Finding and resolving the cause of inaccuracy in the charge controller's set point voltage.
By Julian
Youtuber, shed dweller, solar charge controller aficionado
Good morning all…
Youtuber, shed dweller, solar charge controller aficionado
This site uses Akismet to reduce spam. Learn how your comment data is processed.
Suggestion: if you have a known voltage, which is accurately set to the target voltage, on the battery input during your testing, you could use the reset to read the ADC input and then set that value as your target in the EEPROM. That would eliminate the need to step through values. You could also use that to set a scaling factor to improve the accuracy of the voltage meter function.
Why not just put in a variable resistor to adjust it?
You only need to store the 13.50 Voltage level, no need for more voltages.
The other values could be a bit off, or you could scale them from that singel point and zero.
Hi Julian , your level of inaccuracy may have been as a result of ambient temperature change , you may require that it continually recalibrates itself.
thanks so much for this, alot of good product design advice that im absorbing.. currently an EE student working on designing my own MPPT charge controller.. using alot of your expertise. thank you
i have an idea put the controllers in a project box with the leds on the lid so you can see what status they are at without opening the box
like to see this series on the PWM5 continue
You're a clever guy. Genuinely. That idea about resetting the unit for the calibration is the sort of idea that very few people have.
Shame you didn't make more money from this.
Btw I know I'm slow to comment on this vid. I'm slowly going through your back catalogue of videos.
I'm hooked 🙂
Great Video, hope to see more on these. Video should also be added to your playlist.
Thanks for sharing your experience. I faced such problems myself when I tried using my product for commercial use. Some of the points mentioned in the video are really useful and worth learning from. Although I learned much from my own mistakes, I wish Julian would have uploaded these videos 4yrs ago!!
I don't get it. You are faffing with +/-0.1V but you have no temperature compensation and are using 13.5V as your float voltage anyway. Temp comp is comparably huge!! (-4mV/'C/cell e.g. the difference between charging at 0'C and 25'C is 0.6V). Along with this, you should be insisting to the customer that the controller shares the same temperature as the battery i.e. so the temperature monitored is that of the chemistry of the battery. Or have I missed something?
this video isnt in the pwm5 playlist
fyi not added to your playlist mate.
From what I've seen the Arduino voltage reference is inherently inaccurate, +/- 5% worst case.
As Kedar Nimbalkar mentioned the TL431(it's cheap) & arduino AREF pin should improve a lot of the inaccuracy ( but increase part count). Using known accurate TL431+ 1% resistor should make a difference.
You should consider some pic with Internal voltage reference (like the PIC12F615)…it has channels to 0.6V or 1.2V reference which can be used to indirectly measure the actual Vcc…this is something very accurate indeed. I have test it on a 16F690 and is surprisingly accurate. Basically you take an adc sample of a known voltage and you can calculate in software the resolution.