SN65HVD21A: TIA/EIA-422-B, TIA/EIA-485-A and ITU Recommendations V.11 and X.27 requirements

Kirill,

As I understand, V.11/X.27 does no longer in use. The latest version is V.11 published in 1996, in which the electrical characteristics is similar to TIA/EIA-422-B. SN65HVD21 meets most of TIA/EIA-422-B. The only difference is Ios (bus short current). RS-422 requires 150mA, while SN65HVD21 provides 250mA. I don't this would impact most of applications. Please let me know if you have more questions.

Regards,
Hao

Hi Hao,

Yes you are correct V.11/X.27 is deprecated and no longer in use, at least X.27, however V.11 published in 1996 still in force.
I made the same conclusions by comparing datasheets for SN65HVD21A and SN65ALS176 which is meet the requirements of TIA/EIA-422-B, TIA/EIA-485-A and ITU Recommendations V.11 and X.27. I have also spotted that test conditions are different between SN65HVD21A and SN65ALS176 for Ios parameter, Vo=-20 to 25V for SN65HVD21A and SN65ALS176 have multiply test conditions, where Vo=0V (as mentioned in V.11) was measured too. The most interesting thing here is that the method for driver short-circuit test is different, see Figure 9 of SN65HVD21A datasheet and V.11 (see attached), there is no schematic diagram for parameter measurements appearing in SN65ALS176 datasheet, I guess it is because it complies to V.11, so the test methods were taken from V.11 recommendations.
There are also differences between above mentioned ICs in driver switching characteristics section, I believe it is due to SN65ALS176 supports higher data transfer rates.

Is there an evidence for SN65HVD21A somewhere on TI, where Ios measurement was done using V.11 test circuit?

We have to comply to IEC 61162-1 and IEC 61162-2 standards which are both refers to the ITU V.11/X.27 recommendations. The simplest way would be to provide our customers an statement of compliance based on transceiver datasheet or other document from TI, where would be clearly standing that the SN65HVD21A transceiver is comply to the mentioned standards.
Can TI provide such of document?

Best regards

Kirill Kornuta

Kirill,

The test conditions of different standards are different. I summarized the major difference between RS-485 and V.11/X.27 in the table. 

	RS-485	V.11/X.27
Vod	1.5V with 54Ohm	2V with 100Ohm
Ios	250mA to -7V to 12V	150mA to GND

In general a RS-485 transceiver can meet V.11/X2.7 spec as well. But unfortunately many devices don't characterize for both test conditions, such as SN65HVD21A. It would need test program and quality change for collaborating new characterization items. Is there any specific reason you chose SN65HVD21A? You could look at THVD1500, which specifies 2V Vod with 100Ohm load and 100mA Ios to -7V to 12V.

Regards,

Hao  

Hao,

Thank you for your answer and comparison table. We have had public holiday Friday last week. 

The HW design was done by other people, not me, and I guess they chose SN65HVD21A because it can handle 15V and more for both differential and common modes indefinitely. I was able to find their questions about our spec. see here: https://e2e.ti.com/support/interface/f/138/t/557820?tisearch=e2e-sitesearch&keymatch=SN65HVD21A

THVD1500 seems meets all criteria, except 15V on the bus. 

I've measured Ios as described in V.11 with disappointing results Ios =165mA for SN65HVD21A.

I can see TI recommends THVD1550 as alternative for SN65ALS176, but Ios does not characterized for Vo=0. 

Can you please propose another transceiver that will meet all of requirements?

Best regards

Kirill

Kirill,

Thanks for your information. Unfortunately I could not any parts meet all of the requirements. TI has some multi-channel line drivers that meet V.11, such as AM26C31. But they don't have -15V to 15V common mode range either. However I'm curious where this 15V requirement comes from. V.11 only needs -7V to 7V (smaller than RS-485).

Regards,
Hao

Hao,

15V requrement is comming from IEC 61162-1, see attached:

The result of Ios measurement I did yesterday is really disappointing and incomprehensible: I can see it starting with 165 mA, after 20-30 sec.the current falls to 150.8 mA while IC heating up, then after 15-20 min. it falls to 150.3 mA, and it is only 0.4 mA that prevents us from a possibility to be compliant with V.11. I believe a poorer PCB design would help lower the current further as it will allow higher temperature of transceiver. The Ios current is measured by fluke 179 and 175, which are both show the same result. Besides the identical results from both multimeters, they are calibrated.

Because of very strange results (150.8 - 150.3 mA), I'd like to ask you to consult and clarify that behavior with development department and the team which is developed that transceiver SN65HVD21A.

It could be that I measured it wrong (the test setup is wrong) in this case I would like to know what I'm doing wrong. The measurement was done directly at the module output terminals without any additional cables.

Thank you,

Best regards

Kirill

Kirill,

I think the maximum voltage on bus is more related to the max fault voltage of the bus pins, which is listed in the abs max voltage session in the datasheet. THVD1500 has 18V fault voltage.

Talking about short current test, it's a little tricky. I'm not sure the test would be done in the system. I suggest you gradually change the short voltage to ground during the measurement. If the change is too fast, it could create some bounding in the current. On the other hand, if the change is too slow, the high current might generate some heat to make the device enter thermal shutdown mode.

Regards,

Hao

Hao,

I think you are right +/-15V is related to the max. fault voltage on the bus pins (either between them or to the ground), but I'm not sure if THVD1500 can handle it indefinitely, when we talking about absolute max. parameters for IC?

I'm not sure what did you mean about gradually changing the short voltage to the ground. I'm sure I measured correctly, I'm just not sure if this test must be done with some length of cable or just directly on the terminals. Either V.11 or http://www.ti.com/lit/an/slla070d/slla070d.pdf does not specifying that. Logically this test must be done without any cables, as short current can be tripped while mounting a cable to the terminals, where terminal B and C are very close to each other and a accidentally a wire could be hooked up f. ex. from B to C. I'm sure this test must be done continuously because this test ensures that the device is not destroyed by excessive current flowing through the output stage.

I just do not understand in case of my measurements are correct, why TI developing an IC that must be backward compatible with TIA/EIA-422B as it is meets TIA/EIA-485A which is newer, but based on TIA/EIA-422B, where Ios is about 160mA and not f.ex. 140mA? This can indicate that I do something wrong, but what???

(Here is the quote from the Analog Devices AN-960: The TIA/EIA-485-A standards are similar to those described in TIA/EIA-422-B, and the values used to specify the drivers and receivers in TIA/EIA-485-A standards are specified so that it can meet both standards.)

Again please contact some developer at TI to discuss that or tell me how I can contact some transceiver developers on TI to discuss that. Our customers deadline is on 3rd of June, so this is very urgent.

Meanwhile I ordered some other transceivers, so I can measure them too. The one of them is THVD1550 as I understand it is from the same series of THVD15xx but it can take +/-15V at any bus terminal, but Ios is not specified for Vo=0.

Best regards

Kirill

Kirill,

Yes, THVD1500 has +-18V abs max voltage on bus pins. It should be able to handle the 15V maximum voltage on bus required by IEC 61162-1.

SN65HVD21 was designed to meet TIA/EIA-485 Requirement. The device is also compatible with TIA/EIA-422B, but not necessarily compliant with TIA/EIA-422B. Actually since the difference between two standards are minor, many customers are OK with this compatible in the applications.

Again, if you want to have an RS-485 transceiver that is guaranteed to pass TIA/EIA-422B, THVD1500 might be a better choice for you. The specifications on the datasheet are tested rigorously under all kind of variations (process, temperature, voltage).

Please let me know if you have any further questions.

Regards,
Hao

Hi Hao,

Thank you reply, as mentioned before, I have ordered some sample of different transceivers. I will do some tests to realize if I can use another one instead of SN65HVD21A.

Regards,

Kirill Kornuta