This site is now in read-only archive mode. Please move all discussion, and create a new account at the new Victron Community site.
I just noticed the voltage display via the app in my blue solar is quite off.
It tends to show battery voltage about .2V higher than what I measure right at the terminals of the charger. That is comparable to my worst Chinese charger...
Someone else got similar findings?
It might be related to the recent update?!
You say the battery voltage at battery is .2v higher then what mppt says. But was it the other way, battery was .2v lower?
What wire gauge and one way wire distance do you have? what is your battery voltage?
If you have nominal 12v and .2v off, thats almost 2%. If this is due to wire issue, you might have to 1/2 the AWG to get it within .1%, no?
Imho in 12v system going any less then 2/0 is asking for trouble.
no, this was another user's complaint.
The inaccuracies in my unit were probably due to 5V power supplied.
How are you guys with voltage deviations connecting to the charger? I suspect now it is my non-Victron Prolific TTL device connected to the MPPT's VE.direct port which is supplying power on the 5V side to the blue solar that might be causing the deviation.
O wow, did you test with and without the connection?
so maybe the external voltage influences the internal reference voltage?
Hi Boekel and all,
we have not spend any hardware costs or design time on making that port robust / other-use proof. Its very likely that feeding the port externally, as well as drawing too much power from it, influences the reference voltage.
The VE.Direct port on our devices is designed to be a super low cost port; so we can have it; even on devices where 999 of a 1000 times nobody is going to use the port.
As also written elsewhere (its about using a non-Victron RS485 or VE.Direct USB cable):
*Victron - Using Unsupported Accessories*
Hey all, we can’t stop you from using self-made or self-sourced cables and adapters; and making money on cables and such is not important for us; and I understand the appeal of saving costs; however I’m not happy with it; and recommend heavily against it. Here is why:
(See next comment)
(Continued from previous post)
we cannot make any guarantees about functioning; and in case of issues; don’t ask Victron for help. Not about issues with the cable or meter, but also don’t ask us for help if the system works awkward in whatever other way; for example the complete ESS being unstable. Thats of course not likely to happen; but also not impossible;
For example, I’ve recently spent quite some time on debugging a customer system with VE.Direct USB cables that didn’t work, to in the end found out that they were self made, and and communication didn’t work bi-directional. Only RX worked, TX didn’t and that lead to very weird error messages that required multiple top R&D people, and myself to spend time on it before discovering the root cause. A major waste of time; those people can far better spend there time improving our products or making new ones.
Very frustrating.
(continued on comment 3)
Another example: we’ve tested multiple RS485 cables ourselves; and there are many that don’t work properly; they loose packets now and then. So from the outside it seems to work; to then in reality not work really well.
And now this example ;-); I’ve been worried about voltage accuracy; to in the end find out thats its likely caused by using a non-victron thing on a Victron-port.
So on topic:
if you go connecting non victron things to Victron ports; disconnect it before coming to conclusions or asking for help.
So on conclusion; thanks Joe for bringing up this use of the port; me, nor the software and hw engineering designer had thought of that.
Thanks!
- mvader
I wouldn't even consider making my own ve.direct cables, not worth my time.
But I'm glad @JoeOffgrid found the cause and also posted about it so we can all learn from it.
One of the reasons to always ask for photo's of the installation when people ask for help.
Please remember there were too issues.
First I wondered why time for absorption is so short and started measuring voltages indepentently... This happens to be the case whether or not the TTL device is connected.
Btw. My Chinese charger came with a very handy USB socket or was it even two...?
I used it to recharge my phone often. My suggestion would be to have an USB built into the charger. HW costs less than 20ct I guess!
Still need to verify when I have a chance.
It was a just a question. Wondering what the other people with issues are using to connect...
Hi all. in this (lengthy) thread, two units were reported to be inaccurate in their battery terminal voltage measurement. Both have been investigated; and there are no issues with voltage accuracy.
Joe received a new unit from us; his unit was sent to us; tested found to be Ok.
Dean also reported an issue; this was looked into; and also there is no problem with the unit. It measures just fine; unit is Ok.
For all the details; happy reading!
I’m considering this case closed now. Thanks for your help.
Have a good day, Matthijs
Does the delta fluctuate according to the current going through to the battery?
Hard so say. Now I only have a small flat ponsitioned panel and could not generate more than 2A at ideal conditions yesterday.
Now it is too clowdy.
Both voltage readings fluctuate, but I cannot say for the delta.
No difference that I could measure between Victron terminal and battery! The connection is good.
Yes, I think there is a relation. The bigger the current be bigger the delta.
Due to the bad weather I have very little solar power to play with...
@JoeOffgrid
Victron units are well known at this point for having poor accuracy across the MPPT range.
At this stage, the response from the company seems to be "this is normal, live with it", and the response from much of the forum community is generally to attack people who bring it up too much (just so you're aware).
A friend of mine (@Solvan) recently returned his unit because of poor current accuracy (and other issues), and when testing with his vendor, found their entire stock was subject to varying degrees of error. His posts were subsequently deleted and I believe he has now been banned from posting.
In short - nothing you can do. This is the Victron standard of quality. I live with it and rely on an accurate battery monitor to compensate (at least for that problem).
See also:
https://community.victronenergy.com/questions/7861/innaccurate-and-poor-absorption-on-smartsolar-an-u.html
https://community.victronenergy.com/questions/322/inaccurate-dc-input-current-readings-from-multiplu.html
Hey; that we say that inaccurate voltage measurements are normal is false. And that this is our standard of quality is false as well.
Above two linked posts are about a current measurement. For the mppt current measurement being wrong I’ve asked for units to be returned to us for analysis: we measured lots of units; and have not been able to reproduce it.
But, no units were returned / no-one has let me know the rma-number.
I’ll put my real answer in an answer rather than this comment.
If this is the standard, I'd rather use the the dirt cheap Chinese unit.
It would not charge my lead acit properly like that!
Actually, the charger is going from bulk to float very quickly. I set max absorption to 5h and absorption time is like 5 minutes.
I really hope it is a bug in firmware 5.2. At least I did not notice before I updated.
Regarding short absorption times - that would be the OTHER big MPPT problem. See
this and this and this and this .... nutshell version is that in trying to be 'smart' about how they calculate abs time, they're actually very dumb...and yet lack the ability to be told "Just hold absorb for x minutes or until y amps" which, as you point out, even the cheapest ebay jobs do now.
I've been waiting through several firmware updates with no change. Victron just don't seem to care, they're more focused on making fancy looking web apps and slick LCD displays. I've run my unit at float=absorb, and manually set the voltage down if needed.
Not true either; I’ve replied quite seriously a few weeks ago; and we’re working hard on this internally.
You’re own report has been very valuable in that.
I am about to share the document (and soon also test firmware) on how to improve this with the few interested people here on the forum.
I dont know what happened to make you so negative, but I’m here to change that the other way around. Just give me some time.
Hi @JoeOffgrid, 0.2V deviation is indeed too much. I’ll contact you from the helpdesk email to resolve it further / ask for serial number details etcetera.
I hope the following might help some users. Firstly thanks to the Victron team for their assistance here its been very helpful. Secondly I'd like to point out that the conclusion I've reached regarding voltage accuracy has all been tested on cloudy days and I want to confirm my findings on a sunny day.
My belief (like others) that the Victron MPPT 100/50 controller over stated the battery voltage was due to measuring the voltage at the battery and finding it lower than what the controller was reporting.
If anything I was expecting to see a higher voltage at the battery due to any voltage drop, but after testing the voltage at the controller terminal I found this was in fact higher at the controller than the battery.
There are three main aspect of the system worth considering here.
Because the controller is generating a charge the voltage drop as a result of resistance in the BMS drops from the controller to the battery. Hence by testing the voltage at the controller (as the controller software does) you get a higher voltage reading.
Perhaps a more accurate label here would be MPPT Controller output voltage rather than battery voltage.
One way to address this problem would be for the controller to test the voltage directly on the battery (ie bypassing the BMS) and this is exactly what the additional Smart Battery Sense Victron module does if its installed.
Otherwise the work around is to increase the voltage settings in the controller to compensate for the loss. This will still report the wrong "Battery Voltage" but at least the batteries will get fully charge. Obviously this has its dangers especially as the value can change with current (ie V=IR), but measure the actual install resistance (or voltage drop) or with some small incremental tweaking the results are pretty good.
