SN75176A: Temperature issue

With the driver active, I'd estimate the junction temperature to be about 20 °C over ambient; this should be no problem. Is it possible that some output is shorted? Please show the schematic.

The SN75176A is a very old device and has a high idle power consumption. A more modern device like the THVD1520 (or SN65LBC176 as through-hole) would run cooler (and be more robust). But I do not have enough information to predict whether this would fix your problem.

Hi David,

I agree with Clemens assessment. 

Additionally, I'm wondering if the heating may be causing the device to weaken it's drive capability (VoD becomes lower as temp rises) which at 1.2km may make the receiver unable to properly read the data packets correctly. 

Is it possible for you to get scopeshots of the A/B signals on both the transmission node and the receive node? (Make sure to do a math function of A-B to see the differential voltage).

Do you have a schematic we could look at?

-Bobby

Thanks to you both. The manufacturer won't release the schematic for the circuit so I have had to work out some of this via reverse engineering it. It is opto isolated with a DC-DC SIP module for power. The only thing connected on the A + B lines that I could see was some biasing resistors. No TVD etc. 

Bobby, I did manage to scope the signal from the other end at the affected device when it had stopped working but don't have any captures. There was a nice clean signal at the end that failed. The baud rate is only 9600. The levels were about 1V less than what I was seeing on the other end. The peak to peak was 4.2V so all good. When it fails, there is no transmission from the device. I didn't scope the output of the transceiver as I wasn't able to connect the scope to the isolated ground. 

I should add that the manufacturer of the device informed us recently that a previous client in the past had a similar failure. No other information was given.

I am trying to convince the client to swap out the SN75176 for a later device with higher common mode voltage and temperature range. I have proposed to build a test circuit a put this in a controlled over to confirm failure of the original and that the alternative continues to work, although I won't be able to simulate the 1200 m link so easily. 

When the transceiver is isolated, then there should be no common-mode problem (i.e., the transceiver's ground should be the cable ground).

From your description, I'm not even sure whether it is the SN75176A itself that causes the problem, or some other device nearby, or a bad solder connection. You indeed have to replace the chip to be sure.

Ignoring the capacitance/inductance (which is probably not a concern at 9600 baud), you can simulate the voltage drop over long cables with resistors.

David McLaughlin said:

There was a nice clean signal at the end that failed. The baud rate is only 9600. The levels were about 1V less than what I was seeing on the other end.

If thats the case, you don't need to probe the transmission side. 4.2V swings should be good enough to ensure the receiver actually reads/interprets the data correctly.

I also agree with Clemen's comment. I'm not sure whether or not the root cause is the SN75176A just yet. Do you have any additional SN75176 devices on hand that you could swap out (remove the current one and replace with a new one). 

It could be possible the device was damaged or the soldering on the device may not be good. 

Clemens Ladisch said:

Ignoring the capacitance/inductance (which is probably not a concern at 9600 baud), you can simulate the voltage drop over long cables with resistors.

Clemens is correct. At this distance and data rate. You can just simulate a voltage drop across the cable if you know the cable's resistance/meter. 

Have you tried probing the device's D, DE, and Vcc pin when the issue occurs? 

-Bobby

Hi Bobby, we did swap out to a brand new CPU card and the same fault appeared after heating up time had passed sitting in the workshop. This is why I am believing the fault is the SN75176 as it was the only part we aimed the heating at and the air duster to carefully cool it down.

These are processor modules for a subsea system and mounted in sealed enclosures at 1 atmosphere and pressure rated for depth. The same CPU Is used at the surface so identical RS485 drivers etc.

There are fans internal to keep the air moving. Normally in water they are cooled by the surrounding sea water but on deck they can warm up. It has failed whilst they have it powered up and doing their system checks. 

Hi David,

Any way you can probe the DE and Vcc pins when you apply heat to the device on the transmission node?

You could also try to replace the device with THVD1520 and just see if the problem still occurs. 

-Bobby