Hyperion EOS 1420i NET3 14s Balance Charger...

[dnmun]

I have looked over the board with a fine tooth comb.
No components look off at all.

Just dead on the inside, like a goth kid..

what did you do to analyze your failure besides look at it?

did you test the mosfets or look for voltage on the output of the voltage regulator?

 

[neptronix]

I don't have the skills to really do that.

Do you want to buy the dead unit off me for i dunno.. $20 shipped? maybe you can discover what ails these things.

 

[neptronix]

Or how about i ship it to you, and if you can fix it, i'll pay you to post up the fix ( this is the most common mode of death ) and send me the returned unit.

I paid $125 for it, so there is some room for me to pay you to get it revived.

 

[dnmun]

i have not been able to fix the hyperions. zippie sent me his and i could not fix it. but there may be something simple wrong with it and mostly this is just a way to educate people who wanna fix their stuff.

 

[Ratking]

have you ever repaired anything before? do you know what a transistor is? your pictures are not useful, just too big for the page. you should always use VGA settings for pictures here. most of them at 4110 n-channel mosfets and the 4905 is a p-channel mosfet.

I got a bachelor degree that says that I should be able to
I've been changing obvious bad components and done some easy work with mosfets before, but never done extensive troubleshooting without a schematic.
But I will try to find out where the voltage stops and see if I can check the mosfets and voltage regulators in circuit and give some feedback.

 

[dnmun]

if you stand up the transistors, you can see there are some that are schttky diodes, and some are mosfets. the 4110 are n-channel mosfets and the 4905 are p-channel mosfets.

test the 4110 with your diode tester by putting the red probe on the drain leg and the black on the source leg. your diode tester should read open circuit. then reverse the probes with black on the drain and red on the source and it will tell you the forward bias of the body diode.

for the p-channel mosfets, 4905, you do the opposite. put the red probe on the source leg, and the black probe on the drain and you should see open circuit. then put the red probe on the drain and the black on the source and you should see the forward bias of the body diode.

the mosfet legs are gate, drain, source from left to right.

both of those schottky diode are called common cathode because the outside legs are the anode side and the center leg is the cathode. you can test them with the diode tester also but i doubt if they failed. more likely that the 4905 has failed, but check them all first.

if you look in the corner where the input voltage come onto the pcb you will see the 7805 three terminal regulator. go google the data sheet for LM7805 so you understand what it is doing. middle leg is ground, left leg is voltage input, right leg is voltage out. it should read 5V with respect to the ground when the charger is powered up on those input wires. so plug in the power supply to test for the 5V output of the regulator.

 

[Ratking]

if you stand up the transistors, you can see there are some that are schttky diodes, and some are mosfets. the 4110 are n-channel mosfets and the 4905 are p-channel mosfets.

test the 4110 with your diode tester by putting the red probe on the drain leg and the black on the source leg. your diode tester should read open circuit. then reverse the probes with black on the drain and red on the source and it will tell you the forward bias of the body diode.

for the p-channel mosfets, 4905, you do the opposite. put the red probe on the source leg, and the black probe on the drain and you should see open circuit. then put the red probe on the drain and the black on the source and you should see the forward bias of the body diode.

the mosfet legs are gate, drain, source from left to right.

both of those schottky diode are called common cathode because the outside legs are the anode side and the center leg is the cathode. you can test them with the diode tester also but i doubt if they failed. more likely that the 4905 has failed, but check them all first.

if you look in the corner where the input voltage come onto the pcb you will see the 7805 three terminal regulator. go google the data sheet for LM7805 so you understand what it is doing. middle leg is ground, left leg is voltage input, right leg is voltage out. it should read 5V with respect to the ground when the charger is powered up on those input wires. so plug in the power supply to test for the 5V output of the regulator.

Thank you, you made it almost too easy for me Lets hope some of them are dead at least. I am a bit worried since there is no smell or visual damage. It just stopped to work, from perfect operation to dead next time I tried to use it.

 

[dnmun]

there is only going to be just one part that is broken and not lots of them. the one i have here to look at is all covered with smoke so it is hard to see anything.

 

[RVD]

I did a few searches in this thread but didn't see the problem that I observed last night.

I was doing a storage charge on some 6s 5ah lipos. Most of the cells were around 3.95v and when doing a storage charge to 60%, it would discharge (as expected) and finish up at around 3.82-3.83v/cell. This is fine and expected.

However, there was one time during this process when I moved the battery slightly while it was charging and plugged in and it gave me a battery output connection error. I simply restarted the storage charge and it seemed ok.

However, a short time later on another battery, I did the same thing but noticed that it was charging rather than discharging as soon as the storage charge started even though the cells were at 3.9+v. I checked the menus and it said that the battery is only around 40% charged so the charger is thinking that it needs to charge up rather than discharge.

This kept happening a lot to a lot of batteries where the % reading was inaccurate causing the storage charge to charge rather than discharge.

After some frustration, I just used discharge mode to get the cells down to 3.82v and just stopped it there. However, the cells did not go through a balance.

Any idea what is going on here? I didn't try letting it storage charge for a long time to see if the % reading gets accurate later...e.g., I wonder if the charger will still think it's only at around 50 or 60% when it's charged up to 4.1v, etc...which can get dangerous if it keeps charging up to my 85% cutoff.

 

[dnmun]

@ratking: i gave instructions on how to test the mosfets that involved testing with the diode tester. i had done this before to the charger zippie sent me. had forgotten, now i remember why i forgot. the numbers do not make sense when measured while in the circuit.

there are resistors and capacitors tied to the legs of the mosfets so the meter readings do not conform to what i had said to expect. i switched over to using the ohmmeter to measure the drain/source and gate/source and found numbers that reflected the fact that the other components are masking a test of the mosfets while in the board. that makes it very hard but i persevered.

i kept looking at the body diode measurements and i found one of the 4110 mosfets has a very very low value. .24V when i expect it should be about .45V as it is for the other 4110 mosfets.

so i have decided that this mosfet is the one that has died, and shorted so the forward bias is a very low value because of the short. that is my assumption. i also can see discoloration of the metal tab on the very end as though it was overheated. clue #2.

would you post up a picture of the inside of your box showing the location of the 3 twin wires with 2 pin plugs on the end?

2 of them go to the fans and one goes to the display. i lost track of where they should be plugged in. i think i know but would like to see a picture inside.

anyway, i am gonna replace this one mosfet and see if it will work. it is the 5th package from the left side, just to the right of the little thermistor bulb sticking up there in the middle. will post up later, but post up a picture for me showing the locations of the 2 wire plugs.

edit: nope, not that. i cut out the mosfet and measured the body diode and it is ok. .45V. i think the .243 reading is from one of the schottky diodes next to it but i do not see how it is wired up in parallel with the mosfet.

so i destroyed the pcb now trying to unsolder the cut off legs from the board so i am now out of options to test this one.goddam parkinsons makes it impossible for me to hold the goddam soldering iron and i ruined the surface plating on the pcb, don't frocking care now anyway..

 

[Ratking]

I am writing on an Ipad, so I will keep it short. The measurement are all over the place. I de soldered one 4110 fet and it was shot. Infinte resistance between d and s. I tried to measure other fets aswell but it seems that several of them are destroyed. The voltage regulator does not produce any voltage either. I will take another look tomorrow

 

[dnmun]

drain to source resistance should be infinite. when you think about how the n channel mosfet works, the drain is always connected to the higher voltage and the source is connected to the lower voltage. so if it had anything less than infinite resistance, then current would flow all the time.

because of the way the mosfet is constructed using diffusion of the dopants from the surface and in different concentrations to create the channel in the substrate that is the transistor, then that dopant profile creates a p-n junction in the silicon between the source and the drain, through the substrate and it is called the body diode.

i look for the body diode measurement myself, other people just use resistance to tell if the mosfet is shorted.

the reason the numbers drift all over the place when you put the probes on the drain and source legs is because there are big capacitors to smooth out the current spikes as that mosfet switches on and off very fast. i think the schottky diode is next to it because it may be there to help absorb the induction spike when the mosfet switches off and the voltage on the mosfet can be reversed. that allow current to flow through the mosfet when it has high resistance and that causes the mosfet to overheat and fail. i think they call then flyback diodes.

anyway, i was seeing numbers between 1.6-1.9 and up and it would continue to climb as the current from the diode tester charged up the capacitor connected to the legs of the mosfet. so that is not useful as a way to test them. and your mosfet that has infinite resistance from drain to source is ok. the body diode should measure about.45-.46V as i recall.

 

[Ratking]

drain to source resistance should be infinite. when you think about how the n channel mosfet works, the drain is always connected to the higher voltage and the source is connected to the lower voltage. so if it had anything less than infinite resistance, then current would flow all the time.

because of the way the mosfet is constructed using diffusion of the dopants from the surface and in different concentrations to create the channel in the substrate that is the transistor, then that dopant profile creates a p-n junction in the silicon between the source and the drain, through the substrate and it is called the body diode.

i look for the body diode measurement myself, other people just use resistance to tell if the mosfet is shorted.

the reason the numbers drift all over the place when you put the probes on the drain and source legs is because there are big capacitors to smooth out the current spikes as that mosfet switches on and off very fast. i think the schottky diode is next to it because it may be there to help absorb the induction spike when the mosfet switches off and the voltage on the mosfet can be reversed. that allow current to flow through the mosfet when it has high resistance and that causes the mosfet to overheat and fail. i think they call then flyback diodes.

anyway, i was seeing numbers between 1.6-1.9 and up and it would continue to climb as the current from the diode tester charged up the capacitor connected to the legs of the mosfet. so that is not useful as a way to test them. and your mosfet that has infinite resistance from drain to source is ok. the body diode should measure about.45-.46V as i recall.

Hello,

The problem whas that several of the mosfets was infinite or open both ways. meaning measured both d-s and s-d. Like for the diodes, they appeared to be blown too, since they measured infinite resistance for one of them and the other 850 ish ohms. Measured in the blocking direction

 

[mrpk]

I see this charger is now discontinued - can someone tell me a suitable replacement - im a bit of a greenhorn when it comes down to lipos and i don't want to get it wrong.
Im building my first ebike - have all the parts and now find out that the charger i had identified as the one for the job is no more...
My battery setup is one from i copied on this thread.... on page 9
Thanks in advance.
PK

 

[motomech]

For 12S and under, we like the Thunder 1220. Unfortunately, it's out of stock at the moment.

http://www.hobbypartz.com/75p-1220-charger.html

For 13S and 14S, this is the only balancing charger;

http://www.amazon.com/Thunder-Power-RC-TP1430C-Multi-Chemistry/dp/B00681QVOY

 

[mrpk]

Thank you motomech,
Is that 12s unit ever in stock though? I find more and more recently that websites just leave items on display as out of stock but then never re-stock.
Im ready to buy a charger but just not getting any luck finding one at the moment.

 

[Alan B]

[strike]I have a Hyperion 1420i I'm not using. Drop me a PM if interested.[/strike]

 

[wesnewell]

I see this charger is now discontinued - can someone tell me a suitable replacement

You might want to consider a bms and bulk charger.

 

[docnjoj]

My Hyperion 1420 still seems to work but it is fussy on the trike battery. I use the Thunder 1220 and it is quite stable. If it ever dies I would consider another or if not available, the 1430. At 185 bucks it is less than half the cost of the fancy Graupner units. I don't know of any others. I am trying to use a BMS on my wifes new Tumich Panasonic cells, however with a regular 2 amp Ping Charger. It is voltage adjustable.
otherDoc

 

[mrpk]

Thank you all very much for the replies - makes a bit change to come onto a forum and be so welcomed.
AlanB - Thank you for your kind offer - i have attempted to email you but the message seems stuck in my out box for the moment - let me know if you get it.

I have looked at the info on BMS that is also suggested but at the moment i just want to get the bike running - i may certainly look into it further in the future however.

 

[Alan B]

Got your PM. Replied. Was away so slow to reply.

 

[Blackssr]

Can anyone recommend a repair facility in the US the can repair the balance port on an Hyperion EOS 1420i NET 3 Balance charger? I have been using this for 4 years on my ebike pack. It has performed without issue until last week. One of the balance ports reads zero. The pack is still good as I have another charger at home that still works. I just require both chargers for longer commutes.

 

[creep_pea]

Hi

My 1420i decided to burn it's IRF4905 MOSFET so I've stripped it down and also found a small surface mount item missing (capacitor maybe), it's labeled C30. Does anyone know if they come missing or if not what is supposed to be there?

Here is the missing part, it's the very bottom left on the circuit board:

And here is the blown IRF4905 MOSFET, I assume I will be fine replacing it with this Infineon IRF4905PBF P-channel MOSFET, 74 A, 55 V HEXFET, 3-Pin TO-220AB https://uk.rs-online.com/web/p/products/5409799/:

 

[fechter]

I wouldn't worry about the C30 capacitor. It is common to find empty spots on a board.

There is a chance that when the FET blew, it took out the gate driver. This is sort of hard to test, but if you remove the bad part and measure between the holes, it should not look like a short between any two.

The other way to test is to simply replace the part and see if the charger works. If the new part immediately smokes, then the driver or something upstream is bad.

 

[creep_pea]

Hi

Thank you, I'll swap that part and hopefully it will work again.