TC Charger Fixable?

How do you store and manage your electricity?
User avatar
4Springs
Site Admin
Posts: 813
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: TC Charger Fixable?

Post by 4Springs »

coulomb wrote: Thu, 06 Jan 2022, 08:10I think it's worth doing the extra tests with the jumpers before re-installing.
Thank you!
I thought that there was some extra testing to be done on the 50V, but I could not find it today. The key word I needed was 'jumpers'! I was skipping over that section of the forum, thinking that it was for something else.

I made myself some jumpers as per the instructions:
PXL_20220106_010501884.png
PXL_20220106_010501884.png (596.95 KiB) Viewed 228 times

It was reasonably straightforward to follow the instructions for the tests, and all voltages measured as expected.

Next thing is to re-assemble.
I noticed that I have three pieces of ceramic to put behind the semiconductors on the central heatsink, presumably for electrical insulation. However, there are four groups of semiconductors. I think that the rectifier is the one that doesn't need the insulator?
User avatar
coulomb
Site Admin
Posts: 5095
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: TC Charger Fixable?

Post by coulomb »

4Springs wrote: Thu, 06 Jan 2022, 10:07all voltages measured as expected.
Excellent!
I think that the rectifier is the one that doesn't need the insulator?
The bridge rectifier doesn't need the insulation. The output rectifiers (usually 4 TO-220 diodes) DO need the insulation.

The bridge rectifier comes with its own insulation.

Use the thermal paste sparingly. I've been chipped before for advocating the use of too much of the stuff. It's just to fill in the tiny gaps, rather than to enhance metal to metal conduction, as some seem to think (and I used to, long ago 👶).
MG ZS EV 2021 April 2021.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
Patching PIP-4048/5048 inverter-chargers.
User avatar
4Springs
Site Admin
Posts: 813
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: TC Charger Fixable?

Post by 4Springs »

Putting the charger back together was a bit of a fiddle.
I had the PCB in and out about 10 times, trying to get the PCB at the PFC inductor to actually touch the octagonal heatsink. It eventually went in when I removed the inductor's bolt (rod?). That gave me a clue as to what was the problem, and I was able to make it fit by straightening the bolt. Unfortunately I managed to strip the thread at the end of the bolt, so the nut wouldn't engage at the crucial part where it starts to tighten down the cover. So I trimmed the cover (round PCB material) to gain a millimetre or two, and was able to tighten it up nicely.
Here it is with the heatsink clamps on, glands attached and new silicone sealant.
PXL_20220107_043622977.png
PXL_20220107_043622977.png (954.76 KiB) Viewed 193 times
And here is a picture of it about 24 screws later:
PXL_20220107_045530024.png
PXL_20220107_045530024.png (692.34 KiB) Viewed 193 times
One black screw left over. I wonder where that was supposed to go?
The aluminium pieces in the heatsink are how I attach my fans. I have four fans and a thermostat that screw into those brackets. If the heatsink exceeds 40 degrees Celsius then the fans come on.

I can't test it in the car quite yet unfortunately. The front of the car is up on ramps, running in some new brushes on the motor. This means that the back of the car, where the charger goes, is too low to work on. I need to run the motor for a few more hours until I can take it down off the ramps under its own steam.
User avatar
4Springs
Site Admin
Posts: 813
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: TC Charger Fixable?

Post by 4Springs »

I spent the afternoon under the car today, re-doing the circuitry that surrounds the enable line from the BMS to the charger. I'd jiggery poked it to make it work for my small, dumb charger, which meant that it wouldn't work with the TC Charger. I also need it to work a different way again when my new charger arrives. Rather than making a temporary test circuit to test the charger out of the car, I decided to do a new circuit in the car to suit the new charger and the TC Charger.
So after I'd done that, I had a working enable line to check this TC Charger. I installed it in the car, and turned it on. It worked! Well, at least the lights came on to say that it was working, I didn't have a current meter to see if it was actually delivering much power. Then it made a funny noise. Was that the charger making that noise? Not sure, the lights are still on. I turned it on and off a couple of times with the enable circuitry, to make sure that that worked, and it did.
Then, I think during the third time I had it on, it made a definite bang noise, and all lights stopped. It was running for perhaps 60 seconds. Oh well...

I've removed it from the car and taken the lid off to have a look. This time I can see damage in the output stage! Q1 & Q2 are missing pieces. None of the components I replaced look bad, including the fuse on the input stage precharge resistors. But I guess I'll get out my circuit diagram again tomorrow and start testing again...
User avatar
coulomb
Site Admin
Posts: 5095
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: TC Charger Fixable?

Post by coulomb »

4Springs wrote: Mon, 10 Jan 2022, 15:52It was running for perhaps 60 seconds. Oh well...
Oh. Sorry to hear that.
But I guess I'll get out my circuit diagram again tomorrow and start testing again...
Well, if you're up for it, I DO have one last set of MOSFETs...

If you do, I'd recommend replacing some capacitors as well: C38 (large one near the MOSFETs, as good a quality as you can get), C2, and C46 (small blue ones).

But I'll completely understand if you've had enough of it.
MG ZS EV 2021 April 2021.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
Patching PIP-4048/5048 inverter-chargers.
User avatar
4Springs
Site Admin
Posts: 813
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: TC Charger Fixable?

Post by 4Springs »

Q1 & Q2 are physically damaged. The MOSFET part of them measures ok, but the diode part does not. The associated resistors R32 & R31 are open circuit.
Q3 & Q4 measure ok, but I guess they should be replaced. Their associated resistors R12 & R18 measure ok.
C38 could perhaps have a slight bulge?
PXL_20220110_224615926.jpg
PXL_20220110_224615926.jpg (162.94 KiB) Viewed 138 times
I did notice this last time, but didn't think it was bulging enough to look 'bad'. I've never seen a bad one before, so I just did some googling, and perhaps I should have replaced this one before. It measures as about 12 uF in circuit.

All the components I replaced in the AC input section seem to be ok. Should I try powering it up using the current limited 50 V supply?
coulomb wrote: Mon, 10 Jan 2022, 17:49But I'll completely understand if you've had enough of it.
In for a penny, in for a MY GOD THAT CAPACITOR IS $192!!
Ahem. Perhaps you should send me an email with a price for those MOSFETS, and I can decide from there.
User avatar
4Springs
Site Admin
Posts: 813
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: TC Charger Fixable?

Post by 4Springs »

OK, that first capacitor I looked at was a bit of an outlier...
Here's a more reasonably priced one: Electrolytic Capacitor
@coulomb said to look for one of good quality, but I don't know how to evaluate that.

Here is a potential candidate for C2: Ceramic Capacitor
As for C46, I'm not sure what type it is, or what voltage rating it requires. Perhaps mica 500 V?: Mica Capacitor
User avatar
coulomb
Site Admin
Posts: 5095
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: TC Charger Fixable?

Post by coulomb »

4Springs wrote: Tue, 11 Jan 2022, 09:02 OK, that first capacitor I looked at was a bit of an outlier...
Here's a more reasonably priced one: Electrolytic Capacitor
@coulomb said to look for one of good quality, but I don't know how to evaluate that.
Yes, long life capacitors do cost surprisingly much. I guess that they need more careful design and manufacture, and hopefully, some extra testing and/or quality control, to ensure that the extra price paid is actually worth it. That one seems good, although the datasheet says that the design life according to some standard is 2000 h, but the "useful life" (a pure marketing term?) is 5000 h (for the up to 450 V parts, which this one is; 500 V parts only get 3000 h). But without a lot of effort, the others are probably claiming marketing hours as well.

I have one cheaper United Chemicon capacitor; $15.05 for a 3000 h part from element14. I'll understand if you want to get the (supposedly) 5000 h part. I suspect that the United Chemicon part will actually be the better quality part, but I'm no expert on capacitor endurance. My part has a shelf age of several years, but that is all at room temperature, so it only counts about 1/256th as much (8 doublings between 25°C and 105°C).
Here is a potential candidate for C2: Ceramic Capacitor
That one is fine; I have about 98 of them here.
As for C46, I'm not sure what type it is, or what voltage rating it requires.
It should be another high voltage single layer ceramic, 220 pF. I have 19 in stock.

The MOSFETs I have are $10.43 each + GST at RS; not much cheaper at Mouser. This seems to be a "chip shortage" price; I don't remember paying quite that much, though it's certainly possible.

I'll wave my usual restocking fee, this week only! Avoid the rush! :roll:

[ Edit: added URL for the STP28NM50N MOSFETs. ]
MG ZS EV 2021 April 2021.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
Patching PIP-4048/5048 inverter-chargers.
User avatar
4Springs
Site Admin
Posts: 813
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: TC Charger Fixable?

Post by 4Springs »

I've just set up with the 50V DC power supply again, to see if I can see anything else wrong before I order parts. It appears I'm on a time schedule, as @coulomb 's fire sale is for this week only!
The MOSFETs Q1-Q4 are removed from the circuit, as is the capacitor C38.

The lights don't flash! This is the no jumpers test, with 50VDC supply limited to 0.5A. I'm pretty sure that the lights worked last time I did this test. There should be a red flashing LED on the daughter board, and the main status indicator (red/green LED) on the main PCB.
I measured for 15V on the daughter board and found nothing.
I found the Low Voltage Supply circuit diagram, which should provide this 15V, but I'm afraid I don't understand very much of it. Should it be supplying 15V when I have those components removed?
User avatar
coulomb
Site Admin
Posts: 5095
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: TC Charger Fixable?

Post by coulomb »

4Springs wrote: Fri, 14 Jan 2022, 13:58 It appears I'm on a time schedule, as @coulomb 's fire sale is for this week only!
:twisted:
The lights don't flash!
OK, that's not good.
This is the no jumpers test, with 50VDC supply limited to 0.5A. I'm pretty sure that the lights worked last time I did this test. There should be a red flashing LED on the daughter board, and the main status indicator (red/green LED) on the main PCB.
Yes. The latter should be flashing some error code, with plenty to complain about: mains voltage is very low, battery not present...
I found the Low Voltage Supply circuit diagram, which should provide this 15V, but I'm afraid I don't understand very much of it. Should it be supplying 15V when I have those components removed?
Well, with 50 V, there is no need to panic just yet. The threshold for the little "Viper" chip to start working is somewhere in the fifties of volts DC at the AC input; I used to use 50 V, but now I use 52 V (2 x 26 V) because it seems to more reliably get the Viper started. Some Vipers might require a few more volts. [ Edit: I've edited the troubleshooting page to make this clearer. ] I'd be very surprised if 60 V (the limit for common dual power supplies) was not enough.

So: More voltage, Egor! ⚡

As to how the circuit works, it's a standard flyback circuit. The chip starts oscillating and sending pulses to the output, which drives a transistor (internal in this case) which drives a "transformer" that powers the chip off a winding via D9 and C41. The latter is helpful for when the AC input happens to be near a zero crossing. I don't pretend to understand the circuit around Q9, but it never seems to play up, and it seems to be mainly to do with transient stability or perhaps shaping the pulse of current into the multi-winding inductor. The same winding that powers the Viper chip also generates the 15 V supply.

I have heard of Viper chips failing before, but they're pretty tough critters, so the failure rate is quite low. Hmmm. I seem to have 5 in stock, and a label that seems to indicate that I bought 5 at once. Very low failure rate! :D

Occasionally other power supply parts will fail, e.g. open circuit inductors. I have a few inductors in stock, it looks like three varieties. The power supplies are generally not too hard to fix.

Let me know if you want to go ahead; I can throw in a Viper chip and set of inductors for free. But only if you order now! :twisted:
MG ZS EV 2021 April 2021.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
Patching PIP-4048/5048 inverter-chargers.
User avatar
4Springs
Site Admin
Posts: 813
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: TC Charger Fixable?

Post by 4Springs »

coulomb wrote: Fri, 14 Jan 2022, 16:15 So: More voltage, Egor! ⚡
Still no lights, took it up to 60 V or so.
coulomb wrote: Fri, 14 Jan 2022, 16:15 As to how the circuit works, it's a standard flyback circuit. The chip starts oscillating and sending pulses to the output, which drives a transistor (internal in this case) which drives a "transformer" that powers the chip off a winding via D9 and C41. The latter is helpful for when the AC input happens to be near a zero crossing. I don't pretend to understand the circuit around Q9, but it never seems to play up, and it seems to be mainly to do with transient stability or perhaps shaping the pulse of current into the multi-winding inductor. The same winding that powers the Viper chip also generates the 15 V supply.
Hmm...
The Low Voltage Supply diagram that I have is difficult to read. I found it attached to the first post of this thread. Do you know of a clearer version? I'm trying to figure out what voltages I can measure while running off DC 52 V. With your explanation I think I've sussed it out a bit better.
So the Viper should have 52 V between pin 3 and pins 5-8 (which are all soldered together onto a piece of copper). It gets this through L1. Good, I measure 52 V there.
The Viper should then set up an oscillation across pins 4 & 5 of the L1 transformer, and change this voltage to about 12 VAC. This is rectified by the circuit around D11 to give the 12 V supply. So I should be able to measure 12 VDC on L10 or D11. I tried this, and measured 52 VDC. I tried measuring L10 and D11 (in circuit, power off):
L10 - closed circuit (multimeter on Ohms).
D11 - closed circuit both ways (multimeter on Diode Test).

The oscillating voltage should also go through the transformer L1 to produce 15 VDC, which is then rectified by the circuit around D13. I should be able to measure 15 V on D13 or L11:
Voltage = 0 VDC.
L11 - closed circuit.
D13 - open circuit one way, closed the other (i.e. looks ok).

So if I'm right, it seems that the circuit around the Viper is not producing the oscillating 12 V that it is supposed to, and 52 VDC is going through. D11 has died, which I assume is a symptom, and not the cause. D9 appears to measure ok, as does R17. I can't find D12, but there is a D14 which I reckon is the part, and that measures ok.
Does this look like the Viper is dead? Anything else that I should test here?

Is it likely that having 52 V, and perhaps even 385 V at one stage, would have killed anything else on that 12 V section? There is 160 Ohms across the start relay coil, which sounds ok. Perhaps I should replace C42 & C22. I'm not sure what else the 12 V goes on to do...
User avatar
coulomb
Site Admin
Posts: 5095
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: TC Charger Fixable?

Post by coulomb »

4Springs wrote: Sat, 15 Jan 2022, 07:43 Still no lights, took it up to 60 V or so.
Ok.
The Low Voltage Supply diagram that I have is difficult to read. ... Do you know of a clearer version?

No, that's it. Kenny likes his pencil and eraser.
So the Viper should have 52 V between pin 3 and pins 5-8 (which are all soldered together onto a piece of copper). It gets this through L1. Good, I measure 52 V there.
The Viper should then set up an oscillation across pins 4 & 5 of the L1 transformer, and change this voltage to about 12 VAC. This is rectified by the circuit around D11 to give the 12 V supply. So I should be able to measure 12 VDC on L10 or D11.

Good so far.
I tried this, and measured 52 VDC. I tried measuring L10 and D11 (in circuit, power off):
L10 - closed circuit (multimeter on Ohms).
D11 - closed circuit both ways (multimeter on Diode Test).
The inductors will all measure zero ohms on a multimeter; that's OK.
But the rest is bad. It implies that the "transformer" winding from pins 3 to 4 is open circuit. There must also be a short between say pins 1-3 and 5. My notes say that from pin 1 to pin 3 should be 3.2 or 4.2 Ω. Can you verify if 4-5 is open, and if there is a short or low resistance from 5 to 1, 2, or 3?

If so, you'll need a replacement transformer (after fixing some other problems). You can't buy these new. I have plenty of half-dead chargers though. But see below.
Does this look like the Viper is dead?

Hard to say at this stage. It looks like the "transformer" is dead.
Is it likely that having 52 V, and perhaps even 385 V at one stage, would have killed anything else on that 12 V section?
I think you mean the 15 V section. Yes, at least 52 V, possibly the full 385 V. That powers things like the PFC chip, MOSFET drivers, and so on. That makes it a very difficult repair job, and if so I would not recommend continuing.

To make this decision easier, here is a simple test. Set your power supply to 15 V, and apply it across C42. Observe polarity, of course. You can use the connector on the left side of the control board if that's more convenient. That should power up the left 2/3 of the control board, including the main PWM chip, U14. Check for 5.1 V at pin 2 (Vref), with respect to non-digital ground (e.g. pin 12 of that chip). If you don't get 5.1 V there, then that chip has fried. Similarly pin 11 of U2 should also be 5.1 V. If either of these major chips is fried, then repair is not recommended, sadly.

For kicks, you could try applying 12 V across C44 and see if the micro is working. The LEDs should flash. Again, if that doesn't happen, it's a very difficult repair.
MG ZS EV 2021 April 2021.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
Patching PIP-4048/5048 inverter-chargers.
User avatar
4Springs
Site Admin
Posts: 813
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: TC Charger Fixable?

Post by 4Springs »

coulomb wrote: Sat, 15 Jan 2022, 08:37 But the rest is bad. It implies that the "transformer" winding from pins 3 to 4 is open circuit. There must also be a short between say pins 1-3 and 5. My notes say that from pin 1 to pin 3 should be 3.2 or 4.2 Ω. Can you verify if 4-5 is open, and if there is a short or low resistance from 5 to 1, 2, or 3?

I think that pin 5 of the 'transformer' (which I gather, due to the quote marks, is somehow not a transformer), is the one closest to the edge of the PCB, on the side that has 5 pins.
Pin 5 - 4 = 0.6 Ohms.
Pin 5 - 1, 2, 3 = open circuit (unless I use the diode tester)
Pin 1 - 3 = approx 3 Ohms
I think you mean the 15 V section. Yes, at least 52 V, possibly the full 385 V. That powers things like the PFC chip, MOSFET drivers, and so on. That makes it a very difficult repair job, and if so I would not recommend continuing.
Hang on, both sections are marked as 12 V on this diagram. I measured 52 V at L10 - is this the 15 V section? I thought this was the 12V section, and the L11 part was 15 V.
User avatar
4Springs
Site Admin
Posts: 813
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: TC Charger Fixable?

Post by 4Springs »

coulomb wrote: Sat, 15 Jan 2022, 08:37For kicks, you could try applying 12 V across C44 and see if the micro is working. The LEDs should flash. Again, if that doesn't happen, it's a very difficult repair.
Did this and this section is working. LEDs are flashing.
To make this decision easier, here is a simple test. Set your power supply to 15 V, and apply it across C42. Observe polarity, of course. You can use the connector on the left side of the control board if that's more convenient. That should power up the left 2/3 of the control board, including the main PWM chip, U14. Check for 5.1 V at pin 2 (Vref), with respect to non-digital ground (e.g. pin 12 of that chip). If you don't get 5.1 V there, then that chip has fried. Similarly pin 11 of U2 should also be 5.1 V. If either of these major chips is fried, then repair is not recommended, sadly.
Ok, let's try that.
The three pins of that connector are reading as connected to each other. Not surprising, since D11 is cactus. I need to remove it before I can do this test.

...D11 is now removed and the +15 V and GND are still connected (about 8 Ohms between them).
I tried the test anyway, with current limited to 100 mA. The power supply indicated current limiting. Increased current to 600 mA, same result.
I think that with D11 removed, the Viper chip and related circuitry can't be the source of the short. So it must be C42, C22, or something on the control board.
User avatar
coulomb
Site Admin
Posts: 5095
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: TC Charger Fixable?

Post by coulomb »

4Springs wrote: Sat, 15 Jan 2022, 11:01 I think that with D11 removed, the Viper chip and related circuitry can't be the source of the short. So it must be C42, C22, or something on the control board.
Sadly, yes. It's most likely one or more chips on the power side (left 2/3) of the control board. I've attempted a few repairs on chargers like this, and it's a lot of work. I got one or two going, though.

If you're curious about the extent of the problem (like which chips are the problem), you could leave the current limit at about 200 mA and feel for which chips get hot, or use a non-contact thermometer to minimise finger blisters. I seem to recall also using a multimeter on millivolts, checking the voltage drop across PCB supply traces. U14 has a 10 Ω resistor (R42) that you can check for resistance and use its voltage drop to calculate the rough current consumption.

Other usual suspects are U12 (actually a quad NOR gate), and U15/U16 (MOSFET gate drivers).

Looks like I miss out on my fire sale. :( Sorry for the bad news about your charger.
MG ZS EV 2021 April 2021.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
Patching PIP-4048/5048 inverter-chargers.
User avatar
4Springs
Site Admin
Posts: 813
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: TC Charger Fixable?

Post by 4Springs »

coulomb wrote: Sat, 15 Jan 2022, 12:25 Sorry for the bad news about your charger.
Ah well, it was fun while it lasted.
This was also a catalyst to learn a bit about CAN. I've ordered a new charger that talks CAN, and I had the idea that I'd build a CAN interface for this non-CAN TC Charger. I'd use an Arduino to interpret the CAN messages and adjust the 1-5 V enable input on the charger to suit. That would let my charge controller control the new and old chargers via CAN.
I guess I can shelve that project now, too.
Post Reply