and I see the device has passed both curve trace and bench test. See results from the bench test below. As expected, the TRIG pin charges up to 2/3 of the supply voltage, then discharges down to 1/3 of the supply voltage.

I see in your scope capture the TRIG pin charges up to 2/3 of the supply voltage, then discharges all the way to 0V. This TRIG voltage is also increasing very slowly, and looks like a linear increase in voltage over time instead of the expected exponential response of the RC charging circuit.

This voltage charging behavior looks like the charging current is limited to ~9.2μA. See simulation below showing both the normal RC response and the current limited response.

RC_Charging_Current_Limit.TSC

In order to better understand this, we will need to monitor the TRIG voltage, the Output voltage, and the VDD supply voltage at the same time. Can you take this measurement again using 3 scope probes to show all three signals on the same timing window?

After you capture this measurement, you can disconnect the load from the TLC555 and measure again with the load disconnected. To do this, you can simply remove the 10Ω R20 resistor.

Thanks,

Zach

Dear Zach

Good day! For your further request, we are still waiting for customer verification result, please be patient and do not close this case, thank you!

Dear Zach

Customer recheck the waveform as your order, please check the following:

ch1: VDD ch2: TRIG,ch3: OUTPUT

Before remove R20

After R20 removed:

Is it what you want? any comment please let us know.

BR

Carols

Hi Carols,

Thanks for the further detail.

From this result I can see the power supply voltage appears to be very stable at ~7V and the behavior does not appear to be related to the output loading of the 555 timer.

I can also see that the output does pulse low at the end of the charging interval, when the TRIG pin reaches 2/3 of the supply voltage. This behavior was not visible in the previous waveforms.

I am still unsure why the TRIG pin is discharging all the way to 0V, this pin should discharge to 1/3 of the supply voltage. It also stays low for a period of time before it begins charging again, which is also abnormal.

One thing I notice is that the charging time is significantly different from the previous wave capture. See below, the previous scope showed about 50ms of charging time and now we can see less than 25ms of charging time.

Can you provide any insight on why this charging time is different? Were there any changes to the test setup? What is the current limit setting on your power supply? Is the power supply approaching its current limit? If you leave the circuit on for an extended period of time, does the charge time change?

Can you provide a clear photo of your entire board and test configuration including the connections? Please also include a close image of the DUT on the PCB.

I see in your original post that an ABA test was performed. Can you describe in detail the steps of your ABA test and the results?

Thanks,

Zach

Dear Zach

Good day! I received some of the answer you want, here I share you in advance.

 Can you provide any insight on why this charging time is different? Were there any changes to the test setup? What is the current limit setting on your power supply? Is the power supply approaching its current limit? If you leave the circuit on for an extended period of time, does the charge time change?

A: Please check the following video, the charging time was always changing within 10~60ms, while the good unit was a stable charging time about 25ms.

And their setup condition has no change. Current limit was 4A while the actual current was about 60mA, very low compare to the limit.

VID_20240402_140828.zip

and now they are thinking about this was the root cause why the lamp could not be on continuously. 

Why such kind of problem will occurred and it was following IC ABA test? I need your advice.

The other question will let you know ASAP

Hi Carols,

I'm not sure I understand the following:

Carols Zhu said:

A: Please check the following video, the charging time was always changing within 10~60ms, while the good unit was a stable charging time about 25ms.

In your initial post, the good unit had a charging time of ~2ms which is consistent with the schematic you shared previously. See simulation below.

Did the schematic change since the original waveforms were shared?

What schematic was used to capture the waveform showing 25ms charge time? Was this waveform captured with a known good unit or the unit that was suspected bad?

Regards,

Zach

Dear Zach

So sorry I misunderstand customer message, the good unit charge time was 2ms, please check the actual video, no change for the schematic.

Hi Carols,

Thanks for the clarification.

It is very interesting that the charging time varies dramatically in the first video.

I am still interested in the following information.

1.) Can you provide a clear photo of your entire board and test configuration including the connections? Please also include a close image of the DUT on the PCB.

2.) I see in your original post that an ABA test was performed. Can you describe in detail the steps of your ABA test and the results?

3.) There is another test that can be performed for correlation with the quality report. See below the test configuration that was used by the quality team. Are you able to configured the suspected bad unit in the configuration shown?

Regards,

Zach

Dear Zach

We are performing  Q3 now, please still not close this issue.

BR

Carols

Dear Zach

So sorry to keep you waiting, we have the newest feedback as below:

We check the IC with the same test configuration  as below, and there was no difference between Good sample and NG sample.

But we found the same phenomenon as customer (charging time changing) when we change VDD from 15V to 7V.

You could check it in the attachment，in sheet [reference circuit(7V) ]TLC555QDRQ1_waveform.xlsx

what was your opinion about this ?

Hi Carols,

Thank you for collecting this data, your report is very clear.

I see it is the supply voltage, not the circuit configuration, that determines if the abnormal behavior occurs. I believe this is why the quality team was unable to detect the issue.

Are you able to determine at which supply voltage the operation is restored to normal? If this DUT is a failed unit, we should be able to monitor a very low IQ when the abnormal behavior is occurring and normal IQ when the supply voltage is high enough to maintain normal operation.

Are you able to monitor the device's IQ at supply voltage steps from 2V up to normal operation? For example: (2V, 3V, 4V, ..., 10V, etc). This will help us to understand the failure mode.

Thanks,

Zach

Dear Zach

Thanks for your kindly support, we want to confirm following in advance.

・Iq checking point was designed as below picture, is it OK?

・We will monitor the following waveform, any other needed?

　VDD、VTRIG、VOUT、IVDD

Best regards

Carols

Hi Carols, 

Zach is out of the office today. He will return to the office tomorrow at which time he can address the questions. 

Best Regards, 

Chris Featherstone

Hi Carols,

My apologies for the late response. This current measurement point is acceptable.

Strictly speaking, this is will measure the IQ of the device as well as the bias currents into the reset, discharge, thresh, and trig pins in addition to the charging current of the RC network. However, the bias currents are low compared to IQ and the charging current can be calibrated out as we know the voltage at the VDD and TRIG pins.

I would like the voltage at the DISCH pin to be monitored as well. If you only have 4 channels available on your oscilloscope, feel free to monitor the discharge pin instead of the VDD pin. From your previous results, VDD is expected to be constant while using standard bench equipment, so the VDD voltage can be verified on the supply itself without the need to use an oscilloscope channel. You may simply note the VDD voltage that was applied for each scope capture so that we can compare to the IQ measured.

Looking forward to your results.

Thanks,

Zach

Dear Zach

Thanks for your support, we have tried to measure waveform using your idea, but defect phenomenon dissapear.

We will show you what we did in the following, please tell us your idea or advice about this.

1. Situation when the defect disappeared.

We use the reference circuit with defect IC, change VDD from 2V and raise it by 1V every time.

When VDD=4V, we could check that defect phenomenon start occurring, but when VDD=9V, defect disappeared.

After that, we decreased VDD to 7V, but IC waveform still had no problem.

2. After defect disappeared, we try the following action, but defect still could not re-appear.

1). Push the top mold of IC

2). Heat up and cold down IC

3). Resolder to customer board to check the IC waveform.

Hi Carols,

I don't think that there is any more support over e2e that can resolve this issue.  If you are able to reproduce the problem, the please re-submit the device to the quality team, and include a reference to this e2e post.  We can help the quality team evaluate the device given the information that was given on this post.

I'm going to close this post, I'll work with the quality team for the next steps here.

Regards,
Mike