I was given a Dyson DC59 cordless stick vac. I was told it was working but probably needed a good clean up, they had been given a new one and this was surplus to requirements. I pulled the trigger, 10 seconds later it died and the battery started flashing red. “Oh, it hasn’t done that before, probably needs charging, I haven’t used it for over a year”. Long story short, the literature says a red flashing battery (32 flashes) is a dead battery and a new one is required.
The difference between ‘free’ and a hundred dollar battery replacement was enough for me to decide to see if the battery really was dead. It wouldn’t charge so I pulled the casing off the battery exposing the 6 18650 cells. The voltages ranged from 3.9V to 3.6V so I set up a jig with a ‘dead’ USB battery bank charger and charged each to 4.2V, settling to a satisfactory 4.1V overnight. No luck, still flashing.
Further research revealed many had followed this path and gone into great depth. It turns out the battery management circuit, once it decides a battery is dead, it is a final decision and recharging, rebalancing or replacing the cells did not change this. This is a dubious design decision. A more serious flaw is that the so called battery management circuit included an IC that could rebalance cells but this feature was not implemented. All it does is monitor the state of cells, and relay the data to a microprocessor. Once the difference between cells was sufficient the pack was declared ‘dead’. The full (very long) story is here.
This is sending a lot of quite useful battery packs to landfill. There is a possibility that the microprocessor can be reset with the right tools. I have the tool but it is not something in general use. So I will try. However it does not fix the fact that the cell balancing is not implemented. The circuit board includes pads to mount the components, but they are missing. I believe this is a serious flaw. The pack could have double the lifetime for the want of 20¢ of parts.
So don’t throw away those ‘dead’ Dyson battery packs. If you know someone technical, they may have a use for 18650 cells. A few people have the equipment to test the cells and find in some cases the cells are still giving close to their rated capacity. There are also after market replacement battery management circuits. It is a messy job to replace it, but a few have tried and had success.
I have a similar understanding of the battery protection provided in some/many laptop battery packs. Once a cell is determined to be faulty the pack needs to fail to safe. The protection circuit is designed to prevent against a poorly performed replacement.
Yes, it has a very poor outcome in sending to scrap a perfectly useable pack for the loss of a single cell. At what cost might a professional repair be provided?
In this case, there may not be a failed cell. Because the module does not implement cell balancing, what happens is one cell may have a slightly higher loss than another, so it determines when charging occurs. The cell with the least loss has the highest voltage and determines when the battery is deemed ‘charged’. As these get further apart, the usable lifetime of the battery is shortened even if every cell is OK. Eventually the two voltages are so far apart (it seems about 0.3V is the limit) the microprocessor deems the pack dead, even though every cell may be fine - just in a different state of discharge.
In every battery, each cell will be slightly different. In many battery chemistries it is possible to ‘balance’ the cells by slightly overcharging. In car batteries this is done by bringing a 12V battery to about 14.4V every now and then. Cells that overcharge generate a bit of gas while the ‘weakest’ cell gets fully charged. In a sealed battery the gasses recombine and are reabsorbed. Nickel Cadmium batteries are similar. Lithium batteries explode if they are overcharged so this avenue is not open. So balancing is done by a sophisticated IC that during charging bypasses some of the charging current to all cells except the one with the lowest voltage. When the lowest voltage cell ‘catches up’, the next cell is picked up until one cell reaches 4.2V. If this scheme is used on every charge/discharge cycle the cells remain balanced unless one cell goes bad.
My point is that this scheme was not implemented on a quite expensive battery pack, and it is only luck that a pack lasts as long as it does.
And if there was one dud cell (unlikely, given the above scenario), there is no point replacing it because once the management system determines the battery is dead, there is no (simple) way to reset it. People have tried. People have more success replacing the battery management board than by replacing cells.
Why these batteries are made this way is a mystery. People have determined that the circuitry for battery balancing is on the board, the components (resistors - ten a cent) are missing.
Are you able to provide evidence this is how it is? EG image of the PCC, diagrammatic circuit details, other sources.
Do Dyson offer a 2 year warranty on their batteries? It seems odd that Dyson would risk a battery pack being recharged several hundreds of times over 2 years, without some method of of charge equalisation.
Apologies, it’s important that this technical detail aspect is able to be verified. Most consumers could reasonably expect a lithium rechargeable battery to last more than 2 years. Perhaps 3 or more without significant degradation based on other devices. It’s not readily apparent from the top hits on the internet that Dyson has designed a fault into their product, or that it is not fit for purpose?
One person reverse engineered the pack and drew the circuit. It is part of the long story I linked to in my first post. I am not sure if people need to be members of eevblog to see expanded diagrams so I add them here. In the circuit there is a section ringed in blue marked ‘not fitted’. I verified that my battery pack circuit is very similar (looks the same to me) to the one shown and indeed there are pads for the components marked, but no component.
Someone has gone to considerable effort. We’ve a spare unused battery pack for a V8. The original is 2+ years old and still going great.
If ever it gives up the above will be handy reference. RS Components currently offer the BMS chip and an evaluation kit, ($$$$), from which the following links to the application notes and chip specs. There are several alternative circuit configurations provided. All utilise the balance connections.
The configuration of the BMS in the Dyson example irrespective of cell balancing still serves to prevent over charging of the pack and against a faulty cell. Assuming all other Dyson packs are the same, what has been lost? Lithium cells are best not stressed by fully charging. The way Dyson charges will limit the pack to the charge capacity of the weakest cell. If all cells are healthy does it make a measurable difference to run time?
Is this how Dyson has designed all product sold into Australia?
There is certainly a point to be made concerning difficulty of repairability of the battery packs, for what often may be a single cell failing. That is independent of the lack of implementation of balancing, impact ???
Towards the end of the eevblog story, another person experimented with changing the EEPROM data of the microprocessor. Writing new data to 2 bytes removed the 32 red flash lockout condition. I was able to follow the instructions (requires a PICkit 3 programmer or similar) and indeed it does unlock the processor. I have reassembled the battery and put it in the vacuum cleaner. A quick test confirms the cleaner motor is running. More testing is needed to see if there is any real life in the battery. I shall leave it for a while then report back.
For safety it’s essential when one cell in a lithium pack fails the battery is shut down.
The open question is the extent to which any of these failures are attributable to the design and quality of the BMS. We don’t have sufficient product information to know with any certainty. On the basis the protection correctly identifies a potentially faulty cell the product is performing as designed.
Is the real problem cells which are not of equal quality? A professional test of all the cells from a failed pack, along side a reliable history of the charge discharge cycles is one way to gain a better understanding, in that single instance. I wonder what percentage of Dyson battery packs fail within warranty or too soon after?
When a battery pack fails early it’s likely the remainder of the cells can continue to be used once the faulty cell is correctly repaired with a suitable replacement. Not so practical for most of us at home, (as noted @meltam). If the battery pack has some years of steady use, perhaps not so worthwhile, if all the cells are of similar quality. IE near to end of life.
I’ve amended the topic title to ask a question rather than making a statement critical of the Dyson product.
In more complex battery packs, failures of one or more cells is accepted in the design. The battery is designed to continue to function reliably with one or more failed cells. EG Tesla BEV’s. For a home battery powered tool or appliance there are too few cells in a battery for a similar strategy to be useful.
Presumably the two bytes of code in the PIC micro controller you ‘zapped’ has disabled calling a section of code. Or it could be a some actual data used for a compare in calling a section of code. You do not say.
I do hope that as Lithium Ion batteries can and do catch fire that you keep a close watch on your recharging.
Maybe the engineers put that battery management software protection, that seems stupid, there for a reason. That would be a commercial secret no doubt.
But I admire your skills with electronics, and tenacity in hunting down a problem.
I don’t know. I was an IT person, and my hobby involves programming microprocessors so I can guess. It is likely the control program is written in a high level language (C, Python) and the two bytes are a single value, may even be a flag to say “battery is dead” and nothing more. Unfortunately the program memory is not readable (by design, a security feature) so we don’t know what it does.
It would seem the battery pack is not 100% healthy, although what I am seeing may be due to the battery management software rather than the cells. The vacuum cleaner runs for about a minute and then the lights indicate the battery needs recharging. However, if it is run for 30 seconds, rest for 10 seconds, run for another 30 seconds, and so on, it is quite happy to run for quarter of an hour (at which time I quit, it still seemed to be working fine with no change of performance). The battery pack didn’t warm (to touch, I don’t have a better way to measure). It charged normally, again not warming.
This means the device is not going to be useful as a whole of floor vacuum cleaner. But for crevices, cleaning the sofa or the car, or other spot cleaning it is ideal. 30 seconds is quite long enough for these tasks. This is the way I intend to use it, so I am satisfied.
I speculate the problem arose because the device was put away in a cupboard for over a year with the battery in an uncharged state. Had I put it on a charger before testing it, this whole saga may have been avoided. Too late now. Maybe there’s a caution for other people - if a device has been unused for some time, charge before use. Most new devices say just that - charge before using. This may be the reason.
Our Dyson stopped working suddenly when it was just under two years old. I assume the battery has died, but Dyson requested we send the whole thing back to them, so we did. Now we aren’t sure how much longer it will take till we get our vacuum back. It’s been gone for a couple of weeks now with no word from them.
For my battery, and I think for many models (after looking extensively online), there is a mechanical connection between the battery and the trigger. There’s not much that can fail in that connection so if the trigger is pulled and there is no indication from the battery LED, then the battery is almost certainly the problem.
If the battery gives flashes on its LED, it can be for a variety of reasons. Needs charging, overheating, etc. documented in troubleshooting. In that case, the battery may not be the problem.
Not much help to you now. This might help people that like to do their own repairs. The battery is easy to remove and replace, two screws. Also the battery can be charged on the charger while removed, all the electronics are in the battery. Might be worthwhile every year for people that have a spare battery (Lithium cells have a good shelf life but they do have a small self discharge).
It all happened suddenly. Up till then we could vacuum the whole house with around 40 to 50 percent of charge left. The last time I tried to use it, it only ran for a few seconds and flashed up a charge image. I stuck it on the charge and it showed 100% then went dead. Dyson sent out a new motor and when that didn’t work, they just asked for the whole lot including parts to be sent back to them. Never once did they say anything about the battery being an issue.
This would not be the first example of Dyson building obsolescence into their products.
I bought a Dyson vacuum and paid another $120 for a turbo cleaning head.
The turbo cleaning head wore out after what I considered very little usage, and the local salesman told me that the bearing was worn out, it was a common problem, and I would need to buy a new turbo cleaning head at $120.
I refused to pay up, and then the salesman told me a quick and cheap hack to fix the problem.
I asked as to why Dyson did not share this quick fix hack?
I was told that Dyson would rather sell you another turbo cleaning head for $120 than tell you a quick fix for the problem.