Leakage current from USB killing the fuse?!

[vapealone]

I got the bad feeling that it is some leakage current from the USB killing the fuses.

Details here, but the bottom line is that USB power can directly reach output power circuit. (with or without fuse/battery)

And if it reaches without any shut down mechanism a circuit powered simultaneously by full battery input and USB at the same time doesn't sound too promising to me.

Sorry, for starting a new topic about it, but my previous one with this finding remained mostly ignored.
If I am right, please look into it.
If I am wrong, I do apologize for double-topicing. My only excuse is that I want to build a super tight fit triple 18650 mod and the last thing I need to replace the fuse weekly.
Not to mention that any external fuse between the board and the batteries wouldn't help either unless I bridge the on-board one first.

Edited for clarity:

1. By default, USB behaves as it should.
2. Unintentional power up only happens when some rotating torque is applied on the socket
3. Happened with me accidentally when I have tilted the case w/USB plugged in and it put has some tension on the socket/board
4. It is one specific point/position when it happens with a very narrow margin. When I realized what happened found it hard to find the spot again for double-checking.
5. It is a 3D printed plastic case which is probably more flexible than an alloy one.

[retird]

hum, I wonder, as I've vaped a reference device since early June and much of the time I vape while it is charging via the computer USB. Each time I press the fire button the charging stops and continues charging again after releasing the fire button. After more than 3 months of doing this daily I've not had a single fuse let go. Just giving my experience vaping while plugged into charging USB.  No issues here.

[vapealone]

dc601 said:

hum, I wonder, as I've vaped a reference device since early June and much of the time I vape while it is charging via the computer USB. Each time I press the fire button the charging stops and continues charging again after releasing the fire button. After more than 3 months of doing this daily I've not had a single fuse let go. Just giving my experience vaping while plugged into charging USB.  No issues here.

Yep. That is the normal way. But as I have mentioned in the linked post there is a specific position of the USB when it bypasses protection: I have pushed the plug slighly upward accidentaly and this little tension moved something inside powering up the whole board the way it shouldn't have.

[VapingBad]

With the USB plugged in and the battery & balance tap disconnected you will see the normal screens without the battery on the lhs, are you seeing the battery on the lhs?

As I said in the linked thread this is needed to program the board before a battery is connected and probably other checks in production and I don't think it is anything to worry about.   It does bypass the fuse, I don't see that a 25 A fuse on a USB connection could be viewed as protection and IIRC the fuse is mainly to protect the battery anyway.

[vapealone]

VapingBad said:

With the USB plugged in and the battery & balance tap disconnected you will see the normal screens without the battery on the lhs, are you seeing the battery on the lhs?

As I said in the linked thread this is needed to program the board before a battery is connected and probably other checks in production and I don't think it is anything to worry about.   It does bypass the fuse, I don't see that a 25 A fuse on a USB connection could be viewed as protection and IIRC the fuse is mainly to protect the battery anyway.

It is not the battery readout.
It is the output power.
Try to check the atomizer resistance with battery unplugged (both main and charging plugs)

When I do it w/o battery I got a big '?' for there is no input battery power that should power up the output circuit. (Battery voltage readout comes from balance charging ports)
When I do it with the method described above some miraculous current powers up the output circuit.
And this is it.
Battery readout, chip readout, charging has nothing to do with it. The only question is the actual output under load.

[VapingBad]

So is that with the battery disconnected and the balance taps connected?

I don't know the answer for you, but not clear if the balance plug is connected,

[vapealone]

VapingBad said:

So is that with the battery disconnected and the balance taps connected?

I don't know the answer for you, but not clear if the balance plug is connected,

I didn't want to abuse the board more but I will do some test when I got home. It is happening for sure when battery disconnected an ballance charging is on. Confirmed for my battery is permanently disconnected by a blown fuse

[VapingBad]

I see, the path would be through the battery monitoring/charging controller chip I saw someone post a link to the spec sheet for that a couple of months ago on ECF IIRC.

[vapealone]

VapingBad said:

I see, the path would be through the battery monitoring/charging controller chip I saw someone post a link to the spec sheet for that a couple of months ago on ECF IIRC.

In that case it is hopefully controllable via firmware, isn't it? Or it is just wishful thinking?

[Mad Scientist]

vapealone said:

[QUOTE=VapingBad]I see, the path would be through the battery monitoring/charging controller chip I saw someone post a link to the spec sheet for that a couple of months ago on ECF IIRC.

In that case it is hopefully controllable via firmware, isn't it? Or it is just wishful thinking?[/QUOTE]

I'm not sure that it's a normal operating condition to have the B+/B- to the battery disconnected but the balance plug connected and USB charging connected.  Am I missing something?  That's what I'm picking up from the thread.

Is that the concern?  That when the B+/B- is disconnected but the balance plug and USB are connected it acts weird?

[vapealone]

When I have experienced this, my battery was connected, however the fuse was blown so B+ actually can be considered unplugged and only B- was present (if at all, dunno)
Tonight, I will Check all the possible combinations

[John]

This seems like it would be a pretty good theory, but everything you're seeing is normal and won't/cant pop the fuse. I'll walk you through it:

The USB can absolutely provide power to the battery terminals. Otherwise it couldn't charge. However, if the charger isn't active then it can only provide up to USB voltage to the battery, and there's a diode in the path, so as long as the battery input (after the fuse) is greater than 5V power doesn't go that direction without the charger on. 

With the charger on (and the fuse intact), what the charger... does... is generate a controlled voltage from the USB input that is higher than the battery voltage, and then feed that into the battery. Basic battery charging. 

It uses the tap to read the battery voltage, so if your fuse is already blown but you have the tap connected, it will see the battery there and try to charge it. Won't be able to, of course, because the fuse path is gone, so the power will get dumped into the overvoltage protection diode. This will heat the board, so I don't recommend it long term, but it won't pop fuses. 

In this scenario, as long as you're asking for less power than the USB port can supply, it will quite happily fire from the generated battery voltage for a short period of time. Basically the input capacitors acting like tiny batteries. 

In no circumstance can any of this blow the fuse, though. No USB port on earth can supply the 25 amps it takes to pop a fuse. So I'm not 100% sure where the question is. You're describing what would be normal operation in a degraded state (battery positive not connected, battery negative connected, taps connected, USB connected)

Hope that helps
John

[vapealone]

John said:

This seems like it would be a pretty good theory, but everything you're seeing is normal and won't/cant pop the fuse (...)

Hi John,

Thank you for your detailed answer, However, I fear that you missed my point.

I haven’t got any concerns about the built current/leakage/polarisation protection on the board as long as designed functions/connections concerned. Neither with charging or USB current per se.

I do have serious concerns about some alternative current path that was never intended to exist.

In my case, an otherwise open circuit can be closed by flexing the USB socket a bit (via the mail plug)

In other word, the USB socket acts like a push button switch or more properly a flex button switch switching power on a specific circuit that unlikely to be powered up this way.

The effect is the same as probably all of us experienced with our earphone cords. When one of the wires inside got broken the stereo headphone became mono. Should you have found the right spot and the proper amount of flexing required you could temporarily close the open circuit and get the stereo back.

In real life I have experienced the following:

1., Battery attached (both power and charger connectors) fuse blown, USB not attached:

• The thing is dead.

Note: Apparently your logic circuits are hooked to power in behind the fuse  which makes sense.

2., Battery attached (both power and charger connectors) fuse blown, USB plugged:
    
She looks OK and even charging happily.

Device Monitor: doesn’t show any strange things save for the missing cold ohm.

Atomizer Analyzer: couldn’t find anything, gives a big question mark.

 
Note: Apparently your logic circuit(s) couldn't (or not supposed to) power up atomizer for checking its resistance.

3., Battery attached (both power and charger connectors) fuse blown, USB plugged and slightly flexed:

Device Monitor: mostly the same with occasional error message here and there, still no cold ohm

Atomizer Analyzer: finding and measuring the atomizer as per normal.

 
Note: It doesn't work with all of my cables but with the one I have used. Some cable just gives error message when flexed.
 
Bottom line:

I can switch the power on for the Atomizer Analyzer by simply flexing the USB and closing some circuit that you more likely didn’t design to be closed this way. Or at least I hope so, and it is not that you referred as normal to 

My humble request:

I would greatly appreciate if you could have a close look on the Atomizer Analyzer function both hardware and software side (if needed) and explain me which circuit can possibly get closed by flexing my plug.And let me repeat myself: it is not an easy trick. Unfortunately it is my high end default bypass cable that can do the trick, my other cables are apparently weaker than the case/bard/socket combo and a simple error message is the max they can get. 
But either way, the 3 different readouts above with the corresponding current events are modulated via slightly flexing/ not flexing the USB port. And it shouldn't happen. 

Thank you in advance

[VapingBad]

It seems to me that the main issue is you have some damaged USB cables.  You are assuming that the analyser readings are due to voltage going to the 510, well I imagine the voltage used it the analyser in the one second polls has probably come from a capacitor and not straining the USB, that it is not a problem for the board or the USB anyway and it could easily be the data lines being corrupted when you strain the cable that cause "?" and "error" to appear.  I am not sure what the analyser should be doing from just USB or USB and no battery+ so I could be wrong.

Thanks for the info John, the part about how the heat is generated charging with a blown fuse explains a couple of post I have read in the past.

[vapealone]

Well, I think you don't get it. All of the cable reads/charges the board just fine when it is not flexed. The stronger cable does make the Atomizer Analizer working when flexed, weaker ones just disturb it. Please think of the earphone cable. It is just that in my case the effect of the flexing is happening on the board: some circuit is getting closed that is normaly shouldn't. It has nothing to do with the cable as such. If anything, it is more likely the socket itself that can short out or something. If the cable was damaged there was no proper readout/charging when not flexed.

[VapingBad]

You are right I don't get it, as I read your posts you are saying it is flexing the cable causes these symptoms and also saying it is not the cable causing them.   The insulation inside cables can get damaged by tight bending over time and can be anything from conductors getting too close to each other causing interference to shorts between conductors.  In your case the simplest explanation is that a damaged cable is interfering with the data transmission.  You do say that doing it with different cables causes different symptoms and to me that makes a problem with the board or socket far less likely.   Not trying to be argumentative and have found this thread interesting, just not sure you actually have a fault.

[vapealone]

I am applying T=F x l rotating torque on the socket using the plug as a lever as I described above This unfortunate move can be described as flexing or rotating or moving upward or prying or whatever.
But this is it: 
 

To rule out the option that an upward torque of given magnitude applied on the socket causes the problem we got a very simple solution:

Please confirm that a fully functional board tethered via a fully functional USB can and always should read resistance via  Atomizer Analyzer when the fuse is blown.

If this is the case, than it is my cable(s).
If not, it is the socket.
There are no more options.

[VapingBad]

Can't check exactly that scenario, but with the main battery plug disconnected and the balance still connected it shows the question mark in atty analyser.

[vapealone]

VapingBad said:

Can't check exactly that scenario, but with the main battery plug disconnected and the balance still connected it shows the question mark in atty analyser.

Thanks for that 
Don't flex the cable/pry the socket then!
Because that is what I have feared that it shouldn't.
And I still can get it work which is a bad news because it means I can short something out.
OMG

[VapingBad]

It is a jump to think that anything bad is happening that will shorten the life of your board, especially as I don't believe any damage has resulted.  I bet if you put a USB metre like a Charge Doctor on it you will not see over 5 W being drawn.

[vapealone]

VapingBad said:

It is a jump to think that anything bad is happening that will shorten the life of your board, especially as I don't believe any damage has resulted.  I bet if you put a USB metre like a Charge Doctor on it you will not see over 5 W being drawn.

Lol
You eally get me run hard for my money
I like it Helps thinking.

In our case it is the potential difference that counts.
So lets do a bit of a maths For the sake of simplicity let DC/DC conversion efficiency 100% and there is no voltage drop in the system.

Let:
Ppp(preheat power)=200W
R(load, i.e atomizer)=0.2ohms

Then
Vload(atty)=sqrt(200W*0.2ohms)=6.32455
Iload(atty)=sqrt(Ppp/R)=31.623A

and

Vout(from DC/DC converter)=V(load)=sqrt(200W*0.2ohms)=6.32455
Iout=Ppp/Vout (from DC/DC converter)=I(load)=sqrt(Ppp/R)=31.623A

Vin(battery)= 11.1V
Iin(battery)= depends as per below

Now the scenarios.

1., Normal usage:

Ppp=Vload*iload=Vout*Iout=Vin*Iin=200W -> Iin(battery drain)=200/11.1=18.02A
So far so good, we are safe)

2., USB shorted out and the shorting current is working against batteries due to the potential difference
Let Vusb=4.77545V<5V (for the sake of simplicity)
Now, to get the required potential difference on the load you will need higher Voltage:
DV=Vload=Vout-Vusb=6.32455V -> Vout=6.32455+4.77545V=11.1V

Iload=Iout=31.623A*

Now Ppp=200W<Pout=Vout*Iout=Vin*Iin=351.0153W

And

Iout=Iin=31.623A (battery drain)*

And

Kaboom

So, it looks feasible

*for the sake of simplicity. Probably different, depending on the additional resistance in the system that I don't know, but I beleive that in general, the maths holds

[Spirometry]

Broken dna200 + rma = new dna200

[VapingBad]

lol, you don't need to do the math, just saying that seeing the analyse function working in your case doesn't necessarily mean the full operation is available, unless you have measured it happening.  Like Spirometry said RMA it.

[vapealone]

VapingBad said:

lol, you don't need to do the math, just saying that seeing the analyse function working in your case doesn't necessarily mean the full operation is available, unless you have measured it happening.  Like Spirometry said RMA it.

It is rmaed. But I still want to know what happened and why to avoid it in the future if possible

[Mad Scientist]

vapealone said:

[QUOTE=VapingBad]lol, you don't need to do the math, just saying that seeing the analyse function working in your case doesn't necessarily mean the full operation is available, unless you have measured it happening.  Like Spirometry said RMA it.

It is rmaed. But I still want to know what happened and why to avoid it in the future if possible [/QUOTE]

Well seems the good news is if it was a design issue with the board you'd see more folks reporting the issue on this forum.  You seem to be the one and only.

Do you know what it may have been that caused the fuse to blow?  Drawing > 25 Amps from the battery is supposed to be an unexpected fault condition.  There have been several reports of blown fuses.  I wonder if there are really that many careless builders (doubtful?) or if there's a trap for the unwary somewhere of doing x = blown fuse.