SN74AUC125 behavior when short duration pulse on OE

I can not tell wat the voltage levels are. What is Vcc and what are the input levels?

Pulse width and Fmax are not usually specd on parts without a clock. You can look at a clocked part in the same family to get and Idea of these limits.

Hello, Chris

Thank you for your response.

The part is used at Vcc=2.5V and the input signal level is the same.
The short duration pulse on OE is approx. 4.5nsec.

I do not understand why the incorrect output is generated (output=High
even though OE=Low and input=Low) if the timing is not a point of concern.   

Could you advise what can be possible cause of this issue ?

Thank you in advance.
Best regards,

RY

The plot does not make sense. Even when th output toggles, it toggles about 50ns after the input. This part should have a propagation delay of less than 2ns.

Can you tell me the the markings on the device?.

Hello Chris,

I will reply on behalf of RY.

Thank you for pointing out about the large propagation delay.

I confirmed about the large delay to the customer.

In actual, the output was not at Y pin of the SN74AUC125.

Input  -->  SN74AUC125 --> RS422 driver --> RS422 receiver

The output signal in the waveform is captured at RS422 receiver.

(The customer could not capture waveform at Y pin)

This is the reason of the large delay.

Best regards,

Rak

Hi,

I got additional information from the customer.

Please find the attachment.

As shown in the attachment, there are voltage-dividing resistors at OE.

Originally, with 2.2KOhm, the customer observes the output error.

But, changing 2.2KOhm resistor to 1.0KOhm, the error does not happen.

The peak of the short duration pulse on OE becomes lower by 0.23V

when using 1.0KOhm.

From this test result, is there anything you can suppose as the cause

of the "no output" failure ?

Thank you in advance.

Best regards,

Rak

I am not sure what is suppose to happen when the pulse goes High but when the OE is low it is active. lowering the ouput level of the pulse probably keeps the input below the threshold.

Here is what I think is happening. With the 2.2K the OE level is just above the threshold sensing a high putting the part in high impedance. The level is not a good high though so the part could oscillate for a while keeping the part disabled. When they use the 1K the OE senses the input as a low and the part continues to operate properly.

However both the input levels into the OE are not valid highs or lows. They violate vih and vil. The part is not guarranteed to operate correctly when your input level is out of spec. If Vcc is 2,5V (Vih is 1.7V) and (Vil is 0.7V). The OE input should be above 1.7V for a valid high or below 0.7V for a valid low.   anywhere in between it may not function correctly. If they want the output to continue to follow the input then they should keep the OE input low. Below 0.7V

Hi,

Thank you for your reply.

I understand that threshold level is the possible cause of the different

result between 2.2Kohm and 1.0Kohm resistor.

But, I have another question about this symptom (for 2.2Kohm case).

Please find the attachment. I wrote my question in red.

Since the pulse on OE is for short duration in comparison with input

LOW pulse,  I think output should become LOW soon after the end of

OE pulse.

But, actual result at the customer is…. the output is kept high “entirely”.

I do not understand why output does not become LOW after the OE short pulse....

What can be the possible cause ?

Best regards,

Rak

As you know you can make an oscillator by holding a CMOS device at the threshold level and I think this is what is happening here. You can even see the oscilation on the OE and it continues to oscillate through the low input signal causing it to be ignored. If you look at the OE and the input you can see oscilation after the OE pulse that slowly settles. To me this says that something internal on the part is oscillating and locking it up until the oscillation settles. I suspect that you could also see some oscillation on Vcc or gnd during this oscillation.

 2 thing could be happening with the stronger pulldown. The pulldown holds the signal lower which may get it away from the threshol and it may settle the oscillation quicker.

Have them check Vcc and gnd and see if they see them wiggle during this time. If they do then they might put a better bypass on the part to correct it. . Where is the bypass located and what value is it? It is best to have it as close to the Vcc pin as possible.