Daly BMS - a potential cause of battery fires?

[WheezyRider]

I've heard of some people saying they've had bad experiences with Daly BMS products then recently I came across this over on ES:

DALY BMS Teardown & Failure Review (16S 40A)

Particular Model 16S 48V 40A LiFePO4 ....... (simple, no uart) Exact link where I bought 15 of them ..................(not prime, shipped from China) https://www.amazon.com/gp/product/B0876ZT5M1 I have already posted all the positive things I have to say. The BMS are amazing in every way...

endless-sphere.com

It's quite worrying. If what he says is correct for all Daly products then it means there are potentially thousands of unsafe battery packs out there.

Makes me wonder what brand of BMS to get. Most are not marked with a manufacturer.

 

[Bikes4two]

• @Nealh has had failures from Daly BMS - whether failures of the sort described in the ES article I know not.
• I think he now uses BMS from Ann Power (link) or is it BeMuchSafer (link)?

 

[guerney]

Read elsewhere that the "manufacturer" of HEYO BMSs is also Daly:

" For the software HEYO refers to the Daly website.
With the Sinowealth software you can read the manufacturer from the unit, and it shows DALY - DALY. "

https://diysolarforum.com/threads/daly-clone-of-a-clone-daly-bms-vs-heyo-bms.12473/

 

[Bikes4two]

• There are 7 pages to that thread so far and it was started in Nov'22 and the last post was late Dec'22
• As 'interesting' as the guy's tear-down is, the fact that others have not come along since Dec'22 to report similar problems does indicate that either the poster's use case was unusual and/or he had maybe a faulty batch (15 units) of BMS
• And the BMS was for 16s, 40A, so clearly a power user.
• All in all if I had a Daly BMS I'd not loose any sleep over the possibility of the same fault occuring in my use case - I'd be more concerned at the general unreliability of the BMS (as reported by @Nealh ) and I'd certainly not be buying any fire extinguishers based on that one report.

 

[Nealh]

The DALY type I have seen in Red and Blue.
My DALY's were Red and two that failed within 6 months of each other the std 3 wire BMS with P-,C- & B- , something on the discharge side failed and all I got was some residule voltage reading of less then 0.5v. They were 30a 12s with balance, no switch.

I have a third DALY BMS which is working and has been good for two summers on the UTE with tsdz2. this BMS is same port 30a 12s , no switch.
I used Moli p42a's for a 12s 2p config.

I used ann power before with switch and they seemed good , but then one year my 48v battery of PF's failed. I checked it one month and all was good , the next month it read 0v and each cell group was dead at 0v. I changed brand because I wanted a narrow small footpring BMS for my bottle battery.
After the DALY failed I looked into other BMS and saw the BeMuchSafer ones with no balance but have protection for charge/discharge, I bought in case it was needed an active balancer to wire in but have not yet needed to.

Using A1 genuine HG2's in 12s2p I have no need to balance except when the DALY failed a couple of groups went a bit lower so can only assume bleed resistors bled them down.

 

[Nealh]

I don't think the DALY's are a cause of battery fire as there is no evidence , all that appears to happen is a component gets hot and fails and this failure shows up on the discharge side by the look of it as one gets no discharge voltage bar a few millivolts.
The battery pack is protected as the BMS shuts down under failure so no charge or discharge can occur .

 

[WheezyRider]

• There are 7 pages to that thread so far and it was started in Nov'22 and the last post was late Dec'22
• As 'interesting' as the guy's tear-down is, the fact that others have not come along since Dec'22 to report similar problems does indicate that either the poster's use case was unusual and/or he had maybe a faulty batch (15 units) of BMS
• And the BMS was for 16s, 40A, so clearly a power user.
• All in all if I had a Daly BMS I'd not loose any sleep over the possibility of the same fault occuring in my use case - I'd be more concerned at the general unreliability of the BMS (as reported by @Nealh ) and I'd certainly not be buying any fire extinguishers based on that one report.

Yes, the guy writing the thread is rambling and the whole thing is very drawn out. But, he is someone who knows something about electronics and circuit design. To summarise what he is saying, it seems Daly BMS circuit design (for this model of BMS at least) is poor, uses poor quality components and where there should be components to suppress damaging transients like TVS diodes - these are missing. Hence if high voltage spikes come from running inductive loads like motors, there is a good chance of the MOSFETs being damaged.

There seem to be quite a few people who say the LVC does not work properly on Daly BMS, or there are faults with balancing and even over charging prevention. The OP on this long thread at ES doesn't think the short circuit, over current protection would work well either.

When a battery pack is potentially a bomb if not managed properly, it's vitally important to get a BMS that is robust and is going to do its job.

 

[WheezyRider]

I don't think the DALY's are a cause of battery fire as there is no evidence , all that appears to happen is a component gets hot and fails and this failure shows up on the discharge side by the look of it as one gets no discharge voltage bar a few millivolts.
The battery pack is protected as the BMS shuts down under failure so no charge or discharge can occur .

I hope that's the case, and we've not just been lucky so far. If a pack does incinerate itself, is anyone going to go down to the BMS circuit level to see what was at fault?

 

[WheezyRider]

• @Nealh has had failures from Daly BMS - whether failures of the sort described in the ES article I know not.
• I think he now uses BMS from Ann Power (link) or is it BeMuchSafer (link)?

They look interesting, although it's a shame they don't seem to do higher current BMS models. I like the addition of the on/off switch to the BMS. Also, they come with tails already attached, which is nice.

 

[guerney]

Using A1 genuine HG2's in 12s2p I have no need to balance except when the DALY failed a couple of groups went a bit lower so can only assume bleed resistors bled them down.

Are all bleed resistors switched off when one fails? I admit, with all the talk of charging 42V battery packs to 41V, I had been partially charging for some months, hoping to extend battery pack lifespan... but aren't bleed resistors heating elements? Those don't last if used for long periods of time. Therefore if the cells in your battery pack don't have 100% consistent chemistry, and you've generally been partially charging, surely when you go to fully charge until the charger light goes green, those bleed resistors will be used for longer to balance the pack at 42V, than if you had fully charged and balanced every time?

I've gone back to always charging and balancing at 42V. I'm hoping none of my cell banks have become very unbalanced by partial charging, and may check.

 

[saneagle]

Are all bleed resistors switched off when one fails? I admit, with all the talk of 41V, I had been partially charging for some months, hoping to extend battery pack lifespan... but aren't bleed resistors heating elements? Those don't last long if used for long periods of time. Therefore if the cells in your battery pack don't have 100% consistent chemistry, and you've generally been partially charging, surely when you go to fully charge until the charger light goes green, those bleed resistors will be used for longer periods of time to balance the pack, than if you had fully charged and balanced every time?

The current through the bleed resistors isn't very high. They get warm when bleeding, which is their job to convert excess charge into heat. The one in front of me now has 100 Ohms, so at around 4v, that makes 40 mA current, which is 0.16 watts. per resistor. That means about 15 minutes to drain down a single cell from 4.2 to 4.18v. Other's I've seen had higher values. I'm pretty sure I've seen 200 Ohms, which would bleed at 20mA. They vary in size but most are quite large so must be at least 1/2 watt rating.

AFAICS bleed resistors don't fail, but if water gets in, the transistors that switch them on can get stuck or the water itself can give a path to ground for any of the cells. Years ago, when we used to use Ping LiFePO4 batteries, they had LEDs on the bleed resistors, which lit up when the transistor switched on so you could see which ones were active until all the lights would be on at the end of the charge.

 

[guerney]

I used ann power before with switch and they seemed good , but then one year my 48v battery of PF's failed. I checked it one month and all was good , the next month it read 0v and each cell group was dead at 0v. I changed brand because I wanted a narrow small footpring BMS for my bottle battery.

Is it safer to replace our BMSs every year or so? Was looking at this ANNPower 10S 20A, which has a switch:

9.26￡ 33% OFF|36v Li-ion Battery Bms, 10s 36v 20a Lithium Ion Batteries Bms, With Balanced Function And On/off Switch. - Battery Packs - AliExpress

Smarter Shopping, Better Living! Aliexpress.com

www.aliexpress.com

 

[Bikes4two]

Is it safer to replace our BMSs every year or so? Was looking at this ANNPower 10S 20A, which has a switch:

Why on earth would anyone want to go through a replacement of a BMS (that in itsself has an element of risk to it in the dismantling and re-assembling of a battery) when there isn't any solid evidence that it is necessary?

 

[guerney]

Why on earth would anyone want to go through a replacement of a BMS (that in itsself has an element of risk to it in the dismantling and re-assembling of a battery) when there isn't any solid evidence that it is necessary?

It doesn't look too difficult:

I'd need a pair of these myself, because my eyes aren't what they used to be (they used to be a better pair of eyes):

Magnifying Glasses Head Band Adjustable 2 LED Light Magnifier with 5 Lens Loupe | eBay

Find many great new & used options and get the best deals for Magnifying Glasses Head Band Adjustable 2 LED Light Magnifier with 5 Lens Loupe at the best online prices at eBay! Free delivery for many products.

www.ebay.co.uk

" If it aint broke, don't fix it! "

I may buy that ANNPower BMS to keep as a spare, because I don't want to be without use of my bike for 2 weeks, while a new BMS gets delivered from China.

 

[WheezyRider]

I've gone back to always charging and balancing at 42V. I'm hoping none of my cell banks have become very unbalanced by partial charging, and may check.

If you have a reasonable quality pack and the cells were balanced in the first place, it's unlikely the cells would have gone much out of balance, if at all. It takes hundreds of cycles for decent cells to start to drift. Even if your cells have gone significantly out of balance, you'd probably start noticing a drop in range. So even if you did a balance once every hundred cycles you're likely to be fine.

 

[guerney]

If you have a reasonable quality pack and the cells were balanced in the first place, it's unlikely the cells would have gone much out of balance, if at all. It takes hundreds of cycles for decent cells to start to drift. Even if your cells have gone significantly out of balance, you'd probably start noticing a drop in range. So even if you did a balance once every hundred cycles you're likely to be fine.

LG MH1 cells, a 22A Chinese BMS with everything visible written in Chinese apart from "22A". I've noticed a drop in fully charged "balanced" voltage: Straight off the charger the voltage is 41.8V. 30 minutes later it's 41.7V, same voltage at 90 minutes, and that's where it stays. About 4,000 km since the end of December 2019 from new. When new, it maintained a voltage a smidge higher than 42V. I might give it a really long balance charge next to a open window on a pizza paddle, covered in four cheeses. This could attract flies.

 

[WheezyRider]

LG MH1 cells, a 22A Chinese BMS with everything visible written in Chinese apart from "22A". I've noticed a drop in fully charged "balanced" voltage: Straight off the charger the voltage is 41.8V. 30 minutes later it's 41.7V, same voltage at 90 minutes, and that's where it stays. About 4,000 km since the end of December 2019 from new. When new, it maintained a voltage a smidge higher than 42V. I might give it a really long balance charge next to a open window on a pizza paddle, covered in four cheeses. This could attract flies.

Those are good quality cells. If the pack was put together properly and the cells matched to start with, I would expect it to go hundreds of cycles without going out of balance. End state voltage doesn't mean very much, it's the capacity profile that matters. If the pack is giving more than 42 V off the charger, that isn't good - but then, your DVM is not calibrated, so you have no way of knowing what the actual voltage the charger is putting out.

 

[guerney]

If the pack is giving more than 42 V off the charger, that isn't good - but then, your DVM is not calibrated, so you have no way of knowing what the actual voltage the charger is putting out.

I didn't have a DVM when the battery pack was new, I was relying on the voltmeter of the DPC-18 display. How far off can my £8.95 DVM possibly be when measuring the charger's output? 1V? 8V? I might try searching locally, to find somewhere I can calibrate my cheap DVM using a calibrated power supply.

Those are good quality cells. If the pack was put together properly and the cells matched to start with, I would expect it to go hundreds of cycles without going out of balance

I should throttle about (with my new throttle) on the flat somewhere, to see how much of the 38.5kph/23.9mph no-load speed my Dahon Helios 250W Bafang conversion retains unrestricted, when wheels meet tarmac, now that I've increased the controller's amp limit from 15A to 18A.

 

[WheezyRider]

I didn't have a DVM when the battery pack was new, I was relying on the voltmeter of the DPC-18 display. How far off can my £8.95 DVM possibly be when measuring the charger's output? 1V? 8V? I might try searching locally, to find somewhere I can calibrate my cheap DVM using a calibrated power supply.

A cheap DVM can easily be out by 1V on the 200V range. If you have a manual, it may give an accuracy value, but how much you can trust it? Typically, 1% is the quoted accuracy for a lot of DVMs. At 42V, that could give anything from 41.58 to 42.42 V, which is the difference between not balancing and over charging.

Another thing to bear in mind is whether the charging circuit on your BMS has a blocking diode. If so, the charger will need to output more than 42V to compensate for the junction loss of the diode. You can tell this by looking for voltage on the charging port of the battery when the charger is disconnected. If you see battery voltage there, then there isn't a diode and the battery gets what the charger is putting out.

 

[guerney]

1% accuracy is pretty poor, but this cheapo DVM fulfills it's original function of enabling the detection of alarming sudden increase of voltage from the charger, should it malfunction - IIRC @Nealh mentioned that one day, he found his charger putting out 88.8 Gigavolts or something, and understandably, ever since he's measured the voltage of his chargers before every charge. For convenience, I thought about soldering a voltmeter inline between the charger and the battery, but I'm wondering if this one's circa 20ma power consumption (or some other factor) could impede the balancing phase a little somehow.

Voltmeter 0.36'' or 0.56'' 7-150V Digital LED Display 2-Wire Voltage Panel Meter | eBay

Find many great new & used options and get the best deals for Voltmeter 0.36'' or 0.56'' 7-150V Digital LED Display 2-Wire Voltage Panel Meter at the best online prices at eBay! Free delivery for many products.

www.ebay.co.uk