Setec CCS to Tesla Adapter

[Boeingpilot]

I wonder if it’s a clocking issue. When you hook up the serial port, then you are slowing it down?
BTW Setec just sent me v130. I’ll let you all know how that works by Monday

 

[cwerdna]

Right. There could be a timing issue/race condition that is "corrected" when something is altering the timing of the adapter. And, for all we know, there could be some code path on the adapter that is only exercised when it's attached to a computer.

 

[Boeingpilot]

Well today was disappointing.... Tried again at the BTCPower (Blink) charger. Back to the isolation faults at the charger with v130. So, then I said, let's downgrade back to v129 and try. Again, isolation faults (even though it said it wrote, who knows if you can downgrade?) So, more emails and logs heading back to Setec.

Maybe they will offer me a job as a compatibility tester?

 

[cwerdna]

Randy Sinn: Not sure if you've been able to keep up with this entire thread but at Setec CCS to Tesla Adapter, I alluded to at least 8 DC FC CCS manufacturers w/equipment in the US.
....
I can't speak to all of EVgo, but the DC FCs I've used with them are BTC Power. But, they do have some from ABB (e.g. PlugShare - Find Electric Vehicle Charging Locations Near You, PlugShare - Find Electric Vehicle Charging Locations Near You and PlugShare - Find Electric Vehicle Charging Locations Near You).

If I missed any Combo1 DC FC manufacturers with equipment in the US, please chime in.

I was informed via private message on MyNissanLeaf that in Chicagoland, EVgo also uses some Efacec. Thanks to the tip, I found a few examples:
PlugShare - Find Electric Vehicle Charging Locations Near You
PlugShare - Find Electric Vehicle Charging Locations Near You
PlugShare - Find Electric Vehicle Charging Locations Near You

So, EVgo for CCS is known to use at least BTC Power, ABB and now Efacec.

 

[Reddy_KW]

This is definitely a crowd sourced beta test program. I’m hoping to get another firmware update to test tomorrow. I’m still on v129.

 

[miimura]

This makes me wonder, what if you didn't hook it up to a laptop but a regular power bank?

Pure speculation, but when the adapter is plugged into the laptop it may technically be always on or something like that which may affect the operation of the adapter.

Even though it "works", it may still be useful to send the data to them as they may notice something in the logs that could explain the failure and allow them to fix it without having anything plugged in.

This thought occurred to me too. Someone up-thread also said to make sure to charge it fully through the USB port before using the adapter.
Trying it with a USB power bank instead of the laptop would be a useful data point at a station that doesn't work with nothing connected to the USB port.

 

[NickFie]

I wonder why it works while connected to the laptop and instead of without it.

Looks like two posters beat me to it.

Two possibilities to explain why the adapter works in more scenarios with the logging laptop attached:

1. Laptop USB connection provides continuous, well-conditioned power from a relatively high-capacity battery. You could test this with a USB power brick and USB cable.
   
   SETEC fix would be to provide better power to their adapter. They could probably emulate whatever Tesla's CHAdeMO adapter does to power its brains from the low-voltage signal & probe lines.
   
   
2. Clock or signal bus data rate could be a bit too aggressive for many CCS chargers. Logging introduces a bit of processing overhead. That slows the cadence to the point the CCS charger can cope.
   
   One can safely assume that nearly any product with a long design - develop - manufacture - first ship cycle will have barely-adequate CPU speed, slow and overstuffed RAM, possibly insufficient storage. The cost-control team corner the engineers by asking, "Can you guarantee the product will fail if we build it with the next slower/smaller version of that component?" This is aggravated by inevitable software growth over a very long field life. Yeah, we're talking about YOU, MCU-1.

SETEC seems like a pretty sharp and motivated team. Hope these suggestions help move things forward. I'm getting interested in CCS adapter for trips in Supercharger-sparse areas.

FWIW - I spend a lot of my work day with this sort of knotty problem. One of my favorites was diagnosing failing production line motor start capacitors from an interesting pattern of TCP misbehavior on some equipment associated with the line. Seems like an obvious possibility in retrospect, sure had to connect some interesting technology dots to explain intermittent data failures during line startup.

 

[Boeingpilot]

Setec has sent v132..... will let you all know what happens

 

[Reddy_KW]

I just tested V130 on a Signet V1 Electrify America station and a Efacec QC45 station. No problems and I did not have to have USB connection to laptop.

 

[jmbloom_m3p]

I just tested V130 on a Signet V1 Electrify America station and a Efacec QC45 station. No problems and I did not have to have USB connection to laptop.

They are iterating so rapidly. That is great. They clearly believe they can solve it, there is a market, or I doubt they would have this kind of turnaround. Helps to have paying testers! Thanks all.

 

[Boeingpilot]

They are iterating so rapidly. That is great. They clearly believe they can solve it, there is a market, or I doubt they would have this kind of turnaround. Helps to have paying testers! Thanks all.

I'm just happy they are actually trying to support it. I've gotten more support for a $660 adapter than Tesla has given towards my constantly locking up and rebooting MCU (attached to a $80k car....) Just saying......

 

[KJD]

Setec has sent v132..... will let you all know what happens

Looking forward to your test results with the latest version.

 

[ItsNotAboutTheMoney]

The writing was on the wall for CHAdeMO even before EA deploying chargers because the only BEV that used it was the Leaf and the PHEV Outlander and maybe a few thousand Tesla owners with adapters. Everything else was using CCS and most deployments already saw CHAdeMO limited to 62.5 kW with CCS going higher. So to me it makes complete sense for EA to only include one per location to support the small number of users.

At the end of 2019 there had still been more CHAdeMO cars sold in the USA than CCS cars,

U.S. BEV Sales By DC Fast Charging Capability: 2019

158k v 144k, with CCS getting 23k closer in 2019.

What's really annoying about it is that it's just a cable and protocol. They could have put 2 CHAdeMO in and gone back and replaced the cable later. It's not like the cables don't ever need replacing.

 

[Boeingpilot]

More updates! As you may remember the most problematic CCS station I've dealt with is the local 50kw Blink (BTCPower). It had only worked once, and even that time the handshake was very long.

Setec forwarded me v132 firmware a couple of days ago. Loaded it up and went over to Blink station. It connected properly and immediately. The handshake time was significantly reduced (as if instead of doing retries trying to get it to work it actually did its handshake the first time).

There have been some reports on this thread that one software fix may cause incompatibilities on a previous working station. I am planning a short trip in the next week and am going to try to survey different stations to see that a - we haven't gone backwards (though the only other station I've had success at was a EA Signet unit) and b - see what else is out there. I'll report when I've done.

Would a separate thread ONLY with the following be useful? Vehicle - network - type station - success or fail - software version (if known)

A spreadsheet would be better, but not sure how we keep that up to date. The other option is if people on this thread want to PM me their results, I could tabulate, and generate a link to a shared google drive spreadsheet. Thoughts?

 

[ItsNotAboutTheMoney]

More updates! As you may remember the most problematic CCS station I've dealt with is the local 50kw Blink (BTCPower). It had only worked once, and even that time the handshake was very long.

Setec forwarded me v132 firmware a couple of days ago. Loaded it up and went over to Blink station. It connected properly and immediately. The handshake time was significantly reduced (as if instead of doing retries trying to get it to work it actually did its handshake the first time).

There have been some reports on this thread that one software fix may cause incompatibilities on a previous working station. I am planning a short trip in the next week and am going to try to survey different stations to see that a - we haven't gone backwards (though the only other station I've had success at was a EA Signet unit) and b - see what else is out there. I'll report when I've done.

Would a separate thread ONLY with the following be useful? Vehicle - network - type station - success or fail - software version (if known)

A spreadsheet would be better, but not sure how we keep that up to date. The other option is if people on this thread want to PM me their results, I could tabulate, and generate a link to a shared google drive spreadsheet. Thoughts?

A Setec Wiki thread could be useful.

 

[pb2000]

We've gone from hardly anything working to mostly everything working in two weeks, so unless progress stalls out, it may be a moot point.

It might still be worth putting something together listing max charge speeds for each vehicle and sub variant and anything confirmed not working after Setec moves from beta to stable firmware releases.

 

[chronoreverse]

The fact Setec is making a very apparent effort to have this fully working is really promising. Good thing they have that USB update port!

 

[mociaf9]

Don't forget, this is a piece of 3rd party hardware. I certainly won't be that surprised if some future minor vehicle firmware change that Tesla implements via an update ends up breaking (accidentally or otherwise) whatever Setec gets working. The fact that it is taking so many revisions for the adapter to consistently work is evidence of how finicky and particular these charging systems are.

 

[Boeingpilot]

Don't forget, this is a piece of 3rd party hardware. I certainly won't be that surprised if some future minor vehicle firmware change that Tesla implements via an update ends up breaking (accidentally or otherwise) whatever Setec gets working. The fact that it is taking so many revisions for the adapter to consistently work is evidence of how finicky and particular these charging systems are.

Not so sure of that... the issues have been the variability on chargers, not Teslas. We've had reports of successes with M3s and MSs. (Not sure if any MYs or MXs have checked in). Yes Elon & Co. could change something, but I doubt it. The adapter basically shows itself to the car as a Chademo adapter. At his point for Tesla to change anything would (or should I say could) break the Chademo adapter.

Remember, when the M3 first came out it was incompatible with the Chademo adapter. Can't imagine Tesla would do something to make the Chademo not work

 

[BritneySpears]

Not so sure of that... the issues have been the variability on chargers, not Teslas. We've had reports of successes with M3s and MSs. (Not sure if any MYs or MXs have checked in). Yes Elon & Co. could change something, but I doubt it. The adapter basically shows itself to the car as a Chademo adapter. At his point for Tesla to change anything would (or should I say could) break the Chademo adapter.

Remember, when the M3 first came out it was incompatible with the Chademo adapter. Can't imagine Tesla would do something to make the Chademo not work

I'm not familiar with how the comm works between the car, adapters, and station but if the Setec is just emulating a Chademo would a dumb, pass-thru adapter work if placed between the station CCS1 and Tesla Chademo adapter? The J1772 receptacle-to-Tesla plug for a/c charging seems to work this way.