turnigy accucel 8150 old charger repair

whatever

100 kW
Joined
Jun 3, 2010
Messages
1,297
TURNIGY ACCUCEL 8150 is model name

So I was using an old turnigy charger the other day on a 7s 10ahr pack ( 24v nom), and I think I put the wires back to front, got a bigger than usually spark, a loud crack and smoke from inside the charger.
Charger got stuck on reverse polarity on screen, and simply wont work anymore.
So opened it up and I see one fet is blown ( obvious crack ) and suspect the one next to it also blown ( 2 fets furthest to left in pic below). I'll detail this repair as I did an icharger repair recently
https://endless-sphere.com/forums/viewtopic.php?f=14&t=49661&start=50
Below shows the case and inner board
turnigy.jpg

one blown fet indicated by red arrow ( suspect the one next to it also), there is a small black electroytic cap in between the middle of all the fets, are they supposed to be black? I can remember only seeing red ones.
edit: postcript: at this point I had assumed all these 8 lots of t0-220 were mosfets.....wrong!
L7805cv is 5volt supply and I suspect nfa14 are schottkey diodes!

turningy1.jpg

Its been a good charger, reliable over the years but not used daily. These were cheaper than the ichargers and looking at the fets is one reason I can see why, the fets are a mix up of all different ones.
Here is the list of what they are from left to right:
irfz44n( blown), irf3205, nfa14, irfz44n, nfa14,irf4905, L7805cv and irfz44n
Theres only 8 fets and they are almost all different.
So i'll go through similar repair procedure to the icharger and document what I find.

Plan to replace all the fets with irf3206 genuine from same aliexpress source as documented in icharger thread linked previously.
 
L7805cv unless I've read it wrong might be a voltage regulator, I assumed they were all fets.
 
A bit more on the turnigy repair. Fortunately the screen has removable type connectors so easy to take off the screen makes work a bit easier.
A pic to show the 7 mosfets and one 5volt regulator and their positions.
mosfet details.jpg
It appears quite a mismatched bunch of fets on there, I would assume like types are usually next to each other, I could be wrong, but seems they just have a bucket of fets of similar ratings and just used whatever at hand.
If anyone knows differently would be interested on info.
edit : postscript:
Please note nfa14 may not be mosfets, I cant find datasheet ( maybe schottkey diodes?? I realised this once finding hobbking charger has two schottkey diodes rather than mosfets on the 8 lots of t0-220 packages)
 
Mosfet irfz44n furthest on left is blown, there is a visible crack in it, and you can smell its burned internally.
blown irfz44n.jpg

This repair I'm trying to test the fets in situ, last repair I did I just took out all the fets and replaced them ( icharger), this time trying to save work time. All the fets appear ok using multimeter testing resistance, except for the last one on the right gives strange readings. So at the moment suspecting that one also but not sure as yet.
First one the left have removed it as its obvously blown, it has blown so that there is continuous contact/current flowthrough the gate to the drain leg ( left and centre leg). No current flow through drain/source legs.
 
There is a voltage regulator on the turnigy as shown previous diagram, its an L7805CV, so its supplying 5volt to some components. Since last repair was icharger208b and it had all mosfets on this portion of the board, I had at first just assumed all these components were mosfets ( icharger had 5volt supply underneath the screen).
So its quite different layout to the icharger 208b which is same vintage/era.
(postscript: similar to imax b6 in layout)
Pic below shows the regulator up close, can see its middle and right legs have very thin traces, so definitely is meant to be regulator there ( not mosfet).
volt regulator 7805.jpg
 
Just for anyone who is interested the processor it atmega32A
atmega 32A.jpg
 
So I decided to just replace the one obviously blown fet, used an irf3607 genuine ( secondhand fet ), and fingers crossed, unit turned on and is charging again. So problemo solved for now.
edit: postscript:
turns out that was very lucky decision, as not all these to-220 components are mosfets
One thing I did notice is that there is no spark when initially connecting to a 7s ( 24v nom) lithium ion pack, which indicates the main capacitor is not charging up, normally there is a pretty decent spark on all the chargers I use ( all older models similar to this turnigy), so it will mean opening up again and finding some decent capacitor to replace original one. Probably wont hurt to replace all the electrolytic caps.
For now though it fixed.
Pic below shows replaced mosfet furthest to left, I needed to add a little bit of tabbing wire as the high amp trace was blown off the board when old fet failed
turnigy fix1.jpg
testing repaired unit:
turnigy fix2.jpg

Also I felt the 5v supply ( no.7 position looks like a mosfet), without any load the 5v supply was slightly warm to the touch.
 
Just some thoughts on these older style chargers. I think they are quite the piece of engineering, considering all the functions they can perform and generally reliable in my experience. In my view they are works of art.
 
went to change main cap in charger, only have higher voltage one, too big..........just a thought if no spark when connect charger, simpler fix without opening would be just put another cap across the output wires on outside of charger should work also.
 
Food for thought: Depends on what the cap is intended to do. If it is there to filter the input voltage for noise or voltage ripple, then it won't matter if it's inside or outside. If it's there to reduce ripple currents on the internal conductors, FETs, coils, etc from the device operation, it may not do the intended job if it's not in the original place.

I don't know enough about the design to say anything about what they intended, though.
 
thanks amberwolf, I decided to compare icharger 208b and turnigy accucell 8150 ( basically same function/size chargers 8s 150watt approx) in terms of input and output caps.
Icharger probably being the better of the two I'm assume the design might also be a bit more advanced.
Icharger input cap 36v 1000uf ( rubycon) output cap 50v 1000uf ( rubycon)( also a ceramic cap 250v 105J)
turnigy input cap 25v 2200uf ( nichicon) output cap 63v 470uf ( huang)( no ceramic)

some pics to compare locations/details
Icharger input cap close to input wires
icharger input cap.jpg

Icharger output cap close to output wires ( ceramic right inbetween output wires)
icharger output cap.jpg

Turningy input cap close to input wires
turnigy input cap.jpg

Turnigy output cap a bit further from output wires but close to middle of fets
turnigy output cap.jpg
 
I dont think it would be a problem to add the caps on the outside of the charger probably quite close to the case would be best position. I am certainly concerned that there is no spark on the turnigy output wires when connected to battery, pretty sure if I tested that cap it would be very low capacity, think I will take it out and replace.
 
I managed to squeeze into the box a 100v 470uf capacitor across the output wires, the capacitor that didn't spark when battery leads attached still had some capacitance ( wasn't shorted) so left it there, new capacitor just adds additional capacitance, getting a nice spark now. So that project done.
Onto next project is to compare hobbyking eco8 charger to turnigy and icharger of same ilk.
hobbyking eco8 teardown/mods here:
eco8 hobbyking charger repair
 
Last edited:
Noticed repaired turnigy getting hot whilst charging 24v nominal at 1 amp, also the charge voltage was higher than an identical pack that sits in series, let it go for a while, but decided to play it safe. Turns out the inductor was getting too hot, I'm suspecting the large capacitor I added across the output wires might be the culprit.
Have taken that out and no heat apparent when turned on ( the 5v supply gets a bit warm), haven't yet tested in charge mode will do shortly, but at this stage it appears the charger doesn't like the additional capacitor ( too my surprise).
I will just replace the suspect capacitor with one of same ratings.
 
removed the additional large cap I put on the output wires, and charger went back to normal, so I can say puting additional caps on the output wires ( on inside or outside the box) probably not going to work.
I then thought I would try the charger on discharge mode ( which I never use) just for the heck of it, I've read some info where discharge mode seemed to be when fets blew.
The very thin wires I used started smoking and charger said "short circuit" or something to that effect.
I was discharging a 24v nom. lilo pack at 0.3amp, all fine, put it up to 1amp whilst discharging and thats when the short circuit occured.
Cant see any physical signs of damage to mosfets, so will have to remove and test all the fets to find which one failed ( assuming it is a fet failure). The 5v supply to-220 package is fine.
 
So I'll just go back over the to-220 packages on this charger.
Pic below shows all packages are charger was before any failures.
mosfet details 1.jpg

Next pic shows which ones are not mosfets:
mosfet details 4.jpg

Next pic shows the mosfets, and which ones have blown.
mosfet details 3.jpg
In the pic above initially fet 1 blew ( in charing mode) I replace that mosfet with irf3607.
I also added some an additional capacitor across the output wires, last time I went to use the charger there was an issue, not charging correctly. Removed that additional capacitor and went to test charger.
All well.........but decided to try in discharge mode, see if any weaknesses there, and indeed fet 2 blew.
So thats where i'm up to.
Tried to locate possible blown fet using multimeter whilst fets in situ, without success. Also thought maybe a capacitor might be shorted but couldn't test either of those two big caps either since everything just showed up as shorted.
So removed all the mosfets ( left the two shottkey diodes and 5v supply on board), turns out the second fet from left ( pic near top of thread), had blown. Its irf3205. So the mosfet that I replaced before ( furthest on left) was irfz44n, replaced it with irf3607, that one was still ok.
So in discharge mode it looks like second fet from left might be switching the current ( just guessing).
I shall replace the irf3205 with irf3607 shortly, replace the fets and test again. Probably wont test in discharge mode though, as I only need it for charging, dont wont to risk blowing another fet, unless in charge mode.
 
glad to say the turnigy lives again, replaced all the mosfets, I just put all the same kind, irf3607 since I had enough and they were all tested as working. I am too afraid to try it in discharge mode, which I dont use anyhow, so will just use it in charging mode as I usually do.
One thing I did learn recently is that in circuits in general it appears to be capacitors which do decay over time, the capacity drops the rate depending on the initial quality I guess. I've always wondered at what rate do components decay. Mosfets I have no idea if they decay over time, they seem to have a long lifespan.
So in future I will be also suspecting capacitor failures and thinking of them as something to keep an eye on.
This charger only has a tiny spark when I first connect it to a 24v nom. pack, have ordered some replacements cap do be done at later date ( only talking about the two large caps on the board).
I anyone is interested I have simple technique for cleaning out solder from the holes through the board where the mosfets go ( they fill with solder when removed).
I just get a safety pin, stick it on the hole filled with solder with some light pressure down, put a little solder on the iron, touch that onto the bottom of the safety pin needle where its touching the solder in the hole and the pin will easily go through the hole, taking out all the solder with it.
Scrap the solder off the safety pin before withdrawing it through the hole.
Wiggle the safety pin in the hole firmly just too make the hole big enough for the mosfet legs to push through, its usually a tight squeeze, there is metal inside the holes which needs to remain in place, its only the solder needs pushing out. So only requires soldering mosfets from bottom of the board........hope that makes sense, works extremely well.
 
Electrolytic capacitors have a rated lifespan at a rated temperature. There's a number of good pages about it
https://www.google.com/search?q=capacitor+lifespan+vs+temperature+chart
but the manufacturer pages for a specific part they make will be the most accurate, based on their own test data. This is one of the general info pages from one manufacturer
https://chemi-con.com/capacitor-life/


This is a useful appnote for our purposes, since power supplies, chargers, and controllers all have similar usages and internal conditions:
https://www.xppower.com/resources/blog/electrolytic-capacitor-lifetime-in-power-supplies

This is a generally useful article with a little more accessibility to the average person
https://www.allaboutcircuits.com/news/calculating-the-lifespan-of-electrolytic-capacitors-with-de-rating/


FETs and other semiconductors should last much longer; like most things temperature cycling factors in, and how far within the ratings they are used (the closer to the edge, the more likely a failure can be).
 
thanks amberwolf for the positive assistance. I shall read through those links you posted.

So tested the charger on a depleted 24v 10ahr battery, everything worked apart from the torroidal coil on the board getting extremely hot. So there is some other issue going on apart from the fets.
Though I have identified the two schottkey diodes in to-220 packages on the fet section as mbr2045ctg, I was using the top letters NF A14 as the identifying letters but its the lower letters mbr2045ctg which identify those components. they are 20amp 45v rated, 10amp each leg.

schottkey diodes.jpg

it is possible during installing new fets, that I damaged the schottkey diodes legs, as they bent quite a few times and I will check that if I have weakened the legs too much or solder joints. So the plot thickens.
The diodes connect directly to the torroid coil so its somewhere to start for diagnose new issue.
 
Just after reading amberwolfs google search links, capacitor cooling does seem to be a relevant way to extend electronics lifespans. Very interesting indeed. So capacitors are in some way a weak link in electronic boards it appears.
 
aha..........just realised that I replaced all the fets with same type irf3607 which are N chanel, and one mosfet on the board is a P chanel, the irf4905, which is third from the right is P chanel, so that may well explain my problem with the torriod coil ( inductor getting hot), if nothing else damaged fingers crossed it might live on.
If the original irf4905 that I took out is ok, I will put that back in place and try again.
Hope I got it right this time

mosfet details 5.jpg
 
initially I was thinking that all the different mosfets were just different ones that the manufacturer had on hand, but it really in the case they are all different for good reasons. I now understand a little bit better that alot of design and thought went into this charger.
Just to make it a little more concrete here is list of the original mosfets on the board and their ratings.
Now i'm guessing their ratings in particular the amperage also has significance in the design process.
I would hazard a guess the irf3205 with highest current rating experiences the highest current load, its rated at 110amp, the irf4905 rated at 74amps probably has a bit less load, and the three irfz44n fets the lowest load being rated at 49amps. I'm not too worried using all irf3607 being rated at 80amp (75v) as replacements, I'm only using the charger at 1amp so should be well within spec, I doubt will be any heat detectable from the fets ( of course the irf4905 I will replace with original). If this charger does work again after this, I would say its a pretty damn robust charger......if it works!
mosfet details 7.jpg

One other thing I learned is that the discharge current for this charger is quite low rating, when I tested in dishcarge mode I turned up the amps from 0.3amps, all was fine, to 1amp when there was a failure.
At 1amp on a 24v nom lilo pack its pretty close to the max spec of 25watt max discharge power rating of the charger.
 
so i've found a good source for the exact circuit dia of this charger is here:
https://03874331818329849343.google...mIEaru1EpTfn44POw-EhnROd-2JT5zOlRDqD48la6c9q0
and a very good discussion on finding a charging fault ( from an expert on these chargers)
https://groups.google.com/g/cheali-charger/c/akMlGo5ZtiE?pli=1
I would like to post the circuit dia but probably need permission from the source, if the links above dont work at some stage I have saved those files on my computer if any needs.

So from circuit dia linked above, if charging problems the items to look at are
Q15 : irf4905
Q14: irfz44
L1: torroid coil inductor
C7: 470uf
Q22 and check r150 is correct ohmns ( should be below 0.035Ohm)
D2 and D3 are the schottkey diodes: part number MBR2045CTG
r150: wire shunt should be 0.35ohm or less ( not sure which shunt wire is r150 there are only two on the board near each other)......Q22 links to it
 
Back
Top