DIY Electric Car Forums banner

61 - 80 of 121 Posts

·
Registered
Joined
·
1,437 Posts
Discussion Starter #61 (Edited)
41.3 is too high for a bad pack with unknown CL voltages. How did it get so high?

What are the CELL voltages? Please measure and post, then put a load on it to drop the voltage down to 38, then see if it will work on the OEM charger.

i think the 3Ahr packs are a mirror image of the 2.6 and 5 packs. Here i have labelled all the CL (+). i think the FETs are reversed also, the Discharge FET should be the one closest to GND. Does this look right?
 

Attachments

·
Registered
Joined
·
6 Posts
Hi everyone

Battery Repaired !!!!

In my previous post , I informed you that the i2C signals was constantly produced in loop. I thought it was strange that the PIC dialogue constantly with the OZ8940. it consumes current.
I have a second battery which works perfectly. I open it and make measures.
On a good battery, the I2C dialog only lasts 3 seconds when you press the button to see the capacity.
I decided to monitor the I2C dialogue when you plug it on the charger.
When the battery is charging, the I2C dialogue is constant in loop.
And stop when you unplug the battery.

I supposed that my faulty battery thought it always connected to the charger.
that's why the I2C dialogue was in loop.


I unpluged the battery from the charger and made some measures.
Thanks to the diagram made by Kenny.

I measured the output U2_pin21
On the good battery i measured 3.3 V
On the bad battery, i measured 0 V
While T1 was disconnected and 0V measured on the input U2_pin 16.
OV was mesured on the base of Q1, so the transistor should be cut-off.
The transistor Q1 was defective. I replaced it, but the result was the same, except the voltage on the base of Q1 went to 0.6 V.

I decided to measure the T1 input.
On the goog battery , i measured 0V.
On the bad battery , i measured 1,8 V

The only way to have a voltage to the unconnected T1 contact, is a failure on Q3, or Q6 and Q7.
Replacing Q3, the voltage on T1 returned to the normal 0V, and output U2_pin21 became 3.3 V

Connected to the charger, the battery charges !!!

To summarize, when you have a defective battery:
- Make sure all cells are balanced, and more 3.2V
- You have to measure T1 contact. You have to measure 0V between T1 contact and (-) , otherwise you have a problem on the T1 circuit.

There may be other source of failure....
Very nice! I was doing more I2C diag... and I expect to receive a datasheet for the chip soon (O2 only sells to OEMs, but I happen to work for one). I replaced Q1 and Q3 with some 60v MOSFETs with 20v gates. I'm not surprised they went bad... the design is not good IMHO. The gate of Q1 can be exposed to half of the full battery voltage (max 21-22v, higher than most gate/base voltages), through the 100k/100k resistor divider. Ideally there will be a zener across R6, or R6 would be something less, like 50-80k. I will probably adjust this on my board.

When I get my charger back I'll test, but now T1 sticks at 0v. We'll see.

@Oric 1 what does the T1 signal look like when it's charging? Wondering how I can fake it for quicker testing.
 

·
Registered
Joined
·
1,437 Posts
Discussion Starter #63 (Edited)
Did you get my PM about the datasheet?

T1 in the charger sends out 5V and it monitors and controls that output to be around 1.5 V when a battery pack is connected. See post #2 on the first page for more details.
 

·
Registered
Joined
·
6 Posts
Did you get my PM about the datasheet?

T1 in the charger sends out 5V and it monitors and controls that output to be around 1.5 V when a battery pack is connected. See post #2 on the first page for more details.
I just replied. Has anyone done a schematic of the T1 circuit in the charger? I'm sure it applies that 5v through a resistor, since the battery can override it to 0v. My pack doesn't enable the CHG FET by default, but I'm thinking it wants to see a charger connection to do so. Or, I may still have a broken pack.

Edit: I should have re-read your post before asking. I'll just wait for my charger to test. Can anyone confirm if they can charge a pack through the B+/- terminals with nothing connected to T1? Such as via a benchtop power supply.
 

·
Registered
Joined
·
1,437 Posts
Discussion Starter #65
You can charge thru B+ and GND, which bypasses the FETs, using a bench top supply. But now YOU are the BMS, so don't walk away and let it overcharge.
 

·
Registered
Joined
·
6 Posts
You can charge thru B+ and GND, which bypasses the FETs, using a bench top supply. But now YOU are the BMS, so don't walk away and let it overcharge.
That's not what I'm asking; I know I can do that already. I'm wanting to test the BMS repairs (replacing Q1+3) before I get the actual charger back.

Edit: got the charger, still not fixed, despite T1 sitting at 0v after replacing Q1 and Q3. I may need to observe the behavior with a good battery (I'll have to find one) on T1, but after replacing Q1+3 the I2C traffic stopped being constant.

Edit #2: the T1 signal starts at 0v (before plug), rises to 2.5v (on plug) for ~3 seconds, then back to 0v (still on plug) and the charger errors (red+green blinking). Anyone able to do the same scope trace on a good battery?
 

·
Registered
Joined
·
1,437 Posts
Discussion Starter #67
Q1 and Q3 are npn BJT marked "1D", MMBTA42. Q2 is the pnp compliment marked "2D", MMBTA92.

When plugged into charger, the T1 line is regulated to 1.5 VDC by the charger. The base of Q3 will see half of that, 0.75 to turn ON.

Then the U2p16 can be used to test the status of the big Charge and DisC FETs, and U2p21 is used to detect the results of this test.

U2p16 goes Hi to make the test:
If FETs are OFF then the Q7 can't turn ON to pull T1 low, and p21 stays Lo.

If the FETs are On, then Q7 turns ON and pulls T1 low, and p21 goes Hi.

Here is a quick sketch of the T1 output of the charger:
 

Attachments

·
Registered
Joined
·
46 Posts
Very nice! I was doing more I2C diag........

@Oric 1 what does the T1 signal look like when it's charging? Wondering how I can fake it for quicker testing.
Sorry, i have no measurement on T1 when the battery is connected to the charger, but Kenny did a great job.

Otherwise, be careful with the connections on the I2C.
I wanted to monitor the I2C dialog when applying an electronic load to my 2,6 Ah battery, to perform the cells discharge curve.

When I disconnected the I2C, the battery is blocked. The SDA line is permanently kept at zero. The OZ8940 chip has been damaged.
Plan to use an I2C isolator.
 

·
Registered
Joined
·
6 Posts
Sorry, i have no measurement on T1 when the battery is connected to the charger, but Kenny did a great job.

Otherwise, be careful with the connections on the I2C.
I wanted to monitor the I2C dialog when applying an electronic load to my 2,6 Ah battery, to perform the cells discharge curve.

When I disconnected the I2C, the battery is blocked. The SDA line is permanently kept at zero. The OZ8940 chip has been damaged.
Plan to use an I2C isolator.
I'm using very high-impedance probes with a Picoscope to watch the I2C bus.

Kenny DID do a great job, that's for sure. I'll be back on it when I have a bit to play around with them again. Thanks Kenny.
 

·
Registered
Joined
·
46 Posts
I'm using very high-impedance probes with a Picoscope to watch the I2C bus.

Kenny DID do a great job, that's for sure. I'll be back on it when I have a bit to play around with them again. Thanks Kenny.
Incredible ! I use also a picoscope .



The problem wasn't the probe, but my arduino.

My electronic load is an Itech IT8512A

When i disconnected the I2C from battery to arduino, my computer make a sound, and i realized that the battery didn't work. The OZ8940's SDA line is locked to 0V. No answer when pushed jauge button
Maybe a ground problem between my electronic load, and the ground provided by USB to the arduino.
I remember reading an I2C problem on the OZ890. They recommend using an I2C optical isolator.
 

·
Registered
Joined
·
1,437 Posts
Discussion Starter #71
Sorry to hear that the OZ chip has blown, but maybe you can buy some from one of those online chip suppliers that show up when you do a datasheet search.

i was thinking your issue might be a ground level difference when first reading your post about this.

The picoscope has a USB, as does the arduino, and so does the Itech load? And does the USB ground connect to the mains ground? And the PicoScope BNC barrel is connected to the USB ground?

What about the Itech load path, is it isolated or connected to the mains ground?

When you connected the Itech load, was it to the battery terminals B+ and B-, or did you by-pass the FETs and connect to CL1- ?
 

·
Registered
Joined
·
46 Posts
Yes, i found IC on Aliexpress....
I also found an I2C isolator. Maybe useful for future use.

The Computer, the USB, the Arduino and the electronic load had the same ground.
Maybe a problem with the electronic load.
I don't know if the inputs are floating.
The inputs was connected directly to the battery on B+ and B-.

We have a lot of problem with this BMS PCB.
It could be interesting to redesign a complete BMS.
But it's a real projet, a big job.
Or find an Aliexpress compatible seller who agrees to sell only the BMS.
 

·
Registered
Joined
·
2 Posts
Wow guys, Great Work.

I've been waiting for some info on these 40v Ryobi BMS's. I wish I could contribute to the project. Im just a DIYer Electronics guy. I've recently purchased probably more then 80+ of these 40v Ryobi Batteries. About half of them can be repaired by charging and balancing the cells with a Cc/CV power Supply and cycling them a few times to make sure the BMS and Cells are functioning Properly. Out of all the ones I've been able to repair successfully and sell, just one buyer has contacted me with a problem after using it for a month.

the other half of the batteries, I disassemble and recover the 18650's that checkout after cycling and testing for self discharge for my DIY power-wall.

❓❓❓
The one question I have is, has anybody figured out how to charge Ryobi battery packs without a Ryobi charger? Without opening the packs? I use Ryobi 18v cordless tools at work and as well as these 40v packs for diy projects and only have 1 charger for each.

What is the T1 and or T2 terminals on the BMS's looking FOR? Is it possible to use a resistor or a small voltage input on T1 or T2 to get the BMS to allow Charging on the Batt + and Batt - terminals?
❓❓❓

Unbelievable that These 40v Ryobi Batteries Sell for upto $160 plus and how often they fail because of there poorly designed BMS!!!! Tens of thousands of the 2.4Ah packs have been sold for just $6 until they went out of stock just last week... I've been selling them for $30ea or 4 for $100 and I can fix them fast enough.
 

·
Registered
Joined
·
1,437 Posts
Discussion Starter #74
Post #67 has the T1 schematic inside the charger, and post #2 has the T1 circuit inside the battery.

T1 is a Test Terminal between the battery and charger, where either side has a way to test the other, and either side can shut itself off if conditions are not satisfied. There are processors in both and the firmware codes are not known, so we don't know how often this test is run during charging.

There is no safe way to bypass the Test without opening the cover that i have discovered.
 

·
Registered
Joined
·
46 Posts
As mentioned by Kenny, it exist a logic between the charger and the battery.
The T2 terminal is not used by the charger. On the 2,6 Ah battery, this pin is not connected. All is made by the T1 terminal.
I think in normal condition, it's the BMS's battery that stop the charge.
It's not easy to measure the T1 signal while connecting on the charger.

The Battery know his capacity. I recently changed the cells of a defective 2,6 Ah. A 2,6 Ah battery is composed of 20 cells 1,3 Ah assembled in 10S2P.
I replaced them by 2,5 Ah cells.
After 1h charge, the charger starts to blink. I need to unplug and replug the battery to finish the charge.

You are lucky to have found batteries so inexpensive.
Here in France, the 2,6 Ah are sell 150 $, and the 5 Ah 200 $
 

·
Registered
Joined
·
2 Posts
@kennybobby

@Oric 1

Thx for you quick replies. So im assuming the the same problem would also be true on the 18v battery packs also? With making a diy charger instead of having opening them... They also have T1 terminal ...

I Have a couple of 18v Chargers that got wet, and the the DC-DC switching IC "TOP256EN" blew in both of them... in one of the Chargers I replaced the IC with a new one but when I plug it in it makes it strange noise for a second but still will not charge any batteries. I'm good at repairing pcbs so I know I replaced it correctly but something else must be wrong that is not noticeable


@Oric 1 - yeah that's a little more expensive than we pay here in the US for new batteries. I don't understand how they're able to charge so much money and people are willing to pay when these batteries fail so often. just that one business was able to sell tens of thousands of these defective 40 volt batteries. thankfully I was able to make a bunch of money since they were only $6 each and like I said half of them were repairable and the other half I extract the sales and use them myself. I'm waiting for more to become available. if shipping wasn't so expensive it would almost be worth buying them and having them shipped overseas to you for $6 a piece you can't beat that
 

·
Registered
Joined
·
46 Posts
I don't understand how Ryobi can sell products that are so unreliable.
Just a look on ebay to see a lot of defective batteries
 

·
Registered
Joined
·
1,437 Posts
Discussion Starter #78
It's all about the jack($)

Oh they are reliable--for Ryobi.

The patented BMS will shutdown to avoid any situation that might result in a fire and the legal claims to follow. It reliably saves them money.

Failed packs cost the users much money if they want to keep using the tools. It reliably brings them lots of money.

The needless waste of resources is offensive to the DIY spirit.
 

·
Registered
Joined
·
46 Posts
Wow guys, Great Work................
❓❓❓
The one question I have is, has anybody figured out how to charge Ryobi battery packs without a Ryobi charger? Without opening the packs? I use Ryobi 18v cordless tools at work and as well as these 40v packs for diy projects and only have 1 charger for each.

What is the T1 and or T2 terminals on the BMS's looking FOR? Is it possible to use a resistor or a small voltage input on T1 or T2 to get the BMS to allow Charging on the Batt + and Batt - terminals?
.........
One more difficult with Ryobi : :mad:

A scope probe on T1 terminal while charging
It exist a dialog between the charger and the battery.
I don't know which way this dialog takes place.





10 sec before end of charge
 

·
Registered
Joined
·
1,437 Posts
Discussion Starter #80
Well that is a very interesting scope capture, and explains why the T1 terminal connection is labelled "COM" on the charger board.

i'm going to guess that the communication primarily originates in the charger due to the blocking diode D10 on the "COM" line shown in the hand-drawn schematic. That line gets controlled by signals from the IC3, "ABOV2011" pin23 which drives the base of Q9. So it can assert both Hi and Lo states.

On the battery side, T1 can only get pulled Low by the action of pin 16 of the PIC; it cannot assert a Hi. Maybe it pulls Low to assert a marking pulse or RTR (ready to receive) back to the charger?

In either case it appears that pin 21 of the PIC is the intended recipient of the message.

i wonder if these signals are status or commands somehow used for voltage or current control?

i can't find any info or datasheet for an ABOV microcontroller chip with that marking. We would need the firmware from that chip, plus the PIC firmware to really understand.
 
61 - 80 of 121 Posts
Top