PV optimized Volkswagen charging w/o additional hardware

 

This time it’s not about photography and – to limit your expectaions – it is still “Work in Progress“

End of May we got our 1St electric car: a VW e-UP!

One question from the beginning on was how coordinate charging with our solar panels in the most effective way.

Our solar system is orientated in East/West direction (35%/65%), 9,35 kW peak and has a 5 kWh Li-battery produced by SENEC. Details can be found in my one year report (sorry: it’s German only) https://klaus-macht-bilder.blogspot.com/2020/05/ein-jahr-solar-eine-erfahrungsbericht.html,

Initial idea was to get a SENEC-Wall box with excess charging function – so I applied for NWR subsidies and ordered it November last year. Now – after 7 month there is still no timing for delivery and assembly.

Since that order there were several things that happened:

• I learned how to install and operate Home Assistant (at least a little bit)
• Experiences of users of the SENEC system were very disappointing. Due to the operation via the SENEC sever and communication in time intervals of app. 5 minutes seem to limit the effectiveness significantly à after consultation with our solar technician it was possible to change the order an ABL wall box without the (expensive) SENEC integration.
• Due to the situation owning an EV but no WB I had the opportunity to collect first experiences with the VW granny charger over a normal plug with 2.300 W max. charging capabilities

1^st (temporary) Solution: a SHELLY WiFi Plug with up to 16A

-       This is connected to the home grid – which means that the LI-battery is utilized in the charging process too

-       The data acquisition from the SENEC battery system is based on the Home Assistant SENEC integrations (HACS) – and limited by the SENEC refresh rate too.

-       Further I was able to read yield data from our SMA inverter via ModBus in Node Red (separate German post at https://klaus-macht-bilder.blogspot.com/2021/04/pv-anlagedaten-in-home-assistant.html)

-       Control via HA and Node Red:

o   Definition of an HA Helper ´Solar-Laden 220´. When ON the system should run solar controlled, otherwise the Shelly-Plug should run unaffected

o   Definition of start conditions: SENEC SOC > 60% AND solar generation > 2.500 W

o   Definition of charge cancellation conditions (all OR):

• SENEC SOC < 40%
• PV generation < 1.800 W
• Purchased electricity > 1.000 W

o   The hysteresis limits the number of switching operations – as a result overall quantity of purchased electricity is minimal (at least during summer)

o   The parameters can be adopted to specific needs

EV Dasboard showing PV & EV Data and charging condition

2^nd (Work in Progress) solution: Use of the Volkswagen „We Connect“ interface

Idea is to control the charging process from the “vehicle side”

-       For the first year after buying a new e-UP! the WeConnect-App is free of charge – later there will be an annual fee.

-       In Home Assistant there is a HACS integration for VW-We-Connect – via this the charging process can be started and interrupted similar to the SHELLY plug in solution one

-       This was the part that works already so far and probably will work with a wall box too

Node Red implementation

     Now I come to the open issues:

-       The number of switch operations seems to have limitations

-       One it stopped working and in the WeConnect-App I got the message that switching is not possible because of too may operations – re-use was possible after an engine start of the car only (reported at https://github.com/robinostlund/homeassistant-volkswagencarnet/issues/308)

-      

To reproduce this failure mode I establish a HA-Automation that ran in a 5 Min-ON/5 Min-OFF mode continuously  - even after 3h I was not able to reproduce the failure mode.

  

Meanwhile I found that the failure seems to be on the App-side only. When I started WeConnect on my mobile after the test (which seemed to be ok from the HA side) I got the same error massage I got before.

In the VW-App the maximum current can be limited to 5 – 10 -13 -16 A steps. It would be great if this could be set by the HA-Integration too. I opened a Github-Issue on that already https://github.com/robinostlund/homeassistant-volkswagencarnet/issues/307

     During the failure reproduction (above) there was one unexpected behavior which could have a interference with this current setting function: I set the car profile to Max-Current=5A and used my granny charger – so charging started at ~ 1.1 kW (5A). At that point Home Assistant still reported Max-Current=16A. After a HA initiated forces update this value stayed a 16A. When the Automation was started the real charging current rose to 10A (capacity of the granny charger) and the sensor-value for Max-Current stayed at 16A à it seems that the HA-Values can be transferred to the car somehow, but I am lacking the capability to find out how. 

     After the test I am not sure whether I changed the APP back to 5A and actualized it or whether it was on the 5A all the time - in that case it may be possibel thet the operation via HA-switch is bypassing the vehicle settings  - in that case this would be bad for a wall box transition of this process. I will try to repeat this part of the test.

 

I will  link this to the relevant platforms where I discussed parts of this already and hope to get constructive feedback – I will keep you updated here.

 

 

 

 

 

 

 

 

 

 

 

 

Update 18 May: 

• It seems that the charge control by HA overrules the settings from "Charging Locations" (German menu: Ladestandorte - in the case above this was set to 5A max charging current) 
• The settings for "Instant Charge" (German meue: Sofortladen) from the Settings/e-Manager are kept and not overruled
• Meanwhile I extended the Node-Red flow by a switch WeConnect/Shelly-control and implemented a Shelly EM in the supply line to measure the charging process independant of the Shelly Plug.

Camera flashes and gravity

I use a  Pentax-Flash AF-540FGZ since 2007 - some advantages inside the Pentax system are:

• High power
• Reflector movement in 2 planes
• Automatik Modus independant of the  P_TTL-System (for fast moving eye lids and reflecting elements
• 2 Slave Modi (TTL and "just follow")

2010 g<rvity hit first time: The flash dropped down to the ground from a tripod. Afterward the battery door did not lock any longer because the plastic closing elements were broken. 

A initial repair ba adhesive tape did not work because battery pre-tention peeled ist off after some time.


Solution was the fixation of Aluminum elements with 2-komponent-glue and a small milling tto the battery door. 

 

This solution worked reliable for nearly 6 years and needed to be re-glued after 6 years the first time.

Recently there was the next impact: The flash dropped down tgether with a Q10er from ahight of a little bit more than one foot  and the connection shue brake.   This assembly is just fixed by 4 small screws and an electric connector. A try to get spare parts as a organ donor by a forum or Twitter failed.

At  Pentax-Service März the spare part was available for ~48€  and after assembly the flash worked agaiin.

Conclusion: Gravity is effective for camera flashes too - bit that is no reason for scraping it or buying it new. Some thing could be done better by the manufacturer. For example  the shue of Cactus V6 triggers are metal reinforced.

Firmware update of the Sigma 17-70 Contemporary using the #Dock

1St Experiences with #CineStillFilm #800T

It started at the 2014 Photokina in Cologne. At the booth of CineStillFilm I had seen that film the first time - and heared about the Kickstarter Project for 120-Type-Films.

At home I placed some Dollars at Kickstarter and later I found 135-Type-Films available at Fotoimpex, So I ordered 3 of these films and lodes the one to my Pentax Super A.

Meanwhile the Kickstarter-Project failed but my initial film is exposed and devoloped.

The use of tungsten film with 800-ASA is very limited but the results show that this film has it's qualities.

All these pictures have been taken with a M35/2,0 w/o any tripod - but see yourself:

 

The results are vey impessive and night athmosphere is shwon in a unique way. 

Especially with sufficient exposure (related to 800 ASA) the results are absolutely great- but when taking picture at night some areas of the pictures are underexposed/black. These ares show significant blue grain that can be reduced by using lightroom CA-Correction after sanning (some of the tungsten calibarion is lost by this adaptation).

Conclusion: RECOMMENDABLE - It's a pitty that the Kick-Starter-Project for 120-Type Film failed.


PS: All pictures are shown at my  Homepage

Indian Summer and NYC

In October we traveled the NE of the USA - during the next weeks I will report in my German Blog and will show pictures - You are invited to follow this blog by using a translation.

My travel report starts HERE

Modification of a #BG4 for #PentaxK3

This week I got may new K3 to support the old K7 and K01.

I found some descriptions, how the modify the BG4 in a way that it works on a K3 - so I took my tools and started work:

Stage 1: What is different (by Scanner & GIMP)

Stage 2: Handle Modifications (Supplier Khalia)

Stage 3: Ready :-)

#Pentax #6x7: Broken chain

I love using my old #Lunasix.......especially since the coupling chain of my Pentax 67 is broken since Saturday :-(

#Pentax #K01 #Firmware update solves #Sigma #Ring-flash issue

As reported in March (sorry - just German) the collaboration between Pentax K01 and Sigma Ring-flash EM-140DG did not work - the flash fired always full power and didn't care about the TTL-settings. This fault is gone after the latest update to Firmware 1.04 that was released last week by Pentax.

Isle of Man TT 2012

The TT 2013 is just over and I have already prepared the last racing photos from last year.

I want to guide you arround the Mountain Course with 8 stops: 

Start/Finish

Braddan

Ballaugh Bridge

Gooseneck

Bungalow

Creg-Ny-Baa

Governers Bridge

Victoria Road

I hope you liked the 60.73 km (37.733 mi) trip arround the course.
More pictures are  here.

Conversion of a Pentax K24/2,8 to 'A'

There is a lot of great  K and M lenses around - and they would even be better if the are "A"!

Therefore I thought about a way of conversion........

Starting lens was a K 24/2,8 and to get some of the necessary spare parts I bought a cheap F35-80 at ebay.

Initially I tried to fit the F-back-ring to the K-lens. The positive surprise was: It fits! 
It is just necessary to put the right contact pins to the right positions.

To find out which are the right pins Bojidar Dimitrov's Pentax K-Mount Page is a good source. It shows that 4 pins have to be inserted and 2 positions need to be isolated (011*10). 

After recognizing that the insertion for spring loaded contact pins is much more difficult than the isolation of positions I changed my plans and decided to use the original ring as a starting point. The F-ring was used as a drilling jig only.

This should work with one contact pin and 2 holes/isolated points.

  But during the first camera trials it shows poor results:
- On Pentax P30 and Z70 the cameras recognized "A" - but apertures are  1,4...22
- On Pentax K10d and K7 the lens was not recognized at all :-(

Applying one additional hole didn't help in the expected way. After some discussions with a friend from German Pentax Forum DFN I decided that one of the holes was too much.

After soldering the excessive hole and isolated the last drilling it works: Apertures form 2.8 to 22 can be adjusted by the camera!  

 
The first tests on my K7 showed the expected results.

Exposure is OK at 2,8 and needs to be adjusted by minus 0,5 step for each step the aperture is closed (means F8 --> - 1,5). Without this correction it causes over-exposures the more you close the aperture. It is caused by a lack of linearity in these old lenses but I think that is something I can use.

During the tests I found that the spring loaded pin didn't like any movement of the aperture ring. Using one of the longer screws from the F-lens blocks the ring @ F22 now. 

The surprise came when I tried the lens on my Pentax K01:

Right exposure was possible only at F8 and above.

What is the reason/A hypothesis:
If you look at the K01 behavior you can see that A/F/FA/DA-Lenses have partly closed apertures during view and measurement mode depending on the light conditions. This may cause in the result that aperture values cloes to the settings during the measurement are well exposed and wide open shots are underexposed.

 If you have additional/other ideas your input is appreciated!

Update September 2013: Due to the liniarity failure and the limited advantages of the modifications I rebuild the lens now using a back ring of a defect K28/2,8. Parallel I got a manual Sigma A lens 

Therefore the K24/2.8 is on sale now!