Electroquest 30 Amp Charger difficulties - Maximum Battery Bank Size?

[alan_fincher]

3 hours ago, Dr Bob said:

....then we will have to disagree.

I do not think that a gauge that on one hand claims to be 'highly accurate' but has the following specification is any better than my limited skills.

• Accuracy For State Of Charge (Charge): +/- 10%
• Accuracy For State Of Charge (Discharge): +/- 5%
• Accuracy Voltage: +/- 0.5V DC

 

47 minutes ago, Dr Bob said:

On the Force 4 web site.

https://www.force4.co.uk/merlin-smartgauge-battery-monitor.html#.WgCpQIZpE0o

Yes, that's why I was attempting to raise the awareness of how accurate the SG could be. When I look a the volts in the morning, I want to see it accurate to at least ten times better. Not checked any other sellers site though.

 

24 minutes ago, Mike the Boilerman said:

From the Force 4 website, describing the Smartgauge

• Accuracy Voltage: +/- 0.5V DC

Has this always been the case or is it new? I'm thinking now if I returned mine as faulty, I'd just get them back again stating they are within specification. 

If it is a recent slackening of the spec, I wonder if it is in response to an avalanche of returns. 

 

 

 

 

24 minutes ago, WotEver said:

It appears that Force4 don’t understand the difference between 0.5% and 0.5V. 

(It’s a difference of 0.44V)

Neither. It’s a mistake. 

Cactus, however, know how to type

http://www.cactusnav.com/merlin-smartguage-battery-monitor-p-11874.html

However, as far as I can see, the actual supplies manual does not specify the accuracy for the voltage reading.

What it does say is this.....

Quote

Although the battery voltage is displayed to a resolution of 0.05 volts (0.1 volts in 24 volt systems),internally it is measured and dealt with to a finer resolution.

So equally you could argue that Force4 have the decimal point in the wrong place, I suppose.

[WotEver]

Just now, alan_fincher said:

So equally you could argue that Force4 have the decimal point in the wrong place, I suppose.

Nope. That would simply replace mixing up percentages and volts with mixing up accuracy and display resolution instead. 

[MtB]

1 minute ago, alan_fincher said:

Although the battery voltage is displayed to a resolution of 0.05 volts (0.1 volts in 24 volt systems),internally it is measured and dealt with to a finer resolution.

 

So just in case anyone missed it, when my two SGs display battery voltage the difference between the two (connected to the same battery) is 0.7v. 

A leetle more than the claimed 0.05v I thnk!

[alan_fincher]

Just now, Mike the Boilerman said:

 

So just in case anyone missed it, when my two SGs display battery voltage the difference between the two (connected to the same battery) is 0.7v. 

A leetle more than the claimed 0.05v I thnk!

Yes,

But I still don't understand why you keep telling us that, rather than getting them fixed.

[WotEver]

Just now, Mike the Boilerman said:

So just in case anyone missed it, when my two SGs display battery voltage the difference between the two (connected to the same battery) is 0.7v. 

A leetle more than the claimed 0.05v I thnk!

Yes Mike, yours are out of spec. You know it, and we know it. 

And you’re making the same mistake as Alan above, a display resolution of 0.05V is NOT the same thing as an accuracy of 0.05V. 

[MtB]

Just now, alan_fincher said:

Yes,

But I still don't understand why you keep telling us that, rather than getting them fixed.

 

You missed this part of my post:

"So just in case anyone missed it"

Besides, thw two are not mutually exclusive. Should I get them fixed, I'll continue telling the board about my godawful experience with factory-faulty Smartgauges.

[alan_fincher]

1 minute ago, WotEver said:

And you’re making the same mistake as Alan above, a display resolution of 0.05V is NOT the same thing as an accuracy of 0.05V. 

I didn't actually intend to state the two were the same.

However in my head if the resolution is 0.05V, but it says internally it "is measured and dealt with to a finer resolution" tha statement seems odd if the accuracy of measuring the voltage were less than that.  Why work at a finer resolution than you know you are actually measuring to?

It surely shouldn't be hard to hit an accuracy of at least +/- 0.05V in a quite pricey bit of kit, particularly if that accuracy is absolutely fundamental to how good the information is that then gets presented to the user.

What I don't understand is why Gibbo chose to only show .00 or .05 numbers, and not those that lie between, or has someone already worked that out?

[MtB]

7 minutes ago, WotEver said:

Yes Mike, yours are out of spec. You know it, and we know it. 

And you’re making the same mistake as Alan above, a display resolution of 0.05V is NOT the same thing as an accuracy of 0.05V. 

Ah ok, so you're saying the display resolution is 0.05v, but the accuracy is actually +/-0.5v as started on the Force 4 website after all?

Thanks for clearing that one up.

 

[WotEver]

6 minutes ago, alan_fincher said:

However in my head if the resolution is 0.05V, but it says internally it "is measured and dealt with to a finer resolution" tha statement seems odd if the accuracy of measuring the voltage were less than that.  Why work at a finer resolution than you know you are actually measuring to?

It isn’t less. The accuracy of the measured voltage on any SmartGauge not owned by Mike is far better than 0.05V. My guess is that it’s at least 0.001V but the figure isn’t published, so I could be wrong. 

Why did Gibbo only display to 1.5 decimal places? I don’t know, but I suppose it’s because it’s quite a sufficient resolution for the intended non-technical user.

Edited November 6, 2017 by WotEver

[system 4-50]

11 hours ago, alan_fincher said:

I didn't actually intend to state the two were the same.
... What I don't understand is why Gibbo chose to only show .00 or .05 numbers, and not those that lie between, or has someone already worked that out?

I have advanced the theory before without hearing a contradiction that if you could see the voltage that the SG was working on to the precision that it does that then you would be able to reverse-engineer the SG (more easily) and take advantage of Gibbo's work without paying him for it.

MTB, is the reason you have not taken your SGs back that you have opened the boxes?

[nicknorman]

11 hours ago, WotEver said:

It isn’t less. The accuracy of the measured voltage on any SmartGauge not owned by Mike is far better than 0.05V. My guess is that it’s at least 0.001V but the figure isn’t published, so I could be wrong. 

Why did Gibbo only display to 1.5 decimal places? I don’t know, but I suppose it’s because it’s quite a sufficient resolution for the intended non-technical user.

I think the accuracy of the measured voltage depends on how well it was set up at the factory. But once set up, it seems to be remarkably stable and not prone to such things as temperature drift.

Why 1.5 decimal places? I think because the way it measures voltage is subject to some noise and if there were more less significant digits they would be constantly changing a bit. Displaying the value to 1.5 decimal places is a way of stabilising that without just slugging the value with a large time constant that would make the response to sudden large voltage changes seem sluggish.

[Dr Bob]

36 minutes ago, system 4-50 said:

I have advanced the theory before without hearing a contradiction that if you could see the voltage that the SG was working on to the precision that it does that then you would be able to reverse-engineer the SG (more easily) and take advantage of Gibbo's work without paying him for it.

MTB, is the reason you have not taken your SGs back that you have opened the boxes?

You may have something there. If you recorded the data (viz voltage Vs Soc) over a 12 month period, you could then produce a model that could attempt to predict SoC from V for your system. You then need to do this 100 times or more on other systems to get a more 'global' model. Far easier than starting from scratch. You couldnt do this with a resolution of only 0.05V.

[MtB]

39 minutes ago, system 4-50 said:

MTB, is the reason you have not taken your SGs back that you have opened the boxes?

 

No, it's because I'm not sure why I would need them to be accurate. 

Why does anyone need a  Smartgauge anyway, given it can't tell you when the batteries are at 100%? 

[Dr Bob]

Marinsuperstore is also showing the voltage accuracy at +/-0.5V. Surely these suppliers wouldnt be copying each others data/specs? Does this point the error (and it must be an error) at Merlin?

http://www.marinesuperstore.com/batteries-power-chargers/battery-management/merlin-smartgauge-battery-monitor-standard-12-24

• Accuracy For State Of Charge (Charge): +/- 10%.
• Accuracy For State Of Charge (Discharge): +/- 5%.
• Accuracy Voltage: +/- 0.5V DC

Tony's suggestion earlier that the 0.5V should be 0.5% as the accuracies should use the same 'units' ie %. I am not so sure. When I quote accuracies on data I usally use the units in use - therefore voltage accuracy would be +/-0.5V, +/-0.05V etc and SoC as % (as that is the unit used for SoC). Possible they just have the decimal place in the wrong place.

[Dr Bob]

10 minutes ago, Mike the Boilerman said:

 

No, it's because I'm not sure why I would need them to be accurate. 

Why does anyone need a  Smartgauge anyway, given it can't tell you when the batteries are at 100%? 

It is interesting that NN advised 'rowland al' on another thread (current at the time of this post) -

 to buy an ammeter based battery meter to understand his battery problem rather than a SG. Good choice.

[MtB]

2 minutes ago, Dr Bob said:

Marinsuperstore is also showing the voltage accuracy at +/-0.5V. Surely these suppliers wouldnt be copying each others data/specs? Does this point the error (and it must be an error) at Merlin?

 

 A more cynical explanation would be that Merlin are well aware that some of their products leave the factory calibrated wrongly, and have loosened the manufacturing specification to account for the range of different calibrations going out the door. 

My two are 0.3V low and 0.4V high, for example. Three could be quite a large population of incorrectly calibrated Smartgauges out there, given that any individual user is unlikely to:

1) Think there is any reason to test the accuracy of calibration their SG in the first place

2) Have an instrument known to be correctly calibrated to carry out the test

3) Understand the critical importance of correct calibration anyway, if the SG is to display even vaguely accurate results.

 

Slackening the claimed accuracy will allow Merlin, should I ever send them back, to just tell me mine are within specification.

[Dr Bob]

3 minutes ago, Mike the Boilerman said:

1) Think there is any reason to test the accuracy of calibration their SG in the first place

 

You are right on calibration. When you buy a piece of kit you assume it is fit for purpose. For power it is helpful a volt meter is accurate to 2 places of decimals as the difference between 80 and 90% SoC (if you believe the tables) is 12.50V and 12.62V. For an ammeter it only has to be +/-10% accurate to get the understanding of what is going in or out. The demands on accuracy on a volt meter are much higher. For a system that relies on voltage (either an SG or an ammeter based system) - then some sort of check should be done against 2 or 3 other volt meters if a calibration problem suspected.

[smileypete]

Why not get one of those cheap Ebay 5 digit voltmeters? from about £4 delivered from a UK seller.

https://www.ebay.co.uk/itm/5-Digit-Red-LED-DC-0-4-3000-33-000V-Digital-Voltage-Meter-Panel-Voltmeter-v/181719077914

https://www.ebay.co.uk/itm/Red-0-36-LED-5-Digit-DC-0-33-000V-Digital-Voltmeter-Voltage-Meter-Car-Panel-UK/192057429317

There's a few different versions now, but going by the pics the above ones are similar to the original ones.

Edited November 7, 2017 by smileypete

[WotEver]

41 minutes ago, Mike the Boilerman said:

Why does anyone need a  Smartgauge anyway, given it can't tell you when the batteries are at 100%? 

In order to tell you the SoC under discharge silly. That’s what it’s for. 

35 minutes ago, Dr Bob said:

Tony's suggestion earlier that the 0.5V should be 0.5% as the accuracies should use the same 'units' ie %. I am not so sure.

You can be sure. What I wrote is correct. I even supplied a link to probably the UK’s biggest supplier of the product to show you. 

[MtB]

Just now, WotEver said:

In order to tell you the SoC under discharge silly. That’s what it’s for. 

 

I suppose that would be mildly useful ... but it only gives a rough idea. Just as my BMV702 gives a rough idea, and I had to buy the BMV702 anyway as the SG could not be trusted. 

[WotEver]

30 minutes ago, Dr Bob said:

It is interesting that NN advised 'rowland al' on another thread (current at the time of this post) - to buy an ammeter based battery meter to understand his battery problem rather than a SG. Good choice.

Yes, because his problem wasn’t knowing when to start charging (or stop discharging) his problem was not knowing when to stop. So Nick recommended the best tool for the purpose that the OP required. 

Just now, Mike the Boilerman said:

 

I suppose that would be mildly useful ... but it only gives a rough idea. Just as my BMV702 gives a rough idea, and I had to buy the BMV702 anyway as the SG could not be trusted. 

But the BMV tells lies that accumulate and get worse with time unless you regularly charge to 100% and reset it, which I know you don’t. 

[Dr Bob]

5 minutes ago, WotEver said:

You can be sure. What I wrote is correct. I even supplied a link to probably the UK’s biggest supplier of the product to show you. 

I cant remember what you wrote, but accuracy is often written in the units used via a voltage accuracy of +/-0.05V. Yes, the 'biggest supplier' may have used % not V. Doesnt mean it is correct. It could be either way. What we need to see is what the manufacturers of the kit have as a spec. You seem to be one of the more 'switched on' people on here wrt info. Have you seen a voltage spec from Gibbo/merlin?

[WotEver]

2 minutes ago, Dr Bob said:

I cant remember what you wrote, but accuracy is often written in the units used via a voltage accuracy of +/-0.05V. Yes, the 'biggest supplier' may have used % not V. Doesnt mean it is correct. It could be either way. What we need to see is what the manufacturers of the kit have as a spec. You seem to be one of the more 'switched on' people on here wrt info. Have you seen a voltage spec from Gibbo/merlin?

Why do you keep questioning this? IT IS CORRECT. It’s not a guess, it’s a fact. 

As to Merlin’s published spec, yes I’ve seen one. I just went looking for it but they’ve revamped their website and I can’t find it right now. However the document they supplied stated:

within 5%SoC during discharge, within 10%SoC during charge, and voltage within 0.5%.

Cheers,

Tony

[nicknorman]

It doesn't really matter what the voltage accuracy is. If you want to measure voltage super-accurately, use a DVM. And a proper, calibrated one, not one with lots of digits that mislead you into thinking its more accurate than it is!

The point is that since the device works only on voltage, the voltage measurement needs to be accurate enough only to meet the SoC accuracy figures. If the voltage is less accurate than that, that error will dominate and the SoC readings will be out of tolerance. If the voltage measurement is more accurate than it needs to be to meet the SoC accuracy, it just means that other sources of error/uncertainty dominate (such as the vagaries of what actually goes on in a real-world LA battery). So there is an optimum accuracy required to calculate the SoC within the 5 and 10% figures.

[nicknorman]

Anyway all this silly arguing is getting rather tedious. So I'm going to take charge and say: Dr Bob, you are banned from ever buying a Smartgauge. It is far too clever for you to understand how it could work and anyway, you don't need it. Never Ever buy a Smartgauge or you'll have me to answer to.

There, that seems to have solved the problem.