Robo-Tank v6.7 is Ready

Have you placed the jumper leads for the ac bars on the main controller. These need to be placed before the ac bars work. You can also place a safety float switch if you want the backup.
Yes, I have a fully running Robo Tank on version 6.6 that I attempted to upgrade but couldn’t. It still works fine on 6.6. I took a second card and cleaned loaded it on a second raspberry pi and robo tank. The 6.7 will not control either power bar on either controller. The 6.6 card works on both.
 
I don't think the problem is the jumpers as those only affect 2 outlets. Any chance you have Bookworm running? I'm wondering if that's the cause. I'm just formatting an SD card with it now and will test. I have a feeling that's the cause as mine is working ok.
Not that I know of. I literally wiped the SD and started clean.
 
Not that I know of. I literally wiped the SD and started clean.
I just tested bookworm and the sysfs system has changed or no longer there. They made a lot of changes in bookworm, sadly this is one of them. If you run the following command it'll print out what OS you are using.

grep VERSION_CODENAME /etc/os-release | cut -d '=' -f 2

It will likely say bookworm, if so you need to reinstall everything but this time in the Pi Imager program you need to select the version I circled below.

Screen Shot 07-03-24 at 05.07 PM.PNG
 
I don't think the problem is the jumpers as those only affect 2 outlets. Any chance you have Bookworm running? I'm wondering if that's the cause. I'm just formatting an SD card with it now and will test. I have a feeling that's the cause as mine is working ok.
It is running bookworm. I am reloading it.
 
Hey Rob,
I'm having trouble with the dosing getting stuck and running a handful of extra doses and writing lots of duplicate log entries and not dosing any of the dosing schedules after that. When I tried to look at the log in the app is locked up. When I looked at the log in the database I found pages and pages and pages of duplicate log entries. It seems to run good for 1 day and then goofs on the second...
I've no idea how to fix this..? Please help : )
Thanks

Screenshot 2024-07-04 070239.png
Screenshot 2024-07-04 081651.png
Screenshot 2024-07-04 081205.png
 
I just tested bookworm and the sysfs system has changed or no longer there. They made a lot of changes in bookworm, sadly this is one of them. If you run the following command it'll print out what OS you are using.

grep VERSION_CODENAME /etc/os-release | cut -d '=' -f 2

It will likely say bookworm, if so you need to reinstall everything but this time in the Pi Imager program you need to select the version I circled below.

View attachment 741
Rob, this was the issue with the power bar.
 
Hey Rob,
I'm having trouble with the dosing getting stuck and running a handful of extra doses and writing lots of duplicate log entries and not dosing any of the dosing schedules after that. When I tried to look at the log in the app is locked up. When I looked at the log in the database I found pages and pages and pages of duplicate log entries. It seems to run good for 1 day and then goofs on the second...
I've no idea how to fix this..? Please help : )
Thanks

View attachment 744
View attachment 746
View attachment 747
Hi @GarrickG very sorry this happened, last night I found a slight issue with the scheduling but I think this is different. Thanks for the pictures, I will load up your schedules and dig through it tonight and find what's going on.

Make sure you set the "Minimum Delay Between Doses" setting for each dosing pump, this will make sure it can't dose too often in case something like this happens.

As for the log locking up that's due to so many records it's trying to fetch. I've since added filters and pages for the logs so that won't happen anymore. Once I get this scheduling fixed I'll release another update.
 
Yea I had some original install issues and I believe you assisted.

I actually have a second complete build as a backup and to test the 0-10 light control. (If you remember I had flicking issues and you suggested to load reef PI and I opted to buy a complete second controller for future projects.)

I can try a fresh install on the blank pi. Is there a way to pull the database from my 6.6 robo tank and restore it to the new build?
1720130222745.pngI just installed Reef PI and it is still flickering. Using a Ni Crew controller is it not having the issue.
 
View attachment 748I just installed Reef PI and it is still flickering. Using a Ni Crew controller is it not having the issue.
In reef-pi can you go to the "configuration" tab, then the "drivers" subtab. You'll see the pca9685 driver possibly called "extra ports" if you used my install script. Edit the driver and change the frequency to 1500 and save. Then go to the "admin" subtab and press the "reload" button. Now test the light again and the flicker should be gone.
 
Hey Rob,
I'm having trouble with the dosing getting stuck and running a handful of extra doses and writing lots of duplicate log entries and not dosing any of the dosing schedules after that. When I tried to look at the log in the app is locked up. When I looked at the log in the database I found pages and pages and pages of duplicate log entries. It seems to run good for 1 day and then goofs on the second...
I've no idea how to fix this..? Please help : )
Thanks

View attachment 744
View attachment 746
View attachment 747
Here's some files to download for the dosing schedules, I made some changes. I'm running your schedules and will be monitoring. Everyone running v6.7 should do this.

One useful thing, a bit of a pain for now, but if you view the schedules and expand it, the "Next Schedule" shown is calculated exactly as the backend so maybe after it doses just check that the next schedule is what's expected. I am going to have a page showing next coming schedule for each piece of equipment. To make life easier maybe empty your log, in the database if you view the logs table, then click the "operations" tab and that top and scroll to the bottom and click "Empty the Table" link.

Here's the commands to update. This won't affect anything when I release v6.8.

sudo wget -O /var/www/html/cpp/schedulesDosing.h https://www.robo-tank.ca/app/updates/v6_7_doseScheduleFix/schedulesDosing.h

sudo wget -O /var/www/html/js/settingsDosing.js https://www.robo-tank.ca/app/updates/v6_7_doseScheduleFix/settingsDosing.js

sudo systemctl stop robotank.service

g++ -o /var/www/html/cpp/robotank /var/www/html/cpp/robotank.cpp `mysql_config --cflags --libs` -lwiringPi -lpthread -lquickmail -lcurl

sudo systemctl start robotank.service


After you do this press CTRL + SHIFT + R a few times in the browser to reload the cache.
 
I'm trying to build in fail safes for things, for the dosing pumps I added the "minimum delay between doses" setting for that reason however I can see that's not going to work for everyone. @GarrickG is a good example. Now I'm thinking of running an independent query on the schedules that will add up the total amount being dosed that day. Then as it doses throughout the day it will track how much has been dosed, once it hits the total amount expected to be dosed that day via schedules it will no longer dose and will send out an alert saying the system is trying to dose more than expected.

I also have a "minimum delay between doses" setting with the custom rules however this only overrides the main delay counter if it's set. I'm going to change the custom rule dose delay to only monitor custom rules. I think that would be more useful as you could say I don't want any more doses from a custom rule for X minutes after a custom rule ran. I'll still leave the main dose delay as that will be useful for some.

Let me know your thoughts, if you think this is good or anything else I could do? Obviously it's very important overdosing can't happen.
 
It appears mine has crashed. After the dosing went crazy this morning and dumped 1500 of calcium into my tank. It now says its DEC 31 1969 and all of my sensors stopped responding. "UPDATE" After I ran the commands apply the update that Rob Posted yesterday everything came back online.

sudo wget -O /var/www/html/cpp/schedulesDosing.h https://www.robo-tank.ca/app/updates/v6_7_doseScheduleFix/schedulesDosing.h

sudo wget -O /var/www/html/js/settingsDosing.js https://www.robo-tank.ca/app/updates/v6_7_doseScheduleFix/settingsDosing.js

sudo systemctl stop robotank.service

g++ -o /var/www/html/cpp/robotank /var/www/html/cpp/robotank.cpp `mysql_config --cflags --libs` -lwiringPi -lpthread -lquickmail -lcurl

sudo systemctl start robotank.service






1720213666014.png
 
Last edited:
It appears mine has crashed. After the dosing went crazy this morning and dumped 1500 of calcium into my tank. It now says its DEC 31 1969 and all of my sensors stopped responding. "UPDATE" After I ran the commands apply the update that Rob Posted yesterday everything came back online.

sudo wget -O /var/www/html/cpp/schedulesDosing.h https://www.robo-tank.ca/app/updates/v6_7_doseScheduleFix/schedulesDosing.h

sudo wget -O /var/www/html/js/settingsDosing.js https://www.robo-tank.ca/app/updates/v6_7_doseScheduleFix/settingsDosing.js

sudo systemctl stop robotank.service

g++ -o /var/www/html/cpp/robotank /var/www/html/cpp/robotank.cpp `mysql_config --cflags --libs` -lwiringPi -lpthread -lquickmail -lcurl

sudo systemctl start robotank.service






View attachment 752
Oh no, I'm very sorry to hear that. :( What does the log show?
 
Back
Top