0

I have an issue with the following schematic:

enter image description here

When I am doing a hot swapping on the RS422 (Local console) connector X1200, my fuse F1200 blow up. I have a TVS diode to protect against overvoltage on power line (+36VDC) against ESD. However I don't have TVS diodes on Tx, Rx, gnd lines. Can it be the issue ? How could I prevent my fuse F1200 to blow up if the user is doing a hot swapping ?

I already check the fuse sizing and the typical current used is about 1 A, so it can't be a fuse sizing issue.

Additional information: VSL is +36 VDC and its power coming from an external power supply. This power is sent directly to pin 6 of connector X1200 in order to power supply an external circuit which is a console (containing a display unit, touchscreen, ssd.... (small computer)). This power supply is received by the external circuit then buffered. In the following picture you will see reception of the power supply by the external circuit.

enter image description here

Thanks for your help, Regards,

Armand
  • 21
  • 4
  • How are the grounds related in the different devices you are connecting/hot swapping? – Math Keeps Me Busy Mar 24 '21 at 15:08
  • 1
    I am bothered by the ferrite bead you have between pin 5 of your DB-9 connector and ground. Please read the answers to [this question](https://electronics.stackexchange.com/questions/92014/can-i-use-ferrites-for-the-ground-connections). – Math Keeps Me Busy Mar 24 '21 at 15:33
  • Ground between both devices are connected together with a SMD Common Mode Line Filter (Würth Elektronik, ordercode 744273222) – Armand Mar 24 '21 at 15:46
  • When you plug into the console, is it powering your circuit shown above or vice versa? – Andy aka Mar 24 '21 at 15:51
  • 2
    wow, 1.5A *slow blow*! This has nothing to do with ESD. Most likely you have huge bypass capacitors somewhere. If it was male plug I'd also consider accidental shorting of the power pin to the ground or shield, but you seem to have female socket there. TVS protection on external comm lines is always a good thing, but again, it has nothing to do with that fuse. – Maple Mar 24 '21 at 16:27
  • Can you describe more, is this the device that receives power from another device, or will this device give power to another device? – Justme Mar 24 '21 at 16:54
  • @Maple : I have a 1.5 A slow blow fuse because I receive +36VDC on "VSL" pin and I power supply my external circuit through pin 6 of this socket. My external circuit need 0.9 A maximum to work. Question is: if I have a ESD on one of the connector pins, normally the TVS diode should conduct and then the fuse should not blow ? The connector is femal socket yes. – Armand Mar 25 '21 at 08:27
  • @Justme : "VSL" pin is +36VDC external power supply output. I am power supplying external circuit through pin 6 of connector X1200. My circuit need 0.9A max. – Armand Mar 25 '21 at 08:29
  • What is the external circuit? Does it have a lot of capacitors so it draws a huge inrush current that blows the fuse? – Justme Mar 25 '21 at 08:32
  • @Justme The external circuit is a console which is containing: a display unit, a resistive touchscreen, ssd, microcontroller, audio speakers, external output com ports... The +36 VDC is buffered into the external circuit, so by capacitors do you mean directly on pin 6 on the external circuit ? – Armand Mar 25 '21 at 08:50
  • @Justme I just updated my issue description to show you how the power supply is receipt by the external circuit on pin 6. – Armand Mar 25 '21 at 09:01
  • Small computer? that's where your huge bypass caps are. By the way, I believe all those inductors also play their role in this. Normally slow blow fuse wouldn't even notice inrush current. That is what it is designed to do. But inductors act as sort of a throttle, keeping high current long enough for fuse to blow. – Maple Mar 25 '21 at 09:04
  • @Mapple: thanks for your answer. What do you suggest in order to prevent the fuse to blow if the issue is due to inductors ? What kind of tests should I do to characterise that ? – Armand Mar 25 '21 at 09:29
  • Normally this happens when you have big bulk caps of many hundreds of uF, but those caps on the picture are too small to blow a fuse. I would rather suspect something else. The problem isn't likely on the board with the fuse, but on the external board. – Lundin Mar 25 '21 at 10:07
  • 1
    Other problems not related to your main problem: 1) the ferrite(?) on the signal ground does no good but might cause ground problems. 2) you need the same resistor solution on your Tx differential pair as on the Rx one. 3) I don't get what R1200 and R1202 are good for. Do you have another board-to-board on the left side of the schematic where you expose UART to connectors? If so, and R1202 is for ESD, then why isn't there a similar resistor on the other line? 4) Your TVS is too weak for industrial use. Rule of thumb: 600W for industrial, 1500W for automotive. – Lundin Mar 25 '21 at 10:20
  • As for ESD, the TVS naturally only protects the supply pin, not the other pins. Having ferrites or series resistance will help though, so you shouldn't need TVS on each line unless you have very tough requirements. – Lundin Mar 25 '21 at 10:22
  • @Armand let me clarify - the inductors might contribute by delaying high current long enough for fuse to blow. But the actual problem is high capacitance somewhere. If you have access to that "small computer" of yours, check its power supply lines. You can also try powering it directly from VSl via fuse only, without all that intermediate wiring. Finally, fuse sizing **must** consider I2t. To my surprise Littelfuse did not list it in datasheet, but digikey puts it at 3.65, which is very low for slow blow part. Try any suitable fuse with I2t above 25, see if it helps. – Maple Mar 25 '21 at 18:14
  • Thanks all for your answers. I found that the operator was doing a short circuit between the power supply pin of DSUB-9 cable and shield when the machine was switched on. This is why the fuse was blowing. I checked the consumption of the fuse and it was ok in different test case scenarios. For the rest of the comment, I keep that in mind in order to improve the current design. Thanks – Armand Apr 22 '21 at 07:51

0 Answers0