The disappointing child with Rev 1856 / Rev 1860 brings new issues

Started by bmrc, April 30, 2018, 10:40:56 PM

Previous topic - Next topic

ClassicCrazy

We - some of the customers have volunteered to be Beta testers. And we the customers often ask for new updates and Midnite does them. We often beta test those updates too before the general release. Sometimes some stuff breaks in the process that neither Midnite or the volunteer beta testers uncover.  Like I mentioned before  , you could buy the Chinese controller and see if they have an open forum that you can communicate with other users and the engineers. And see if they will respond to requests for new features and back up their products. 

There are so many different installations , wiring techniques, add on equipments, how could every single scenario and installation get beta tested ? 

Larry
system 1
Classic 150 , 5s3p  Kyocera 135watt , 12s Soneil 2v 540amp lead crystal for 24v pack , Outback 3524 inverter
system 2
 5s 135w Kyocero , 3s3p 270w Kyocera  to Classic 150 ,   8s Kyocera 225w to Hawkes Bay Jakiper 48v 15kwh LiFePO4 , Outback VFX 3648 inverter
system 3
KID / Brat portable

bmrc

I certainly agree the Chinese product would lack any support. It would also be about 1/3 the price.

Quote from: ClassicCrazy on May 31, 2018, 04:19:24 PM
There are so many different installations , wiring techniques, add on equipments, how could every single scenario and installation get beta tested ? 

They couldn't. However, this doesn't mean Midnite shouldn't or couldn't perform basic testing for every new firmware release. For example - while testing COMM, the test script should include a sanity check for serial data in each mode; with a WBJR, and BTS. Baseline testing like this is pretty standard in the software development world. Some companies even start coding from testing - https://en.wikipedia.org/wiki/Test-driven_development

Highflyer

BMRC,

I have been using the Kid charge controller from the beginning.  I love it for what it does.  For you to say that Midnite does not test their software is laughable.  I get you are mad because of a problem you are having and you want to find someone to blame, noted.

Please spend your time doing something productive.  I have asked for and had Midnite add several features to the Kid.  I have found several bugs as well.  Just the fact that I can talk to the engineers about their code and product is several steps above most other companies. 

Yes, I have had to reset one of my Kids once or twice over the last four years I have had them.  Show me a windows computer that does not need to be rebooted from time to time.  I have had more things than I can count blue screen.  But for me, Midnite has listened and corrected every issue I have seen to date.   

In short, if they build it, I'll buy it.

If you want to spend your money on other equipment, best of luck.  That's more Midnite equipment for the rest of us. 
Brian

The one thing is the one thing

smanners

Nothing like a robust discussion lol...

I suppose the primary objective of a charge controller is correctly charge and manage batteries. That is it’s primary objective. Everything that it does on top of that is just an additional feature. If those additional features cause it to fail at its primary job then the product has real problems. Having a charge controller lockup so that it no longer charges your batteries is a significant failure. Batteries are expensive and if they get damaged because of this then you lose faith in the product and feel “ripped off”.

Implementing new features to make the product more marketable is great and I love my classics and what they can do but they have never let me down. My view on all of this is build in redundancy. For remote systems where you cannot easily monitor or physically access split your charging process into two. Two smaller solar arrays rather than one big one. Two smaller charger controllers rather than one. That way if the worst happens and you loose a controller for whatever reason you still have one system hopefully feeding your batteries. Add in low voltage cutoff or some form of remote alarms and you increase your chances of surviving a failure.

I worked for many years in Papua New Guinea for Oil Search and this is how we approached extreme remote power systems for monitoring and telemetry of mission critical systems.

Great fun.

KyleM

Quote from: smanners on June 04, 2018, 07:09:32 PM
Nothing like a robust discussion lol...

I suppose the primary objective of a charge controller is correctly charge and manage batteries. That is it’s primary objective. Everything that it does on top of that is just an additional feature. If those additional features cause it to fail at its primary job then the product has real problems. Having a charge controller lockup so that it no longer charges your batteries is a significant failure. Batteries are expensive and if they get damaged because of this then you lose faith in the product and feel “ripped off”.

Implementing new features to make the product more marketable is great and I love my classics and what they can do but they have never let me down. My view on all of this is build in redundancy. For remote systems where you cannot easily monitor or physically access split your charging process into two. Two smaller solar arrays rather than one big one. Two smaller charger controllers rather than one. That way if the worst happens and you loose a controller for whatever reason you still have one system hopefully feeding your batteries. Add in low voltage cutoff or some form of remote alarms and you increase your chances of surviving a failure.

I worked for many years in Papua New Guinea for Oil Search and this is how we approached extreme remote power systems for monitoring and telemetry of mission critical systems.

Great fun.


We are in an age where superheroes exist and your super power is Common Sense never let anyone tell you it is not!

Redundancy in any system should be question number 3 or 4 that anyone asks  when designing because nothing is perfect, everything breaks at some point.

Thanks for reminding us to use that lump on our shoulders for something besides a hat rack.

Kyle

smanners

Lol thanks...

I like the midnite products it’s a real engineers bit of kit made with a genuine love of the science. You can see it from the outside to the in. I love this type of gear. However I’m not a midnite die hard. It has to be fit for purpose. I use the classics on my primary house solar charging but I use Enerdrive for my smaller applications (its Australian made sorry). Victron also make good gear if you like top notch plug and play. But I find Victron a bit boring for some reason. Midnite gear is interesting. Just please make sure that with all these cool features you try to isolate those processes from interfering with its primary job. I don’t mind that the network stack on my classics lock up because I just set a system restart each night. It’s not mission critical that monitoring works 100%, the charging side always works and that’s all I really care about. I have had issues with SOC occasionally resetting itself to
100%. No big deal, I just add in a safety net of an intervolt voltage sensing relay to ensure batteries cannot be discharged past a certain voltage when the signal from the aux interface does not work.

Most importantly remember your spending a lot of money so try to have fun with it.

ricardo

i think the attachment that Mario had made for 1863 has been edited out..  can we get a heads up on ETA on a release of the fix for the debug flag on the PC mode? ..

Or can I volunteer to beta test the fix .. would love to get data sooner than later.

ClassicCrazy

system 1
Classic 150 , 5s3p  Kyocera 135watt , 12s Soneil 2v 540amp lead crystal for 24v pack , Outback 3524 inverter
system 2
 5s 135w Kyocero , 3s3p 270w Kyocera  to Classic 150 ,   8s Kyocera 225w to Hawkes Bay Jakiper 48v 15kwh LiFePO4 , Outback VFX 3648 inverter
system 3
KID / Brat portable

ricardo

Quote from: ClassicCrazy on June 11, 2018, 01:30:14 AM
There is a 1860 1.1 on the Midnite page - maybe that is the corrected version
http://www.midnitesolar.com/firmware.php?firmwareProduct_ID=4

Someone mentioned a few pages back that 1860 1.1 is 1860 with the windows updater.  (plus the date of that release is april which is before this bug was known) -- I think the FW with the fix is called 1863 (unless they make another cut)

ClassicCrazy

oh - okay - wonder why they took 1863 down ?  I usually download all the new copies of firmware but for some reason I didn't get that one - I just looked in my download folder and don't have it.

Larry
system 1
Classic 150 , 5s3p  Kyocera 135watt , 12s Soneil 2v 540amp lead crystal for 24v pack , Outback 3524 inverter
system 2
 5s 135w Kyocero , 3s3p 270w Kyocera  to Classic 150 ,   8s Kyocera 225w to Hawkes Bay Jakiper 48v 15kwh LiFePO4 , Outback VFX 3648 inverter
system 3
KID / Brat portable

ricardo

given that midnite doesn't seem to keep a historic download page for older FW .. keeping all the releases yourself is a good practice.

FNG

Mario found a bug in the load output that effected PWM divert. He pulled it until he gets it fixed. And yes I agree on a page for "Archived" firmware that sounds like a great idea. I will ping the folks over at Midnite and see if they may consider that.

ricardo

for those of you who are beta testers for new firmware, is there a formal process for requesting to be in a beta test program .. or how do you get on that list?

I would like to get the next FW sooner rather than later so i can finish up my monitoring dashboard.

ClassicCrazy

Quote from: ricardo on June 12, 2018, 11:21:42 AM
for those of you who are beta testers for new firmware, is there a formal process for requesting to be in a beta test program .. or how do you get on that list?

I would like to get the next FW sooner rather than later so i can finish up my monitoring dashboard.

Maybe give Midnite a call or fill out a ticket offering to test out the new firmware so they know how to get in touch with you. Often  they just post on the forums with new firmwares for anyone to try out in beta before they post on their main download page.

Larry
system 1
Classic 150 , 5s3p  Kyocera 135watt , 12s Soneil 2v 540amp lead crystal for 24v pack , Outback 3524 inverter
system 2
 5s 135w Kyocero , 3s3p 270w Kyocera  to Classic 150 ,   8s Kyocera 225w to Hawkes Bay Jakiper 48v 15kwh LiFePO4 , Outback VFX 3648 inverter
system 3
KID / Brat portable

bmrc

It looks like the charger (with 1860) has frozen again. I won't know for sure until I visit my property again, but the charger isn't spitting out any serial data anymore.