Develop a QCA7000 board?

[celeron55]

I feel I should remind people there's also the Chinese 56-pin case that ZombieVerter uses. It might have too many pins, but it's easier to live with too many than too little. And you'll be surprised how fast you run out of pins.

Do we have any statistics about how many AC chargers would be happy with just a mirrored CP signal as 0-12V PWM? I've heard the Outlander OBC is.

[asavage]

. . . the Chinese 56-pin case that ZombieVerter uses.

ZombieVerter enclosure kit links:
https://openinverter.org/wiki/ZombieVerter_VCU#The_enclosure_kit_links

Direct link:
https://www.aliexpress.us/item/2251832671457223.html?spm=a2g0s.9042311.0.0.39f24c4dWOmGPE

There are at least two disadvantages to this enclosure. There does not appear to be any documentation available for it (CAD, etc.), and the lid is grooved for a gasket, but no gasket is available.

I have one, it's nice enough, but I don't know if it's a compelling product.

[muehlpower]

I prefer a housing that I can easily get here and also the appropriate connectors individually. Two individual connectors are also useful, e.g. to make one for the charging socket with LED, locking actuator, PE/PP, control button and fuel flap switch and the other for the connection to the car with 12V, KL15, CAN Charge enable, HV Contactor control etc. Deutsch DTM or Molex (Cinch)

[uhi22]

The next version will have the TE DEUTSCH DTM13 connector. I was also planning, that two connectors into two directions (one for 12V and CAN, the other for the charge port) is perfect, but not all functionality can be easily assigned to the one or the other direction. So I ended up in a "random" pin assignment, also driven by layout. Still some work to do, but at least all components fit. Added also discrete highside drivers for the RGB. The charge port lock motor driver is my biggest question mark, I have no idea about a small, cheap and available driver.

This is a draft:

And this the connector, ~17 Euro on Aliexpress. The cable side connectors are reasonable priced (payed 18 Euros for 4 parts), the enclosure comes on top, did not yet find a good price.

[muehlpower]

Do you think this is suitable for the locking actuator?
https://www.mouser.de/ProductDetail/Tex ... qTkw%3D%3D

The locking actuator is ratet at 3A in 12V version!
https://www.evsegroup.com/electromagnet ... x-product/

[MeKe]

Do you think this is suitable for the locking actuator?
https://www.mouser.de/ProductDetail/Tex ... qTkw%3D%3D

The locking actuator is ratet at 3A in 12V version!
https://www.evsegroup.com/electromagnet ... x-product/

There is the following statement in the datasheet, I think it is not appropriate. only parallel connection is suitable

• 1-A RMS current for each output
– 6-A maximum current for paralleled outputs

[asavage]

The locking actuator is rated at 3A in 12V version!
https://www.evsegroup.com/electromagnet ... x-product/

That web page has two different sets of specs!

Nozzle Lock DSIEC-ELM 01b

Nozzle Lock DSIEC-ELM 02b

Which to believe?

[later: removed a bunch of unrelated stuff about NACS and moved it to a different post + thread]

[catphish]

Do you think this is suitable for the locking actuator?
https://www.mouser.de/ProductDetail/Tex ... qTkw%3D%3D

The locking actuator is ratet at 3A in 12V version!
https://www.evsegroup.com/electromagnet ... x-product/

Alas this doesn't quite add up. The part you listed is 4 x 1A half bridge. In order to be able to reverse the motor we need a full bridge (half bridge on each side of the motor), so that's only 2 x 1A on each side of the motor or 2A total. Not quite enough.

The 6-driver version of this chip would do, but I suspect there are larger full bridge chips available without the complication of paralleling.

[mstegen]

The DSIEC locks come in a variety of different versions.
The earliest models were solenoid operated, they require a 12V pulse @3A for around 0.3 - 0.6 sec
On the newer models a 12V motor operates the locking pin, they require a lot less current. (around 100mA)
There are also locking actuators from Hella: 11122-05 also 12V and driven with a motor.

The one from Hella is most likely the most suitable, as they are fitted with a automotive (TE MCON) connector.

[muehlpower]

Do you think this is suitable for the locking actuator? Now the right link!
https://www.mouser.de/ProductDetail/Tex ... 3Itg%3D%3D

The locking actuator is ratet at 3A in 12V version!
https://www.evsegroup.com/electromagnet ... x-product/

[muehlpower]

I installed this, it is specified as 3A / 36 W. Most sources say so. But maybe there are also differences, e.g. DC motor versus solenoid with peak and hold.
https://www.evsegroup.com/electromagnet ... r-product/

[asavage]

The locking actuator is rated at 3A in 12V version!
https://www.evsegroup.com/electromagnetic-lock-for-electric-vehicle-power-type-2-ev-charging-socket-actuator-wallbox-product/

The linked page above shows wildly disparate specs.

I installed this, it is specified as 3A / 36 W. Most sources say so. But maybe there are also differences, e.g. DC motor versus solenoid with peak and hold.
https://www.evsegroup.com/electromagnetic-lock-v4z-dsi-el-for-iec-62196-2-type-2-socket-actuator-product/

This is a different unit, with more reasonable printed specs.

I mistrust suppliers who can't keep their specs straight (consistent) on a single web page. If they can't be bothered to fix their publc-facing documentation, what other corners have been cut? I will always pay more to a more professional vendor . . . even for the same product! Because they are providing better service overall, and not forcing me to guess, cross-check, and doubt.

[asavage]

I'm working with a start-of-production (SOP) Tesla Model S (NACS) charge port, which uses a fairly large solenoid and bellcrank arrangement that is powered to unlock, and spring return to lock. There is a lock position sense microswitch to provide feedback that a lock or unlock event has occurred. Therefore, it doesn't require a lock motor driver, but does require an economizer, as it requires ~12v/4.5A to pull-in, but only 2v/0.5A to hold-in. It overheats quickly if not economized.

As mine is probably a niche case, I do not expect this PWM (alternatively: peak & hold) functionality for a solenoid-operated nozzle lock to be accommodated by this CCS PCB; I mention it only as -- perhaps -- an edge case.

As more of the charging infrastructure in the North America market move to NACS -- for various reasons -- I think that there will also be more DIY projects that move to using the NACS physical connector, while using the CCS signaling/comms protocol. While my own niche project is rare now, in future it may be more common.

I'll probably use an external PWM timed function to drive this Tesla NACS nozzle lock. [later: or use one of the techniques described here]

More information on this Tesla early Model S NACS charge inlet/port, with lots of pictures, is -->here.

[uhi22]

This sounds like the 6A peak of the DRV8874 should also cover this use case with the solenoid, as well as the motor variant. I will go with it for the next prototype. It is nice small. (Also found how to use the available 3D models of the components, so we have a nice overall picture now)

[asavage]

The DRV8874 looks good, good suggestion by muehlpower! I've had a read through the datasheet. They don't give many clues for choosing the local bulk capacitance value, other than "test your design and adjust accordingly", but others have questioned on the TI forum and gotten, somewhat begrudgingly, "2-4uf/w", and a warning not to exceed 4uf/w.

I like how the datasheet says to refer to a table in the datasheet for guidance . . . and then doesn't have that table

---

Is it worth your trouble to make it possible -- onboard or not -- to drive eg. 20mhz 20khz PWM on the EN/IN1 line? I don't want to enable scope creep -- well, maybe I do -- but having the option of toggling PMODE high and either generating PWM onboard or allowing it to be driven from an external source would make my life easier.

I'll need to drive that pesky Tesla solenoid via peak & hold somehow . . . maybe I should just install a BDC motor instead

[mstegen]

For my EVSE i use the 2EDN7524FXTMA1 (mosfet driver) as Actuator driver. It can drive the DSIEC locks (both solenoid and motor versions) without problems.
Also tested the Hella, and Phoenix Contact actuators with it.

[P.S.Mangelsdorf]

The next version will have the TE DEUTSCH DTM13 connector.

Just want to add that I full endorse using Deutsch type connectors whenever possible. If you get the right tool, the barrel type connector pins are super easy to crimp correctly. No silly tabs to fold over, just: wire in pin, pin in tool, squeeze, done.

[asavage]

Just want to add that I full endorse using Deutsch type connectors whenever possible. If you get the right tool, the barrel type connector pins are super easy to crimp correctly. No silly tabs to fold over, just: wire in pin, pin in tool, squeeze, done.

Pressmaster MCT crimper + dies

Sadly, my main terminal crimping system (Pressmaster MCT aka Waytek) has no dies for crimping Deutsch terminals.

Can you recommend a good quality crimping tool for Deutsch terminals? I usually like to buy top-tier tools, used. I think I'd like to spend no more than USD$200. I like to buy tools, but I do have to limit myself somehow.

I do not buy cheap tools: it only encourages them to make more cheap tools.

[P.S.Mangelsdorf]

Collection_20230813.jpg

Sadly, my main terminal crimping system (Pressmaster MCT aka Waytek) has no dies for crimping Deutsch terminals.

Can you recommend a good quality crimping tool for Deutsch terminals? I usually like to buy top-tier tools, used. I think I'd like to spend no more than USD$200. I like to buy tools, but I do have to limit myself somehow.

I do not buy cheap tools: it only encourages them to make more cheap tools.

Honestly, this one (though fairly cheap) has worked amazing for me. I had also bought one from WireCare when I bought a bunch of Deutsch connectors, and I prefer this one.



Just to note, there are two forms of Deutsch pins. The cheaper have the fold over tabs, the slightly more expensive are the barrel type that use this crimper. I whole-heartedly recommend the barrel type.

[asavage]

No image showed up for me in your post (maybe I have a blocker installed that catches it?) but I extracted Amazon's ASIN from the html, and came up with B07V1C6N2V.

https://www.amazon.com/IWISS-Crimper-Impression-Deutsch-Contacts/dp/B07V1C6N2V

IWISS ICrimp IWD-16 Deutsch crimper

I like the imprinted directions for use, handy for those of us who don't use it daily. I think of IWISS as the top tier of the bottom end of hand tools. I don't own any, but read generally good things about the brand in the reviews, whenever I'm tempted to buy one of their products.

For automotive use, 14/16/18 AWG (2.5/1.5/.75mm²) is about all I think I'd need, so it covers those. Thanks for the recommendation.

[P.S.Mangelsdorf]

No image showed up for me in your post (maybe I have a blocker installed that catches it?) but I extracted Amazon's ASIN from the html, and came up with B07V1C6N2V.

https://www.amazon.com/IWISS-Crimper-Impression-Deutsch-Contacts/dp/B07V1C6N2V

That's odd, I had originally just inserted a link.

[uhi22]

Is it worth your trouble to make it possible -- onboard or not -- to drive eg. 20mhz PWM on the EN/IN1 line? I don't want to enable scope creep -- well, maybe I do -- but having the option of toggling PMODE high and either generating PWM onboard or allowing it to be driven from an external source would make my life easier.

I'll need to drive that pesky Tesla solenoid via peak & hold somehow . . . maybe I should just install a BDC motor instead

The 20mhz is a typo, right? The board will support PWM on both, the contactor outputs and the lock/unlock. I would start with something like 100Hz, because the contactor driver is quite slow, did not have a look to the DRVs speed, but now they both use the same timer of the STM, so they have the same cycle time. Any reason for much faster PWM?

[asavage]

Ugh, yes mhz was a typo, Thanks.

That's good news, and I appreciate the info.

I dragged the 20khz figure from a sample application design exercise from the datasheet:

[johu]

I have soldered the processor to the 2023-07-12 board along with the bypass caps and 3V3 reg. Strangely it only draws 7 mA on the 3V3 rail instead of 13 mA on a BMS board. And it doesn't communicate via SWD. I have only GND, SWCK and SWDIO connected and external 3V3. Checked all power pins and SWD pins for continuity. Any ideas?

[uhi22]

Maybe the reset line is shorted to ground? And the orientation of the controller is correct?