EEVblog #811 - How The Varta 15 Minute Battery Charger Works

[AF6LJ]

I would love to see Dave get in there with a four trace scope and and see what sequence those MOSFETs are being switched in.

Oh, I forgot to mention....

I'd love to see a four trace scope of the four + battery terminals during a charge cycle.  Even just looking at the voltages there relative to each other would tell us pretty much exactly what it's doing with the charge profiles.  Current to each cell and investigating how the mosfets are being switched would be interesting also but even just seeing those voltage curves would be very revealing.

Agreed. 

[HKJ]

Whenever reading about quick charging NiMHs they recommend dT/dt termination and using max-voltage, max-temperature and max-duration for backup since detecting dV/dt is going to be difficult.

That charger probably uses all the methods. -dv/dt is easy at 8 amps.

Don't really understand how it detects if it's an AA or AAA battery though, there must be more to it and all the 1 ohm resistors must have some purpose, but the only reason I can think of is current sensing?

My guess is that they measures the voltage on both sides of the resistors (Maybe measuring the switcher output voltage). To do that they may use the mosfets to switch connections around, but with only a few mOhm in each they will not affect the result significantly.

[tszaboo]

Actually it would be fairly simple to test any battery lifetime with 4C charging. All you need is a good battery tester, set up 4C charging, dv/dt or even cc/cv to fixed voltage which would take about 20-30 minutes and discharge with 1C/2C rate. Measure the coulumbic efficiency, and in an hour you get the expected lifetime.
Or even 200 cycles would be finished in a week and a half. I would do it myself and post results if there wouldnt be that bloody NDA.

[IanB]

Actually it would be fairly simple to test any battery lifetime with 4C charging. All you need is a good battery tester, set up 4C charging, dv/dt or even cc/cv to fixed voltage which would take about 20-30 minutes and discharge with 1C/2C rate. Measure the coulumbic efficiency, and in an hour you get the expected lifetime.
Or even 200 cycles would be finished in a week and a half. I would do it myself and post results if there wouldnt be that bloody NDA.

I recall that user SilverFox on CandlePowerForums tested some Eneloops on a 15 minute charger. He put them through continuous charge discharge cycles. They were still holding up quite well after 100 cycles, but they lost some capacity and also lost some of their low self discharge properties.

I can't seem to find the thread right now to link to it.

[drussell]

They were still holding up quite well after 100 cycles, but they lost some capacity and also lost some of their low self discharge properties.

The loss of low-self-discharge (increase in internal leakage current) is by far the worst result of any kind of fast charging of most types of rechargeables.  Even 1h chargers will severely increase the self-discharge compared to being trickle charged over the cycle lifetime.  I have lots of batteries here that still hold quite a good charge after 10+ years and hundreds and hundreds of cycles but they're totally dead after just a few days off the charger.  That might be a useable situation for a power tool but not so good if you have cells you want to use in a remote control or something. 

Best to keep some cells marked as trickle-charge only for low-drain, long term, low-cycle use and blast the crap out of your other ones, your older ones, etc., whatever, for your heavy repeated charge/discharge use and beat 'em until they just won't charge anymore...  lol

[apis]

Whenever reading about quick charging NiMHs they recommend dT/dt termination and using max-voltage, max-temperature and max-duration for backup since detecting dV/dt is going to be difficult.

That charger probably uses all the methods. -dv/dt is easy at 8 amps.

Yes you're right, the reson charging in the .1C to .5C region isn't recomended is because even dT/dt detection becomes unreliable, but at 4C it should be much easier.

[tombi]

This come up on Dangerous Prototypes recently - here is an appnote from Vishay on dT/dt sensing for fast charging.
http://www.vishay.com/docs/29089/fastappl.pdf

Not clear how much current they are pushing.

[Stonent]

I would think at some point, all this heat is breaking down the chemicals inside.  The external temperature may be quite different than what's going on in the center.

[VK3DRB]

To measure each battery discharge current, the micro might be multiplexing the FETs to measure the current from each battery bay one at a time periodically, and using the single sense resistor. Same for charging. Most the duty cycle is not measuring but actual charging and discharging.

Comments about the board...

1. In general, 45 degree micro is also a good idea for hot air and vapour phase reflow. The reason is the paste screening is generally a lot more uniform. You get less chance of shadowing and therefore less chance of dry joints.

2. No pin one on the micro and the SOIC. Crap.

3. Could have done better in the component designators. Not what I would call artwork. Also some white ink labels going over vias. Sloppy work.

4. Breaks in the high current tracks were a good idea. I have not tried this myself but I assume this tends to prevent solder voids at one end of long tracks due to surface tension during a wave soldering.

5. They used a crappy sleeve fan (probably bronze bushes) that dries out and then wears out, creating a very annoying whirring sound and eventually seizing. So it is clearly not a quality product.

[EEVblog]

Another video uploading now.

[VK3DRB]

As I mentioned, the fan in the charger is crappy sleeve type.

Good products use sealed ball bearing fans. Expensive graphics cards for gamers and CAD gurus and cheapo power supplies often also have fans with cheap bushes. First thing that fails is the fan. You can oil the bearings with sewing machine oil, but that only delays the inevitable... finding a replacement fan.

I own an expensive Pioneer 7.1 channel audio amplifier. Pioneer used a cheapo fan rather than a ball bearing fan. I have replaced the fan twice in 5 years and it is a unique form factor in a tight spot so I have to get the same rubbishy fan each time at never-to-be-repeated prices. How was I supposed to know Pioneers would put a crappy fan in? The product was not old enough when I bought it so I guess there were no failures in the field yet. Sales people won't tell you the product has a crap fan in it. In some stores you have to buy the device sight unseen (cannot take it out of the box.) The only solution is "once bitten, twice shy". That is, never buy Pioneer again. Stores will not allow you to tear down a product to look at its fan. I would not have bought the amplifier if I have known the trouble I would have with its fan. You'd think Pioneer being a Japanese company would know better.

Manufacturers should use quality fans like Panasonic or NMB DC brushless ball bearing fans, and avoid Hu Flung Dung brand fans if they want to build trust in the market place.

It is false economy using cheap and nasty fans - you lose customers. 

[bktemp]

As I mentioned, the fan in the charger is crappy sleeve type.

But does it matter? The typical lifetime of a sleeve bearing type fan depends on its temperature. You can find values in the range of several 10000 hours.
Lets assume the fan fails after 10000 hours. With one charge cycle every 30 minutes you can charge 80000 batteries. Most likely something else will fail before the bearing of the fan. If you charge one set of batteries per day the fan will last 50 years. I doubt many people charge that many batteries in their life.

A pc or some other device running 24/7 is a different story, but in rarely used charger even the cheapest bearing will last long enough.

[Groucho2005]

With one charge cycle every 30 minutes you can charge 80000 batteries. Most likely something else will fail before the bearing of the fan.

Yes, at 8 amps charge current probably all 80000 batteries. 

[AF6LJ]

Another video uploading now.

Good Something to watch while the test equipment warms up this afternoon...

[VK3DRB]

As I mentioned, the fan in the charger is crappy sleeve type.

But does it matter? The typical lifetime of a sleeve bearing type fan depends on its temperature. You can find values in the range of several 10000 hours.
Lets assume the fan fails after 10000 hours. With one charge cycle every 30 minutes you can charge 80000 batteries. Most likely something else will fail before the bearing of the fan. If you charge one set of batteries per day the fan will last 50 years. I doubt many people charge that many batteries in their life.

A pc or some other device running 24/7 is a different story, but in rarely used charger even the cheapest bearing will last long enough.

Your 50 years life from a cheap phosphor bronze bushing is far too optimistic. The lubricant would dry out way before then and start vibrating or seize. Maybe try 5 years, irrespective of the manufacturers' claims. I mostly use my Pioneer amp with Sennheiser wireless headphones, so the internal temperature is low, but the fans failed after only a few hundred hours of use over 2 or 3 years.

There are re-chargers too that get a heck of a lot of use, like in concert halls, churches etc where band mics have their batteries recharged and swapped out regularly. Whilst the 15 minute recharge time is impressive, the recharge life of the cells will be adversely affected compared to the recommended charge rates by reputable manufacturers.

I use a La Crosse BC-900 for recharging my genuine Sanyo Eneloops, all which I imported from a reputable dealer in the USA. The charger is excellent and no fan is required (slow recharging time). After 6 years and hundreds of recharges, the Eneloops are still going strong. Best batteries around in my opinion.

There are fake Sanyo Eneloops made where else but in "copywatch" China and sold on eBay. https://thecounterfeitreport.com/product/524/Eneloop-AA-Batteries.html

[rr100]

Good Something to watch while the test equipment warms up this afternoon...

Not so fast, still traveling through the intertubes I guess...

I wonder how well it does different size (mix of AA and AAA). I ordered one myself for the heck of it, the old BC-900 was starting to have trouble with the LCD (tried to clean it up but seems to be a losing battle). Plus I've a feeling that it fails to terminate charging in time many times.

[AF6LJ]

Good Something to watch while the test equipment warms up this afternoon...

Not so fast, still traveling through the intertubes I guess...

I wonder how well it does different size (mix of AA and AAA). I ordered one myself for the heck of it, the old BC-900 was starting to have trouble with the LCD (tried to clean it up but seems to be a losing battle). Plus I've a feeling that it fails to terminate charging in time many times.

This will be interesting to find out how well it works with mixed cells.

A search online reveals a hand full of 15minute wonders.
I'm not in a big hurry to charge batteries so something like this wouldn't appeal to me.
What does appeal to me is how they go about doing it.

[felixd]

Dave I was wondering while watching 2nd part: Why not to use FLIR to track current flow for discharge and charge, test modes ?
Wouldn't that help to reverse engineer the process? BTW that would also look freaking GOOD to watch how this thing heats up!

I hold my fingers for that shot. Please make 3rd part with thermal images !

[EEVblog]

Dave I was wondering while watching 2nd part: Why not to use FLIR to track current flow for discharge and charge, test modes ?
Wouldn't that help to reverse engineer the process? BTW that would also look freaking GOOD to watch how this thing heats up!

I would have done that, but the iProber is a more interesting tool I haven't shown much.

[oaliey]

When can we expect a teardown of that multimeter?

[oaliey]

If you meant the current probe, Mike had done that.

I am referring to the Brymen BM235. The interwebs is suspiciously lacking any information on this meter.

[SeanB]

Coming soon to an online store familiar to you........

[Svuppe]

That charger probably uses all the methods. -dv/dt is easy at 8 amps.

I found the datasheet of the charger:
http://www.produktinfo.conrad.com/datenblaetter/1100000-1199999/001172036-da-01-en-VARTA_LADER_LCD_ULTRA_FAST_4X_2400MAH_AA.pdf
According to that, the temperature is only used as a backup safety cutoff, when the temperature exceeds 75 degrees C. It doesn't mention dT/dt at all.

I am also curious as to how AAA batteries are handled. If the switchmode is a constant current source, then all those resistors in series with the AAA cells won't make a difference. I may have a look at that myself, as I have the same charger coming in the mail one of these days. I have had the old 15 minute charger, but it is about to give up now (screaming fan, that sometimes won't start until I smack it a couple of times). I didn't know Varta had made a new version until I saw the video.

[apis]

That charger probably uses all the methods. -dv/dt is easy at 8 amps.

I found the datasheet of the charger:
http://www.produktinfo.conrad.com/datenblaetter/1100000-1199999/001172036-da-01-en-VARTA_LADER_LCD_ULTRA_FAST_4X_2400MAH_AA.pdf
According to that, the temperature is only used as a backup safety cutoff, when the temperature exceeds 75 degrees C. It doesn't mention dT/dt at all.

I am also curious as to how AAA batteries are handled. If the switchmode is a constant current source, then all those resistors in series with the AAA cells won't make a difference. I may have a look at that myself, as I have the same charger coming in the mail one of these days. I have had the old 15 minute charger, but it is about to give up now (screaming fan, that sometimes won't start until I smack it a couple of times). I didn't know Varta had made a new version until I saw the video.

Hmm, yeah, maybe dT/dt doesn't work when they use a fan (or maybe it would have a cost a cent extra) and, as HKJ pointed out, at 8 ampere detecting dV/dt should be easy. The battery manufacturers datasheets that Iv'e read usually recommend dT/dt though (and all the backups) as a more reliable method when quick-charging (at 0.8C). But I don't have practical experience, it's just what I've read in datasheets and guides.

I also don't get the purpose of the 1 omh resistors, I assume they are used when detecting if it's an AA or AAA battery inserted but I don't really see how. Maybe they measure voltage drop across to see if there is current flowing through the AAA slot or not? Limiting current to the specified 3 Amps for AAA would proabably be done by limiting the duty cycle like it does when charging more than 2 AAs.

[SeanB]

The resistors on the AAA connectors are likely there for the discharge only, extra resistance so the AAA cell has a lower discharge current. On charge they add an extra voltage and power loss, so the unit likely reduces the constant current for the AAA cell charge, unless they allow you to mix cell sizes in charging. they have to reduce current for AAA as otherwise those resistors will have to dissipate 21W and will cook even with fan cooling.