Help With Power Tools Charger

[johnd]

hello all, this is my first thread and I hope I came to the right place to ask.
I apologize but my electronics knowledge is very basic.

I have a fault Makita DC18RA-T 110V charger that the company refuses to repair.
its schematics and components list are not published ;-(

The problem is a broken Zener(?) glass diode 'D4' in the AC-DC section.
Can you please help me identify or estimate which diode type should I use to try and fix it?

I attach an image of the AC-DC section of it and I marked some components
to help you understand the design.


Thanks a lot for your kind help
John

This post has been edited by johnd on December 20, 2012 08:25 am

[Geek]

Is it across the MOSFET gate and source? It's a protection diode then and 18V oughtta do, as most MOSFET's have a maximum Vgs of 20V.

But you need to find out why it went kaput. I'd test Q11 and Q12 there.

Cheers!

[johnd]

Thank you so much Geek for your reply

I have no idea if it "across the MOSFET gate and source" ;-(
I can upload the PS side if it helps you figure this out

Q11= C3209 YK736
Q12= A4281 YK741
how do I test it?

additional information:
A friend of mine has this charger but with 220V version and he measured the voltage
across this D4 diode, he told me that while charging, the voltage was toggling between 2.74-2.9 and ~3.2v when he measured it on both directions of the diode.

thanks

This post has been edited by johnd on December 08, 2012 07:31 pm

[CWB]

are there any numbers on the good diode ?

the bobbling voltage that was measured is most likely due to the pulse width modulation used to "control" the mosfet(s) .

yep ... there is a reason that zeners let go ...
over-voltage and old age .
sometimes there is failure "just because" ... but this is rare .
i do not see evidence of a long term "runnin' it up against the wall" ... this would be indicated by a discoloration of the board under the zener .

it looks like it grenaded ... most likely from a healthy spike .
this warrants further testing rather than simply stabbing in a part .
the photocoupler may be bad .
simply replacing the zener and firing it up may cause a repairable unit to turn into a door-stop .

to accurately check the two items geek mentioned , it is best to pull the parts .
if you are not hip with soldering and testing semiconductors , i would strongly suggest finding someone who is .

a possible indicator to the extent of damage :
have you checked the line fuse ?
if it is ok , this is a good indication .

[johnd]

thank you CWB

line fuse is OK.
no readable marking on the diode as it is broken ;-(
and it is a single diode in this area as seen on photo.
no burnt signs under the diode

I can desolder/solder the transistors and I have a multimeter.

from your experience do you also think as Geek that the diode is a 18V zener?
Vgs gate source voltage of MOSFET 2SK3683 is +-30 according its datasheet,
Geek wrote that it is 20V and thus the diode should be 18V.

thanks

This post has been edited by johnd on December 09, 2012 06:58 am

[Geek]

Hi,

Most MOSFET's have a Vgs of 20V, there are some that go higher. Not that it takes the whole Vgs swing to turn it on. If it's 30, use a higher voltage diode than 18

[johnd]

Thanks Gregg

I'll order the 27v diode and try
will report later

thanks for your help
John

[CWB]

the main fuse is good ... this is much better than it being open .

"from your experience do you also think as Geek that the diode is a 18V zener?
Vgs gate source voltage of MOSFET 2SK3683 is +-30 according its datasheet,
Geek wrote that it is 20V and thus the diode should be 18V."

without examining the board to "rough out" a mental image of the circuit , or looking at a schematic , it is not possible to say with certainty that the zener should be 17-18 volts .

personally speaking , the SWAG method should be used only as a last resort .

look at diode in the charger that your friend has (that is still good) ... it may have numbers on it (that is what i was sort of getting at in the first place) .

lacking any good numbers that can be cross-referenced ...
by using a suitable power supply , current limiting resistor , some alligator jumper leads (small) and a dmm , it is possible to check the diode in your friend's charger to find the rating .
you will have to lift one end of the diode to do this .



there are a couple of other items to consider :
R103 : (silk screening partially blocked by the red ellipse) no value discernable
R5 : 12 ohm / 5 watt

these are most likely for safety/current limiting (again , no schematic and based on general design considerations) .
the job of these is usually to limit the current in case of component failure .
R103 may open up in the event that the mosfet (Q1) shorts ... this could be the reason that the main fuse did not open .
R5 may be used to limit the current to the switchmode transformer primary winding .

testing mosfets (ie : Q1) depends on their internal construction .
some of these devices have built in components (as it were) that take the place of external components ...
a likely item is a diode that connects across the drain and source of Q1 .
(the data sheet for this device will confirm this either way) .
testing the device is best done by pulling it out of circuit .

as geek mentioned , checking Q11 and Q12 is important .

[johnd]

Thanks CWB

As mentioned, I could not find a way to receive schematics of this pcb, Makita does not publish it nor do they fix this item. ;-(

I'll try to get my hands on a power supply and measure my friend's diode
but this will take some time to do...

R103 writing is 12 oHmJ 1w and it measures on board 10.4 ohm so I guess it is OK.

R5 writing is 0-12oHm 5% 2w 7N2 and it measures on board 7.55 MoHm so I guess it fault...
I have no idea what type is this resistor as it looks like a ceramic one.
can I replace it with standard 12 oHm 2w resistor ?

thanks

[CWB]

the 12 ohm resistor is most likely ok .
this is where a schematic or having the board right in front of me would help .

i have been looking for a schematic for your charger .
there are a lot of references over at "fixya"
a guy asked about a schematic and some smart-*ss said to contact makita !
(ya know , ignorant twits like that ought to be darwin award recipients)
there are no makita charger schematics over there ... i found a dewalt one though !

it is amazing how many people think that schematic / manual / service manual /owner manual are interchangeable words .


when testing that zener calculate the resistor to limit the current to about 100 mA ... the diode looks to be a 250 mW unit (they may even rate it as 500 mW) .
this should put the zener well into conduction and provide a good reading .
too low of a resistance will pooch the zener and cause your friend to become somewhat agitated with you .

[johnd]

Thanks
sorry to nag CWB

what about the "R5" 0.12 oHm resistor?
it looks to be 0.12 and not 12 ! here is a close look on it from both sides


when I measure it on board, it reads now 7.55 MoHm !

should I replace it and can it be a regular resistor ?
I have no idea what type is this resistor

thanks

P.s
I wish I could find the Makita schematics, but I did not
As for the Dewalt schematics that you have found, is it the following Dewalt drawing which is blurred ?, can you pls send the url of what you found?

This post has been edited by johnd on December 09, 2012 01:59 pm

[CWB]

yep ... that is the dewalt schematic .

7.5 megs is a far cry from 7.5 ohms .
it would seem to be open . the best check is to lift one end .
if it is indeed open , then the mosfet has most likely undergone internal silicon to carbon transmutation .
this indicates what took out the zener and also points to the other two transistors and the optocoupler quite possibly being toast .

when the fet shorted this could have put a helluva spike on the input of the swmps transformer and thusly there may be further components on the secondary side that are bad .

i will say this ... the device did not go into the "smoke on the water" mode when it failed .
the safety stuff did it's job .

as for how to check fets and bjts , there are several tutorials (with drawings) on the web about how to do this with a dmm .
the procedure is actually somewhat simple (there are some do's and don'ts) but somewhat "wordy" to explain and as we all know , "a picture is worth a thousand words" .

a point about swmps ...
unless one thoroughly checks the components and replaces all the defective ones in the first attempt at repair there is a chance that all the new parts one puts in will fail within the first .001 second of power-up .

[Sch3mat1c]

Here's Q11, not that it matters:
http://www.datasheetcatalog.org/datasheet/kec/KTC3209.pdf
Methinks Q12 is actually "A1281". Nice little Japanese transistors. Or maybe they're Korean, I don't remember.

Do you measure any resistance (or diode test voltage) across any of the pins on Q1? If it's still okay, then you may find a high resistance (megs?) from G-S (pins 1 and 3, looking at the face of the transistor, reading left to right), or around 0.7V or something like that, both of which are due to Q11 and Q12 being there. It should measure as a good diode D-S (pins 2 and 3).

If it's failed, It's probably a three way short. If it was a really bad failure, it may've passed "well done" and went into the "charred to ashes" phase, in which case it may be an open circuit again. But that usually happens after the magic smoke leaves.

Y'see, semiconductors are made with a very specific dose of magic smoke. Heat creates magic smoke, and semiconductors are manufactured with tightly controlled amounts of heat to get just the right properties. If you overheat a semiconductor, you may create too much magic smoke, in which case the device reads with lower resistance, or it becomes a short circuit. This is a special amount of heat, just enough to cause failure, but not spectacular enough to explode, freeing the smoke. When that happens, the smoke leaves entirely, and with the smoke gone, it usually ends up an open circuit.

Tim

[johnd]

Thanks Tim

what I measure on Q1 is:
S-G -15.5 KΩ - OL V
S-D - 7.66MΩ - OL V
G-D - 1.08MΩ - 0.53 V

[deth502]

you shouldnt have an overload on volts. if your meter is not auto ranging, you need to set it to the next higher setting.

[CWB]

nope ...

first ... these are resistance checks that are to be performed .
this is done in a non-powered state ... ie : unplug the unit and discharge the filter cap with a 1K resistor .

second ...
note that i mentioned removal or lifting one end of the part in question . obviously , with a resistor , lifting one end is all that is needed .
for the person that is not familiar with component checking and the vagaries involved , this is the absolute best method of checking as it eliminates the possibility of another component in parallel with the part you are tying to check producing erroneous results/readings .

third ...
many dmms are equipped with a "diode check" function .
this applies just enough voltage/limited current to safely check the junctions in semiconductor devices .
yes , one can destroy a perfectly good device by checking it "the wrong way" , the old low noise gaasfets come to mind .
(there are plenty of tutorials about the hows and whys of semiconductor checking online)

for the time being i see no real need to have the unit powered up until certain other items are checked/proved out .
this is safer and will help avoid any further damage to the unit .