Ryobi clone battery with odd circuitry
Вставка
- Опубліковано 12 вер 2024
- At around £40 for two of these "5Ah" Ryobi compatible batteries I was expecting them to be fairly low capacity. In reality they're pretty close to the claimed capacity (80%) but the circuitry they use for monitoring the cell voltage is unusual, and doesn't give me confidence in leaving them charging unattended.
In the video I mentioned the potential for the 100 ohm balancing resistors to have a higher voltage across them at full charge current. Theoretically the current would need to drop to around 40mA for the resistors to safely pass the current without the risk of the cell being charged above 4.2V.
Although the cell balancing seems like a good feature, I'd have been a lot more comfortable with these batteries if they had used the simpler single chip unbalanced management chip as used in the cheaper tools. It would decisively detect any cell reaching full capacity and cut the charging current off.
The way the circuitry senses the voltage across the pack gave me doubts about its ability to detect a single cell going into the bypass mode. Maybe I'm just being pessimistic, but there is a lot of weird circuitry involved. All the divider resistors are critical values and even the tiniest manufacturing defect of a missing or incorrect resistor value could have significant consequences.
I spent a HUGE amount of time trying to get my head around the way the circuitry worked. To properly reverse engineer the circuit would require large images, removal of components and lots of time.
On a plus note they seem to have a decent cluster of 18650 cells. The temptation is to reconfigure them as a large parallel group as a chunky 20,000mAh power bank.
If you enjoy these videos you can help support the channel with a dollar for coffee, cookies and random gadgets for disassembly at:-
www.bigclive.co...
This also keeps the channel independent of UA-cam's advertising algorithms allowing it to be a bit more dangerous and naughty.
#ElectronicsCreators
My counterfeit Ridgid/AEG batteries have a really neat feature - when you get to near the end of the charge, it self destructs and blows a big puff of smoke so you know not to keep using the battery! Ingenious!
Well, 2/3 puff smoke. The other one is full of sand... to act as a fire block.
I love this BMS configurtation! There's nothing "random" about those resistors, and in fact this BMS is exactly why I *would* trust this BMS for both safety and cell longevity, compared to one that doesn't do balanced cell charging (the original Ryobi it almost copies, did you say they had a recall due to battery failure?? that's almost certainly why!) I've had significant experience with long series strings of batteries/cells, from itty bitty ones, to truck-battery-sized in grid-connect scenarios, and cell balancing and monitoring is critical to both cell longevity and safety - they're two sides of the same coin. When you recharge, if one cell is at a different state-of-charge (SoC) than its peers, let's say it's at a slightly higher SoC, then it will reach 100% charged before its peers. Which means it will become over-charged when its peers reach 100%. Or if a cell is at a lower SoC than its peers, it will become *over*-discharged once its peers approach 0% charge. It's a matter of "policy" (and warranty obligation!) in a BMS as to what to do in these two scenarios.
In any set of series-connected cells, over time they will "diverge" from each other in their state of charge (SoC), unless you specifically act to keep them at the same SoC, either by stopping recharge of one cell and let the others 'catch up', or even place an extra load on one cell to bring it back down to the same SoC as its peers. There are inherent differences in cells/monoblocks/batteries due to manufacturing variances, and even in environmental differences; for example, if there's localised heating experienced by cells at one end of the string but not near the other end, then those heated cells will diverge in their SoC more than the rest of them.
That circuit across each cell is in two halves: first half is the P-ch mosfet & 'random' resistor that puts a small nominal - but to the point, known & predictable - load on each cell, when it's instructed by the MCU to do so using the N-ch mosfet.
Unlike that cell balance chip that's stacked one above the over (they don't care about the ever-higher voltage up the string, because their 'ground' is virtual, referenced only to the cell it's connected to), this two-mosfet duplicated circuitry has each instance *referenced to ground* (so that the MCU can control and V-measure each one), so the load resistor, relative to ground, needs to be different for each cell, because each cell up the stack is 3.0-4.2V higher than the one below it, when measuring relative to the Gnd of the 'bottom' cell. The P-ch mosfet, on the high-side turns on the load, and it's in turn turned on by the N-ch mosfet & V-divider, which also has a series if different R values so that each cell can be measured 'ideantically' to the others, regardless of where along the string it is.
I suspect that connection back from the power-tool down into the battery pack might be to tell the BMS MCU that the user has just pushed the trgger and to stop doing any BMS operations.
P.S. You also shouldn't "float charge" Li-Ion cells, unlike all older chemistries. You simply have to stop charging once you reach cell max-V. Attemptiong to 'float-charge' Li-Ion cells is one of may paths to premature cell failure (spectacularly or not).
Yup, seems like a reasonable BMS. Balance charging, individual cell monitoring, and a way to completely isolate the batteries from charger or tool
As I mention in another comment, I'm pretty sure the BMS (or cells themselves) will reduce charging current prior to the cells hitting the 4.2V threshold of the balance ICs, so the balancing resistors don't have to dissipate the full charge current. So rather than dissipating a potential 25W at 500mA charge, it will be more like 180mW with 100 ohms across 4.2V. Maybe a bit more if there is still a small charge current applied, but that resistor is only enabled until the cell voltage drops and the IC hysteresis turns the balancing load off. I'm sure the designers tested this pretty thoroughly, including with fault conditions like a shorted cell in the pack (detected and handled by the BMS presumably).
@@KeanM The batteries do that themselves, though, as they near the charge voltage, they pull less and less current. If it switches to the resistor shunt abruptly, I wonder if there would be any weird current spikes.
@@GadgetBoy Agreed. I realised this as I wrote the above and thus included "or cells themselves". Of course a faulty cell or buggy BMS may behave differently.
I'm checking the top comments before actually watching the video for one.
This is another reason I love Big Clives channel - you'll always find insightful and knowledgeable comments from people who have greater insight than a mere layman enthusiast like me has. Thank you for sharing your knowledge! ♥ You have successfully explained to me something I was previously ignorant of.
I love that you've been going back to your older videos to reply to more recent comments.
I get a comment feed that will let me see new comments on any video. I reply to as many as possible.
I'm incredibly hungover, and Clive's voice is the only thing helping.
He has got that soothing tone that I wish was compulsory for A+E doctors to have. :)
it's so good to know how such a 'simple' problem of charging and discharging a cell can be solved in so many different ways! Keep up the investigations going Clive You Rule :)
@@BrianG61UK it does, however show how this approach is executed. That it is suboptimal is another point all together. It can serve as a memento how not to necessarily do such designs, ,thus the educational value of the video is still very positive!
I have been using 4 of these for about the last three years ,I am a builder so they get a bloody good workout , they go great .
So, I'm actually really familiar with battery management, circuitry, and copy/clones. (I've researched an industry wide issue in this market for years)
This battery is actually really impressive. For a clone battery, a group of people somewhere in a Chinese factory deserve a gold star for this one.
There's a possibility that the manufacturer of this battery may provide similar packs to the prominent brands.
I agree. It is far more effort and should perform better than an unbalanced battery for sure.
@@bigclivedotcom I think that's a reasonable conclusion.
It's also reasonable to think that the top of the IC may have been ground off to remove the proprietary identification that would show that the IC shouldn't have made it into a clone battery.
A good example is the SN27545 in an iPhone battery. It's essentially a BQ27545 with some extras in firmware. But the SN27545 is not sold to anyone other than Apple.
If you got your hands on the SN variant, but didn't want anyone to know, a good way to hide what you have done is to remove the top layer of the IC entirely. If anyone asks, it's a BQ.. Depending on the country, something as simple as that could be the difference between a "counterfeit" product and a "compatible" product.
Not even the original manufacturer can tell the difference.
@@BrianG61UK right, but when inspected, one of them is a breach of NDA/IP laws, the other is not
Very interesting! Could you do a genuine Ryobi battery to compare and contrast?
I tried to reverse enginer one and became hopelessly confused
For most people who don’t know or care what’s inside it’ll do the job they expect it to do with no problems. I think sometimes a little knowledge can be a dangerous thing, but we like to know how things work, and inevitably that’s going to lead us to question the decisions that the designers and manufacturers make, with good reason sometimes! Keep up the good work Clive! 👍
Oxygen is a great album.
@@UberAlphaSirus lol... I get it... good one! =]
A side by side comparison of this circuit and the genuine Riobi battery would be interesting.
"The temptation is to reconfigure them as a large parallel group as a chunky 20,000mAh power bank." This would be a great project video!
I bit into the Ryobi line in 2015, All the tools still work great after the first 3 years of heavy usage in restaurant maintenance. I started with 6 OEM batteries, again with heavy usage drilling 1/2" [13mm] holes in concrete, running a 4" [100mm] circular saw that draws high current on the batteries wearing them out. Finally some of the batteries started to drop out with the charger light flashing red [bad battery] so I figured a cell or two had given out in those batteries. I replaced the cells only to find out after all that work that it is the electronics that is the common failure.
The OEM units are just as complicated if not more and I have come to the conclusion that it is better just to replace the entire battery than screw with it and waste time and effort. Scavenger the good cells and trash the rest. A 2.6Ah OEM is around $50, so if you can get this 4.0ah for $20-$30 and it gives service like the OEM then it is well worth it to try the generic ones off of E-Bay.
Edit: Some of the failures was a green lime scale that seemed to leak from the outside via that sensing tab that's away from the power snout and near that rubbed off IC chip.
Interesting, I have a whole bunch of 18v Ridgid stuff dating to 2014, mostly the same batteries. (certainly the same cells inside, made at the same place)
I have yet to kill a single battery out of like 12 or more. Not for lack of trying. Some of the oldest ones only hold about 75% of the power they used to, but otherwise work fine. That's 7 years at least!
I really beat on my tools, killed an impact and a drill and a saw in that time. No batteries harmed!
As I said "...it is the electronics that is the common failure...."
The actual cells held up to the high current usage. I wish there was a way to install a simplified high current 5S BMS board into the battery packs and have them still charge in the factory charger, I have three of those. It helps to have multiple chargers when you have a big job and are burning through battery charge cycles. Example: I had to install FRP on all the walls in the back office/kitchen/prep areas using the circular saw to make the cuts. Another: All 10 restaurants needed several trash cans bolted to the concrete with 3/8 lag bolts, three per can using the hammer drill and a 1/2 inch mason bit. As I recall on the last store I was able to use an extension cord with a 120V hammer drill...Much better. :-)
@@DrHarryT I think it's the ability to not outrun your charge that has made the tools so good in the last few years. I remember when it took 4 hours and you could drill like 3 holes before it was dead! I keep 4 chargers setup with extra fans so I can charge multiples and keep them cool. They're making 6 pack battery chargers now but the stupid things only charge 1 battery at at time....
I wonder if the Ryobi batteries are made to be a little more fussy on purpose? The Ridgid batteries have a lifetime warranty, so they don't want them to break, even though it's basically the same thing with different plastic.
I've been using this exact model of battery for around 3 years now. It lasts for a very long time, even with the RyobiOne hoover.
Yesterday, I started building big pergola in the garden. I used this battery from a full charge. By the time I finished the project today, the battery was down to 2 bars on the LED. That was around 200 decking grade screws.
Are you using the big brushless drill driver that lights up like an alien spaceship? That thing lasts an obscenely long time on one full 5Ah battery if you're driving anything less than about 8mm screws.
The voltage sensing circuitry is very clever! That 100 ohm resistor after the P-Channel Fet is so that the voltage sense is not just the float charge, it loads the battery a bit that way you get a more reliable voltage sensing. Otherwise you'd measure the open circuit float voltage.
Nothing wrong with the BMS setup....it "changes" the resistance as it gets toward the end of charging. It's basically an opamp at that point, but I guess the manufacturer got a deal on the circuits rather than having a bunch of opamp chips everywhere.
That's incredibly cool, I've got a half knackered old Hitachi which still have fantastic batteries but you can power so much more with these Ryobi packs.
We did not need to know you have a rechargeable Hitachi "magic wand" and we needed to know even less that you have.... "half-knackered" it. :3
Hi Clive, I decided to take the easier option and replace the cells in my old Ryobi Li-Ion packs, using cells from Lidl, their 2Ah 20/18V packs which were on clearance at £7.50 each, normally £12.
2 packs are great for converting old 2.4Ah Ryobi packs, upgrading them to 4Ah👍
You do require nickel strip and a spot welder to do the job...
My old Ryobi Li-Ion packs are from 2007 and eventually died.
I'm also going to convert the old 18V Ni-Cad packs to 2Ah using the same cells but also converting the old Ni-Cad chargers using the contents of the Lidl chargers, they arent compatible with the new Ryobi chargers because the old Ni-Cad packs don't have the Ryobi Li-Ion charge circuit, so I'm using Lidl in both the chargers and battery packs. When I get time to do it....
I like the way the number has been ground off. The Chinese like to copy stuff but don't want their copy, copied.
In "The Hardware Hacker" by Andrew 'bunnie' Huang, he explains the Chinese mindset a little bit - anything you get a schematic of is basically 'up for grabs', modification, re-production, and the like. But not for free, they have a honour-based system of sharing and paying in kind or in favours.
It’s a distinct possibility that the reason is not to discourage copiers, but to make it impossible for people - say Ryobi’s lawyers - to find out where they’re sourcing those chips (and presumably shutting that route down).
@@JasperJanssen this was precisely my thought.
@@JasperJanssen but then they could just rub off the Ryobi on the board
Great video. I had fun with an aftermarket 18v Bosch battery a few days ago. I was testing an old 18v planer, and as I did the battery got very hot and burnt the palm of my hand. About ten mins later I checked the temperature with my thermal camera, and that 15 cell NiCd battery at the hottest point was still 184 degrees Fahrenheit. The circuitry in that was no PCB, and looked like just a couple of resistors.
G'day from Australia Dear Clive, regarding that Micro controller FYI the IC is (PIC16F1936 I/SS) on Ryobi Batteries, which is equivalent to Clone IC (HA 1839 I/SS) on Clone batteries.
That was really fascinating, I really do not understand such complicated circuits at all, I'm still at the very beginning, but every time you post videos like these, it makes me a little less scared of working on something harder. I'm not afraid of failure, I'm afraid of not trying. Cheers for the share!
In discrete electronics it's not "I'm not afraid of failure, I'm afraid of not trying" it's "I'm not afraid of something blowing up, I'm afraid of it taking my oscilloscope with it! " :P
I had the same thoughts as techydude. Wow, Clive, you are getting my brain going again. Great work. Thank you for all of the work that you put into your videos. I look forward to them so much.
Clive, you are one of the best reverse engineers on this planet. Love your channel, and your sense of humor .
This is perfect timing for me, having recently purchased some Ryobi tools. I've been tempted to buy some faulty battery's and actually replace the cells
As techy dude says, it's the right way to do it. Take a laptop battery apart you'll find much the same. Lithium cells are not my favourite but I've been working with them the last two years and the make me so nervous
I can't tell if it's the accent, the voice itself, or a combination of the two that make watching these videos so easy and enjoyable.
Not really much of an ASMR person myself but I can totally understand how some people get into certain things that put them in a state of calm and relaxation.
I want a Clive Mitchell voice on my GPS to calm me down while driving.
I've got one of those sitting around. Another problem with it is that the ABS(?) plastic embrittles much faster than I would have assumed. Without direct prolonged sunlight exposure too. The case shattered in multiple places after three years of knocking around in a tool bag.
Battery charging, at least with the one I was using, worked okay. Albeit with the batteries still being okay at it's end of life. So I don't have much input on how it'd fail in practice.
Now I'll have to take it apart again and check the board to see if it's the same or if there are interesting differences.
I have 4 -4AhRoybi batteries that just decided to stop working (not the same day, or same charger) I was hoping I could see what went wrong but never could tell. For the cost of the tool, I have been impressed with their tools for home use. I have definitely put them to the test. The batteries are what cost so much.
I picked up a 40v Ryobi lawnmower with two 40v defective batteries, although they were charging but would not last more than 2 two mins. The batteries are 10S, 20 18650 cells. The BMS looks very similar with similar mosfets on side with heat sinks and 11 soldering points for 10S battery balancing. In each pack, two of the cells at similar location were under charged, which I charged individually to the correct voltage and the pack came to life and recycled a couple of times.
As you mentioned, I think there is a problem with the bms logic not balancing the batteries properly.
The little contact on the battery (not on the three pin stem) is indeed a pin for communication with the charger and tool. The charger can receive information from the battery (that’s how the charger sees high temp or defective battery). A tool can also use that pin to demand a kick (short burst of slightly higher current) which is useful for say if a wood cutting tool gets a little stuck and needs a little more power to make it through.
6:48 I think you made a bit of a error with the HY2213 balancer IC, Its only going to have a max of 4.2 V over the 100 OHM resistor. So 42mA and 176mW. It does not have the full charge current passing through the resistor, most of it is still passing through the cell. The IC just gently corrects any slight over voltage.
Hi Clive, balancing at top of charge is a good thing. Called passive balancing. The 100 Ohm resistors are high, 10 Ohms would be better but the need a higher Wattage rating. Using the different dividers to reference each cell to 0V saves them from placing a Mix / sample and hold circuit, but is limited by the accuracy and resolution of the ADC.
So it's probably not awful, but has been designed to a price.
Strange circuit. Have you measured the current at which the pack is charged?
If the mosfet in series with the 100 ohm resistor is switched on, a current of 4.2 V/100 ohm = 42 mA will flow. By itself it is a good idea to measure a cell under a low load, but the remainder of the charge current will still flow through the cell, unless they momentarily switch off charging. So I don't think they will switch on the mosfet to let the other cells come up to the same level.
I think that as soon as one of the cells has reached 4.2 V, all the charging will stop. They may measure all the cells one by one to see how well they are charged. If one has reached 4.2 V, but the others don't, they may discharge the cell with 4.2 V. The 100 ohm resistor will consume (U^2/R) 0.176 W. When the voltage of the cell has dropped enough (probably 3.8 V), the mosfet is switched off, and all the cells are charged again, until all cells are charged.
Not sure if this is how it's done, it's just my idea.
You are on the right track. Assuming a HT2213-BB3A was used, they will turn the MOSFET on when the cell voltage exceeds 4.2V +/- 0.025V and turn off again once the cell drops to 4.19V +/-0.035V. So it will only discharge the cell a very small amount to keep it safe. Charge current at this stage (i.e. when all the cells approach 4.2V) will have dropped significantly from the max charge current.
The BMS chip would be monitoring individual cell voltages via all those individually controlled resistor dividers and if any cells are not charging properly, or are shorted, it would disable the pack via the back to back power MOSFETs. The BMS doesn't seem to monitor for, or be able to control, the outputs of the HY2213 chips.
Usually these BMSs have the balancing and overvoltage protection separate. So when any cell reaches 4.2v, the balancing IC will enable the Mosfet and start draining that cell, while the others continue to charge (more on the charging current later). If this cell is so out of balance from the others, its voltage will continue to go up, and only when it reaches 4.25v or 4.3v will the main MCU disable the charging mosfet in order to protect the pack. The balancing IC will continue on, and thus that overcharged cell will eventually set to 4.18v or lower, when the balancing stops. This setpoint often is enough for the the main MCU to re-enable the charging Mosfet and so it continues.
Now about charging current: any CC/CV charger will have the charging voltage same as the max pack voltage. When battery is depleted, voltage difference between pack and charged is high, and with low internal resistance it would draw a lot of current, hence the charger limits. However, when the pack is approaching the fully charged voltage, the difference between charger voltage and pack voltage is now minimal, and the charging current will also lower significantly. If all of the series cells are like 4.15v and only one is already at 4.2v, the charging current will already be very small, to the point that maybe only the 42mA balancing current is enough to hold that cell state of charge so that the other cells will catch up. In the very end of charge cycle you will have only tens of mA going into the pack
@@danilolattaro What you say is completely correct. A poor man's single-cell Li-Ion battery charger is a simple lab power supply with a current limit. Set the current limit to whatever the charge current limit may be (usually ≤ 0.5C) and the voltage to be the battery limit (usually 4.200V). Connect to the battery, and you will see the power supply go into current limit (CC). Once the battery voltage hits 4.200V, the current will slowly begin to taper down to 0, and the supply's voltage limit takes over (CV). That is pretty much what a low cost charger chip does, although it adds various safety items, too, such as a timer.
I've used these batteries for several years and haven't had a problem. The only problem is that they are no longer being sold in the US. I need to find a new company to work with. I'm glad they are that robust.
We purchased a deep tissue massage gun awhile back. After only a few uses it no longer would charge or display battery charge level. It was replaced under its warranty but they didn't want its return. Now left with two such guns and only so mush tissue to deeply massage I decided to crack open the faulty device. In its handle were 6-18650 cells, in series. Measuring pack voltage at the output plug I only saw a little over 10 volts. Wrapped up alongside the cells was a similar charge/discharge board as the one in this video, including the thermistor. Part of its connection were metal taps from groups of cells running to the board, I concluded for balancing. Ultimately wanting to check all cells I found two completely dead, the rest OK. Not sure exactly where the fault(s) lay I took the pack apart to use the cells in other projects. I was only able to find a like pack from Alibaba where I sent a bid in for one. Never heard back but was able to find the exact individual cells sold through a lithium-ion retailer and at a good price. Soldered 6 of them together, using the original connector. With no protection or balancing circuitry I charge it cautiously, watching charge current taper to hundredths of an amp. Perhaps a bit extra work to get it up and running but better than throwing it in the bin. Fun project.
The boy stood on the burning deck, playing a game of cricket . The ball flew off the quarter mast and stumped his middle wicket ! BOOM !
I actually get the feeling that this cloan battery is from a real supplier of Robi, using there custom chip (the reason it’s ground down) and board “ROBI18B”. Is there any one with a “real” ROBI whom can pop the cover of there battery? And at the end you told about the recall,… well that mite explain this cloan!
oh boy i have found my lost brother
The chips are common and plentiful
There is a Multimillion dollar bussiness in selling
Clone tool batteries most of the circuit boards are silkscreened with the branding they are made for.
Their custom chip... _their_ battery
@@drkastenbrot 👍 😂
The middle "temperature" sensor I think was the only feedback for the Nicad battery packs. The ryobi lithiums have the extra pins which allows backward compatibility with nicads when using the new chargers. At least here in Canada the battery chargers swing both ways. They are quite impressive.
It was. It was a thermal switch that cut off the charger until it was reset.
@@someguy2741 usually that means the circuitry is in the tool/product and charger instead.
Clive it looks too complex and overdone. I bought the small hand saw you covered and it’s amazing. Got two batteries and cut threw 4x4 without any issues…. Keep these coming beardy man I love it.
A similar teardown of an actual Ryobi charger would be interesting to see. I find that, while there are loads of '18V' tools about, there seems not to be a great consensus on how the charging works. Nor on the balance (pun not intended) between circuitry in the battery and in the charger.
Agreed. I've access to 24-30v dc supply from solar PV & would love to understand or if I could provide a 'green' dc power input into the Ryobi or clone charger to avoid up conversion to AC. Currently using trickle charge with Vmax cutoff of 21.0Vdc into the spike power terminals. Are there any other options?
I think the logic behind the balancing circuit is supposed to be that each cell is supplied with, say, 4V for charging via the V1, V2, etc. shown on the schematic, and when it hits its fully charged voltage, the resistor is brought in to allow the excess few 10ths of a volt to be drawn off so the cell does not overcharge. These seem to be targeted towards LiFePO4 cells, which would provide a simple method of balancing. If they're being done with standard cells all being charged in series that might be a spontaneous home combustion event. I don't think that would work well since the resistor is only meant to just draw away a small excess voltage from a dedicated voltage source for charging, not bypass the full charge current. Edit: If they are being charged in series, then there likely is no balancing while charging due to the low cell impedance. It will still pass the bulk of the current and just overcharge. Maybe after its taken off then it might stay on to draw the cell voltage down to nominal, but obviously if they are normal Li-Ion cells that's not too swift. Although seeing the extra circuitry for the MCU implies that they may be doing it properly.
That thing is surprisingly similar to a real Ryobi 5Ah pack. Similar construction, the connectors and MOSFET setup are exactly the same
That IC might be an Ablic 4-cell battery protection IC. The balancing circuitry seems standard ones we use (although it’s combined in some BMS IC). This is a proper BMS it seems.
I believe the strategy of the balancing ships is shunt top-balancing of the battery after it has reached full charge and/or while the battery is being discharged. I believe the 100 Ohm resistors are not used to bypass charging current. I had a Nissan LEAF EV where it was possible to monitor real time analytics via the OBD-II port, via an app called LEAF Spy. Each cell voltage was displayed separately as a blue bar in a chart with cells being actively balanced shown in red. Cells were actively being balanced while the car was being driven with bars variously turning red and blue.
The controller is taking a first cut at estimating battery health via the measuring taps and it can see something catastrophic happening if the voltage is too far off while charging. The controller also has the thermistor to assure the pack temperature is neither too high nor too low. I would have preferred to see at least one more thermistor. The double MOSFETs are clever, but it seems there should be a fusible link somewhere too. I’d give this battery a grade of B+, but I am not sure I would trust it. For a very impressive battery, the OEM pack from a Dyson cordless sweeper is a great example of good engineering, though it is a complex. The Dyson motor is also very interesting, but that is another story.
My understanding is that you'd generally not use the balancing until you're in the 80% range or above where you'd drop the charge rate down & top the battery off. This is why most decent sized lithium packs / equipped devices like EV cars quote charge time to 80% as it's the handy most flattering metric to quote. The isolation is also a similar strategy but clearly big packs use contactors rather than silicon although there are ever more options becoming availble as the tech becomes more widespread.
Don't discount the possibility that the chip is actually an official Ryobi chop but has had it's markings ground off because it is a stolen or rejected/failed chip, meaning it might be just as dangerous as the rest of the circuit.
The chip could also be salvaged from scrapped Ryobi battery packs. There is quite a large industry to remove valuable chips and 'recycle' them.
Strange that it has Ryobi printed on the board , usually clones don’t go to that level if it can’t be seen .
Possible genuine board taken from ewaste?
@@mikehall3976 Or just a really faithful swiped design.
It looks like it's running that chip as a multi channel comparator. A saftey limit to prevent depleting the battery, and a few other safety functions may be included. I recommend a temperature test, in the back yard. Good luck. 👍
Good morning Sir Clive.
Sending everyone much love, take care.
RIP Sir Clive.
Interesting breakdown of circuit. The extra goofyness may come from the fact that this is a series/parallel arrangement. There are extra commons in each serial group. See the welded commons on both ends of 4 cells. 15:55 to 16.08 is where you can see all the welded tin men welded to the cells at both ends. This brings in more commons that are not at either the + or - of the complete pack. Just a possibility. Thanks for sharing your frustration.
This is completely standard practice. It is referred to as 5S2P, meaning five series, 2 parallel. (Or perhaps 2P5S, both are common.) High capacity batteries are always done this way. EV batteries are a case in point, as are eBike batteries.
I've tinkered with Ryobi li ion batteries and hobby li ion chargers (I gutted an old charger to essentially make an adapter between battery terminals and an XT60.). Using the + and - contacts it'll discharge, but the battery itself won't accept charge unless (inside my adapter) the + and center terminal are bridged with a resistor like you did. Once they are, the hobby charger can charge it.
Keeping it simple, I leave it unbridged, and just discharge to 18.5 V for storage.
I wonder how the real Ryobi batteries circuit board is in comparison. There are a lot components on the boards that I have seen. Thank you for the great videos.
pull off one of those scrubbed chips, sometimes the manufacturer will imprint it on the bottom as well
I've got a half dozen of these marked "6.0AH" (mine are "Homedas" brand, but are identical) and have had good success with them. I haven't had them apart (no need to, Big Clive is in the business) but they seem well behaved, have been thru hundreds of cycles, and seem to out last (cycle time) the 4.0 Ryobi branded ones. They may not be 3000 mah cells, but they are bigger than 2000 mah. Some of my simple power tool batteries - the ones without onboard balance like the Harbor Freight Bauer line, I run on my RC charger and it accurately reports charge taken, but I haven't tried it on the Homedas batteries as I didn't know what's in there. I'll probably just keep using them. For the price, I'd buy them again. Cost is about 1/3 of genuine Ryobi 4ah.
So far, nothing bad to say about them. Thanks for giving them the Big Clive treatment!
Hi clive, saw all your videos, Ive had this same stupid problem of too much circuitry in a black and decker drill battery pack. It sometimes went completely dead only to 'wake up' instantly when put on charger for 1 second. It got very irritating very soon, I basically had to amputate out all that evil smart circuitry. I also added a real fuse because Im paranoid.
It may sound like complete blasphemy but I dont think Li-ION battery packs necessarily need a BMS. I have used tested batteries from the same batch to make battery packs with no BMS at all for donkeys years and never had a problem. Ive even tested cell voltages to be within 50 mV of each other typically after years of use. Your mileage may vary.
I expect people to have a problem with this. Some are even angry.
They didnt use a BMS and their batteries blew up!
The current through the 100 ohm resistor will only be around 40 mA because it sits in parallel with the cell. And the cell voltage will always be in the 3-4 volt range so 4V / 100 ohm is 40 mA.
Lower capacity Li-Ion cells are usually rated for higher peak discharge rates (20A-25A) than higher capacity ones (10A-20A) and have a lower internal resistance. I guess they use thicker plates.
Seems like the clone is better then the original, since no power tools have a balance circuit! Looks quite nice actually
Great video. It would be great to see these side by side with the name brand for comparison. I work on packs alot on my channel and I gained alot from this video. It has gone further than I have with the Ryobi packs BMS. I have worked with the 40V Ryobi packs in the past and if you ever want a deep dive Big Clive, try reverse engineering one of those. It's so hard to trace out and Ryobi also has the very complex BMS board with Charge/Discharge MOSFETs and the 40V packs are known to be very problematic.
i do like that the board even says Ryobi in the corner.
Interesting that the pcb is marked "RYOBI 18B v20161104". That makes it a very bold and audacious clone, if indeed it be.
Through the magic of buying two of them he can already tell you it's a four amp hour battery.
Except Clive doesn't have to act like he's excessively clever when that happens repeatedly.
When Big C said silicone smoo I heard a AvE in the distance yelling *release the smoo!!!*
These appear to be quite similar in terms of design to the genuine AEG (aka Rigid) batteries, although completely different board layout. They don't have the 6 pin cell balancing chips, but have heaps of transistors. We had one board get a bit water damaged and after a few attempts at cleaning it up, the best outcome so far has been angry flashing lights, so not sure if the controller chip is detecting something wrong still or is using EEPROM or something to remember it has had faults in the past.
Excellent analysis as always, Clive. Release the schmoo!
I am sure that within the (330) comments someone must have said about the RYOBI 188 silkscreen in the corner. Is it a "rip off" PCB? Have you seen the circuitry of the genuine part?
It's good to keep the brain active! Keep um coming ... Thanks
Wow. Thanks Big Clive. Very complex and I'm sure an interesting puzzle. Yow.
Good Sunday afternoon to you sir from Wellington Somerset
7:40 not sure I understand your maths there Clive. The voltage might go up to 4.2v per cell, so 100ohm will pass 42mA. It can't drop 50v? 42mA at 4.2v gives a couple of hundred mW, which looks to be about the right size? Maybe the charger can stuff 500mA through the cell, but I don't think those resistors will ever see that.
Have a few clone Milwaukee 4ah packs for about the last three years.
Even with regular use and occasional abuse, you can't tell the difference from genuine.
Running a cheap 12v compressor, one will easily pump up a large flat tractor tyre.
The thumbnails are becoming better by the day.
My first thought when I saw the feedback circuit was a dedicated circuit designed around the scrubbed-off IC.
If the IC is trying to read each battery independently in circuit, resistor values for those independent test points would be different to identify which battery is reporting voltage.
I've just never seen it done with such complexity.
I am wondering if the resistor instead goes across the the individual cell, to help run it back down a bit?? It would then allow for 42mA of discharge current.... (assuming the cell was at capacity of 4.2V) and then allow for it to not charge up whilst the remaining cells are still charging. Might be the way that the newer rapid chargers work. Charge up to 80% and then trickle charge to the end. But, then again, I could be talking out my ass... Keep in mind, I am only at 7:22 in the video and my opinion may change.
Edit: I think that i may be on the right path, and Clive was too, but I think he overspent time looking at it, as according to his drawings, it IS set up to allow for a discharge/bypass, and especially if the charge does finish in the mA range. Also, to allow 50V to go across the resistor, you would have to assume that the cell would be cut off from charging, which it cannot be, as they are all connector together via those metal tabs welded to the battery ends. It may just allow for a subtle difference in the cell that could amount in a nicely put together pair of cells in a larger pack, over time. Even the 42mA (at full charge, of course,) if left to run the whole charge cycle, could allow for a small correction factor during the whole charge, as the battery will still be accepting current through the main charging circuit. But, then again, I may be mistaken.
The charge balance doesn't really get much simpler than having one dedicated chip, transistor and resistor per cell short of having a single chip that integrates multiple cell channels and the transistors in one package to reduce total component count and board space. Nothing weird about the back-to-back FETs either, standard over-charge/discharge disconnect setup. The only "weirdness" is the cell voltage monitoring by the "BMS" chip or whatever it might be but then again, a voltage divider to bring each cell's voltage relative to pack ground within the ADC's dynamic range would need to have different values for each cell. If you don't want voltage dividers, then you need to transfer cell voltage to a floating sampling capacitor and then bring that capacitor down to ADC ground for conversion, which would be even more funky-looking when done using discrete components.
I guess those 100Ω resistors are just making sure all cells are at the same voltage after the charging ends, just slowly discharging them to some safe(r) value.
have you looked in the real roybi packs? they are just as strange.. if one cell gets +-0.05V out from the rest it simply stops charging that cell... its rather annoying
The earlier versions of these batteries were NiMh, not LiPo. I have a set of the One+ tools and I bought in quite early. I still have a few of the blue ones before they changed over to LiPo and changed the color of the tools to the yellow. I just bought 2 of those knockoff batteries recently. Haven't had a chance to really put them through their paces though.
I bought one of these off ebay, first time I put it in my drill, it made a flash from behind the clips on the side to hold it in, not sure why but it's been fine ever since 😂
DeWalt had a problem with their early chargers also. I got two free batteries and a new charger because if you left a battery in the charger too long it set the battery into over charge and would melt down and could start a fire. 1 of my batteries did melt down but I was lucky and it didn't catch on fire. They decided to replace both batteries just to be on the safe side!
Strange that the other comments are from 2 weeks ago but I just got notified that you had posted a new video to UA-cam?
Patreon supporters get pre-publication access.
I wonder if you could build a really simple balancing circuit using optoisolators and maybe series shunt regulator so that when cells reach 4.2V the shunt regulator starts shunting and series with an optocoupler can send a signal to an MCU. Simple but I guess optoisolators aren't trivially cheap?
I would like to see inside a real Ryobi pack if you have one! Awesome video btw
The board has RYOBI188 marked on it. Do you think its a genuine board they have used? Maybe this is a batch they got of the first series boards. What do you think?
I have a bunch of the Ryobi 18 volt tools found one thing if you get them hot under heavy loads they will cut out as a safety and they some times will not even charge till they are cooled off a trip in the freezer on a 90+ degree day will reset them quicker same for the low voltage they just stop it’s interesting I have one of the old ni cad packs from 15 or so years ago that still works ! They had no such circuitry they would run down stone dead .
That's weird indeed. If I understand correctly this is kind of a belt-and-suspenders approach, where both the MCU and the little 6-pin chips can decide to discharge individual chips. I'm sure it'll work just fine but I don't understand why you need both. Also, I don't think those stacked chips communicate anything to the MCU at all. To a degree, in fact, I wonder if the same result couldn't be accomplished with nothing more than a zener diode and a resistor across each cell.
I've been using two 5AH eBay knock off packs for 3 years now with no issues at all. The cost of the genuine packs is extortionate at $149 when I can buy one for $39 on eBay.
I had a 12v "300w" power supply skidmark its main chip... that was astounding the amount of excitement. I wouldnt have thought that much energy would have been present in that chip at any time... but it will have a glorious story to tell in shitty IC heaven. I also think of these as BigClive moments... and was excited to take it apart and find the failure. Using all of my knowledge from BigClive clearly it failed because the power cord had a knot in it and it was preventing the electrons from squirting through. I am SMRT!
Great analysis, I couldn't have sorted that.
I have one of those aftermarket Ryobi clones from a different brand and a few genuine Ryobi batteries and I can confirm that the genuine batteries have a lot more circuitry in them. My aftermarket clone doesn't have the power like the genuine ones have if you compare like for like.
Not sure about that calculation, where did the 50V come from ?
4.2*499 ohm=8.4mA *4.2=35mw (i think i got that right)
Looks like they balance like supercapacitors, but with a multi cell chip.
I think they are sensing the voltage with the weird circuit (stepping through each), and if that voltage exceeds the threshold will turn on the mosfet to discharge the cell.
If you've see those 8 cell testers, they step through and display each cell votage, plus the total.
It would be interesting to see a scope trace of what happens to cell voltage and current, and the "data" on the control/monitor pin during a charge cycle.
About the 100 Ohm resistors... Isn't that pretty normal BMS design? I'm not sure about this battery. But normal BMS chips, while having over and under voltage protection, don't actually regulate the charging process. They only monitor the voltage difference between cells. And if one cell is particularly high, they "connect" a resistor in parallel to bypass some of the charging current and make that cell charge a little slower than the other cells.
Obviously, this is not the best way to do it, as it can only balance while charging. And if the cells voltages are really far apart, the cells won't fully balance until the first cell is fully charged, in which case of course the whole charging process has to be stopped. But you know, that's the cheap way to do it, which works for the most part. And assuming the whole CC/CV charging process is handled somewhere separately, that's fine, I guess.
As to the whole voltage divider, voltage measurement circuit. Yeah, not sure what that's about. I mean, it's nice to see they added a way to shut off the voltage dividers (you know, some circuits just use high resistance voltage dividers that constantly draw a little current from the battery, 'cause it's easier/cheaper). I get your thoughts about them trying to cancel out the voltage of the other cells (that are connected in series). But you know, it's a microcontroller, it could easily subtract the other voltages. Measuring the full voltage would give you a little less ADC resolution of course (for the single cell you're trying to measure), but is that little bit of resolution worth all the extra parts and potential points of failure? Besides, the resistors all have a certain tolerance anyway, will the extra resolution really make a difference?
I'm sure they were thinking something... But yeah, seems kinda unnecessarily complicated.
Or maybe it's also totally genius and we just don't understand it... :D
Anyway, all in all it doesn't seem that bad to me. Certainly not something I would consider unsafe or "untrustworthy".
I wonder whether some of the complexity is down to Ryobi's attempts to prevent use of 3rd party batteries in their tools.
Probably not so much the tools, but maybe the OEM chargers.
No. None of the circuitry has anything to do with that.
It is all cell balancing and charge monitoring.
I know others have mentioned it, but are we sure this isn't an actual Ryobi circuit board?
I mean, it has RYOBI18B silkscreened on it along with a date code of v20161104.
Maybe they grabbed old Ryobi batteries, refurbished them somewhat and then resold them as their own?
That fourth contact, on the board itself, seems to do some kind of communication with Ryobi’s new line of HP tools. I have some that have that additional contact on the tool. Not all ryobi batteries have that, however, just the ones labeled as HP. I’ve yet to find an appreciable difference, however. If you use an HP tool with the additional contact it seems to make no difference whether the battery has it or not. Operation, power, etc all the same. None of the chargers I’ve seen so far have a 4th contact.
Edit: I should send the tool I have to AvE to have him do a tear down and you two could work together to figure out what that contact does. If anything.
Edit 2: I actually did find one of my chargers has the additional contact. It’s labeled as a fast charger. I have 2 others that don’t have that, though.
Neat, I own two of those batteries. I bought them knowing full well that "5 amp hour" when translated from the Chinese means "4 amp hour" and they were a good deal for 4AH. Kind of surprised to see that the circuitry is actually pretty decent.
5 amp usually means 3 at best... You got a bargain!
Could you desolder the battery pack and use an OEM controller board if your Ryobi battery died?
I have a couple identical/very similar batteries to this and they work perfectly and get abused as much as my genuine kit and take the punishment fine. Cleared out my uncles garden using one in a reciprocating saw cutting small trees down. Only stopped when tool became too hot to hold, but the battery was fine and still had plenty charge left.
I think those 100 ohm resistors just divert 4/100, = 40mA away from an overfilled cell. So if the charger charges at 500mA, any overfilled cell will only get 460mA, until all cells are filled. . . . . . There is a Baofang radio that uses a 8500 mAh battery, things are getting very interesting with these new batteries nowadays.
10:45 power FETs have a "parasitic diode," not "accidental diode." It's just the way the required PN junctions stack up to create the device.
With all respect to you, Clive, I think that because you do not understand the way some part of their circuit works is no reason to question its safety (0:43). If that were a valid reason, would you trust your computer, whose CPU is an incomprehensible jumble of billions of transistors? On the other hand, sometimes the cleverest of circuits is the simplest. The bandgap voltage reference is a classic example. Examine its most stripped down schematic (two transistors and a gain stage) and try understanding how it works. A knowledge of device physics is required, which is "the hidden schematic."
Lol common language issues.
Not designed, accidental.
The stacking introduces it, the designer if they could wouldnt have it . Not their by intention, so "accidental".
Chill man ;)
I gave up on these highly questionable 3rd party packs and now I strip the control systems out of them and simply connect a wire to every inter-battery junction, and wire those out to a balance connector which I can then plug in to a RELIABLE lithium ion charger. It just means plugging it into a known balance charger rather than the proprietary one. One then hopes that the board you pulled out of the battery pack doesn't have some signal chip in it which is required for the drill itself to operate. It's all getting very complicated.