First-Time Config of FOCCI And CCS1 vehicle charging

[uhi22]

The logs are saying all the same: No response from charger. The SOC does not matter at this stage. Any chance to measure the CP PWM, in the web interface and with an oscilloscope?

[serh270]

CCS1_vs_CCS2_signaling_circuit_2.png
Shows the differences

Also wondering if this is a change I may have to make. I found these bits of doc in the CCS1 protocol documentation and it seems that in a certain setup, 4.7k is optional? Again, I'm not very knowledgeable in this space so I figured I'd throw these up anyways just to get y'alls thoughts.

[serh270]

The logs are saying all the same: No response from charger. The SOC does not matter at this stage. Any chance to measure the CP PWM, in the web interface and with an oscilloscope?

CP PWM was measuring 0, scope as well. Only non-zero was the 4.5V measurement across CP and PP.

[uhi22]

Does the charger provide 12V between CP and PE if just sticking the multimeter probes into the CCS plug?

[uhi22]

Also wondering if this is a change I may have to make. I found these bits of doc in the CCS1 protocol documentation and it seems that in a certain setup, 4.7k is optional?

There are two different use cases: CCS type 1 and CCS type 2.
The upper picture, with Rs3 and Rs4 and button in the connector, is CCS type 1. For this, the PP pulldown 2.7k inside the vehicle is mandatory.
The lower picture, without the switch in the connector, is CCS type 2. For this, the PP pulldown is optional and can be 2k7 or 4k7 (or not present at all). Thanks for the official data, this confirms the understanding with the 330 ohms to 5V.
You could cross-check, whether you have the 330ohms to 5V, and the 2k7 (or 3k) pull-down on the PP. If you put a multimeter in current mode between the PP and ground, you should see something like 5V/330ohm=15mA.

[serh270]

The lower picture, without the switch in the connector, is CCS type 2. For this, the PP pulldown is optional and can be 2k7 or 4k7 (or not present at all). Thanks for the official data, this confirms the understanding with the 330 ohms to 5V.

So in this case, if the charger is a CCS2 charger (which it is), then the resistance option on the board does not matter i.e. it should be backwards compatible? Idk I suspect that the issue lies somewhere with the PP because it is not detecting connecting but I am unable to test right now. I will update with results when I can.

[uhi22]

To simplify trouble shooting the CP and PP lines, added ResistanceProxPilot to the serial logging.
If the charger is plugged in, we should see something like
cpDuty [%] 5
ResistanceProxPilot [ohm] 1495
The CP PWM duty cycle should be between 4 and 6 percent (5% nominal), and the PP resistance between 1450 and 1550 ohms (1.5kohm nominal).

The software (bin and hex) is available here: https://github.com/uhi22/ccs32clara/act ... 8949329201

[uhi22]

To cover the typical mistakes "CP and PP mixed up or not wired" just added hints in the trouble-shooting chapter: https://github.com/uhi22/foccci#no-comm ... r-possible

[serh270]

Does the charger provide 12V between CP and PE if just sticking the multimeter probes into the CCS plug?

This was showing 0v. Across the inlet, CP-PP was reading <5v.

Just flashed the new clara pushed. Thanks for the update, will test.

Edit: Tested. Charger is still showing 0v between the CP-PP pins on the plug. EVside, CP-PP is reading <5V. I've attached the serial log from the last test. Begins unconnected and then connects plug to inlet. Plugged/unplugged a couple times.

Code: Select all

[534350] cpDuty [%]  0
[534350] AdcProximityPilot  3086
[534350] ResistanceProxPilot [ohm]  10000

ProxPilotResistance showing 10000 the whole time and CP duty 0%.

Code: Select all

 [534920] [CONNMGR] 165 109 0 0 0 0 0 --> 10 

what are these messages indicating? The second element doesn't seem to follow any pattern.

Code: Select all

 [618500] [PEVSLAC] from 4 entering 0 

What is this indicating? Seems to vary 2->4 4->0 0->2

Code: Select all

 [620810] [CONNMGR] ConnectionLevel changed from 5 to 10.

what are these connection level changes?


Charger did not detected that it was plugged in. Ran board in demo-standalone mode since BMS was not connected. PP variant is 2. Any other params I should look at?

screenlog.txt

(36.28 KiB) Downloaded 190 times

[serh270]

Managed to secure access to another portable charger and I seem to have found the source of connectivity-recognition issues. The charger that worked was providing 12V across cp-pp.

resistanceproxpilot danced around 830ohm. CP was still 0 but that's because I hadn't activated the charger (with an rfid swipe). The charger showed recognition of being plugged in. Next step will be to test full charging but I do not have access to the vehicle today. Will post updates.

[johu]

I think the 5% PWM should start as soon as the charger is powered and you have 2k7 pull-down to PE (not PP). When you say "PP" do you mean "PE" i.e. Earth? Doesn't really make much sense to measure between CP and PP

[uhi22]

Yes, there seems to be confusion with the pins. There are three:
1. The PE, protective earth, chassis ground. This needs to be connected to the ground of Foccci. This is the reference point for your multimeter. The single 6mm pin in the center.
2. The CP. Control pilot. This needs to be connected to the CP of Foccci. If measured on the charger side (between CP and PE), this normaly gives 12V statically.
3. The PP. Proximity Pilot or PlugPresent. This needs to be connected to the PP of Foccci. If measured on the charger side (between PP and PE), the multimeter shows zero volt. Changing the multimeter to resistance measurement, should show 1.5kOhm for CCS2. For CCS1 150ohm or 480ohm depending on the button.

And there is also confusion regarding CCS1 and CCS2. The thread topic says "CCS1", also the picture shows a CCS type 1 plug. On the other hand, I read "Qingdao Penoda Portable DC CCS2 Charger. I knew there are differences between CCS1/2 (I did not order this charger)". So maybe we talk about different chargers or have a different understanding about CCS1 and CCS2.

Regarding the messages of the connection manager: Described it in the Clara user manual: https://github.com/uhi22/ccs32clara/blo ... r-messages . Basically the numbers show that Foccci found its QCA modem, and because nothing else happens, it only checks every few seconds whether this modem is still there. But this debug messages won't help, as long as the physical connection is not clear.

[Edit] Regarding "resistanceproxpilot danced around 830ohm": This is an hardware issue. To find and fix this, try the following:
1. Use a multimeter to check the PP resistor of your charger (between PP and PE). This should 150ohm if the unlock button is not pressed. (Or 1.5kOhm if you have a CCS type 2).
2. When plugging and unplugging, you should see in the web interface resistanceproxpilot changing between 10000ohms (means infinite) to whatever your charger has.
3. If nothing helps, disconnect everything, use a resistor (something between 100ohms and 1500ohms) and connect it between PP and ground on the Foccci. The value should be visible in the web interface at resistanceproxpilot.

[serh270]

And there is also confusion regarding CCS1 and CCS2. The thread topic says "CCS1", also the picture shows a CCS type 1 plug. On the other hand, I read "Qingdao Penoda Portable DC CCS2 Charger. I knew there are differences between CCS1/2 (I did not order this charger)". So maybe we talk about different chargers or have a different understanding about CCS1 and CCS2.

I thought the plugged look like a ccs1 plug but the manual provided by the supplier states ccs2. Not really sure, trying to get in contact with them but they are in China. Again, I did not select this charger/provider so I'm trying to figure that out slowly.

When you say "PP" do you mean "PE" i.e. Earth?

The charger that recognized connection today was measuring 12v across CP and PP - prox pilot... which is somehow more confusing now. I did not measure across PE because... I am silly

Regarding "resistanceproxpilot danced around 830ohm": This is an hardware issue. To find and fix this, try the following:

This seemed odd to me, will try to diagnose again next time I am testing.

I think the 5% PWM should start as soon as the charger is powered and you have 2k7 pull-down to PE (not PP)

The way the charger seems to function is that it detects "plugged-in", then awaits for an external user activation before beginning any communication or action. I ran out of time today and couldn't test beyond the recognition point but this will be the next step, to determine when/if CP communicates from the EVSE.

1. The PE, protective earth, chassis ground. This needs to be connected to the ground of Foccci. This is the reference point for your multimeter. The single 6mm pin in the center.

All EV-side charing hardware is external from the vehicle. All devices outside of the HV system are sharing the same PE.

[uhi22]

Just to be sure, "sharing the same PE" means, that the following signals are together:
- Ground of Foccci
- Ground of the 12V supply that feeds Foccci
- PE of the CCS socket
- minus of the multimeter if you measure the PP voltage on car side (around 4.5V if nothing is connected).
- ground of the osci if you measure CP (should be at 0V if nothing is plugged-in, then change to 9V when you plug-in and then change to PWM between -12V and +9V, with 1kHz and 5% duty).

[serh270]

Just to be sure, "sharing the same PE" means, that the following signals are together:
- Ground of Foccci
- Ground of the 12V supply that feeds Foccci
- PE of the CCS socket
- minus of the multimeter if you measure the PP voltage on car side (around 4.5V if nothing is connected).
- ground of the osci if you measure CP (should be at 0V if nothing is plugged-in, then change to 9V when you plug-in and then change to PWM between -12V and +9V, with 1kHz and 5% duty).

Yes, these are all shared.

- minus of the multimeter if you measure the PP voltage on car side (around 4.5V if nothing is connected).

I did see this value.

[serh270]

Opened the charger. Looks like the PP line was disconnected completely from the factory (on both chargers). Supplier seems to have confirmed that this is not intended. Will test again tomorrow and hopefully this fixes the issue

[uhi22]

Could also be normal. The PP resistor should be in the plug and nothing else is mandatory on the PP. Is it possible to take a multimeter and measure the PP resistance while everything is off and disconnected?

[serh270]

Ran new test, no liftoff yet but almost made it off the pad. I've attached the logs from the latest test. Made it deep into pre-charge but then the EVSE would fault.

Could also be normal. The PP resistor should be in the plug and nothing else is mandatory on the PP. Is it possible to take a multimeter and measure the PP resistance while everything is off and disconnected?

All the values were consistent with CCS1. Connection on the secondary charger was consistently recognized.

All the messages in the log file up until the pre-charge seem consistent with a good charge as per the CCS1 ICC documentation. The contactors never closed, nor did I ever measure a voltage across the relay control lines.

Code: Select all

[680780] [CONNMGR] 165 004 388 0 157 319 --> 100
[680900] In state PreCharge. TcpRetries 0
[680930] PreChargm aknowledge received.
PreCharge aknowledge received. Inlet 0V, accu 20V, uMin 150V
[680940] Difference too big. Continuing PreCharge.

image.png (10.74 KiB) Viewed 7655 times

The messages seems to maybe correspond with this checkpoint in the nominal startup procedure. T9 occurs during pre-charge and is the second to last step before "energy transfer stage". I'm assuming that by EV disconnection device, they mean the HV contactors as they aren't specifically mentioned anywhere else. When should the contactors be closing? They were open the entire time which seems incorrect because then there is no load on the HV side? Might make sense as to why the inlet voltage is read as 0.


I know there is the analog input for measure charge-port voltage but from how the wiki described it, it sounded optional. Is this correct? This could also be a wiring issue: Where is this data point being measured/sent to FOCCCI?

Edit: Found confirmation that EV disconnection device is the set of HV contactors. When/how does FOCCI decide to close these?

[uhi22]

All the values were consistent with CCS1

No. In the log we see e.g. [931600] ResistanceProxPilot [ohm] 895, this is for sure not consistent (or at least the explanation is missing).
And it is not consistent that later in the log the ResistanceProxPilot jumps to around 200 ohms. It would be really helpful to know the real resistance, and to find out why Foccci measures these strange values.

If you found solutions for the issues, it would be very helpful to share your experiences, so that other folks are able to learn from your experiences. At the moment the impression is: Sometimes something works, sometimes not, nobody knows why, and Foccci is doing strange things. If the analysis turns out, that Foccci has issues, then we can fix it. But with the current information, there is no chance to improve anything.

The physical voltage measurement is optional, this is correct. If not installed, InletVtgSrc needs to be set to 0 as written in https://openinverter.org/wiki/CCS32Clara, then Clara assumes that the inlet voltage is the same as the voltage reported by the charger.

You try to charge a 20V accu with CCS, this will not work. Your charger reports a minimum voltage of 150V. The Clara will only finalize the preCharge, if the battery voltage is above the chargers minimum voltage, and if the chargers actual voltage is near to the battery voltage.
With your 20V battery, the charger does not provide any precharge voltage (because its minimum is 150V).
This is visible in the log:

[708290] PreCharge aknowledge received.
PreCharge aknowledge received. Inlet 0V, accu 20V, uMin 150V
[708300] Difference too big. Continuing PreCharge.

[serh270]

No. In the log we see e.g. [931600] ResistanceProxPilot [ohm] 895, this is for sure not consistent (or at least the explanation is missing).
And it is not consistent that later in the log the ResistanceProxPilot jumps to around 200 ohms. It would be really helpful to know the real resistance, and to find out why Foccci measures these strange values.

Correction, the measured values (with multimeter) were consistent with CCS1

At the moment the impression is: Sometimes something works, sometimes not, nobody knows why, and Foccci is doing strange things. If the analysis turns out, that Foccci has issues, then we can fix it. But with the current information, there is no chance to improve anything.

The solution to the connection problem was simply using the backup EVSE. I wish I could tell you more but I haven't been able to investigate further as to why this worked. Pretty much where I'm at is, sometimes something works, sometimes not, as you stated. Apologies if this is unclear.

My supplier for the EVSE is very inconsistent and it seems that the information they have been providing me has been incorrect. Again, apologies I can't give more info, what I've posted is what I have/know. Thank you again for everybody's grace and patience.

You try to charge a 20V accu with CCS, this will not work. Your charger reports a minimum voltage of 150V. The Clara will only finalize the preCharge, if the battery voltage is above the chargers minimum voltage, and if the chargers actual voltage is near to the battery voltage.
With your 20V battery, the charger does not provide any precharge voltage (because its minimum is 150V).
This is visible in the log:

This to me indicates that the values being sent to FOCCCI over can are incorrect as the attempt target voltage is 790v. Was unsure as to what these messages meant exactly, thank you for confirmation.

The physical voltage measurement is optional, this is correct. If not installed, InletVtgSrc needs to be set to 0 as written in https://openinverter.org/wiki/CCS32Clara, then Clara assumes that the inlet voltage is the same as the voltage reported by the charger.

I will implement this.

[uhi22]

The target voltage of 790V will also not work. The DIN schema, which Foccci/Clara uses, supports ususally 500V maximum voltage. This limit normally would be visible in the ChargeParameterDiscoveryResponse here: https://github.com/uhi22/clara-logs/blo ... .txt#L2280 but your log file contains a lot of bit errors, so the decoding of the log does not work. I'd propose to use a serial cable which is as short as possible (10 to 20cm) between the Foccci and the Serial-To-USB converter. Or/and try an other Serial-To-USB.

[serh270]

The DIN schema, which Foccci/Clara uses, supports ususally 500V maximum voltage

Is there any way around this? Our stack is 800V at ~100% soc. Possibly some sort of software workaround to fudge/nudge the numbers? I guess an option may be to split up the stack but this is also quite intensive.

I'd propose to use a serial cable which is as short as possible (10 to 20cm)

Noticed these errors as well. I'll find myself a shorter cable

[uhi22]

Well, this could be the trigger to integrate also the ISO schema besides the currenty used DIN schema. Seems possible, but some work to do. Is your charger rated for 800V?

[serh270]

Is your charger rated for 800V?

Yes, up to 1000V (if the doc is correct)

[serh270]

Screenshot from 2024-05-10 10-41-59.png (7.67 KiB) Viewed 7514 times

Found the problem with the incorrect target value. Silly-mistake: Can-mapping was incorrect. Above was the incorrect mapping in which only the first byte was taken in. Below is the message:

image.png (7.77 KiB) Viewed 7514 times

Corrected it to this ^. I don't have access to the vehicle to test charging, but will test when I get a chance. Is there any danger of testing without the cell load on the EV side? Want to test charging hardware but unsure if doing so will cause a problem other than triggering a fault when the contactors close

If I try to test above that 500v limit, it just simply won't charge?