1
\$\begingroup\$

In a cable I carry a 48 Vdc supply line and a RS 485 differential signal pair from a SN75176A transceiver. The 48 Vdc line is protected by a 5 A fuse, and the SN75176A transceiver inputs and outputs are protected by SM712 TVS diodes, mainly for ESD of course.

What would be a suitable protection circuit for the SN75176A for a short from 48 Vdc line to RS 485 data line that may happen within the cable? The 5 A fuse is too slow for protection of the trancseivers.

Edit: Data rate is 800 Kbps

\$\endgroup\$
9
  • \$\begingroup\$ 75156 datasheet link would be useful. But I'm thinking series R (about the line impedance for minimal BW limitation) near the transceiver, and power diodes to supply rail (or zener to GND). e.g. 48V 100R = 0.5A so the supply rail would have to accept backdrive. Keep the R power rating low so it burns out quickly. \$\endgroup\$
    – user16324
    Jan 16, 2022 at 21:49
  • \$\begingroup\$ Have you considered an isolated power supply for the transceiver? \$\endgroup\$
    – Justme
    Jan 16, 2022 at 21:55
  • \$\begingroup\$ @Justme isolated power supply is not possible due to space and cost constraints \$\endgroup\$ Jan 16, 2022 at 21:57
  • 1
    \$\begingroup\$ @F.Heisenberg A fuse blows because the transceiver has been fried and fails short due to the 48V. It isn't that the fuse isn't too slow to protect the transceiver. It is that the fuse will not do anything until the transceiver is already dead. Not to mention 5A about a million times higher than normal transceiver signal currents. Something is very wrong with the way you are thinking about your protection scheme if you expect a fuse to protect the transceivers in any way. Fuses protect power supplies from causing fires when there is a short in the circuit. They don't prevent the short. \$\endgroup\$
    – DKNguyen
    Jan 17, 2022 at 0:38
  • 1
    \$\begingroup\$ Maxim and Linear Tech (both now Analog Devices) have 485 trancievers that will tolerate the 48VDC being applied to the bus. I had a similar issue, but with 24VDC. With ‘standard’ transceivers I used 5V 400W TVS diodes and 100mA polyfuses. The TVS diodes and polyfuse need to be sized so that they don’t melt down in case of an extended fault. As well, I put 10R 1W resistors in series with the bus to the transceiver. Seems to make a MAX485 less likely to fail. \$\endgroup\$
    – Kartman
    Jan 17, 2022 at 1:20

1 Answer 1

3
\$\begingroup\$

Just improve the TVS clamp on the datalines to something that clamps the signal to <10V above or below tranceiver ground (with the maximum fault current applied). Then add some kind of polyfuse in series on the A/B lines. A polyfuse will have low enough resistance (few ohms) to not mess with the bus signals but will increase its resistance high enough when the fault current runs trough it, so you dont burn up the TVS during sustained fault conditions. Just pick the TVS robust enough that it can survive a few amps in the time it would take the polyfuse to do its magic. At 5A shouldn't take more than 1/10ths of a second

Or just get a better tranceiver if you want verified OV protection without tinkering around too much, like others suggested.

\$\endgroup\$
7
  • \$\begingroup\$ If the transceiver comes with fault-protection against short from Vcc to bus line, then TVS will not be needed anymore, right? My guess is that if then such short occurs, the TVS will start to conduct and burn out. \$\endgroup\$ Jan 17, 2022 at 19:56
  • \$\begingroup\$ The TVS should not burn out if there is a mechanism to limit the current during OV. This can be a resistor or, as I proposed, a polyfuse. If you find a transceiver that can work with (or at least survive) voltages of above 48V, then the entire circuit would not be needed. A smaller TVS and resistor for EMI and ESD protection might still be nice to have. In that case make sure these also survive the higher possible common mode voltage under fault conditions. Not sure what the tranceiver Vcc has to do with this? \$\endgroup\$
    – Thijs
    Jan 19, 2022 at 14:53
  • \$\begingroup\$ I mean when i have a transceiver that survives 48V and still use a TVS diode for ESD protection then the TVS diode has to be rated higher than 48V. Otherwise the transceiver will survive the 48V but not the TVS diode, right? \$\endgroup\$ Jan 19, 2022 at 18:46
  • \$\begingroup\$ Yes if you intend to survive continuous overvoltages without the use of current limiting devices like polyfuses in series on the, the ESD solution would need to be rated for that as well. \$\endgroup\$
    – Thijs
    Jan 19, 2022 at 19:11
  • \$\begingroup\$ Afaik polyfuse will still have some leakage current during OV situation. There is a device called TPU by Bournes that cuts off the data bus with integrated FETS, which may be an even safer solution. \$\endgroup\$ Jan 19, 2022 at 20:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.