1
\$\begingroup\$

Ive made my own 18650 cell tester that: Fully charges -> fully drains -> partially charges

It takes 4 cells, measured independently of each other. While testing the drain circuits (charging circuits were not made at this point), all was perfect

After doing the charging circuits, cells 2, 3, 4 work perfect but cell 1 has the drain circuit on all the time (700mA drain) regardless of whether the mosfet gate is high or low.

Ive replaced the mosfet and 10k pull down resistor but still the same behaviour.

The voltage dividers for cell 1 read lower total resistance than the other cells too..

Can somebody look at these diagrams i made for myself while planning the build and point me to a probable cause?

Ive been over it with a fine tooth comb and all the cells wiring is identical!

Thanks in advance

EDIT:

Voltage divider resistances (measured from one end of the 10k to the other end of the 3k)

V divider before shunt

Cell 1: 5.1 ohms

Cell 2: 6500 ohms

Cell 3: 6500 ohms

Cell 4: 6500 ohms

V divider after shunt

Cell 1: 4.2 ohms

Cell 2: 6500 ohms

Cell 3: 6500 ohms

Cell 4: 6500 ohms

enter image description here enter image description here enter image description here

\$\endgroup\$
  • 1
    \$\begingroup\$ You should provide a schematic and double check your resistor values with e.g. a multimeter \$\endgroup\$ – PlasmaHH May 22 '18 at 16:07
  • 1
    \$\begingroup\$ Added that info above. I dont know why the cell 1 voltage divider has drastically reduced resistance. I suppose that is where the current is finding its way to ground when the mosfet is turn off? \$\endgroup\$ – I Used To Be Funny May 22 '18 at 23:18
2
\$\begingroup\$

Here's a suggestion about how you could approach your troubleshooting and a hypothesis about the area of the problem.

You said:

While testing the drain circuits (charging circuits were not made at this point), all was perfect

OK - so at that point in time (and perhaps still now), the "drain circuit" for cell 1 behaved normally.

After doing the charging circuits, cells 2, 3, 4 work perfect but cell 1 has the drain circuit on all the time (700mA drain) regardless of whether the mosfet gate is high or low.

So the change happened after you added the charging circuits. Assuming that the channels really are independent, then we can say the problems started after charging circuit 1 was added.

Hypothesis:

I suggest that the problem is not that the intentional drain circuit through the MOSFET is on all the time. You've already said that the discharging occurs when the MOSFET gate is off.

Based on what you said, I think you can only say that cell 1 is being discharged all the time - but my hypothesis is that the discharge path is not via the drain circuit. After all, that was working OK until you added the charging circuit to channel 1.

Instead, I suggest that the charging circuit (e.g. the TP4056 module) on channel 1 has added an unintentional discharge path.

A few troubleshooting approaches which quickly come to mind:

  • Disconnect the charging circuit from cell 1 (channel 1). Is it still discharging? If not, then the problem isn't in the discharge (drain) circuit, and the discharge path is through the charge components (which is why it stopped when you disconnected them).

    or

  • Swap the charging components between (say) channels 1 and 2. If the unintended discharging moves from channel 1 to channel 2, then the problem is following the charging components. Again, that would confirm the problem is related to the charging components you moved between channels (or perhaps their wiring etc.).

    or

  • Connect an ammeter in series with the MOSFET (not the cell) in channel 1 while the unexpected discharge is happening (i.e. the MOSFET gate is measured as off). I suspect there will be no current through the MOSFET drain-source, again indicating that the discharge path is elsewhere i.e. via the charging components.

\$\endgroup\$
  • 1
    \$\begingroup\$ Thank you for your comprehensive list. I have just completed your suggestions. I completely disconnected all charging components so it is as it was when it was working. However, still a 700mA drain. I then reconnected them and moved to the mosfet and measured the current across drain and source. I measure 250mA with the machine completel powered off. Strange thing about this test though, I happened to touch to the ground rail elsewhere on the device with my finger and current across the drain-source jumped to 500mA. Went back to 250mA once released?! \$\endgroup\$ – I Used To Be Funny May 23 '18 at 0:09
  • 1
    \$\begingroup\$ Following that, I cut the source leg on the mosfet. Measured the drain at the cell once more and it was 650mA. At this point it has to be the ultra low resistance of my voltage dividers, surely? \$\endgroup\$ – I Used To Be Funny May 23 '18 at 0:11
  • 1
    \$\begingroup\$ @IUsedToBeFunny - I have limited time for further comments tonight (it's late in the UK), but briefly for now: (a) "disconnected all charging components [...] still a 700mA drain" So something has changed since your first test, since (AIUI) all was OK on channel 1 before adding the charging components. Either something is now broken, or else the physical connections now, don't match the connections in the diagram. (The reason for the variation in discharge current e.g. 250mA or 650mA or 700mA is unknown right now; the 500mA value might be explainable). \$\endgroup\$ – SamGibson May 23 '18 at 0:38
  • 1
    \$\begingroup\$ (b) "cut the source leg on the mosfet [...] drain at the cell [...] 650mA" That confirmed part of my hypothesis - the discharge (drain) is not through the MOSFET. (c) "At this point it has to be the ultra low resistance of my voltage dividers, surely?" I'm not convinced that the (incorrect) low resistance measured at the voltage dividers is really due to the dividers. I suggest there may be another discharge path to ground i.e. in parallel with the voltage dividers, making their values appear to be low. I wonder if the Arduino connections are involved... \$\endgroup\$ – SamGibson May 23 '18 at 0:38
  • 1
    \$\begingroup\$ Although the diagram shows only 1 channel, I guess that the Arduino pins are also each dedicated to one channel. Therefore: (d) On channel 1: Disconnect everything except the voltage dividers e.g. the charging components that you reconnected, all Arduino pins, the battery etc. Also leave the MOSFET source leg disconnected. After that, based on the diagram, I don't see any other possible parallel path with the voltage dividers, so measure them. Due to the shunt, I expect channel 1 will measure ~6500 ohms. Then reconnect the disconnected parts until the unexpected measurement starts again. \$\endgroup\$ – SamGibson May 23 '18 at 0:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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