Jump to content

Electric Unicycle's BMS problem and solution


hobby16

Recommended Posts

2 minutes ago, Chriull said:

like the mosfets used in the euc's - from the datasheet specs they all should handle driving with the euc quite without problems - but since they have no proper heatsink/cooling they just die....(beside the ones dying by bridge shootthrough's)

I actually found a quite "nice" listing of reasons for mosfets to die here: http://www.talkingelectronics.com/projects/MOSFET/MOSFET.html

WHY MOSFETs FAIL
There are quite a few possible causes for device failures, here are a few of the most important reasons:

  • Over-voltage:

MOSFETs have very little tolerance to over-voltage. Damage to devices may result even if the voltage rating is exceeded for as little as a few nanoseconds. MOSFET devices should be rated conservatively for the anticipated voltage levels and careful attention should be paid to suppressing any voltage spikes or ringing.

  • Prolonged current overload:

High average current causes considerable thermal dissipation in MOSFET devices even though the on-resistance is relatively low. If the current is very high and heatsinking is poor, the device can be destroyed by excessive temperature rise. MOSFET devices can be paralleled directly to share high load currents.

  • Transient current overload:

Massive current overload, even for short duration, can cause progressive damage to the device with little noticeable temperature rise prior to failure.

  • Shoot-through - cross conduction:

If the control signals to two opposing MOSFETs overlap, a situation can occur where both MOSFETs are switched on together. This effectively short-circuits the supply and is known as a shoot-through condition. If this occurs, the supply decoupling capacitor is discharged rapidly through both devices every time a switching transition occurs. This results in very short but incredibly intense current pulses through both switching devices.
The chances of shoot-through occurring are minimised by allowing a dead time between switching transitions, during which neither MOSFET is turned on. This allows time for one device to turn off before the opposite device is turned on.

  • No free-wheel current path:

When switching current through any inductive load (such as a Tesla Coil) a back EMF is produced when the current is turned off. It is essential to provide a path for this current to free-wheel in the time when the switching device is not conducting the load current.
This current is usually directed through a free-wheel diode connected anti-parallel with the switching device. When a MOSFET is employed as the switching device, the designer gets the free-wheel diode "for free" in the form of the MOSFETs intrinsic body diode. This solves one problem, but creates a whole new one...

  • Slow reverse recovery of MOSFET body diode:

A high Q resonant circuit such as a Tesla Coil is capable of storing considerable energy in its inductance and self capacitance. Under certain tuning conditions, this causes the current to "free-wheel" through the internal body diodes of the MOSFET device. This behaviour is not a problem in itself, but a problem arises due to the slow turn-off (or reverse recovery) of the internal body diode.

MOSFET body diodes generally have a long reverse recovery time compared to the performance of the MOSFET itself. 
This problem is usually eased by the addition of a high speed (fast recovery) diode. This ensures that the MOSFET body diode is never driven into conduction. The free-wheel current is handled by the fast recovery diode which presents less of a "shoot-through" problem.

  • Excessive gate drive:

If the MOSFET gate is driven with too high a voltage, then the gate oxide insulation can be punctured rendering the device useless. Gate-source voltages in excess of +/- 15 volts are likely to cause damage to the gate insulation and lead to failure. Care should be taken to ensure that the gate drive signal is free from any narrow voltage spikes that could exceed the maximum allowable gate voltage.

  • Insufficient gate drive - incomplete turn on:

MOSFET devices are only capable of switching large amounts of power because they are designed to dissipate minimal power when they are turned on. It is the responsibility of the designer to ensure that the MOSFET device is turned hard on to minimise dissipation during conduction. If the device is not fully turned on then the device will have a high resistance during conduction and will dissipate considerable power as heat. A gate voltage of between 10 and 15 volts ensures full turn-on with most MOSFET devices.

  • Slow switching transitions:

Little energy is dissipated during the steady on and off states, but considerable energy is dissipated during the times of a transition. Therefore it is desirable to switch between states as quickly as possible to minimise power dissipation during switching. Since the MOSFET gate appears capacitive, it requires considerable current pulses in order to charge and discharge the gate in a few tens of nano-seconds. Peak gate currents can be as high as 1 amp.

  • Spurious oscillation:

MOSFETs are capable of switching large amounts of current in incredibly short times. Their inputs are also relatively high impedance, which can lead to stability problems. Under certain conditions high voltage MOSFET devices can oscillate at very high frequencies due to stray inductance and capacitance in the surrounding circuit. (Frequencies usually in the low MHz.) This behaviour is highly undesirable since it occurs due to linear operation, and represents a high dissipation condition.
Spurious oscillation can be prevented by minimising stray inductance and capacitance around the MOSFETs. A low impedance gate-drive circuit should also be used to prevent stray signals from coupling to the gate of the device.

  • The "Miller" effect:

MOSFET devices have considerable "Miller capacitance" between their gate and drain terminals. In low voltage or slow switching applications this gate-drain capacitance is rarely a concern, however it can cause problems when high voltages are switched quickly.

A potential problem occurs when the drain voltage of the bottom device rises very quickly due to turn on of the top MOSFET. This high rate of rise of voltage couples capacitively to the gate of the MOSFET via the Miller capacitance. This can cause the gate voltage of the MOSFET to rise resulting in turn on of this device as well ! A shoot-through condition exists and MOSFET failure is certain if not immediate.
The Miller effect can be minimised by using a low impedance gate drive which clamps the gate voltage to 0 volts when in the off state. This reduces the effect of any spikes coupled from the drain. Further protection can be gained by applying a negative voltage to the gate during the off state. eg. applying -10 volts to the gate would require over 12 volts of noise in order to risk turning on a MOSFET that is meant to be turned off !

  • Conducted interference with controller:

Rapid switching of large currents can cause voltage dips and transient spikes on the power supply rails. If one or more supply rails are common to the power and control electronics, then interference can be conducted to the control circuitry.
Good decoupling, and star-point earthing are techniques which should be employed to reduce the effects of conducted interference. The author has also found transformer coupling to drive the MOSFETs very effective at preventing electrical noise from being conducted back to the controller.

  • Static electricity damage:

Antistatic handling precautions should be used to prevent gate oxide damage when installing MOSFET or IGBT devices.  But are very reliable once they are soldered in place.

Link to comment
Share on other sites

  • Replies 306
  • Created
  • Last Reply

Chriull, Esaj,

that was quite a beautiful picture lesson in resistors.  Mine are nichrome wires strung between ceramic hold posts and can handle over a 1000 watts for at least 5 minutes of testing out 24 volt DC Recreational Vehicle batteries to determine if the storage capacity is weak  The resistors are about 25 cm long by 10cm wide by 10 cm tall......and they have the option of having a small DC fan hooked up to cool the wires.   So testing the 67.2 vDC system should be a "piece of cake"   I only want to find out if one or more of the 16 cells are weak during high discharge (like 20 amps).  I just want enough time to use the multimeter to find out if a cell is weak compared to his neighbors.

Here is USA a high quality 16S1P battery pack with built in BMS runs about $300 to $350  (that was the quote from Banggood for a replacement battery for the MoHoo).  So I am really incented to upgrading the cells (to 2S) to see if that would stop the "abrupt stop" problem.

I will take the 4 big MOSFETs and ohm out where their SOURCE, DRAIN, and GATE signals are going (and if they are common).  Since the C+/C- connector has full voltage minus the diode drop, and the main power connector has full voltage, I cannot imagine the motherboard does anything but provide charging voltage to the C+/C- connector and taking drive current to the outrunner magnets from the Main connector.  there are no other connections from BMS to motherboard..   I will drain down the battery this weekend and then let the standard connections re-charge the battery.  I will use induction current monitors to see what connector is supplying battery current during the re-charge.  My guess is that it will be the C+/C- connector on the small Deans connector.   If that is true, then the shunting for over-current draw should only be the B- to the P-........if this is the correct way to "leave out those nasty big MOSFETs".

I cannot think of why the 4th MOSFET would be used for over-voltage protection since the charger has that built in and the main board is probably not monitoring that.  My guess is that over-voltage protection would only be needed in you had a fully charged battery and were coasting downhill and wanted to use regenerative charging to act as your brake.  But if it cuts out and disconnects power, then you will have no gyro balance as you fly down the hill for a truly nasty face plant.

    tjcooper

Link to comment
Share on other sites

17 minutes ago, tjcooper said:

Chriull, Esaj,

that was quite a beautiful picture lesson in resistors.  Mine are nichrome wires strung between ceramic hold posts and can handle over a 1000 watts for at least 5 minutes of testing out 24 volt DC Recreational Vehicle batteries to determine if the storage capacity is weak

If you tested the 24 Volt batteries one at a time it where only ~80W power dissipation by the resistor.

17 minutes ago, tjcooper said:

  The resistors are about 25 cm long by 10cm wide by 10 cm tall......and they have the option of having a small DC fan hooked up to cool the wires. 

Wow - quite a piece! I'm looking forward to get the report of your test!

17 minutes ago, tjcooper said:

  So testing the 67.2 vDC system should be a "piece of cake"  

With 67,2V the power dissipation would be ~650W - so ~8 times the power dissipation than with your 24V batteries (if you tested these one by one...)

17 minutes ago, tjcooper said:

...

I will take the 4 big MOSFETs and ohm out where their SOURCE, DRAIN, and GATE signals are going (and if they are common).

Just be aware that this is a life system and an Ohm Meter normaly puts some voltage on the test points, measures the flowing current and shows you the calculated resistance. So if you "ohm out" two points which are fed by the battery (~67V) this will likely destroy your Ohm meter (or just burn the fuse which is hopefully in the measuring path of the meter).

The voltage of the Ohm meter could also turn on/off the Mosfets of the BMS if you ohm-out around the gates.

Imho its better to get the schematics visually and by measuring the voltages at specific points and so work out missing connections...

17 minutes ago, tjcooper said:

...

I cannot think of why the 4th MOSFET would be used for over-voltage protection since the charger has that built in and the main board is probably not monitoring that.  My guess is that over-voltage protection would only be needed in you had a fully charged battery and were coasting downhill and wanted to use regenerative charging to act as your brake.  But if it cuts out and disconnects power, then you will have no gyro balance as you fly down the hill for a truly nasty face plant.

The overvoltage protection of the 4th Mosfet only cuts the charge input (C+, C-) and not the discharge output (P+,P-). As to be seen in the links i provided you a couple of posts ago, these two protection Mosfet (groups) are normally controlled by an IC which measures many different conditions of every cell and depending on the programming/choice of controller IC each of these two protection Mosfet (groups) can cut the charge input and discharge output for various conditions!

So Overvoltage for the charge input is one very common condition so this is how it gots the name. But it can easily be, that this 4th Mosfet also cuts the line once an overtemperature is measured or any other condition the designer of the BMS decided to monitor.

The same with the 3 Mosfet Group for over-discharge protection (cutting the output lines P+,P-). They main trigger for them is normally overcurrent as short-cuircit prevention. But they often also get triggered by single cell undervoltage and again by any other condition the designer of the BMS decided.

 

Link to comment
Share on other sites

Chriull,

sorry for the "AMERICAN NOTIATION":   When I say "ohm out" I mean to trace down the schematicof a PCB layout using a very low voltage ohm meter (like .1 volts DC).  This is done with the circuit in the OFF state so no battery voltages present.  I am only trying to figure out where the three legs of the MOSFETs are going....especially if they are common to each other.  Huge woodworking tasks have stopped me from performing these tests, or even riding the NB.   Maybe next week.  I am hoping that while I run my "power down test" that I can practice shunting the B- to the P- with a 1amp fast blow fuse in-line to see it that does indeed shunt the "over current" protection.

My 7 ohm resistor is sort of a mis-nomer.  It is many different segments of resistors on switches so that you can test 6,12,24,48, and even 96 volt DC battery systems.  The 1000 watts is for the 12 volt connection.   I have put over 500 amps of 12v DC battery current through the resistor to test if battery still had reserve capacity.   It really does get warm when you do that, but 10 seconds is the standard test time.

 

ANYONE,

does someone know of a source for the large diameter heat shrink tubing?   By my calculations for the MoHoo, I will need one piece that is 11.5 inches in circumference and aother that is 13.5 inches in circumference.    I don't see anything at our local electronics supply houses over 3" in diameter (so 10 inches circumference).  Much thanks.

     tjcooper

Link to comment
Share on other sites

23 minutes ago, tjcooper said:

Chriull,

sorry for the "AMERICAN NOTIATION":   When I say "ohm out" I mean to trace down the schematicof a PCB layout using a very low voltage ohm meter (like .1 volts DC).  This is done with the circuit in the OFF state so no battery voltages present.

Do you mean you remove the BMS-PCB from the pack? I don't know of any other way to "turn it off"?

Quote

 I am only trying to figure out where the three legs of the MOSFETs are going....especially if they are common to each other.  Huge woodworking tasks have stopped me from performing these tests, or even riding the NB.   Maybe next week.  I am hoping that while I run my "power down test" that I can practice shunting the B- to the P- with a 1amp fast blow fuse in-line to see it that does indeed shunt the "over current" protection.

If you do figure out the schematic, please post it, I'd be interested to see it  :)

 

Quote

My 7 ohm resistor is sort of a mis-nomer.  It is many different segments of resistors on switches so that you can test 6,12,24,48, and even 96 volt DC battery systems.  The 1000 watts is for the 12 volt connection.   I have put over 500 amps of 12v DC battery current through the resistor to test if battery still had reserve capacity.   It really does get warm when you do that, but 10 seconds is the standard test time.

Nice, all I have for large current testing is Kanthal-wire (or multiple in parallel for large currents).

Quote

 

ANYONE,

does someone know of a source for the large diameter heat shrink tubing?   By my calculations for the MoHoo, I will need one piece that is 11.5 inches in circumference and aother that is 13.5 inches in circumference.    I don't see anything at our local electronics supply houses over 3" in diameter (so 10 inches circumference).  Much thanks.

Search for "shrink tubing" or "shrink sleeve " or "battery shrink tubing" etc. combinations in Aliexpress or eBay or Amazon or something... Here's an example: http://www.aliexpress.com/item/160mm-diameter-102mm-PVC-Heat-Shrink-Tubing-Battery-Wrap-Mould-Parts-ROHS-1-Meter/32557425505.html

 

Link to comment
Share on other sites

3 minutes ago, tjcooper said:

Chriull,

sorry for the "AMERICAN NOTIATION":   When I say "ohm out" I mean to trace down the schematicof a PCB layout using a very low voltage ohm meter (like .1 volts DC). 

Than thats about what i understood. The very low voltage helps to not impact the active components! But how to you ensure this off state:

3 minutes ago, tjcooper said:

This is done with the circuit in the OFF state so no battery voltages present. 

You'd have to remove all the battery cells from the BMS - they are the mighty power providers! And once they are removed you could also easily see and draw the schematics from the hopefully only two sided PCB.

if you don't disconnect the battery cells and ohm out a wrong point you have 67.2 Volts without any real current limitation against the 0.1 Volt provided by your ohm meter - if this does not have a very very fast fuse it will blow up by (the couple of) 100ts of amps flowing for a very short time! That's what i fear and tried to tell you.

3 minutes ago, tjcooper said:

I am only trying to figure out where the three legs of the MOSFETs are going....especially if they are common to each other. 

The three upper mosfets gates are ad it seems from the photos 99.9% connected together - with a real live visible inspection or some nice sharp photos one should be able to ensure 100%. But for this especial case ohming out with a low voltage ohmmeter should lead to ho harm - imho (that's no advice - somhow very low voltage ohm meter sounds expensive!)

3 minutes ago, tjcooper said:

...  I am hoping that while I run my "power down test" that I can practice shunting the B- to the P- with a 1amp fast blow fuse in-line to see it that does indeed shunt the "over current" protection.

That could easily lead to false positives - there are many connections to be made at the BMS which burn a 1 A fuse without probs but not give you any idea if they are the right shunting points not leading to disaster!

 

26 minutes ago, tjcooper said:

My 7 ohm resistor is sort of a mis-nomer.  It is many different segments of resistors on switches so that you can test 6,12,24,48, and even 96 volt DC battery systems.  The 1000 watts is for the 12 volt connection.   I have put over 500 amps of 12v DC battery current through the resistor to test if battery still had reserve capacity.   It really does get warm when you do that, but 10 seconds is the standard test time.

 Respect for this resistor - he's taking over 6kw for 10 secs on 0,024 ohms of 7 ohms - and thats just a 291th of the total "mass" thats a real monster of a power dissipator! 

3 minutes ago, tjcooper said:

ANYONE,

does someone know of a source for the large diameter heat shrink tubing?   By my calculations for the MoHoo, I will need one piece that is 11.5 inches in circumference and aother that is 13.5 inches in circumference.    I don't see anything at our local electronics supply houses over 3" in diameter (so 10 inches circumference).  Much thanks.

     tjcooper

@cranium found one of this to repack the ninebot battery pack - don't know if he is still reading here. Did not read anything from here since quite some time. Maybe he mentioned his source in his "5a fastcharging" thread, which i quoted a couple of posts bfore here (regarding the BMS controller ICs used in nowadays BMS)

Link to comment
Share on other sites

11 minutes ago, esaj said:

Do you mean you remove the BMS-PCB from the pack? I don't know of any other way to "turn it off"?

Ay - you overtook me with my reply ? I feared that once i saw the pop up, that a new repy from esaj is here...?

11 minutes ago, esaj said:

If you do figure out the schematic, please post it, I'd be interested to see it  :)

Yes, please @tjcooper- this would be very helpfull since your BMS could have some different approach! Could help some readers here to prevent a wrong disastrous shunt! Maybe @hobby16could edit this different schematics (if it really is ?) in his first post and his blog for all the others!

Link to comment
Share on other sites

  • 3 weeks later...

esaj, chriull,

I have sadly lead both of you astray by my "American Convention" comments.  By "power off" state, I mean that the BMS is not connected to the unicycle in any way (or to a charger).  To ohm out anything, I first test the DC voltage between two points.  If there is battery voltage there, then since the circuit is not ON (hopefully) I know that I do not have a direct connection between the two points  I do not attempt to "ohm meter" these two points......I now know they are not directly connected.  If the voltage is zero.....or very close to zero, I use my very low voltage ohm meter to see if there is continuity.

So I do not get a complete schematic.  I only know that the FET legs and gates are connected to neighboring points.....hopefully to the outputs of other driver FETs that will trigger the big FETs to go ON when connected to the unicycle.  I am just looking to see points have continuity to other points near the over current section.

I work at Amazon in cameras for tablets, etc.   We do a lot of work with Samsung.  I am hoping to use contacts to ask the Samsung people if I could get "overview" schematics of the power pack.  That would simply the process greatly.  Not sure if they would allow me to share schematic if I get it.

My main goal now is to defeat the "FET short" that does the P- to P+ for over current.  When I know more about the FET connections, I will try to do the shunt that everyone has discussed.  Assuming that does not destroy the battery pack, I will hook the battery pack up for charge.  If that does not blow up the BMS then I would hook system up to unicycle and test "wheel in air" running.  I would guess that could start up overcurrent very quickly if I grab the wheel abruptly while it is in free running spin.  This is the logic flow of how I think I could test the system.

Thanks for the pointers to heat shrink.  I will go down that path later this week.   I got the Ninebot to work for me and my son this weekend.  Finally was able to run for several blocks without loosing control.   I almost feel like a "rider" now.  Amazing how my 25 year old son picked up riding in about 5 minutes.   Took me several hours of practicing over 2 months.  So much thanks for all you advice and help.

    tjcooper

Link to comment
Share on other sites

  • 2 weeks later...

Hobby16, Chriull,

sorry I have been slow to fill in the details of the Samsung 16S1 battery pack.  I have done the voltmeter measurements without connection to EUC.

I want to refer to Hobby16's first image at start of the thread (shown below).  It shows the current sense going into the SOURCE of the T1 FET.   And it shows that the D1 diode also comes out of the SOURCE lead going into X1 FET.   In my BMS, all the FETs are T470 N channel mosfets.  My system has the current source resistor on the DRAIN part of T1 and the GATE going to the P- part of the system.   My system also has 3 diodes going from the C+ (and a dropping resistor) going to the P+ part of the circuit.

All of this seems very logical to me, the Samsung has put what Hobby16 has on the C+ part of the system on the C- part of my system.  Likewise, the B- has the current sensor resistor rather than Hobby16's circuit being on the P- circuit.  Both of these changes still have the T1 shunting the battery when the proper signal comes in.  Likewise the diodes are on the C+ part of the circuit rather than on the C- part.  But the X1 result is the same: when over voltage is detected, the X1 FET will shut down the charging circuit.

So assuming my analysis is correct, is the proper way to "put in the shunt" is to again connect B- to P-?   I think this takes T1 out of operation and the GATE and DRAIN will always be connected?  All 4 of my mosfets have their DRAIN all connected to P-.  I think this makes sense in that Samsung hooked everything up the the + part of the system rather than the circuit Hobby16 has with all active elements and resistors hooked to the - part of the system.

So I should take a 1 amp fuse on a wire and short B- to P-.   Assuming that does not blow the fuse, then I remove the fuse and hardwire B- to P- and then put the BMS back into the EUC and see if it spins the wheel without any smoke rising.  If that works I can take the EUC out for a ride?

I have measured voltages from SOURCE to GATE to DRAIN for both X1 and T1.  I will put those diagrams and voltages in the thread a little later.

I was just hoping the someone could comment on if my SHUNT METHOD seems reasonable before I try it.

   tjcooper

 

 

ScreenDump_Samsung.jpg

Link to comment
Share on other sites

23 minutes ago, tjcooper said:

Hobby16, Chriull,

if you type a @ and wait for the dropdownbox (or start the first letters of the name) and then choose the name from the dropdown the selected member gets a notification.

23 minutes ago, tjcooper said:

...My system has the current source resistor on the DRAIN part of T1 and the GATE going to the P- part of the system. 

I assume you meant Source instead of Gate. With the Gate on P- and assuming P- is the lowest voltage in the system T1 would never conduct...

23 minutes ago, tjcooper said:

 

  My system also has 3 diodes going from the C+ (and a dropping resistor) going to the P+ part of the circuit.

?Dropping resistor? Maybe a charge current sensor?

23 minutes ago, tjcooper said:

So assuming my analysis is correct, is the proper way to "put in the shunt" is to again connect B- to P-?   I think this takes T1 out of operation and the GATE and DRAIN will always be connected?

Source and Drain should be connected.

23 minutes ago, tjcooper said:

  All 4 of my mosfets have their DRAIN all connected to P-.

Source should be on P-?

 

23 minutes ago, tjcooper said:

I was just hoping the someone could comment on if my SHUNT METHOD seems reasonable before I try it.

Not for now... But i assume you just mixed up Drain, Gate Source and in the end it should work out?

 

Link to comment
Share on other sites

@Chriull @hobby16

much thanks for the tip on how to connect to forum people.

I have Nchannel mosfets of the T470 variety.  I looked up the data sheet and the diagram is:

    D

G D S     where the top D is the aluminum plate with the hole in it.  This is viewed from top of unit.  I think your description had that as the D as the Source???   Regardless of notation, I am using the diagram on the left to refer to Gate, Drain, and Source.

I have included a new diagram with more closeup of T1 and X1 mosfets.  You can also clearly see the 3 diodes going from C+ to P+ along with the "dropping resistor" I mentioned.  I measure that there is an 0.8 vDC difference in going from C+ to P+ when nothing on the BMS is hooked up.

You can see on the included photo that I have marked down two voltages which may help you understand the layout of the Samsung Power Pack.  Both are 15.1vDC and are present when pack is not connected to anything.  All of the metal plates of mosfets are all soldered in parallel and go to the P- point.  I call this the DRAIN point.   I have measured the voltage from P- to B- and there is a very small voltage (like 0.001 or less vDC) present.    So I am hoping I can hook the P- to B- with no problems and that will "remove the shunt" from my battery pack.

QUESTION:  on www.dhgate.com which is a marketing arm for Chinese manufacturers, they have a 2600maH 16S1 pack with Samsung INR cells for $72.13 delivered to USA.  This seems cheaper than I can buy the individual 18650 cells from Samsung for 2600maH cells.  Anybody have any experience with Jianglingo0309 out of Shenzen China?

This might be the cheapest way to find out if my "faceplants" came from the MoHoo Samsung Power Pack that I have.

Will see how long it takes to get delivered.

    tjcooper

OPPS...seems I have exceeded by 3.13 MB and need to remove some other files before I can import this one.

 

20160816_195621_drawing_voltagesSMALL.jpg

Link to comment
Share on other sites

1 hour ago, tjcooper said:

I have Nchannel mosfets of the T470 variety.  I looked up the data sheet and the diagram is:

    D

G D S     where the top D is the aluminum plate with the hole in it.  This is viewed from top of unit.  I think your description had that as the D as the Source???  Regardless of notation, I am using the diagram on the left to refer to Gate, Drain, and Source.

Thats correct. You just wrote in the post before:

23 hours ago, tjcooper said:

My system has the current source resistor on the DRAIN part of T1 and the GATE going to the P- part of the system

 

1 hour ago, tjcooper said:

You can see on the included photo that I have marked down two voltages which may help you understand the layout of the Samsung Power Pack.  Both are 15.1vDC and are present when pack is not connected to anything.  All of the metal plates of mosfets are all soldered in parallel and go to the P- point.  I call this the DRAIN point.  

The Drains of T1 are connected to the P- point, The sources are connected via the shunt resistors to B-. ?!

So its still like http://forum.electricunicycle.org/topic/459-electric-unicycles-bms-problem-and-solution/?do=findComment&comment=49928 with the additional secured knowledge of the Drains connected to P-.

Also the Gates of T1 have a sane voltage for switching the mosfets on, which they are, as you could not really measure any voltage over the current sense resistors and the Gate Source part of the Mosfet.

As written in my post of july 6 and 7 if this mosfets are really the AOT470 n Channel Mosfets (http://www.aosmd.com/pdfs/datasheet/AOT470.pdf) as it seems, they are used with a invers Drain Source "part". That's something i do not understand - i would assume its just utter stupidity. The mosfets T1 can like this never cut the power, because if the Mosfets switch off, the invers body diode will conduct and so the mosfets just burn....

So maybe this BMS is just a piece of junk or i am missing something - thats why i asked in june for a detailed schematics. Without that there will be no recommendation on how to shunt (also with the schematics i'm afraid that there will be not shunt recommendation - this bms and the cells should be presumably canceled...). You could go fine with shunting the Drain Source part (including the current sense resitors) or also not - i don't know.

 

1 hour ago, tjcooper said:

QUESTION:  on www.dhgate.com which is a marketing arm for Chinese manufacturers, they have a 2600maH 16S1 pack with Samsung INR cells for $72.13 delivered to USA.  This seems cheaper than I can buy the individual 18650 cells from Samsung for 2600maH cells.  Anybody have any experience with Jianglingo0309 out of Shenzen China?

Mostly they are so cheap because the use used battery cells.

1 hour ago, tjcooper said:

This might be the cheapest way to find out if my "faceplants" came from the MoHoo Samsung Power Pack that I have.

Like with Moho many bad power packs were reported.

So with ordering this from dhgate you have the chance to replace the bad with something comparable or worse...

As you know, be carefull with power packs which could have bad cells especially while charging!

1 hour ago, tjcooper said:

OPPS...seems I have exceeded by 3.13 MB and need to remove some other files before I can import this one.

For pictures its a nice way to upload them to imgur or something similar. If you paste here the "direct link" you get the pictures in your post just like you uploaded them to the forum server.

Link to comment
Share on other sites

@Chriull11

UPDATE:  put a 1 amp fuse on a wire and shorted P- to B-.  Nothing happened.  So I did a solder job on P- to B- and applied the charger to system. Had a temperature gun and nothing seemed strange.  The charge cycle took in 26 watthr (out of 132 for whole pack).  So it seems my shunt did not cause any problems.   Tomorrow I will mount system in a vice and spin the wheel.  I will monitor the heat in the system.  If all is stable I will grab the spinning wheel and jerk it to a halt.  That should be my max current drain.  I will see if gyro still works at that point.  I think that would be the acid test case that over current (from my halting the wheel) does not have the P+ to P- power removed.

   tjcooper

Link to comment
Share on other sites

4 minutes ago, tjcooper said:

Thats not me ;)

4 minutes ago, tjcooper said:

UPDATE:  put a 1 amp fuse on a wire and shorted P- to B-.  Nothing happened.  So I did a solder job on P- to B- and applied the charger to system. Had a temperature gun and nothing seemed strange.  The charge cycle took in 26 watthr (out of 132 for whole pack). 

Thats good, as the t1 part and the shunting should in no way affect the charging.

4 minutes ago, tjcooper said:

So it seems my shunt did not cause any problems.   Tomorrow I will mount system in a vice and spin the wheel.  I will monitor the heat in the system.  If all is stable I will grab the spinning wheel and jerk it to a halt.  That should be my max current drain. 

I'd assume the fuse will alredy blow once the wheel starts up - as the bms will now deliver nominally just ~65W. Just if the fuse is very very slow it could maybe survive this. But for sure not driving/accelerating/balancing.

imho the t1 will take over the power supply without interruption once the fuse is blown.

 

Link to comment
Share on other sites

@Chriull

I am surprised that there are 11 members of your family that all ride EUC.   LOL.

I am afraid I mislead you.  The fuse was just a short term test.  Once it did not blow the fuse I used a 12 gauge wire to

solder P- to B- permanently with no fuse.  I am assuming that no significant current goes through that 12 gauge wire since the B- is

always my "ground" lead from the battery.  If the P- terminal starts to raise up in voltage it would only be a "surge" from the wheel taking a spike in current from P+ and P- during some high current happening when I jamb the wheel or ask for huge acceleration.   Is that logic correct?

    tjcooper

Link to comment
Share on other sites

@Chriull

update: put MoHoo back together with shunted power pack.  Tested "spinning in the air" and no problems. Put on the ground and powered up.  Got gyro leveling.  Took induction DC ammeter and measured the P+ line.  Only draws .1 amps on level ground.  See the image which shows the MoHoo and the ammeter in operation.  When I really torqued down on the petals with my hands keeping MoHoo in the "stall" mode, I was able to get the current up to 0.5 amps.

QUESTION: what would be a typical guess as to current when a 100 KG male jumps on the petals and tries to move forward slowly?  Is this 2 amps?   or the start impulse is more like10 amps?  Any ideas?  This weekend I will repackage the battery and put everything back.  I screwed up and did not tape down battery without the cover on it.  It fell out of MoHoo as I was mounting it in vice the the B+ solder blob hit the aluminum petal.  Major arc and serious smoke.  But nothing got hurt.  I must remember to take every precaution when playing with that battery.  My Bad!!!!!!!

I will try out re-assembled MoHoo and see if I get any "sudden stops" before battery gets to 50%.   Full safety equipment and just run along a wall with no one around me.  Slightly scared to do this because of the 5 face plants that the MoHoo has already given me.

     tjcooper

20160817_231522.jpg

Link to comment
Share on other sites

3 hours ago, tjcooper said:

@Chriull

update: put MoHoo back together with shunted power pack.  Tested "spinning in the air" and no problems. Put on the ground and powered up.  Got gyro leveling.  Took induction DC ammeter and measured the P+ line.  Only draws .1 amps on level ground.  See the image which shows the MoHoo and the ammeter in operation.  When I really torqued down on the petals with my hands keeping MoHoo in the "stall" mode, I was able to get the current up to 0.5 amps.

QUESTION: what would be a typical guess as to current when a 100 KG male jumps on the petals and tries to move forward slowly?  Is this 2 amps?   or the start impulse is more like10 amps?  Any ideas?

Could be something in this area.

From my ninebot one e+ with a 500 W Motor and ~20km/h (real) top speed i have the folowing rough numbers: 

For driving and balancing ~200-300W so this would be with ~65V roughly 3-4A.

For startup, acceleration or going up inclines the max steady current is about 15A (~1kW). Max peak current i recorded was ~48A.

But the ninebot has 2 cells in parallel so it is able to deliver this currents.

With the MoHo (has a 350W motor and does not drive as fast with a "single cell" pack?) the the numbers should be quite lower...

3 hours ago, tjcooper said:

...I will try out re-assembled MoHoo and see if I get any "sudden stops" before battery gets to 50%.   Full safety equipment and just run along a wall with no one around me.  Slightly scared to do this because of the 5 face plants that the MoHoo has already given me.

As small (only always 1 cell in parallel) packs are quite at the limit to drive a wheel be carefull with the aging of the cells. Once one or more cells degrade it could get dangerous charging the pack (also the BMS should cut off charging at under/overvoltage of single cells - but one can't be sure if for example shut off with bad cells at undervoltage conditions...) and on the other side such bad cells can lead to overleans which feel almost like a shut-off, since the power cannot be delivered anymore...

Normally shunting is just a symptom cure and not the removal of the cause... ;( Imho most people around here in the thread shunted because they had bad battery cells. They can drive now without cut-off, but one has to be more carefull with the battery cells after shunting....

Link to comment
Share on other sites

@chruill

Much thanks for your numbers.  Had no idea 45 amps could be demanded.  Like the 10-15 amp range for typical max output.

Took unit to park tonight.  Moving along walls was effortless.  So I headed out on asphalt.  Rides very different from Ninebot.  Much more "squirrelly" and fast turning.  But no shut-downs or any hints of problems.  Only got 10 minutes of slow use.  The tire had gone low in 6 months and I could hear the system "groan" as I started to move so I quit before doing anything more than like 10% drain.  Filled up tire to 48 psi and will try again tomorrow.

QUESTION: with the shunt of P- to B- as described above, does this also mean that the charger would no longer stop giving charge if any of the FETs and ICs that are going to each cell detected that a specific cell had too low a voltage or too high a voltage?    I have a DYI Dr. Charger that I built for this machine which tells me volts, amps, power, and WattHr delivered during charge.  Hopefully if I have a weak cell I would see after a reasonable time that my voltage had not gone up to 67.2 vDC or that my WattHr were higher than my estimated discharge of the battery from a previous use.

Normally I would use my heat gun to see if any cells were getting hot during charge, but that means tearing out the battery and taking off the cover of PCB.  Maybe I will put in a couple of 10K thermistors in the 16S1 pack and seeing if any section shows higher than normal heat.  But I guess my safest route is to charge cells and watch the WattHr going in and final voltage.  Then take a precision voltmeter to touch each of the 16 cells and be sure they are all within 0.1 volts of each other when fully charged.  What do you think?

    tjcooper

Link to comment
Share on other sites

35 minutes ago, tjcooper said:

QUESTION: with the shunt of P- to B- as described above, does this also mean that the charger would no longer stop giving charge if any of the FETs and ICs that are going to each cell detected that a specific cell had too low a voltage or too high a voltage?  

This should only shunt the "overdischarge protection" Mosfets T1, but neither the "overvoltage protection" Mosfet X1 nor the invers voltage protection Diode. So whatever charge protection is realised within your BMS should stay intact.

35 minutes ago, tjcooper said:

  I have a DYI Dr. Charger that I built for this machine which tells me volts, amps, power, and WattHr delivered during charge.  Hopefully if I have a weak cell I would see after a reasonable time that my voltage had not gone up to 67.2 vDC or that my WattHr were higher than my estimated discharge of the battery from a previous use.

Normally I would use my heat gun to see if any cells were getting hot during charge, but that means tearing out the battery and taking off the cover of PCB.  Maybe I will put in a couple of 10K thermistors in the 16S1 pack and seeing if any section shows higher than normal heat.  But I guess my safest route is to charge cells and watch the WattHr going in and final voltage.  Then take a precision voltmeter to touch each of the 16 cells and be sure they are all within 0.1 volts of each other when fully charged.  What do you think?

    tjcooper

Measuring the individual cells from time to time should be the safest choice. Imho bad cells should be noticable from the charging/riding behaviour - but the battery packs in my wheels are too new, so i don't have any practical experience what will happen once they age and how or if this will be noticeable (soon enough...)

Link to comment
Share on other sites

1 hour ago, tjcooper said:

QUESTION: with the shunt of P- to B- as described above, does this also mean that the charger would no longer stop giving charge if any of the FETs and ICs that are going to each cell detected that a specific cell had too low a voltage or too high a voltage?    I have a DYI Dr. Charger that I built for this machine which tells me volts, amps, power, and WattHr delivered during charge.  Hopefully if I have a weak cell I would see after a reasonable time that my voltage had not gone up to 67.2 vDC or that my WattHr were higher than my estimated discharge of the battery from a previous use.

Normally I would use my heat gun to see if any cells were getting hot during charge, but that means tearing out the battery and taking off the cover of PCB.  Maybe I will put in a couple of 10K thermistors in the 16S1 pack and seeing if any section shows higher than normal heat.  But I guess my safest route is to charge cells and watch the WattHr going in and final voltage. 

I would very much doubt the BMS in a cheap wheel would monitor individual cells during either charge or discharge. It will most likely only react to pack voltage and current levels. The BMS has a very limited capacity to shunt current around individual cells that have already reached 4.2V so if one cell is low compared to all of the others the total voltage will probably still reach 67.2 as all the other cells will rise above 4.2v somewhat. This also means you will not see a significantly higher watt hour input. One cell is unlikely to be hotter than the others, more likely one BMS shunt component will be cooler. In practice, some cell failures can result in effective low capacity, so the faulty cell may well, instead, be the first to reach 4.2V. You might see the current flowing into the pack level off instead of continuing to fall however some BMS cannot shunt more than 50mA and it can take quite some time for a good battery to fall to that.

Cells that are beginning to fail do often only develop higher internal resistance so one may be significantly hotter on discharge, and lower voltage immediately after discharge. Left at rest faulty cells can sometimes recover to the same or even higher voltage than the others.

Best indication of cells beginning to fail is an abnormally high voltage sag under use. I.e. If you go up a particular hill at a particular point in your journey (preferably not much mileage after charging it) and it starts reporting much lower capacity or beeping at lower speed, that is the clearest indicator.

Link to comment
Share on other sites

@Chriull, @Keith

thanks for observations.  I will finish three 50% rides of the MoHoo this weekend (Vacation next to the ocean).  I will try to do a voltmeter measure of each of the 16 cells both BEFORE recharge and AFTER.   Of the two measurements done already, every cell is within 0.05 voltsDC of each other.  That will hopefully tell me if I have any weak cells.

QUESTION: I have a lot of background in RC airplane and helicopter LiPo packs and controllers.  We really don't have BMS in these systems, only just an under voltage and an over current tester.   So I wonder what all the microFETs and integrated circuits are that go to each of the 16 cells.  I know and use circuits in airplanes that shunt current from cell to cell to balance out the charging load, but those are just a single IC and single resistor to each cell.  There must be 15 resistors and FETs and ICs that belong to each cell on my BMS.  I just assumed that they also measured the discharge cycle and when either over current or under voltage (and maybe over temperature) was reached, it would send its signal to the gates of X1 and T1 to maybe "pulse" a warning to the mother board that something was wrong and warn rider to get off the wheel.   Is there no logic like that in these BMS units?   If not, it seems like there is a huge part count going to each cell that is really un-needed.   If someone has a schematic of what all those parts are that go to each cell, I would love to look at the circuit.  Much thanks.

   tjcooper

Link to comment
Share on other sites

1 hour ago, tjcooper said:

@Chriull, @Keith

QUESTION: I have a lot of background in RC airplane and helicopter LiPo packs and controllers.  We really don't have BMS in these systems, only just an under voltage and an over current tester.   So I wonder what all the microFETs and integrated circuits are that go to each of the 16 cells.  I know and use circuits in airplanes that shunt current from cell to cell to balance out the charging load, but those are just a single IC and single resistor to each cell.  There must be 15 resistors and FETs and ICs that belong to each cell on my BMS.  I just assumed that they also measured the discharge cycle and when either over current or under voltage (and maybe over temperature) was reached, it would send its signal to the gates of X1 and T1 to maybe "pulse" a warning to the mother board that something was wrong and warn rider to get off the wheel.   Is there no logic like that in these BMS units?   If not, it seems like there is a huge part count going to each cell that is really un-needed.   If someone has a schematic of what all those parts are that go to each cell, I would love to look at the circuit.  Much thanks.

   tjcooper

You will find a discussion and link to the spec sheet for one of the chips that are often used DS-HY2213) below: The only safe way to signal the  motherboard is with a signal wire or wires. 

pulsing the power, like Align Helicopter speed controllers do isn't really an option (it is fairly scary hovering low inverted and suddenly having the heli jumping up and down as well!) Have you taken a close look at the electronics in your LiPo charger as that is where most of the job of a BMS is handled. Only Discharge protection is, of course, handled by the ESC (Electronic Speed Controller) but on a pack, not individual cell basis.

 

 

Link to comment
Share on other sites

@keith

thanks for the two references.   The pulsing I was talking about is a few microseconds in duration and just varies the current power level.  I don't think the rider would ever be aware of the signaling.  My BMS (standard Samsung Power Pack of 16S1 with 134 WHr) has only charge cables and power cables going to mother board.  There are no signaling wires in the system.  So one would have to wonder how the motherboard can sense the lowered voltage and or cell deteriorated state and then signal the rider by beeping and changing the pedal orientation.  In all the face-plants that this unit has given me, the power level was never below 90% full and shutdowns occurred without any serious discharge of the battery happening at the same time.   With my recent tests today, the voltage was at 90% and my measurements show no weak cells at all.  So I would have to guess that the BMS has a method to signal the motherboard....but it could do that with sophisticated voltage and current monitoring independent of BMS.  I am planning on putting a new BMS I just acquired in parallel to the old one (only have about 1 hour of ride time on it).  This should provide enough power to keep system running smoothly without any power shutdown by a single BMS.  My biggest problem is finding the right magnetically attached 16 wire connector that will do a fast release if I have to jetison the second pack.  I am hoping I never need more than 10 amps from each cell in my secondary BMS to keep things running.  I have a pogo-pin pack of 8 bins that should work, but now I will need to join two sets of pogo-pins to get the job done.

 

I will review your chip reference and see if it matches any of the ICs on my BMS.  Much thanks.

   tjcooper

Link to comment
Share on other sites

9 hours ago, tjcooper said:

... My BMS (standard Samsung Power Pack of 16S1 with 134 WHr) has only charge cables and power cables going to mother board.  There are no signaling wires in the system.

Thats the normally used way with eWheels, except Firewheels.

9 hours ago, tjcooper said:

  So one would have to wonder how the motherboard can sense the lowered voltage and or cell deteriorated state and then signal the rider by beeping and changing the pedal orientation.

The motherboard cannot sense the cell deterioration state. It just measures the voltage and the current. The firmware could recognize the (overall) cell state by power sag behaviour under load. But that's imho going to far by now - we can be happy by now if the wheels firmware drives and balances right...

9 hours ago, tjcooper said:

  In all the face-plants that this unit has given me, the power level was never below 90% full and shutdowns occurred without any serious discharge of the battery happening at the same time.   With my recent tests today, the voltage was at 90% and my measurements show no weak cells at all.

With a BMS shutdown occured (restart only with reconnecting the battery or charging the wheel) there are three possible reasons: One cell got undervoltage, an overcurrent situation occured for too long or cell overtemperature (if the temp gets measured by your BMS).

From your reports, if i remember correctly it should not have been the overcurrent which leaves imho cell undervoltage as most probable reason. This can happen also with almost fully charged cells if their internal resistance is high enough, so the voltage sag is under load high enough to reach the low cell voltage threshold of the BMS.

So this could happen with quite normally behaving cells (having the same voltage as all the others, getting charged normally) but by their higher than normal internal resistance the keep the BMS shutting off under higher loads. Or they took from the beginning inproper cells with to high internal resistance for this task.

9 hours ago, tjcooper said:

 I am planning on putting a new BMS I just acquired in parallel to the old one (only have about 1 hour of ride time on it).  This should provide enough power to keep system running smoothly without any power shutdown by a single BMS.  My biggest problem is finding the right magnetically attached 16 wire connector that will do a fast release if I have to jetison the second pack.

Why a sixteen wire connector if you use a second BMS in parallel? Or you mean just putting the sixteen new cells in parallel with the existing ones and using no second BMS? (then you would need 17 wires...)

That would need quite some amount of copper going from the wheel to your new pack... ;(

Link to comment
Share on other sites

@Chriull

I mixed together two plans I have in formulation.  One is to simply use the new BMS which has same connectors as current Samsung BMS.  The second plan was to take 16 of the 18650 Sony/Samsung cells and wlre them in parallel to the current 16.  And yes, that makes 17 wires.  I assume no more than 10 amps comes from the set during normal ride that could "complement" the original batteries.  That way I could get by with 16 gauge silicon wire for this "backup \ power" task   So really very little copper is needed.  The important thing is the "quick release" connector(s).  Found some on Alibaba but nothing here is USA.  Still looking.  Another approach is to "Y" the two BMS together, but still I think each one could open at a poor time.

Tried putting MoHoo on a rotating 3" pipe to see if I could "fake" riding the system.  Hard to stop "overspeed" with it spinning too fast.  Will try a rotating exercise platform later this week.  Goal is to drive MoHoo under power until it is under 50% power (never been that low) and see if the cut-off from BMS happens.  That would spare me doing it with change of face-plant.

   tjcooper

Link to comment
Share on other sites

  • esaj unpinned this topic
  • esaj unfeatured this topic

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...