Turnigy MAX80W Teardown
I have been using a cheap IMAX B6AC clone for a while now which works ok. Accuracy is not very good but charges my lithium ion cells ok. But I felt I needed something a little better made and I don't need the AC part so 12V DC input is fine.
I looked around and found that the Turnigy MAX80W charger should suit my purposes. And it can do higher power (80W vs 50W of the Imax) so i can charge larger battery packs faster.
Once the unit arrived (very quickly I should say - ordered friday night and it arrived monday!) tested it up and verified it is working then proceeded to take it apart.
This is the charger on the operating table.
With the endplates removed and two screws on top, the top plate slides off.
The screws on the outside are all stainless steel.
Back of the top cover.
I don't like these membrane switches but it was the only charger with the specs that suited my needs.
It appears that the PCB was assembled with real tact switches but they were broken off probably after initial testing of the bare PCB.
The membrane buttons line up on the PCB switches. I could see myself rigging some push buttons and pegs when the membrane switch fails in the future.
The fan does not have any labels on it but it is powered from the 10-15V DC input via a transistor switch.
This appears to be an alternative input port. It is in parallel with the barrel jack input power.
80 watts going through a barrel connector is a no no. I shall rig a new connector on those pads for high power use.
The DC input is reverse polarity protected using an 18A mosfet wired as a low loss diode.
Here, we see the balance resistors, some SMD current shunt resistors and the battery jacks.
The 8pin device is a TPC8052-H 18A mosfet for battery reverse polarity protection.
The discharge circuitry uses this IRF730 mosfet as the pass device.
The dc-dc topology is a 4 switch buck boost with a single inductor.
The input side uses an SM4025P SMD Dpack mosfet (could not find datasheet) and a SR56 5A 60V schottky diode.
The output side uses a TPC8052-H 18A 40V mosfet and this MBR1060CT dual 5A 60V schottky diode.
Back of the main PCB.
There are in circuit programming test points or some external communication lines. Not really sure but there are RX and TX pins so...
The control circuitry are all on a separate sub board and there is a large square hole on the main PCB to accommodate the parts on the solder side of this sub board.
Top of the main PCB with the LCD module desoldered.
Other side of sub PCB with the main uC which is an Atmega128.
Here is the deans connector that I will put into the alternate DC input.
I did not have a suitable large gauge copper wire so I used copper shield braid from a coax cable.
The connector installed.
The dual power input mod works and you can only use one at a time. When a barrel connector is inserted, it is used and the added input is disconnected.
This works fine except at 12V and 80 watts, that would mean 6.67A of current will flow through the NC contact of the barrel jack.
Here, the barrel jack contacts are bridged so that there is no risk of melting.
The buzzer is a little too loud for my liking so I put tape on it to make it a little quieter. It is still loud enough for the purpose.
The LCD is black character on blue background.
I didn't like it so I just had to do something about it.
Here is the LED for the backlight.
And here it is yanked from its place.
I tried a cool white LED but it still has the blue tinge to it.
I didn't like it either.
I took a 5mm warm white LED and used a grinder to shape it like the original LED.
And here it is. That's better.
With the new deans power input connector, I need to cut a new hole for it.
It was rectangular-ish so I had to use the mill.
And here it is. It protrudes a bit, unfortunately, but it does not interfere with the use of the original barrel connector so it is fine.
The yellowish backlight tint matches the yellow on the front sticker which is nice.
User interface is the same as the Imax. I don't have a 6S pack on hand so I rigged up two 3S packs in series. In the B6AC, the individual cells reads anywhere from -0.08V and +0.03V difference from my multimeter. The MAX80W reads from -0.02v to +0.02V in the different cells. Not the best but acceptable as it should be within ±0.05V or better.
The build quality and heatsinking of the power devices are better. In the B6AC, only the thin bottom plate is used for heatsinking the regulator and discharging mosfet. In the MAX80W, the entire case works as the heatsink while the sides have fins which also helps in cooling plus the fan. We shall see how this performs and so far, I'm quite satisfied with this charger.
There is an additional function which can measure the internal resistance of individual cells in a pack but a connection to the balance port is needed. This is handy for checking battery pack life/quality. It can also measure individual cells but you'd need to make a sort of kelvin connection using the charger output and 2 pins of the balance inputs.
I fabricated a test jig using PCB stock and pogo pins to make contact with a single cell in kelvin connection style.
This was originally designed for the internal resistance measurement function.
But I found an interesting use.
When charging a single cell, simply connecting the battery connections (leaving the balance plugs unconnected) charges the cell just fine.
The individual cell voltage display reads 0.00 as there are no cells connected to the balance ports.
If one cell is connected to the balance port, it reads the cell voltage directly.
The voltage reads lower as there is a voltage drop on the cables between the charger output and cell.
Once the balance port reads a voltage, the main menu automatically uses this value so it reads more accurately as it compensates for the cable voltage drop.
Now that I know it would be a nice kit to have, I made a better one.
Here is the new one. I added a spring to add tension and take a variety of cell lengths.
The connector that fits on the balance plug. Cell positive goes to the middle pin and negative to the left pin in the pic.
The PCB stock I used now is double sided and the pogo pins are mounted on both sides of the PCB.
The two pogo pins on the cell contact.
Added banana plugs and shrink tubed some parts of the arms.
The keyrings are attached into holes on the PCB stock to make it easier to open the jaws when inserting a cell.
With a cell inserted.
Unfortunately, I forgot to consider 16340 sized cells.
So I added a couple more slots and a 16340 cell now fits.
...And an 18350 cell...
...And a 10440 cell...
And a 17670 cell. I don't have a 26650 cell but I bet it should also fit.
25 Oct 2015:
I read reports of the MAX80W fan going noisy after some use and it seems my unit wasn't immune to the problem so I took the fan apart.
It looks like a decent quality one as the PCB is FR4 not the cheap brown Phenolic ones.
Pressed the bushing out and cleaned it with solvent and compressed air to remove the oil that the manufacturer used.
It seems the oil used is not suitable as it becomes sticky or useless after a while and the motor starts vibrating but not going much faster.
I also removed the orange O-ring since the magnetic attraction of the rotor magnet to the armature adds a slight pressure to this O-ring and slows the fan down.
After adding new oil (used for sewing machines) on the bushing and replacing the O-ring with a washer, the fan is put back together.
Now it is screaming fast and moves a bit of air. I have seen it do over 13k RPM but was not able to take a pic of it.
Having used it for a while now, I can remember seeing two bugs in the software. One is when saving presets. Once you have saved a preset on #1, and then edited the other presets to change the type of lithium cell (Li ion. LiFe, LiPo etc) it also alters the type on the first one. So you have to edit the first one again. Also, when I load a preset and go straight to CHARGE mode and run it directly, the operation commences but the fan does not spin. If you load a preset and push the up/dn button and return to CHARGE then operate it, the fan runs this time.
19 Jun 2016:
So, I was planning on getting a couple hundred 18650 cells from NOS laptop packs for a project and was thinking of choosing the best ones from the lot and the seller said take it all home, charge them and pick which ones I liked and return the rest.
And now I have over 300pcs and charging them all up and testing. Since the MAX80W has an internal resistance test function, I don't have to do a full discharge test on each one saving a bit of time.
The jig above works well but I had to keep the cell from rolling away and align the contacts before closing it then testing. It was a bit tedious so I made this new jig specifically for 18650 cells.
Same pogo pins for contacts and same 4 wire kelvin connection.
One contact was unfortunately coated in solder since I just got these from scrap. cleaned it off and seems to be just fine.
In the bottom side, you can see the spring that pulls it closed. Behind the alu plate is a linear bearing with about 1 inch of stroke.
With a cell in place.
I plugged it on the multimeter and sorted the cells by voltage. The smaller stack are ones that read less than 1V. The bigger stack reads between 1-3V, the rest below are more tahn 3V.
Now it is connected to the MAX80W for charging and testing cells.
All the chargers working overtime. (Not shown is the Nitecore i4 charger also fully booked.)
Later on, I decided it works better with the positive terminal connected to the stationary contacts so I swapped the connectors over.
Here it is charging a cell when there are no fully charged ones to test.
If you noticed, the cuts seemed to go over the opening, I only used the mini table saw for the cuts. The micro mill was not used in this project.
Another shot showing the linear bearing.
Page created and copyright R.Quan © 22 Sept 2015.