General > General Technical Chat
LTC4365 nuisance tripping on ground shift
<< < (2/2)
coppercone2:
some high speed DC DC converters specifically recommend isolating the output of the chip ground through ferrite
tom66:

--- Quote from: liteyear on December 21, 2023, 03:46:36 am ---Good food for thought.

Simple enough to try too! I removed one of the shield-ground resistors and replaced it with a wide flat solid bond, and then recreated the fault scenario...

If anything, it was easier to trip. I wouldn't read too much into that, because just the humidity in the room might be a bigger factor. But certainly, no harder to trip.

However! The oscilloscope captures are maybe slightly different? Here's UV in yellow and Vout in blue.

(Attachment Link)

I'd say broadly the same magnitude, but a bit less higher frequency content?

Obviously none of this provides any useful evidence for/against your well made advice, but it was too tempting not to see what happens!

--- End quote ---

The amplitude of the fault looks far lower in that trace - looks like the shield may be working better - but perhaps this is just one capture among many.    Unfortunately the shield will not always be enough - I suggest you will need to add the 10-100pF capacitance on the '4365 sense lines and model your system to determine if the reduced overvoltage transient response is acceptable.  If it is not, you will need to add a TVS or some other form of over-voltage protection.
liteyear:

--- Quote from: tom66 on December 21, 2023, 01:02:19 pm ---The amplitude of the fault looks far lower in that trace - looks like the shield may be working better - but perhaps this is just one capture among many.
--- End quote ---

Yeah sorry I didn't leave a logical trail to compare against. This is the equivalent before bonding the shield. It differs from the trace in my OP because I've turned on the 20MHz bandwidth on each oscilloscope channel.



Blue is the gate pin, not Vout, but otherwise useful for comparison to the bonded shield capture.


--- Quote ---I suggest you will need to add the 10-100pF capacitance on the '4365 sense lines and model your system to determine if the reduced overvoltage transient response is acceptable.  If it is not, you will need to add a TVS or some other form of over-voltage protection.

--- End quote ---

Aye. I don't expect the reduced transient response to be an issue - I'd prefer to rely on TVS's and the like for that, and reserve the OV/UV chip for detecting less transient miswiring / bad power supply type faults. The bigger issue is that any PCB change at this stage is a recall drama. So if I can understand the mechanism, I might have a hope of mitigating it using installation techniques or software recovery.
tom66:
The other possibility to consider is whether you have a crap USB cable.  Unfortunately you can't always control what customers will use, but some cables are made with pathetic foiled plastic for the shield - or no shield at all in some cases. Also, the shield termination in the connector should ideally be 360 degrees around the connector with a continuous bond but often it is just a single wire connection which creates a dipole antenna, negating a lot of the benefits of the shield.  You might be able to insist on specific cables being used; it's worth testing a few sourced from reputable manufacturers (not Amazon/eBay as quality is too variable.)  If you have to send some decent cables out to your customers and insist they use those instead of whatever they have, it could be cheaper than replacing units.
liteyear:
Yeah it's a good idea. These are not consumer products so I have some control over the installation.

But I can easily reproduce the issue with a decent braided cable, and it doesn't seem sensitive to the cable that is used. Besides, I've already dismissed coupled noise as the root cause (it was an early theory) because the trigger is the moment of connection, not so much the presence of noise. So it seems more like a transient ground shift issue than imposed noise. It's the act of joining grounds, not so much the presence of a unintended antenna that's causing all the drama.
Navigation
Message Index
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod