A range of ZAPTEC GO issues

[Engineeronaut]

Hi,

Need to make a power limiter for coming pwr tariff, however I notice that zaptec integration is not working at all as expected.

1. energy report intervals are too long, up to 4 minutes at times so it becomes hard to make a system. Typically 0.2kWh. which is 240 sec at 3kW approx. So not working in practice. Occasionally it may drop to 0.1kWh. Interpolation is hard because randomly the zaptec reports some values and you count double and i dont know the actual power to interpolate with, see further.
2. Power, amps, voltage does not update if not very large changes and on occasion. So they are always wrong. stay flat for hours.
3. The time on reporting an energy quanta is not precise so typically every other is 10% up or down in a "sinus manner", but not always. On stop of charge it can be any time really so. Also sometime after a stop there will come an additional energy update report some minute after the stop. So i get easily +-10% or occasionally even 20% or more. and I cannot sanity filter since I dont know differnce between a change and an error fro reasonable values.

Thereby I cant know power from power value and I cant know power from dE/dt either. Nor from cos phi amp and voltage. So that makes the data from the zaptec not dependable. Thus i cannot find out what is base power and what is load to manipulate. I have the shelly for the other car and it is consistent.

4. Also Meter power is not added when current session is added. so In practice i need to make my own true energy at all times. Since current session jumps back to zero at disconnect but as of all else, I have low confidence. I dont know if this is a waird zaptec setup but i would guess they count up actual energy meter setting.

5. also on occasion voltage values can be set to 100V or some other erroneus value and stay there for long times.

I dont know if this has been always, since i only now started building the power tracker. Maybe this was the cause for som oddities in the fuse, load watch I have earlier. I think that voltage and current changed properly a year ago but not sure. I remember seeing "blanks" in the device monitor view a while back.

I hava a log of all events that i did so you can see true all reports made from zaptec regarding charge anyway. I haven't looked at temp and humidity and so on.

Best regards
/stefan

[PatrickE94]

Hi @Engineeronaut, I guess you share a bit more than intended in the xlsx. I'd suggest an edit to clear it up :)

I have a hard time following your train of though here, and I also know nothing of your setup, but heres my two cents.

The power tariff is essentially kwh consumed per hour, not momentaneous. So if you're reaching 6kWh for the current hour and you want to stay below 7, it's time to shut off charging.

Power management is, in practice, trivially handled by simply figuring out your "base load" for the hour (like heating pumps and other things) and calculate the headroom you have left for charging. No need for real time management.

I'll also just add to this that changing charging settings more frequently than every five minutes throws some cars of the charging session. Especially the chinese ones.

Best regards,
/Patrick

[Engineeronaut]

Hi Patrick. I believed I had selected the page only from sheets, not the whole file, so yes a lot of crap. I made a new one.

The issue however isnt the logic in flows or anything like that. Its the "API exchange" that isnt working properly. The base load algorithm is fine, its the stale data from zaptec that is the problem.

This data I provided is the RAW data dump from events according to the purposely made event tracker to describe this problem only. It has no other purpose than assisting yuo in finding the bugs. Depicted is its flow so you see how it is logged and labeled. As you see in the data the data goes stale just as I decribed and intervals are way too long on energy reports. This has nothing to do with my decision making or what flows i am building later on. I labeled the event type, i e trigger card used. When I send a current limit request or any request I get instant response, so there is no general network comms issue. Also the "0,2kWh" ticks are completey consistent, althouh on occasion shrinking. Even more strange is that when some other event actually picks a change then the other values arent updated. Start-stop always updates all values. Well also some NULL values at times. seem to be in relation to connect and disconnect, perhaps.

zaptecHomey.xlsx

Look at the amps readings and power and so on from zaptec, stale, flatlining. Hours of one hundreth of watt same value, that is not what is happening in reality. columns F to R is raw data dump provided by trigger cards for zaptec. Maybe that was unclear. So you can see, it is not correct, it is stale data. Except col G which is JAVA absolute seconds. I and U constant while P changes, U constant for hours. I constant for hours. That cannot be correct. The total energy counter also very randomly updated by the looks of it.

Alo to repeat my first message point, the actual timestamp for an current session energy update is not correct it implies pwr variation cyclicly every other value to 20% difference. So today I need minimum 16 minutes of constant power to determine the actual power. Also the load watch not matching since amps are wrong, but in practice a trivial issue. Since amps and power is linked by PowerFactor, cos phi also. Power watch is real power only and current is real current for fuse watch. Nevertheless i dont know cos pi, around 0.93.

So what I would like is a higher data pull frequency from zaptec and also to aid you in what to look for regarding the stale data bugs.

Best regards/stefan