Chargepoint CT500 repair

Introductions, general chit chat and off-topic banter.
Post Reply
User avatar
coulomb
Site Admin
Posts: 3165
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Chargepoint CT500 repair

Post by coulomb » Fri, 22 Jan 2016, 23:27

Graeme of Suzi Auto asked me to see if I could repair a Charge Point CT500 that he had acquired. He said on some forums it was known that often the problem is a power supply. I agreed to take a look.

Image     Image

Coulomb Technologies is no relation Image

Image

The unit opens up without drama. You can see a spare connector, which may be for a card reader. At top left of the second photo the gold connector is for an antenna. As part of an official network, these things send data to some central location, which can tell whether the unit is powered, in use, some fault codes, and so on.

The unit appears to be quite well made. For example, see the heatshrink tubing with labels on the cables, and the safety connectors on the contactor:

Image     Image

With the removal of a few more screws, the two main PCBs are revealed, along with the 5 V power supply:

Image     Image

Another view of the main PCB:
Image

It took me a while to figure out how mains was ever supposed to get to the power supply. Eventually, I figured out that these relays are single pole single throw normally open, but with two sets of connections to the contacts. Two are PCB mount, and two are quick connects at the top. Power goes into the quick connects at the top, through the internal connection to the PCB, and from there to the power supply.

Image     Image

When the relays turn on, the bare quick connect tabs on the relays are live, so I covered them with heatshrink, as shown in the second relay photo.

Sure enough, there was connectivity from the input terminal block to the power supply, and some one megohm from active to neutral. Well, one megohm isn't exactly open circuit, so maybe the power supply is working after all. I traced the 5 V output to a connector accessible from the top of the PCB:

Image     Image

I also noticed that there was mains on some pads nearby, so I covered them with a strip of electrical tape (not shown).

The PCB was put back with just two screws, and I powered up the unit. I was surprised to see LED1 flashing. Yet the main display was totally dead. This did not sound good for a simple repair. Just in case, I put a multimeter on the 5 V and ground connections I found. Just as well, the power supply was jumping up and down. Either there was too much load on the power supply, so it was shutting down to protect itself, or the power supply was faulty after all. To find out, I removed the power supply, and temporarily soldered a capacitor I had lying around onto the 5 V and GND connections thus vacated. (So I can get clip leads onto those connections.) I connected a bench 5 V power supply, set the limit to 2 A, and turned it on.

Life! But maybe not as we want to know it...

Image     Image

Upon powering up, it briefly (1-1.5 seconds) shows "CHARGEPOINT NETWORK" then "Loading.". It stays for a long time on Loading. I'd say it's trying to phone home and connect to the real ChargePoint network, without an antenna. Well, once the power supply is replaced, I think I'll leave that to Graeme. He already has one just like it working fine.

The faulty power supply turns out to be a medical quality one (with a medical quality price) that's supposed to have an MTBF (Mean Time Between Failure, also known as Marketing Total B...) of 200,000 - 400,000 hours. Well, if this part is failing all the time, maybe it's not as good as its datasheet suggests. So after I pulled it apart:

Image     Image

It doesn't look so special inside to me.

I found a replacement power supply at Digi-Key and Mouser:

http://www.digikey.com.au/product-searc ... ds=kms15-5

http://au.mouser.com/ProductDetail/TDK- ... rCMN0JA%3d

Element14 doesn't seem to know about it. RS-Online knows about it, and doesn't stock it, but said it was obsolete and suggested a replacement, and gave a side by side comparison of what I wanted and what they could sell me. I thought that was pretty neat:

http://au.rs-online.com/web/cpd/6811376 ... dk+kms15-5

Although their suggested alternative had the correct pin spacing in one dimension, it was 7 mm too short in the other. Nice try, guys, and if I was designing new equipment, that would be good. For this repair, not so much.

[ Edit 1: Clarify bench power supply; why temporarily soldered in capacitor ]
[ Edit 2: Added electrical tape over hot pads. Also added a clarifying sentence nearby. ]
Last edited by coulomb on Fri, 22 Jan 2016, 12:59, edited 1 time in total.
Learning how to patch and repair PIP-4048 inverter-chargers and Elcon chargers.

User avatar
coulomb
Site Admin
Posts: 3165
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Chargepoint CT500 repair

Post by coulomb » Tue, 02 Feb 2016, 17:50

The replacement power supply arrived yesterday from Mouser:

Image

The new one is on the left. It fit perfectly, and when I powered it up LED1 lit solid. 5V measured 4.95 V; 1% low. So I put most of the screws back, but decided on a final test before the cover went back on. Nothing! But, but... I take out the screws, LED works again. But next test, it doesn't.

I'm doing this work outside in the courtyard, and it's a stinking hot and humid day. So I put everything away for the day and tried again this morning. With a clear head, it occurred to me that it had to be the one connection on the transformer that had a capacitor soldered to it. I must have soldered to the capacitor, but it only had an intermittent connection to the PCB pad. When I resoldered the joint, it came good. Phew.

As I was replacing screws and twisting the board carrier, I noticed that a current sensor was very loose in its mounting:

Image

So that sensor was dosed with silicone to hold it in place. It probably won't see much vibration in actual use, but it seems crazy to have everything else so well made and that part so sloppy. I also found a contactor terminal (one of the coil contacts, 1k ohms resistance) that should have heat shrink on it.

I put it all together and powered it up. As Graeme told me, the Loading... goes away after some 25 seconds, and it said it was ready for use. I tested it by plugging in one of Graeme's J1772 to 15 A adapters. There was a loud click as the contactor came on, and the LED came on the soldering iron I had plugged in to the 15 A socket. All done.

Image
Learning how to patch and repair PIP-4048 inverter-chargers and Elcon chargers.

User avatar
Adverse Effects
Senior Member
Posts: 995
Joined: Sat, 01 Jan 2011, 03:30
Real Name: Adverse Effects
Location: Brisbane

Chargepoint CT500 repair

Post by Adverse Effects » Tue, 02 Feb 2016, 18:00

coulomb wrote:Image


hmm i wonder what would happen if you put in 3 micro switches on the points there Image

User avatar
coulomb
Site Admin
Posts: 3165
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Chargepoint CT500 repair

Post by coulomb » Tue, 02 Feb 2016, 18:38

Adverse Effects wrote: i wonder what would happen if you put in 3 micro switches on the points there Image

I'd say that these are generic or "soft" buttons, whose effect depends on where you are in the program. It would depend on the firmware as to what effect, if any, they would have.

Image
Learning how to patch and repair PIP-4048 inverter-chargers and Elcon chargers.

User avatar
weber
Site Admin
Posts: 2413
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Chargepoint CT500 repair

Post by weber » Wed, 03 Feb 2016, 16:17

Well done, Coulomb. And an enjoyable read. Thanks.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

RedwoodEV
Noobie
Posts: 3
Joined: Thu, 29 Dec 2016, 19:42
Real Name: Donald Becker
Location: Los Gatos CA

Chargepoint CT500 repair

Post by RedwoodEV » Sat, 31 Dec 2016, 19:42

I thought I would add some info for the archives.

I've encountered two failures on the CT-500.

The common one is the main power supply, which I consider a manufacturing defect. A second is a noisy, chattering contactor, which is relatively minor in comparison.

The main power supply is the TDK-Lambda branded module pictured above. This same module is available under different brands, with only a different label. Lambda has been the premium (and premium-priced) power supply company for decades. Generally their power supplies are reliable and conservatively rated. I suspect that TDK-Lambda just puts their sticker on this one, like the other vendors selling the same part.

The supply is targeted to the "medical" market. That mostly means that it has a higher voltage isolation rating, rather than necessarily being more reliable. Still, it has a reliability rating that suggests that it should rarely fail.

My analysis suggests that the supplies are failing because of a flawed potting compound. The supply is built in a conventional manner, installed in a plastic box, and filled with potting compound. A potting compound should electrically insulate, conduct heat, and support the components without stressing them. But these supplies obviously use a mixture that shrunk, either while curing (perhaps from bad curing control) or over time (which points to the mixture used). The shrinkage puts stress on the largest components. In this case, the largest high voltage capacitor. This pulls on the leads, which eventually pull out from the base. There can be various symptoms and a progression of failure, but the end result is that the supply can't put out full voltage.

The fix is simple: replace the power supply. If you are flexible about original looks, pretty much any 5V supply can be wired to the board. The original supply is rated at 3 amp output. The base system with display uses only about 800 mA. (Roughly half of that is used for the display, with relatively little power used for the pilot board.) The optional cellular radio, which doesn't seem to be installed above, can use over 1 amp while communicating. There are other options, such card readers, that use less power but still added to the original design power budget.

The main processor board has relays mounted, but the exposed terminals are never energized. They are used for a lower current configuration where those relays carry the charge current, rather than the external contactor. The footprint that you are concerned about having mains power is for mounting a Safety Monitor Board. This is an independent GFCI circuit, separate from charge control and power monitoring, that can interrupt the relay coil current. Without that board present there isn't a way to turn on the relays.

In other markets the Safety Monitor Board is mounted on the Pilot Board. Your local market must not require the EVSE to have built-in GFCI functionality. Perhaps it requires it externally.

Both the main board and Pilot Board have a Cirrus Logic CS5463 SPI energy monitor, with isolated power and signaling. This monitors charge power. They are duplicate circuits, and the one used depends on the configuration.

The lack of design optimization is also apparent in the board-to-board connections. Every connection uses RS232 signaling levels. Rather than use a single supply with lower cost transmitters, there are a slew of MAX232 (EP223EI) chips, each generating its own voltage levels.

The communication to the VFD display board is standard 19.2 KBaud 8N1 at RS232 levels. There isn't any bi-directional message exchange -- the main board doesn't expect any response to its display messages. So you can disconnect it and substitute your own serial communication if you wish. The display messages are encoded as <ESC> C <index> <0x00> Text <0x00>.

The only time the display board transmits a message is when the stop button is pressed. The unpopulated switch positions do not send a message, so there isn't any hidden functionality that can be activated by adding switches.

The wires to the display board are GND, N/C, Rx, Tx, Reset, +5V. This is the same +5V rail as the main power supply, so it's a good place to quickly test the voltage or back-feed power. Note that some other parts of the circuit are operated off of DC-DC isolators.

The second connector on the display board is wired to the red LED, coming directly from the Pilot Board.

The Pilot Board has a similar RS232 19.2K 8N1 serial connection, but the message encoding is more complex.

The cable dangling out the front is a debug connector. Again, RS232 19.2K 8N1.


EVSE, motor controller and automotive network design

RedwoodEV
Noobie
Posts: 3
Joined: Thu, 29 Dec 2016, 19:42
Real Name: Donald Becker
Location: Los Gatos CA

Chargepoint CT500 repair

Post by RedwoodEV » Mon, 02 Jan 2017, 16:23

I should add that it is possible to repair the power module in a pinch.

The large capacitor is the one that usually fails, with the leads being pulled out. It's on top, and is the easiest to expose from the potting compound. Try peeling away the potting compound and replacing it.

It's a 56uF 400V capacitor that is 12.5mm diameter x 30mm long. Since the power supply is operated at the upper end of it's 100-240V range, and the load is lighter than rated, you can use a somewhat lower capacitance replacement if the fit is a problem. Try to use a 400V rated part. The typical advice of using a higher voltage rated replacement is bad. For low frequency applications an electrolytic capacitor should be used at 75%-80% of it's rated working voltage or it will have a shorter useful life and lower reliability.

EVSE, motor controller and automotive network design

hans
Noobie
Posts: 1
Joined: Mon, 04 Sep 2017, 09:53

Re: Chargepoint CT500 repair

Post by hans » Mon, 04 Sep 2017, 10:18

Thanks for the information on the Chargepoint CT500. I have one installed October 2011 that died last week. I called Chargepoint, who told me to replace it. After reading your post and looking at the photos, I took mine apart and wanted to test to see if the power supply was my issue. Since it was asking for 5v and less than a watt of power, I used an iPhone charging square plugged into an extension cord to test, as I didn't want to spend $50 if that was not the problem. It occurred to me that I could just use the iPhone block instead of the replacement supply, so that is what I did. I looked at the printed circuit board that the power supply was on, and the two leads that connected power to the board did not power anything else on the board that was in use. The board may be from a different application and was adapted for use in the Chargepoint? Anyway, I was able to take the two plugs off of the board, plug those on the prongs of the iPhone charging block, solder the two lines from a USB cable that carry the power to the lines on the board that the dead power supply went to, and my charger is back up and working! You can see the charging block in the upper right corner of the photo at this link.

https://photos.app.goo.gl/ypw0GwoP3c7y1XHm2

dmaz
Noobie
Posts: 1
Joined: Mon, 05 Feb 2018, 04:11

Re: Chargepoint CT500 repair

Post by dmaz » Mon, 05 Feb 2018, 04:23

Just wanted to thank Coulomb for your excellent power supply replacement instructions on the Chargepoint CT500. Mine died a couple weeks back. I opened it up and sure enough the power supply was dead (no 5V). I ordered the part on your link from Mouser $62 with shipping/tax. It came in a week and about an hour later I had my charger back in operation. Anyone with basic electronics and soldering skills should be able to do this. Just be careful measuring your 5V output with the cover off and 220V connected. Saved me buying a new $500 EVSE. Thanks. :)

Post Reply