General > General Technical Chat
LTC4365 nuisance tripping on ground shift
liteyear:
I'm using the LTC4365 to protect a device powered by a 5V/2A wall wart, connected via barrel jack.
All was going well, but now the devices are in the field we're starting to see nuisance tripping. I've managed to produce the symptoms in the lab by repeatedly replugging the device's upstream facing USB port (data only, no power) into a hub. Occasionally, the act of touching the USB cable's shell to a new reference (be it a floating hub, an earthed hub, or sometimes just my fingers) will trigger the LTC4365 to disconnect.
Obviously the effect of ground shift is undesirable and trying to prevent it and keeping a more stable ground would be the ideal solution. But also, the LTC4365 seems mighty sensitive, so I'm not sure if I'm barking up the right tree here.
The LTC4365 section of the schematic looks like this. UVth = 4.51V, OVth = 5.51V and C7 provides about 5ms of droop ride-through. The nets on the left connect to the barrel jack, and the nets on the right connect to the device's 5V rail and GND. There's about 50µF of capacitance on that rail.
The USB connection looks like this. VBUS- goes on to be direct connected to system ground. VBUS+ is not used.
And here's my current best effort to measure the disturbance. This occurs when I connect USB to a hub. Yellow is a 10x 350MHz PVP2350 probe directly across the LTC4365's ground and UV pins. Blue is just touching the 5V rail (no dedicated reference clip). Both channels have 300MHz bandwidth set.
This is enough to cause the LTC4365 to trip, and I've confirmed through various other measurements that it is tripping ever so briefly on OV or UV.
Here's SHDN during the same event, and this time blue has its own reference.
But when I try to do the same to capture OV, I couldn't get it to trip! It's as if the ~10pF of the probe is enough to stabilise the OV pin, even though C7 should already be doing that.
Here's the relevant part of the layout. The LTC4365 is highlighted. The barrel jack input is visible below. The dual mosfet is to the left. Red is the top layer, green is the second layer, which is a solid ground pour.
And finally, here's another view of the glitch event just for context. This time yellow is the gate pin and blue is the FAULT pin (which I've pulled with 10k).
Do you think enough noise could be coupled into the OV pin to trip the chip, even though C7 is present? Would 10pF from the probe really be enough to make a difference? I wondering whether the noise could be so high frequency to render the capacitors or resistors inductive? Am I chasing ghosts?
tom66:
I am uncertain where the idea of connecting SHIELD through an RC network, or ferrite beads, or anything else really comes from. The shield should simply be grounded with as low an impedance as possible. My pet theory is it is spread by EMC consultants that want more work! It's an anti-pattern, and it does nothing good.
The only cases where you might have a different connection would be where electrical safety requirements mean you can't have a case connected to the PCB ground. In that case you might go for a capacitive connection, but it's always worse than a direct connection. Most products using USB don't have this as a requirement as USB is itself not isolated.
As for the glitch on OV/UV pins, I think you might want to consider adding a few tens to hundreds of pF on either the OV/UV pins, as it could be that the internal glitch filter is overwhelmed. If it works I wouldn't worry too much about it.
liteyear:
--- Quote from: tom66 on December 19, 2023, 01:24:41 pm ---I am uncertain where the idea of connecting SHIELD through an RC network, or ferrite beads, or anything else really comes from.
--- End quote ---
Not to stray too far into that hornet's nest, but here's the primary reference I used: https://electronics.stackexchange.com/questions/498039/how-to-correctly-connect-a-usb-shield-on-a-pcb/498077#498077
Some interesting excepts there from reputable names claiming some test result improvements. Anyway, my goal was good coupling without being the preferential path for a earthing misadventure, which I've seen other boards succumb too. But I take your point that it's something short of a direct connection.
Thanks for the puffs recommendation. This is bubbling to the top of the list of sensible ways forward. I remain pretty perplexed that the 0.1uF doesn't do the job already, but I guess it's plausible there's enough reactance in the resistor divider to make a difference. Practically speaking, as long as we understand the issue I'm not going to issue a recall at this stage, and this will just go on the pile for the next revision.
In the end I was able to capture event on the OV pin. I did this over and over to be sure, because I find it hard to believe. Here's what it looks like:
Same situation as before, except yellow is on the OV pin and blue is on the gate pin. So it seems vastly noisier than UV, and even rings when the MOSFET opens (and interrupts about 130mA). A lot more noise that I would expect given the compactness of the circuit, but gives strength to the extra puffs proposal.
tom66:
--- Quote from: liteyear on December 20, 2023, 06:13:39 am ---
--- Quote from: tom66 on December 19, 2023, 01:24:41 pm ---I am uncertain where the idea of connecting SHIELD through an RC network, or ferrite beads, or anything else really comes from.
--- End quote ---
Not to stray too far into that hornet's nest, but here's the primary reference I used: https://electronics.stackexchange.com/questions/498039/how-to-correctly-connect-a-usb-shield-on-a-pcb/498077#498077
--- End quote ---
Going to have to poke the hornet's nest here unfortunately! The entire purpose of the shield is a return for stray common mode currents. USB itself is not truly balanced because it uses differential and single ended signalling on the "diff" pairs which creates a common mode current (bad for emissions.) And no matter how well you design a system there will always be a tiny bit of common mode current due to imbalance in the two signals.
By adding any kind of impedance into the shield path, you are making the shield less preferable as a path for these common mode currents to return into. So instead of being conducted straight into the shield as they radiate inside the cable, they will radiate out of the cable and (potentially) become an emissions headache. Wherever you have an emissions headache, you also usually have an immunity headache.
Whilst it sounds like an attractive idea to reduce system ground current through the shield, it's actually not a big deal from an emissions perspective. The near DC / low frequency AC system current through the shield will be balanced with the positive wire in the cable. So the effective loop that can radiate is quite small. It's good systems design to reduce the high frequency current that does flow through the cable, but this is achieved by using chokes and filters on the supply line to the product, not by messing with the shield.
Connecting the shield through an RC network or a ferrite bead is engineering pseudoscience (and yes, it's often found in app notes and in reference designs - but these are not tested to meet emissions requirements.) There are ideas that the shield should not be connected at both ends, too. This is also nonsense. The shield ideally forms a continuous barrier around the product. That's why your PC case is all metal and the USB shields are directly connected to it. Some products don't allow this but at the very least the shield should connect to the PCB ground.
I think you are seeing an EMI effect with the shield. When you touch the shield, instead of that current going down the preferable low impedance path to ground and being dissipated, it is capacitively coupling onto the supply wire and data lines. And in this case, the coupling onto the supply line is causing enough of an upset to trip your overvoltage protection. I can't say for sure that connecting the shield to ground directly will fix issues - but I can say that it is likely causing more issues than not to have it connected through such a network.
liteyear:
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.
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!
Navigation
[0] Message Index
[#] Next page
Go to full version