TC/Elcon 2013 Charger Schematics

How do you store and manage your electricity?
Post Reply
User avatar
coulomb
Site Admin
Posts: 3400
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

TC/Elcon 2013 Charger Schematics

Post by coulomb » Sat, 21 Jul 2018, 19:42

This topic documents the traced schematic diagrams for TC chargers built in 2013 or earlier, and the Elcon PFC battery chargers. It appers that these chargers are no longer available from TC Chargers, but can still be obtained from Electric Conversions (ElCon) in California, USA. These schematics are for the models with the NXP processor, which seem to be the ones with the 7-pin round connector. Basically, if the charger has a 7-pin round connector, then it has the NXP processor and these schematics apply.

This information was oringinally hosted on diyelectriccar.com here:

https://www.diyelectriccar.com/forums/s ... 89470.html

Unfortunately, users such as myself have lost faith that the new owners of the web site are able to maintain its integrity. At the very least, it seems prudent to make a backup here. [ Edit: It seems that this thread is fully restored now. I'll keep these pages here as a backup. Since many posts are still missing, it's not clear if diyelectriccar.com will survive. ]

Many thanks to DIY members KennyBobby and PDove for the original work on which this is based.

Kenny's original post with original schematics
AC Input - DC Boost Suipply (PFC stage).
High voltage DC output section.
Control board schematic as one page; as 4 pages.

[ Edit: added note about applicability and the 7-pin round connector. ]
Learning how to patch and repair PIP-4048 inverter-chargers and Elcon chargers.

User avatar
coulomb
Site Admin
Posts: 3400
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: TC/Elcon 2013 Charger Schematics

Post by coulomb » Sat, 21 Jul 2018, 19:55

This is a project to create troubleshooting and repair schematics for the 240vdc, 5amp version 2.0 board.

Pictures and sketches will be added as available.

Here is a quick basic overview of the electronics inside the box. The main analog power board is screwed to the heatsink. The big power semiconductors are clamped to a central standing set of plates that pass thru a big slot in the middle of the board. The control board is mounted vertically on the edge of the analog board. Left side is AC input power, rectifier, dc boost regulator to 160vdc, and the low voltage supply sections, on the right side is the H-bridge inverter, power transformer and rectifier for the 240Vdc supply,

The voltage and current range of these chargers is primarily determined by the power transformer, either a step-up or step-down, and it is mounted directly to the heatsink thru a big hole in the main board. So you can't take an 80 volt charger and add some trimpots to turn it up to 144, and vice versa.

i will just edit this first post to add and keep it updated--PM me or add a post if you find a mistake or there is something missing, and i can add it to this first post.

top level block diagram
AC input section
Low voltage supply section
High voltage DC output section
Control Board is found in post #2


EDIT: 8/5/2017 Added fresher diagrams modified with Mike's updates, see his notes for details in posts further down in this thread.
Attachments
block_diagram.jpg
block_diagram.jpg (51.83 KiB) Viewed 706 times
15kw_board.JPG
15kw_board.JPG (674.93 KiB) Viewed 706 times
AC_input_dc_boost_mods.png
AC_input_dc_boost_mods.png (227.72 KiB) Viewed 706 times
low_voltage.JPG
low_voltage.JPG (83.85 KiB) Viewed 706 times
HV_dc_output mods3 orig.png
HV_dc_output mods3 orig.png (242.04 KiB) Viewed 713 times
Learning how to patch and repair PIP-4048 inverter-chargers and Elcon chargers.

User avatar
coulomb
Site Admin
Posts: 3400
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: TC/Elcon 2013 Charger Schematics

Post by coulomb » Sat, 21 Jul 2018, 20:01

kennybobby;741865 wrote:Thanks for checking all this Mike and making corrections.
Thanks for the originals.
On U2 there are shown two connections to gnd on the left side of the chip shown in the diagram...
Great pickup! I didn't totally follow what you said, so I traced it myself.

I've modified the control board diagram to suit, and include a modified AC input diagram showing R13.

How does that compare with what you came up with?

Edit: added PFC MOSFET gate resistors, designators for L12 and R25.
Edit: added R15, C10, C3, CON32S 11, fixed designators for R5, C37.
Edit: Removed R25, C34 (on DC output schematic now).
Edit Mar/29: Input relay returns to DC-, not neutral.
Edit Apr/06: Added designators for the two large polypropylene capacitors (C8 and C26)
Edit Aug/04: Added designator for C7, one of the main filter capacitors.
Edit Dec/21: Added part number for Q7, Q8 (PFC MOSFETs).
Edit 2017/Apr/06: Corrected C37, values for C3, C9, C10, C11. R3 seems too big for 2W; marked as "3W?". C37 connects to other side of R13.
Edit 2017/Apr/07: C8, C26 are X2, not X1. Tidied up.
Edit 2017/Apr/25: values for three snubber capacitors were wrong! (C11, C37, C10)
Edit 2017/May/01: C5 -> C100
Attachments
AC_input_dc_boost.png
AC_input_dc_boost.png (227.72 KiB) Viewed 712 times
Learning how to patch and repair PIP-4048 inverter-chargers and Elcon chargers.

User avatar
coulomb
Site Admin
Posts: 3400
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: TC/Elcon 2013 Charger Schematics

Post by coulomb » Sat, 21 Jul 2018, 20:05

PStechPaul;508298 wrote: There is also something weird with the PNP transistor in the COMP circuit. It looks similar to the slope compensation or current limit connections, but they use NPN transistors. The three diodes in series (D12 and dual D9) also seems odd.
Agreed with the oddities, including the snubber on the DC/DC. Perhaps for another day.

I'm going to post my updates to the DC output section, since I'm trying to sort out the snubber for the main MOSFETs. I think PDove was going to do a nice schematic in Eagle, but it may have stalled for time. KennyBobby, perhaps you would like to include this one in the post with the others near the start of this thread. I'd keep the original, in case I've stuffed up or inadvertently made something worse to read.

I'll also add a link to my desaturation protection partial schematic, since all schematics belong in this thread.

[ Edit: here is a partial list of updates: swapped Q1 and Q2, added 1R0 gate resistors, added capacitors C2, C46, C43, C15, and the 0.022uF, CB28 is not connected, asterisked many parts that vary with voltage and power options, and reworked the output relay contact part of the circuit. Added the MOV or whatever it is after the rectifiers D1/D4/D5/D6. Moved R29 to before the final common mode choke. ]

[ Edit Mar/01: Designators for C38, D17; value for C46 ]
[ Edit Mar/01 (2): fixed output relay contacts, shorted inductor. Thanks, PStechPaul! ]
[ Edit Mar/03: A few more designators (R2? -> R2, added L4, L9, large output inductor is definitely L(?) :(). Added CB13 and CB20. ]
[ Edit Mar/04: C46 goes to ground, not to transformer B. ]
[ Edit Mar/09: Had Q1/R32/CB21 swapped with Q2/R31/CB18 (had swapped the wrong parts, making it worse than original) ]
[ Edit Mar/11: CB28 is indeed the connection from battery minus to DGND! ]
[ Edit Mar/11 (2): R25 and C43 are after the DC rectifiers (had them after the AC bridge rectifier). ]
[ Edit 2017/Apr/14: One of the C43s -> C34; values and designators for more capacitors. ]
[ Edit 2017/Sep/23: Corrected CB28; R22 range. ]
[ Edit 2017/Sep/24: R22 now 1-20 mΩ. ]

The snubber appears to be just RD (Resistor Diode) with no C (Capacitor), which is a bit unusual, but perhaps not unheard of. Perhaps C46 is part of the snubber arrangement, or perhaps it is attempting to make a resonant circuit, but I somehow doubt that.[ Edit: more likely it's to limit dV/dt, so the MOSFETs don't turn themselves on through Miller (drain to gate or reverse transfer) capacitance when switching off with high dV/dt. ]

The decoupling capacitors seem to be the 220 uF electrolytic C38, which would presumably be pretty useless at high frequencies, and C2, which is just a very small ceramic capacitor. That all seems a little light for suppressing ringing to me, but I'll be the first to admit I'm no expert in this area.

The charger I'm repairing at the moment, despite having tested it switching at 48 V and with 365 V on the DC bus (but not switching), ended up blowing one pair of MOSFETs immediately that it started charging for real. So I'm highly suspicious of the snubber circuit, at least on this charger. I'll test for ringing at 48 V on the DC bus once I replace the chain of parts that died (likely 2 x MOSFET, 1 x driver, possibly 1 x NOR gate, likely 2 x 1R0 gate resistor, will replace the diodes across the 10R gate resistors since they're very hard to test in place).
Attachments
HV_dc_output mods2.png
HV_dc_output mods2.png (244.61 KiB) Viewed 711 times
Learning how to patch and repair PIP-4048 inverter-chargers and Elcon chargers.

User avatar
coulomb
Site Admin
Posts: 3400
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: TC/Elcon 2013 Charger Schematics

Post by coulomb » Sat, 21 Jul 2018, 20:12

While I procrastinate over this charger (I can't find anything wrong with the capacitors), I thought I'd post my accumulated changes to the daughter board schematic. I had to save it at 75% size to get it under the 2.38 MB file size limit.

[ Edit: a big thanks to KennyBobby and PDove for the original schematic; despite the changes I made, the original was remarkably good; without it, fault finding on these chargers would not have been possible. ]

The main changes:
* Updates to the desaturation protection circuit, already published, but included for completeness.
* Added some designators, such as D4, D7, U12.
* Corrected some component values, mainly around U13 pin 5; R60 10R -> 100k; R73 3 kR -> 499 kR.
* Added some calculated voltages, mainly around U13 pin 5.
* Corrected some CON32S connections, e.g. 13 and 20 don't go to GND, but to the full bridge outputs.
* Added a few missing components, such as C58 (near Q7, upper left).
* Renamed 12VDC_Aux to 15VDC_Aux, reflecting the actual voltage better.
* Redrew several transistors as dual diodes.
* Made connections to the 7-pin round and 5-pin rectangular connectors more obvious.
* Added some text suggesting function, e.g. what the jumpers are for, when the comparator outputs are high, and so on.
* Added some missing connections, e.g. to the SYNC and VC pins of U14. Pin 13 of U13 connects to pin 12.
* Moved C50 from across R63 (U2 pin 13, near top left) to pin 19.
* Added a few test points, e.g. T44, T45, T33, T34.

Edit Mar/06:
* R107 goes to CON32S pin 5 now, for current sensing.
* Updated some resistor values: R107, R108, R110, R101.

Edit Mar/11: CB28 is indeed the connection from battery - to DGND! (Later changed to CB27.)

Edit Mar/18:
* Fixed R69, R75 go to source, not ground.
* Added part number for D5,D6 (also used in power supply)
* Tentative value for C40, C43: 3.3 nF

Edit Mar/28:
* Fixed PFC OVP threshold to 425 V

Edit 2017/Feb/23
* Added TP29; C38 and C46 definitely 100nF; D4, D7 LGE -> BR6.

Edit 2017/Sep/23
* CON32S pin 28 is a no-connect
* R23, R10 vary with voltage and/or power
* Added note re shunt across R10 via R20.

Edit 2018/Mar/13
* U2 powered by 15VDC_Aux (thanks, KennyBobby!)

Edit 2018/May/24
* Added R103, moved R98, C58; values for C56, C58.
Attachments
TC15_control_board_mods sm.jpg
TC15_control_board_mods sm.jpg (2.08 MiB) Viewed 710 times
Learning how to patch and repair PIP-4048 inverter-chargers and Elcon chargers.

User avatar
coulomb
Site Admin
Posts: 3400
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: TC/Elcon 2013 Charger Schematics

Post by coulomb » Sat, 21 Jul 2018, 20:26

I decided that I needed a new print of the control board schematic. The last one I printed on 4 pages, and cut and taped them together into something like an A2 sheet.

To avoid that work, I've split the control board into 4 parts, so I can just print the 4 pieces and staple together. The four parts are attached. Depending on feedback, I may decide not to update either the large schematic or the 4 page version in future.

This version incorporates some changes around the PFC MOSFET driver transistors (Q6, Q7 of the control board; note that one of the PFC MOSFETs on the main board is also designated Q7). I ended up replacing Q6 and Q7 on the control board when not necessary, because it looked like they were both low resistance base to emitter. With the corrected schematic, you can see that you expect 44 Ω from each base to each emitter. To test the transistors, you have to remove R98 first, or test with at least about 20 mA of current.

I hope I have it right now; it seems odd to have the two identical networks R103/C58 and R99/C56 in parallel.
Attachments
ControlBoardPage1.jpg
ControlBoardPage1.jpg (501.07 KiB) Viewed 707 times
ControlBoardPage2.jpg
ControlBoardPage2.jpg (516.59 KiB) Viewed 707 times
ControlBoardPage3.jpg
ControlBoardPage3.jpg (475.96 KiB) Viewed 707 times
ControlBoardPage4.jpg
ControlBoardPage4.jpg (459.7 KiB) Viewed 707 times
Learning how to patch and repair PIP-4048 inverter-chargers and Elcon chargers.

User avatar
jonescg
Senior Member
Posts: 2368
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Re: TC/Elcon 2013 Charger Schematics

Post by jonescg » Sun, 22 Jul 2018, 20:58

Mike, I'm grateful for your contributions here, and I can only really treat it as a learning experience. But If I follow the basic layout correctly, the mains input is rectified and smoothed with an LC circuit. This DC bus is then switched by four transistors (FETs in this case I believe) and the AC is transformed through T26 to the appropriate voltage depending on turn counts. The output is then re-rectified and ultimately charges a battery of the nominated voltage range.

I can't see any additional current limiting circuitry in there - I presume this is because the drivers board has an input consisting of a millivolt-range across R22? And it can adjust the frequency of the switching accordingly?
AEVA National Secretary, WA branch vice-chair

User avatar
coulomb
Site Admin
Posts: 3400
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: TC/Elcon 2013 Charger Schematics

Post by coulomb » Mon, 23 Jul 2018, 07:17

jonescg wrote:
Sun, 22 Jul 2018, 20:58
Mike, I'm grateful for your contributions here, and I can only really treat it as a learning experience.
Actually, this is mainly old information, archived here because the owners of diyelectriccar have lost a year's worth of posts (roughly last financial year), and don't seem to be competent enough to get them back. You can't search the site except through Google, and the garage section has disappeared completely. I posted my contributions there because the initial hard work was done by two Americans, but now I can't trust that the information will be available into the future.
But If I follow the basic layout correctly, the mains input is rectified and smoothed with an LC circuit.
It's also a boost converter, producing a DC "bus" of around 400 V. Note switching transistors Q7 and Q8. When the mains is instantaneously at say 100 V, it is boosted 4:1 but with only a third the current when the mains is instantaneously at 300 V, thus making the input current a good, in-phase sinusoid. This is often called the "power factor correction" or PFC stage.
This DC bus is then switched by four transistors (FETs in this case I believe) and the AC is transformed through T26 to the appropriate voltage depending on turn counts.
Exactly. The transformer ratio is the main difference between say a 24 V model and a 360 V model. They use the same basic topology, and the exact same printed circuit boards, for all the different voltage models. There are enough differences between a 1500 W model and a 2000 W model to justify a different main PCB, however. A 2500 W model seems to be basically a 2000 W model with a fan on the main heatsink, different firmware, and more prayers.
The output is then re-rectified and ultimately charges a battery of the nominated voltage range.
Yes, with more filtering to get rid of the switching ripple, which could also cause EMI (ElectroMagnetic Interference) in a vehicle.
I can't see any additional current limiting circuitry in there - I presume this is because the drivers board has an input consisting of a millivolt-range across R22?
Yes, R22 is the current "shunt", which gives feedback to U14 about what current is going into the battery. U14 is a "current mode" controller. Any voltage control is achieved by the microcontroller noticing a need to change the voltage, and it does so by adjusting the current set-point. This set-point is communicated via OCB (one of the PWM outputs).
And it can adjust the frequency of the switching accordingly?
It adjusts the duty cycle; the frequency is fixed. A higher duty cycle causes current in the magnetics to increase; it's all about the speed of ramping up and down an inductor's current. I won't pretend to explain it here, as I'm sure I'll get it wrong and confuse readers.

So for an inexpensive unit, there is a lot in there: two power conversion stages, genuine isolation (one brand of EV charger remains stubbornly unisolated), much filtering, and a flexible architecture that covers a wide range of voltages. They combine 1.5, 2, and 2.5 kW units into chargers from 1.5 to 8 kW. For example, a 3 kW model has two 1.5 kW units in one box (paralleled mains input and battery output), and an 8 kW model has four 2 kW units in one box. Paralleled units have slightly different firmware, so one is a master and the others are slaves.

My main interest in these chargers has been the firmware. I was able to use my disassembling skills to produce a workable assembly language source code for the firmware. So we can adjust the output voltage and "curve" of a charger, perhaps changing it from 120 V lead-acid to 130 V lithium iron phosphate. They can also be changed from stand-alone to "CAN bus" versions, and vice versa. This can all be done by the user, with some inexpensive hardware, and unfortunately a lot of software expertise. This is all documented in my third major charger related thread on diyelectriccar, to be copied over soon.

[ Edit: added sentence about Q7, Q8. ]
Learning how to patch and repair PIP-4048 inverter-chargers and Elcon chargers.

User avatar
coulomb
Site Admin
Posts: 3400
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: TC/Elcon 2013 Charger Schematics

Post by coulomb » Wed, 25 Jul 2018, 07:22

The diyelectriccar pages, for this thread at least, seem to be restored. I'll continue to copy them here, at least for now.

I had somehow convinced myself that the connection from battery negative to digital ground (DGND) was somewhat tenuous, about 3kR comes to mind. While tidying up one of the diagrams, I decided to check where CB27 and CB28 actually connected to on the main board, even though I had in mind that there was no connection on the control (daughter) board. But I found that indeed it was to pack negative via CB27, and indeed CB27 and connect to the digital ground plane of the daughter board (DGND), which indeed connects to the 5-pin connector and elsewhere.

Sigh.

My apologies if I led anyone astray by that. I've updated the schematics (the two relevant ones are a page back now).
Learning how to patch and repair PIP-4048 inverter-chargers and Elcon chargers.

User avatar
coulomb
Site Admin
Posts: 3400
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: TC/Elcon 2013 Charger Schematics

Post by coulomb » Wed, 25 Jul 2018, 07:24

Here's another thing that took me a long time to get straight in my head.

Several posts back I noted that DGND on the daughter board was connected almost directly to battery negative, only via half of L4 and the shunt resistor R22. The latter is only 1-10 mΩ.

This means that the negative input of the shunt amplifier op-amp (first half of U6) is connected to DGND via a 100Ω resistor. Since the shunt has negligible resistance compared to the shunt amplifier circuit, it also means that R10, which has a value from 3 kΩ to 6.2 kΩ depending on the model, is shunted by R20, another 100 Ω resistor! They go to a lot of trouble to make the positive and negative inputs have the same resistance to DGND, so how does this work?

It turns out that the op-amp isn't particularly bothered. Yes, there is only 2-3 mV at the op-amp inputs, but it amplifies these by the gain to place the output at about 100-200 mV above DGND (about twice as high as it would be if CON32S pins 24 and 27 weren't connected together, it seems).

It's something to keep in mind if you're checking the shunt op-amp. A moderately common failure mode is for battery voltage to end up at the current shunt, sometimes burning PCB traces, and often one or both of the 100 Ω resistors (R20 and R21), so one might get caught up with this unusual arrangement.
Attachments
Shunt amplifier 2.png
Shunt amplifier 2.png (246.08 KiB) Viewed 650 times
Learning how to patch and repair PIP-4048 inverter-chargers and Elcon chargers.

User avatar
coulomb
Site Admin
Posts: 3400
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: TC/Elcon 2013 Charger Schematics

Post by coulomb » Wed, 25 Jul 2018, 07:25

After years of working with these chargers I've finally figured out what the resistor and diode are for, across the output relay.

Before the output relay is energized, there can be a substantial difference in voltage between the battery voltage and the DC bus (charger output) and its capacitors. They prefer to use a cheap AC rated relay, and besides it's just hard on the contacts and the capacitors to uncharged capacitors across an EV sized battery. So before they close the relay, they make sure that the MOSFETs are off for half a second, so that R2 can approximately equalize the capacitor and battery voltages.

But then, when you first connect the battery, there will be this capacitor charging current, which would still cause an unpleasant, though modest, spark when connecting the battery. So they put in diode D12 to prevent this. But now the charger output has to be higher than the battery voltage before the output capacitors discharge down to the battery voltage.

So in RELAY_SET, they spend the first 1.5 seconds asking for one fortieth (2.5%) of the maximum charge current, which will ramp the output capacitors from zero to somewhere near the maximum charger voltage. From 1.5 to 2.0 seconds after calling RELAY_SET, they set the charge current to a small negative current that guarantees that the MOSFETs will be off. This allows the capacitor voltage to "free wheel", allowing D12 and R2 to discharge the capacitors to approximately the battery voltage. At 2.0 seconds, the relay is energized, shorting D12 and R2 so that the full charge current can go to the battery.

When disconnecting the output relay, the output current is ramped to zero so that the MOSFETs are again turned off before de-energizing the relay. When the contacts open, they will be switching essentially zero current.

All these long delays mean that some of the main loop code as to be replicated in the various RELAY_SET loops. So they've gone to a bit of trouble to get this right.

It's a quite neat system. The only problem is that you might receive a mild shock through D12 and R2 if you touch the charger output when the relay is supposed to be "off". But even that should never happen, because the output voltage won't ramp up if no battery is detected. With no battery detected, it stays in "stage zero", blinking the LEDs in warning, and doesn't call RELAY_ON at all.

The value of R2 in the schematic, is shown as 100K*. The asterisk means that the actual value will depend on the exact model of charger. 100K is for a quite high voltage charger (around 300 V), and lower voltage models will come with lower valued resistors so that the larger valued capacitors can still equalize voltage in the given half second.
Attachments
R2D12 across output relay.png
R2D12 across output relay.png (49.48 KiB) Viewed 650 times
Learning how to patch and repair PIP-4048 inverter-chargers and Elcon chargers.

Post Reply