So, I installed the MNWBJR today. Can't say anything yet on long term SOC% correctness. But I do have some feedback:
It seems that not connecting it directly on the shunt but on some twisted pair shielded wire is not without risk of problems. It gave me a persistent difference of -14 Amps with reality. So if my Mastervolt CSCP (http://www.mastervolt.com/automotive/products/system-panels/panel-cscp-combi-system-control-panel-12-24v-dc-digital/) was saying 0 Amp, WB claimed -14. If my Mastervolt claimed -20 Amp, WB claimed -34. Actually, it wildly fluctuated from -9 to -19, but the fluctuation centered around -14 Amp
The shielded twisted pair was not a single run, there were two RJ45's in there. I suspect some electrolytic offset caused by the contacts on those RJ45 are to blame. So I went with direct mounting on the shunt, despite the tight and impossible spot where the shunt was mounted. After I mounted direct, the WB and the CSCP were very much in agreement on current out and in. (I do not have actual SOC% comparison yet)
So feedback 1: take care with wired remote mounting.
----
Item two: the idea to directly mount the WB and its form factor are very creative and as proved above also very good for stability and precision. But really, the device is horribly flimsy, showing a bare back, and those separate spacer rings are horribly difficult to mount when you do not have good access to the shunt.
I feel it should really have been completely enclosed (not bare back) and in proper strike-proof PVC or something, not in a flimsy (polycarbonate? Acrylic?) half-cover with bare back and clips that break of even easier. And those spacer rings should be held in place by said encasing, even if only by temporary friction, so you don not have to hold 4 rings and two screws and some washers at the same time because it is impossible to do one and then place the second screw's spacers later.
So feedback 2: for future revisions, a better encasing please.
----
Do not misunderstand me: I suspect I will be very happy with the WB and SOC% based Aux1. It's what I missed on the Classic even before I bought it (yes, I read the entire manual before I bought the Classic). And I did manage to mount it without damaging anything, not even a single clip. But really, compare the WB to the case of the Classic itself. No contest.
Thanks
Thank you Cybermaus...
No, you cannot get good accuracy with the shunt placed very far off of the shunt because there is
some current going through the wires that power the WB Jr.... That is probably why you have the
difference. It might work better though if you were to use larger diameter wire going to the WB Jr.
to the shunt.
Can't say that we will be changing the packaging but feedback is very appreciated.
We did think of making the cover a screw capture arrangement but it doesn't work very
well especially if you need to add another pair of wires for another shunt monitor in
parallel with the WB Jr.... You cannot reliably add another pair on the shunt side of
the WB Jr. Only the outside of the WB. Jr.
boB
To put your wiring into perspective remember that their is only 10ths of a millivolt across that shunt.
I cant see it being much fun to try to attach the WBjr while the deltec is mounted. You really have to remove the shunt, and mount the WB on the bench.
Kinda agree that it seems vulnerable. I wrapped a bunch of tape around mine before mounting it. But once in place and with good cable managment its not going anywhere.
Quote from: zoneblue on February 14, 2014, 06:56:29 PM
I cant see it being much fun to try to attach the WBjr while the deltec is mounted. You really have to remove the shunt, and mount the WB on the bench.
Nah! that didn't work for me 'cos the WBjr when mounted on the Deltec means you cannot get at one of the mounting bolts for the shunt into place. The WBjr case covers it over >:(
As my shunt was also nearer the bottom of my MNDC box I could not easily get at the small brass bolts that hold the WBjr to the shunt. After lots or swearing and balancing just the bolt on finger tip and gently rotating to try and just grip the thread I eventually had to resort to not using washers and a right angle bent small flathead screwdriver.
I fiddled and faffed with this for over an hour, and when completed the double single malt was well deserved.
Of course SOD's law (aka Murphy's law) applied and I had the main connections to the shunt the wrong way around and the app reported -ve current when it was really positive. Alt least that was easy to correct even though I had to cut a cable and recrimp a 70mm (wire size mm2) lug. Getting the large crimp tool up to the MNDC was a real contortion act and needed more double single malts to aid recovery.
dgd
Ha! For a wee thing a fair bit of assembly :)
Yeah i must have put the deltec mounting screw in first, then fitted WB, then just tightened it from the rear, and called it good. The back plate comes out of that cabinet at least.
Gee. Sorry it's such a hassle to install that thing !
We went through this here quite a bit. Just saw (at work) the screw underneath the
PCB and for some reason, we didn't think it was going to be a problem in most cases.
Can't remember why now.
boB
@boB : current through those wires? Not just high impedance measurement? Oh wait, you feed only through one wire (with some one wire protocol to boot) so it needs to sinc through the shunt. Tricky.
@dgd : That is your shunt? I don't want to sound belligerent, but man, that is *oceans* of space compared to what I was doing yesterday, even if you do have to use angled screwdrivers. I'll see if I can take/post a picture later today.
Anyway : at least though my shunt had metric screws in them, the screw position and distance was identical to the US one the manual referred to. Imagine also having to deal with that.
Quote from: boB on February 15, 2014, 02:27:23 AM
Gee. Sorry it's such a hassle to install that thing !
We went through this here quite a bit. Just saw (at work) the screw underneath the
PCB and for some reason, we didn't think it was going to be a problem in most cases.
Can't remember why now.
boB
And here I thought it was just me and my install. :o
Glad to see it was not just me. Not really but I just don't have the fine motor skills I once had so it was a fumble fingered ballet getting mine in and now I don't feel so klutzy knowing it was a bit of a chore for others. My Epanel is pretty busy inside cable wise.
Well worth it in the end, however.
Tom
I really like my wbjr!
I thought I would install it without taking too much stuff apart... An hour later after the spacers had made it to the floor many times and wound up behind wires and wedged in nasty spots, I admitted defeat.
I had every screwdriver I own under 12" involved in this project.
At that point I took stuff apart, and was done in 5 minutes, which is where I should have started.
Wade
Quote from: toothy on February 15, 2014, 03:31:34 AM
An hour later after the spacers had made it to the floor many times and wound up behind wires and wedged in nasty spots, I admitted defeat.
I had every screwdriver I own under 12" involved in this project.
At that point I took stuff apart, and was done in 5 minutes, which is where I should have started.
Wade
Sound advice there. One of those "short cut" gone long scenarios.
I definitely think it was worth the effort but if I had to do it again I would take it apart further to get it done.
I really need to get to bed.
Tom
Many/most of the early WBjr applications will be retrofits into existing, and sometimes installations working while the WB is being installed -- the case here. Retrofits will always be more difficult.
But, stepping back a bit, and trying to think of all that the mechanical designer is faced with ... there are SO MANY different configurations in service, and some differing ones still to come. There are not too many ways to make the WB petite, enclosed, and still attach to a Shunt in a workable way.
Believe that the WBjr is a perfectly adequate solution for many, many of us, it is very cost-effective, and did not take five or more years of design/redesign for MN to bring this great addition to many of our systems (am guessing on the design time).
I was quite happy that the WBjr fit a 10 year-old Xantrex DC Conduit Box with about 3/16" from the side wall of that box to spare. Will admit that I was concerned that some of the spacers might just sling themselves into an abyss and never again be found, but, so far, all has gone well with installations.
Had been thinking about a small plastic sheet thingie that might help retain the lower spacers/standoffs, and still be removable after the screws have begun to mate with the Shunt's threads. Bit there are still so many variables in different installations that even that is not a simple 'solution".
THANKS MidNite for another useful, great product, that just gets better and better with FW additions to the original functionality.
Opinions, Vic
Quote from: cybermaus on February 15, 2014, 02:39:42 AM
@dgd : That is your shunt? I don't want to sound belligerent, but man, that is *oceans* of space compared to what I was doing yesterday, even if you do have to use angled screwdrivers. I'll see if I can take/post a picture later today.
ok, so one person's puddle is another person's ocean...
With the limited real estate in the MNDC box, I had no choice but to mount the shunt/WHjr combination so that the WBjr was underneath the shunt. It was the need to hold up the WBjr and still manipulate the brass bolts in the 'puddular' space below the shunt, with back bent, neck contorted, spit trying to steady the bolt balancing on pinkie finger and eyes losing focus in the confined space that made the task so exciting...
Anyway it was worth the effort finally seeing some SOC% stuff the next day. If this business was easy then everyone would be doing it ::)
dgd
PS
I had already decided the MNDC was too cramped and bought an Epanel, aluminum, the exact same size as the MNDC by just taller. It has the shunt already mounted and some wiring done and it looks as if the WBjr will just fit on top of the shunt. Should be easier to do.
Pity is that to get this larger box I had to pay for all the unwanted AC switches and wiring for genny inputs, etc. I wish it were possible to get epanals without all this AC paraphernalia but I suppose thats a whole nuther thread subject
Hi dgd,
Was not trying to argue or similar ... The first installation of the WBjr here was not a slam-dunk either, but some of the issue here was that I had placed one toroid choke kind of in the way, and there were numerous CC + and return cables run at the back edge of the Conduit box very near the Shunt.
Successive installations seemed to come easier, and easier.
Was just trying to make the somewhat obvious point that retrofits are often going to be considerably more protracted vs a new installation, from scratch. And the implied thing, that usually, once installed, the WB will just sit there and work quite well.
I am one who was BEGGING and begging MN to ship the WBjr sooner rather than later. It just keeps getting better, and all, for a very, very reasonable price. Given all of the variables in Shunt location, and cable dress of systems in the field, doubt that much more engineering time would have made the WBjr very much easier to install on existing systems. Again, just opinions ... Vic
hi Vic,
I agree that its the retrofitting of devices such as the WBjr, that makes life complicated. When I look at the MNDC layout I would have done it differently if the shunt and WBjr were there at the planning stage. 20/20 hindsight is just wonderful but all experience is good.
As you say it gets easier as you do more of these installations.
I also realise if I hadn't been such a cheapskate in thinking the MNDC was sufficient instead of doing the job properly in the first place with an Epanel, then I would not have had these problems.
As the pilot in Mad Max said 'No matter where you go, there you are' - so I mostly brought these issues on myself >:(
dgd
Aw dgd, where would you be without a little excitement once in a while :)
Quote from: zoneblue on February 16, 2014, 04:02:33 PM
Aw dgd, where would you be without a little excitement once in a while :)
on my sun lounger in a quiescent state of inebriation satisfied in the knowledge there is order in the universe, God is in heaven and my MN stuff makes abundant electric power and hot water. 8)
What and never get the hydraulic crimpers out ever again? Youd miss the sweet jingle of lugs, you know you would.
So after a few days running with cloudy days battery/SOC% slowly dropped.
2 evenings ago, it finally reached 50% and switched to shore power. Overnight it dropped another 2%, but that is possible as I do still have about 1 amp usage even when the inverter is not active. Also, my Mastervolt manager was a little more pessimistic as it was at 43% in stead of 48%, but I may need to check all the Mastervolts' settings.
---
So far so good. But now the problem: Yesterday was a very sunny day. And yet, it barely charged. 10Amp peak, about 5% SOC back. A day like yesterday, even when winter with low sun, should have been able to put 20% (subjective estimate) back into the bank with 10 Amp average.(equally subjective estimate)
I noticed some people reporting some stuff in the beta thread (http://midniteforum.com/index.php?topic=1670.0), there is even already several newer builds. But I do not fully grasp all the items reported or the fixes applied.
So my question: is there already knowledge about something in the latest production release that could cause bad performance? Should I try some of the beta code?
Also I saw that even the 'production release' had a new build number, without there actually being any addition to the 'whats changed' description. Makes me a little extra unsure, updates without descriptions? Was the previous production release 1759 a dud?
PS; Today is another lousy winter-day, so today is not a good day to test or measure individual cables and PV's anyway.
Quote from: cybermaus on February 18, 2014, 07:47:27 AM
Also, my Mastervolt manager was a little more pessimistic as it was at 43% in stead of 48%, but I may need to check all the Mastervolts' settings.
Firstly SOC *is* rocket science. Both Ah devices are dealing with a fickle chemistry, make a broad range of assumptions and use 'good enough' algorithims. Never rely completely on any SOC product. Always keep an eye on rest battery voltage, and get to know its nuances. That and charge state, absorb duration etc are the real acid test of whats happening.
QuoteYesterday was a very sunny day. And yet, it barely charged. 10Amp peak, about 5% SOC back. A day like yesterday, even when winter with low sun, should have been able to put 20% (subjective estimate) back into the bank with 10 Amp average.(equally subjective estimate)
As a rule of thumb, if you cant fully charge your bank on a single sunny day the battey to panel ratio is out of whack. Is that usual or you are saying your system doesnt seem to be charging right? Which controller are you using, and whats the rest of your gear? Sig needed.
Quote
I noticed some people reporting some stuff in the beta thread (http://midniteforum.com/index.php?topic=1670.0), there is even already several newer builds. But I do not fully grasp all the items reported or the fixes applied.
So my question: is there already knowledge about something in the latest production release that could cause bad performance? Should I try some of the beta code?
Dont worry about firmware at this point, nothing in recent times (or ever, in the case fo the classic) affects basic charging functionality. Time to do some testing and diagnostics.
Oh yeah, PV to battery ratio is probably not ideal when seen from Solar point of view.
Historic reasons. Its a 100 year old river barge converted into a houseboat, still sailing but lived on permanently (by my sister and her partner)
Battery bank is sized for living off at least 48 hours (with reasonable comfort) without going lower then 30%, and charged by either the main engine (a WWII slow running Lister) or shore power (when available). This was all setup some 25 years ago though the bank is renewed every 8 years or so..
Solar was added only last year, and the amount of panels is limited to what was possible to place on the decks without harming the classic look and feel of the barge. There are rules to what you can do and still be designated 'original' by the various relevant organisations and (even more important) peers. I was arguing to fit 8 panels, but that was rejected on aesthetic grounds. And well, its not my boat.
Not only that, but the panels are angled wrong (13 degrees, should be 30) and there are always sailing boots passing (shadows from masts, its amazing what even a narrow mast shadow can do to a panels efficacy). And when sailing, who knows where South is going to be, though during winter they are on a fixed place.
In specifics:
Six 250W panels making 1500W in total on 24V-480Ah. So no, one sunny day will not fill them up. It will take two days in summer.
And the sun is still low, it is still winter, but I thought the 5% SOC gain really too low. I had expected 20%
Anyway, if this is not likely caused due to the firmware change, I then I will not start doing unneeded updates. Maybe there was more shadow then I had thought. I best keep an eye on the matter though, and start checking the panels and wiring the next sunny day.
Thanks
PS:
Quote from: zoneblue on February 18, 2014, 02:13:10 PM... are the real acid test of whats happening.
:D Was that pun/reference (http://www.comparestoreprices.co.uk/images/sy/sykes-pickavant-battery-hydrometer-tester.jpg) accidental or intended.
Quote from: cybermaus on February 18, 2014, 03:28:23 PM
Oh yeah, PV to battery ratio is probably not ideal when seen from Solar point of view.
Historic reasons. Its a 100 year old river barge converted into a houseboat, still sailing but lived on permanently (by my sister and her partner)
Battery bank is sized for living off at least 48 hours (with reasonable comfort) without going lower then 30%, and charged by either the main engine (a WWII slow running Lister) or shore power (when available). This was all setup some 25 years ago though the bank is renewed every 8 years or so..
Solar was added only last year, and the amount of panels is limited to what was possible to place on the decks without harming the classic look and feel of the barge. There are rules to what you can do and still be designated 'original' by the various relevant organisations and (even more important) peers. I was arguing to fit 8 panels, but that was rejected on aesthetic grounds. And well, its not my boat.
Not only that, but the panels are angled wrong (13 degrees, should be 30) and there are always sailing boots passing (shadows from masts, its amazing what even a narrow mast shadow can do to a panels efficacy). And when sailing, who knows where South is going to be, though during winter they are on a fixed place.
In specifics:
Six 250W panels making 1500W in total on 24V-480Ah. So no, one sunny day will not fill them up. It will take two days in summer.
And the sun is still low, it is still winter, but I thought the 5% SOC gain really too low. I had expected 20%
Anyway, if this is not likely caused due to the firmware change, I then I will not start doing unneeded updates. Maybe there was more shadow then I had thought. I best keep an eye on the matter though, and start checking the panels and wiring the next sunny day.
Thanks
Not likely a firmware issue. My money is on shading. Try this. Go into the Mode and switch it to Legacy P&O it will deal with shading much better than solar mode.
Ryan
Did you update this Classic previously?
I would check the Input and output current limits in the Limit menu. In should be 99 and Out should be the Max.
Also whats the VOC of the array?
The array is in 6 strings of single panels. Though I can relatively easily switch them in 2 threads of 3 configuration, Wiring is in place for both.
I have specifically configured them in 6 parallel panels to reduce shadow impact. Each string is terminated with a diode. If one panel shades, it does not pull the entire string low (or at least, the entire string is just the one panel). There is a bit of extra diode and cable loss of course, but it seems to work well. Or at least, worked well last year.
I specifically got panels of VOC of 42 Volt and MPP of 36 Volt for this reason, so even a single panel is still some distance above the 28.8 volt battery bank.
I did try last year with both the 2x3 and 6x1 config, and the 6x1 seemed to work best.
---
Yes, I did update last week, for the SOC%/WBJr. I did do a factory reset, as advised. I can check those settings next week, I am out of the country at this moment.
I agree the 1 panel per series makes the most sense. Like I say try Legacy P&O mode instead of solar I bet you see a big improvement
Ryan
Well, as stated I am out of the country, so did not do anything yet. But I did read the relevant manual section. And if you have some time, I would not mind some further confirmations/explanations.
Also you may want to review the manual, there is some duplicate/difference between pages 34 and 62
Quote from: page 34Solar
This is the default mode for PV systems and has a very fast sweep (typically1/2 second or less) that will
re-sweep at user adjustable sweep intervals, unless the Classic finds that it needs to do a sweep on its own
because of changing conditions. The timed sweep interval is user adjustable and is in units of
minutes. SOLAR mode is typically best for PV systems, especially if there is partial shading at times
during the day. The Classic will show a message of "PV SHADE" if it thinks the PV array is partially
shaded (if this feature is enabled).
SOLAR mode is best suited for shaded or un-shaded PV arrays that are at least one nominal voltage above
the battery voltage. For severe partial shading or PV arrays with nominal voltage equal to battery voltage,
you may also want to try Legacy P&O (Perturb and Observe) MPPT mode.
Legacy P&O (Perturb and Observe) mode is a slow tracking mode similar to the Micro Hydro mode but
with the difference that it is slightly faster and will shut off if the power source goes off.
It has 2 settings that are user adjustable. Sweep Interval is the time between mini-sweeps, in minutes, and
sweeps around the present (i.e. the last found), MPP Voltage. The range of this sweep is determined by
the Sweep Depth user adjustment and is expressed as a percentage of Watts that the sweep started
from. For example, if in Legacy P&O mode, the Classic was outputting 500 Watts and the Sweep Depth
percentage was set for 10%, (50 Watts), the sweep will bring the input voltage DOWN until the output
power drops down to 450 Watts, then will sweep UP in voltage until the power drops again down to 450
Watts and then go back to the newly found MPP Voltage, waiting for the next sweep.
Quote from: page 62Solar
[word for word the same]
Legacy P&O
Legacy P&O (Perturb and Observe) mode is a slow tracking mode similar to the Micro Hydro mode but
with the difference that it is slightly faster and optimized for Solar. Legacy mode can be very usefull for
arrays that experience heavy shading issues as well as PV arrays that have a low voltage. If the array has a
VOC (Open Circuit Voltage) of less than 125% of the battery voltage Solar mode will not work as well so
Legacy will be a better choice.
I am interpreting "one nominal voltage above the battery voltage" as "just over twice the nominal battery voltage" ? So I guess between 200% and 125% of battery voltage is a gray area, but below 125% definitely go legacy. And with all the shading and just over 125% that I have, also go legacy. Still, solar mode seemed to work well last year.
The manuals descriptions seems to make sense. Actually, one of the reasons I was pushing for 8 panels is so I could organize them as 4x2 and have a 72Vmp, with 4 strings / 'shading blocks'. It's also why I selected the 200 model, so I could potentially also try 2x4 in that case. But as stated, we have 6 panels, and unless I have a really good reason to get 2 more, aesthetics means it will stay at 6 panels.
I do want to remark that though random shading is an issue, because of the diode wiring, the array does not actually reacts in a typically shaded manner. If one panel gets some shade, it drops voltage, but due to the diodes, the overall array voltage stays pretty high. The shaded panel of course also tries to match that, so it completely falls of at the steep end of the curve and drops current. It may be better if the Vmp would be slightly lowered, so all panels keep up current at a slightly lower Vmp
Is that what the P&O would achieve better? Text seems to indicate the Solar MPPT is more aggressive, maybe too aggressive, and the P&O less aggressive in how far it sweeps the Vmp. I will follow advise and try the P&O, I am just fishing for a better understanding of what it is I will be trying.
---
I found another document with "Watt-Voltage table" (http://www.midnitesolar.com/pdfs/classicWatt-Voltage_Table.pdf)
In it there is not even a discussion of any Vmp below 70V and even 70Vmp is
na for the 150 model. What is up with that, is the Classic really not ideal for single panels, and am I pushing the (low end of) the envelope?
---
PS : not critical, but I do have a self made shunt installed pre-diode with each of the 6 panels, so I can see per-panel potential and current.
Quote from: cybermaus on February 22, 2014, 02:53:23 AM
I am interpreting "one nominal voltage above the battery voltage" as "just over twice the nominal battery voltage" ?
I have also wondered about this.
Check out Midnite's string calculator. I set up a 1 panel array on a 12 volt system, and then reduced the Vmp until the calculator alerted me that the Vmp was too low. IIRC, I got down to 14 volts.
--vtMaps
Quote from: vtmaps on February 22, 2014, 05:03:45 AM
Quote from: cybermaus on February 22, 2014, 02:53:23 AM
I am interpreting "one nominal voltage above the battery voltage" as "just over twice the nominal battery voltage" ?
I have also wondered about this.
Check out Midnite's string calculator. I set up a 1 panel array on a 12 volt system, and then reduced the Vmp until the calculator alerted me that the Vmp was too low. IIRC, I got down to 14 volts.
--vtMaps
Yes, you basically have the right idea on 2wice the nominal voltage of the battery but PV modules used to be rated
with battery voltage numbers before all this grid tie terminology took over. There was a 24V PV module or a 48V or
12V module designation. That is kind of what I meant. So for a 48V battery system, you would choose a 60V PV array.
Now it's a bit different. Number of cells instead of battery voltage pretty much is what I have been seeing.
boB