# Replacing 4.8V NiCd battery with lithium @10 Amps vaccum cleaner (4.6V)

#### mmocz

Joined Apr 26, 2017
14
Hi,

I got a handheld vacuum cleaner and its runtime is quite bad with the included batteries. So I was thinking, hey this will be easy, I can surely come up with something to upgrade it. But I found one problem.

My initial plan was to replace 4.8V NiCd with an 18650 batterypack (1S, 2S) and than either boost or buck it to the desired voltage. The problem is that the vacuum cleaner is running at 10 Amps so none of my small boost converters (rated for 2A) will do. I can get a larger boost converter, but thats kinda useless due to its size. I checked the motor and it is rated for 4.6V.

So is there an another good solution to get 4.6V from lithium batteries which would be able to handle at least 10A?

Thanks a lot!

#### Alec_t

Joined Sep 17, 2013
12,257
Have you checked the sizes of available buck converters, to provide 10A when run from a 2S battery?

#### Bordodynov

Joined May 20, 2015
2,916
See

#### mmocz

Joined Apr 26, 2017
14
Have you checked the sizes of available buck converters, to provide 10A when run from a 2S battery?
Well I have checked aliexpress and ebay and all I have found are buck converters with large heatsinks. I need sth a bit more portable.

Thx Mr. Bordodynov for that diagram, I am gonan check it out later as I am a bit new to electronics and it will take me some time to understand whats happening there.

#### Janis59

Joined Aug 21, 2017
1,355
NiCd has 1V2 on cell, Lithium has 1V85 but them are produced in pairs, thus You can work or with 1+1/2 Li standard cell or use a two or one cells with invertor. As You have a high amper*hour need, my advice is to look for LiFePo4 batteries (typically 60 000 recharge cycles!!!!) like these or similar:
https://www.ebay.com/itm/6pcs-3-2V-...134942?hash=item2396d1009e:g:a8IAAOSwXf1aT08K
https://www.ebay.com/itm/3-2V-100-A...287028&hash=item239760508f:g:uVgAAOSw61haWNw7
Only problem You shall get then - its that charger You shall build Yourself: the price for specialized intelligent charge controlller IC is <<1 USD, the price for intelligent charger is 200-400 USD.
Ouch, ah there the real money is growing well?

#### MrSoftware

Joined Oct 29, 2013
2,015
How about use a mosfet and a timer to PWM the motor? Use a battery that gives more than 4.8v and adjust the PWM until you get the desired average current.

#### philba

Joined Aug 17, 2017
960
How about use a mosfet and a timer to PWM the motor? Use a battery that gives more than 4.8v and adjust the PWM until you get the desired average current.
Hmmm, that may work well. Kind of crude but would have better efficiency than any buck/boost converter. The motor will probably draw 50% (7.4/4.8) more current with a 2S LiPo/LiIon battery so duty cycle would need to be about 2/3. Need to check to see if the motor is overheating, though. Also, will need a pretty beefy battery pack to deliver 10A average.

#### mmocz

Joined Apr 26, 2017
14
How about use a mosfet and a timer to PWM the motor? Use a battery that gives more than 4.8v and adjust the PWM until you get the desired average current.
This sounds really interesting and maybe doable even for somenone with my skill level! Thanks a lot

I will probably try out to use a mosfet with a 555 chip to create the PWM signal.

#### sghioto

Joined Dec 31, 2017
2,398
Probably just needs a new battery. Is the vacuum new? Are the batteries new?
Steve G

#### mmocz

Joined Apr 26, 2017
14
Probably just needs a new battery. Is the vacuum new? Are the batteries new?
Steve G
Youre right, but wheres the fun in that? And I can get better run time from the 18650.

So far I have put together a NE555 and IRLZ44N cirucit which "bucks" the voltage the way I need and soon I will try it with the needed 10A, I am looking forward to what will burn out. But so far this looks like a really good idea, thanks Mr. MrSoftware.

#### sghioto

Joined Dec 31, 2017
2,398

#### ian field

Joined Oct 27, 2012
6,539
Hi,

I got a handheld vacuum cleaner and its runtime is quite bad with the included batteries. So I was thinking, hey this will be easy, I can surely come up with something to upgrade it. But I found one problem.

My initial plan was to replace 4.8V NiCd with an 18650 batterypack (1S, 2S) and than either boost or buck it to the desired voltage. The problem is that the vacuum cleaner is running at 10 Amps so none of my small boost converters (rated for 2A) will do. I can get a larger boost converter, but thats kinda useless due to its size. I checked the motor and it is rated for 4.6V.

So is there an another good solution to get 4.6V from lithium batteries which would be able to handle at least 10A?

Thanks a lot!
SLA batteries are easier to get close to the original voltage - there used to be Cyclone cells that were cylindrical SLA, but I haven't heard of them for years - probably not in a compatible size and usually have 2 spade tags on top instead of contact at each end.

#### ro0ter

Joined Jul 9, 2013
14
given the context (retro-fitting a dead vacuum cleaner with li-ion batteries), I would also have a question... My vacuum cleaner uses 3xNiCD batteries (which were properly disposed). It was regularly drawing 4A, but the batteries were so dead that they only hold up to 10 seconds... I tried to add a 2P configuration of Samsung ICRs (ex-Makita power pack, first two cells were killed by battery circuit, I managed to recycle the other 8 ones as they were in perfect condition: ~1400ma, so close to its original 1500, battery pack was 5s2p).

Some questions are in bold, help is much appreciated. Currently Im only at theoretical level and no, I dont own a scope...

My question is... Same as above.. but I have the following plan (for 1 year already... not much spare time with baby1, he doesnt give me a rest... and baby2 is on her way ):
* ATTiny85
* BSO110N03MS - RDsOn is at most 17mO for 3.2VGS... 8.5A@25*C@4.5VGS/6.6A@90*C@4.5VGS continuous, source input capacitance max 1500nF, output capacitance max 520pF etc (see datasheet for all details)
* 1 x beefy toroidal inductor from dead computer smps (smooth out square wave from mosfet to motor)
* 2 x 0R05 1206 smd (current sense)
* 1 x r-c filter (smooth voltage drop over current sense resistors for AVR ADC feedback, differential mode, 20x gain)
* 1 x china li-ion microusb battery protection circuit + charger (uvlo is the most important feature here) - http://s.aliexpress.com/nyiUf2uM for instance

So charging is straight-forward, no need to insist here.
Next, the AVR is fed from the 1A china charger/protection circuit. It wont ever trip the over-current protection, its a simple AVR... However, when the UVLO engages, AVR will die and mosfet will switch off.

The motor is fed directly from the battery through the mosfet (in parallel with an extra reverse voltage protection (flyback) diode) and through the bulky inductor (to spare the motor from the square PWM wave, although motor is another huge inductor.. but still....). Again - if AVR is dead, the motor stops (mosfet is turned off).

I am planning to sense the current with the AVR (20x differential gain mode) and ajust the pwm duty cycle as required. I had made some calculations based on the mosfets turn-on, rise, turn-off and fall maximum times (7.8ms, 4.4ms, 9.5ms and respectively 4.4ms) and I can run it safely at 22mhz (wont go above 8mhz/16mhz as its useless and practically impossible with this microcontroller... moreover - I would need a driver).

I am also planning to use internal temperature sensor, therefore I must thermally bridge the mosfet with the AVR and sense its temperature. Too hot and I will reduce the PWM frequency (or duty cycle?). What should I do here? Should temperature still increase and I`ll turn off the system... Would the AVR suffice as a mosfet gate driver? Should I add a ~200ohm resistor in series just to limit the harmonics?