nicknorman, on 24 Jan 2016 - 8:33 PM, said:
I'm doing a monitored comparison discharge between the Smartgauge and the Mastershunt tonight, and was going to post the results on my "smartgauge accuracy" thread but I though I would add a bit about the Mastershunt since it seems to be a relatively rare beast, and this thread about battery monitoring seemed the more relevant one, so I'm going to put it all here. We started today with around 85% Soc, cruised for a good few hours and tied up with both the Smartgauge (SG) and the Mastershunt (MS) showing 100%. The tail current for our 450AH bank of T105s was 3.7A = 0.8% of capacity at 14.8v, battery temperature 21C. The batteries are 2 years old and the MS tells me that the total charge taken out is the equivalent of 40 full cycles down to 0%SoC ie 40*450 = 18kAH. As far as I can tell they have the original capacity, in fact better than original since they need 50-100 partial cycles to fully "bed in". The MS tells me that the average discharge is 121AH ie down to 73% SoC. That would imply around 149 partial cycles (ie 149 days out of the marina) which is probably about right. Charge Efficiency Factor (ie proportion of charge put in, that can be taken out again, or if you like the reciprocal of the extra charge needed to be put in to replace charge taken out) is calculated by the MS and currently showing 95% The Smartgauge is configured thus: Battery type: 1 That's it. The Mastershunt is configured thus: Battery capacity: 450AH Battery type: Wet Nominal voltage: 12 Amps = full: 1.5% (ie the tail current when the counter will reset to 100%) Float voltage: 13.24v (voltage must be above this for the counter to reset to 100%) Absoption voltage: 14.4v (not sure what this is for) The mastershunt is a shunt + 500A fuse that does all the stuff mentioned above but doesn't have a display. You can connect it to a PC (inconvenient) or a Masterview Easy (ME) touchscreen display, which is what I have, by means of the Masterbus data network. Data is sent from the MS to the ME in digital form, ie all the processing is done in the MS itself. The ME is "plug and play" with other Mastervolt products including our Mastervolt Combi 12/2500-100. When you plug the Combi and the MS into the ME they are both detected and the display looks like this: Masterview devices.jpg You can press on any of the 3 items to configure them to your heart's content, eg changing the Combi's charge voltage and max current to whatever you like. You can also see all the parameters output by the devices, eg SoC from the MS and loads more. The MV has a "favourites" page which you can configure, taking data from any connected devices and merging it onto one page. Max 6 items. Mine is set up like this: Masterview Favourites.jpg Top row, SoC, Battery voltage from the MS Middle row, Current flowing into or out of the Battery, Battery temperature from the MS Bottom row, current going into or out of the Combi, Inverter output (or throughput) current (ie at 230v) Later (tomorrow) I'll post the results of the comparison discharge between MS and SG. Oh and just to mention, I need to put my glasses on to read the ME. I don't need to put my glasses on to read the SG.
nicknorman, on 25 Jan 2016 - 7:22 PM, said:
So of course one is then stuck with the figures at the end of the discharge, to start the charge process. The Smartgauge was clearly significantly under-reading the SoC at this point. On this graph, note that the current axis is on the right ie it started at around 165A before rapidly falling off.
I couldn't quite decide whether to continue the charge until fully charged or not. If I had have done so, I think you can see by the shape of the graphs that the two figures were converging but since the starting point for the SG was bad this didn't seem particularly useful. I therefore decided to terminate the charge prematurely in order to demonstrate one of the SGs major advantages which is that when it does go wrong, it self corrects without any user intervention. More of that in the next post.
I think one of the things this exercise has demonstrated to me is just how complex battery chemistry is. It is easy to think of a simple model but the reality is complicated. For example, when I started to recharge the current was 165A, rapidly fell off to 125A, but then for some strange reason increased back to 130A. What is going on!
nicknorman, on 25 Jan 2016 - 7:32 PM, said:
What was interesting was the jump the SG took when I bunged on 200A of discharge current for the 2kw kettle and then 100A for the toaster, much more than the drop in the MS, and this was where it went from over-estimate to under-estimate.
If you note the point about the final rested voltage, I think it shows that the MS was pretty correct, the SG was under-estimating.
nicknorman, on 25 Jan 2016 - 7:37 PM, said:
Well when the charging finished I had 84 on SG and 93 on MS. Just over 2 hrs later I now have 84 on the SG and 85 on the MS. So it is doing exactly what Gibbo said, it is holding its reading on discharge until the actual SoC passes the indicated SoC when it will become synchronised and track down - hopefully! I'll wait a bit longer and then do another graph.
nicknorman, on 25 Jan 2016 - 10:31 PM, said:
Well its all getting a bit weird now. I stopped charging at 90% on the MS and 84% on the SG. 4 hours later the MS is on 77% and accordingly I have taken out 13AH. The SG, however, is still on 84%. Which one is right? I switched everything off (missed the news on telly!) and quickly the voltage recovered to 12.65 on the Smartgauge, on my DVM and on the Combi's voltage measurement. To 12.56 according to the Mastershunt. So according to the Voc from the MS and Trojan's table of voltage vs SoC, the SoC is 83%. According to the other devices the SoC is 94%. And yet I've taken out 13% capacity (aka 58AH). As I said, what is going on? Anyway at the moment I believe the SG more than I believe the MS - which has obviously suffered from a cycle without reaching 100%.
So, which reading is correct?
Were any of the readings by a calibrated valid certified device? I suspect not.
If not any or all of them could incorrect. The Smartgauge only reads 3 ½ digits so 12.65V could easily be 12.6V or 12.7V. As for the Mastershunt, Voltage accuracy ± 0.6 % ± 1 digit and State of charge accuracy: ± 1% whereas the Smartgauge State of Charge is ± 5%, Voltage Accuracy ± 0.5%
So looking at the voltages taking into account accuracy: Which are being put forward as rested battery open circuit voltages after a relatively short rest. The battery industry recommendation being 24 hours. The longer the rested the higher the voltage.
The smartgauge voltage assuming a correctly calibrated box could be 12.71V
The Mastershunt assuming a correctly calibrated box could easily be 12.63V
A difference of 0.08V
Looking at State of Charge taking into account accuracy:
Smartgauge SoC could be 79.8% but it has not changed since the end of the charge, possibly, it has not re-synced
Mastershunt SoC could be 78%
A difference of 1.8%
Interestingly the Mastershunt at the end of the charge showed 93% (was this the value at the same time that the smartgauge showed 84% immediately or as the charging was stopped) and at this point 4 hours later, the Mastershunt showed 77%, which is 16%, used and the Smartgauge 84% nil used. Ahs used is declared as being 13Ah, (I am assuming this is a reading from the Mastershunt and not a calculated figure) as a percentage of the declared available battery capacity 90% (at end of charge) of 450Ahs, 13Ahs is (13Ahs/405Ah =) 3.11%. Now is the Mastergauge incorrectly set or is the battery capacity less than originally declared, say down at well that would be 100Ahs.
If we use a later figure give for Ahs used, 58Ahs how does that affect the figures. Mastershunt at the end of the charge showed 93% (was this the value at the same time that the smartgauge showed 84% immediately or as the charging was stopped) and at this point 4 hours later, the Mastershunt showed 77%, which is 16%, used and the Smartgauge 84%, nil used. Ahs used is declared as being 58Ah. As a percentage of the declared available battery capacity 93% (at end of charge) of 450Ahs, 58Ahs is (58Ahs/418.2Ah =) 13.86%. Now that would make the Mastershunt to be within 2.14% of the expected reading. Is the battery capacity less than originally declared, say down at 418.47Ahs.
Basically this tells us that either of the boxes could be right and does not show either box to be correct or incorrect, the differences are too close to decide. If the charge level had been taken back to the starting level i.e. to “The tail current for our 450AH bank of T105s was 3.7A = 0.8% of capacity at 14.8v” or the battery SoC been check by the Hydrometer that the owner has; we might have got something out of this information. Instead of something where apparently the tests were biased towards the smartgauge.
nicknorman, on 26 Jan 2016 - 12:02 AM, said:
HIGNFY was a repeat! Anyway by 23:30 the SG was on 81%, the MS on 72%. Switching everything off for 5 mins gave 12.55v on the DVM, 12.57 on the Combi, 12.58 or so on the SG, 12.47 on the MS. Even taking the lowest of these gives an SoC of 77%, taking a mean of the values gives over 80% and either of these puts the SG nearer than the MS. The latter is clearly going out of sync after just one cycle of failing to reach 100%. So in some ways this has become a victory for the SG since its errors are not cumulative. It may bounce around the correct SoC a bit but it never departs too far unlike the MS whose errors are boundless given enough cycles. But then again what is SoC? From the results tonight I am more convinced than ever that it is very hard to define or measure by any means. Graph of post-charging SoC showing the SG biding its time until the actual SoC passes the indicated SoC. Discharge2.JPG
Now what does this tell us
I note that this time there are no Ahs used readings.
The voltages were taken after 5 minutes resting and I would suggest are inaccurate because the battery is still settling and tell us nothing. Again, which of the instruments were calibrated and certified?
What were the State of Charge readings when the voltages were taken?
The Smartgauge has an accuracy of +/-5% and that means that the 81% read could be 76.95% and is still re-syncing.
The Mastergauge has an accuracy of +/-1% and that means the 72% reading could be 72.72%
Voltage wise the Mastergauge could be 12.395V or 12.54V
Voltage wise the Smartgauge is only readable to 3½ digits so 12.58V must be a guess as it flicks up and down, it was either 12.55 or 12.6V could be 12.49V or 12.66V
Whichever way you look at it the voltage readings are both within spec
Which is right I do not know, the voltage readings because of the shortness of the resting period tell us little or nothing. I do think these two gauges are very close.
I think readings done the next morning when the smartgauge had had a chance to fully re-sync together with specific gravity readings from the batteries would have shown results that were a lot more accurate and interesting.
Taking the same voltage reading with four different readings with four different meters does not mean that any of the readings is correct. Just that you have four different meters that read the voltage differently.
Think someone quoted a Chinese Proverb something like, “the more meters the more confusion”, and that is so true.