Battery Monitors and Capacity

[MtB]

39 minutes ago, nicknorman said:

 

Of course they were. As well you know, this is because they are both way out of cal in opposite directions. But it pleases you to have them like that, so we don’t mind!

 

Indeed, but new readers here might not know that is how they seem to be sent out from the factory these days, and checking the calibration on installation is essential before relying on a SmartGauge. The manual does not tell you this. 

[Dr Bob]

27 minutes ago, Jon57 said:

Working on 4000 cycles of the batteries they should cost about 36pence a day plus the cost to recharge (time and diesel)

looks good to me

....then add on the cost of capital. Big up front outlay with the benefits recoupped over 10 years. The reason I will not jump to these now is that battery technology advances will speed up significantly with the advance of electric cars and I bet there are far better options in less than 5 years time - so half your big initial capex expense will be wasted.

[Robbo]

9 minutes ago, Dr Bob said:

....then add on the cost of capital. Big up front outlay with the benefits recoupped over 10 years. The reason I will not jump to these now is that battery technology advances will speed up significantly with the advance of electric cars and I bet there are far better options in less than 5 years time - so half your big initial capex expense will be wasted.

I don’t think battery technology will significantly speed up to be honest (I think it may actually slow down).  If cars use lithium batteries more and more the cost of may actually go up.   In 5 years time if you installed lithium’s they would have paid off for them selves anyhow, and if you always wait for the next big thing you’ll be always waiting.

[mross]

3 minutes ago, Robbo said:

I don’t think battery technology will significantly speed up to be honest (I think it may actually slow down).  If cars use lithium batteries more and more the cost of may actually go up.   In 5 years time if you installed lithium’s they would have paid off for them selves anyhow, and if you always wait for the next big thing you’ll be always waiting.

Why do think this way?  When technologies become popular, costs drop.

[Jon57]

5 minutes ago, Dr Bob said:

....then add on the cost of capital. Big up front outlay with the benefits recoupped over 10 years. The reason I will not jump to these now is that battery technology advances will speed up significantly with the advance of electric cars and I bet there are far better options in less than 5 years time - so half your big initial capex expense will be wasted.

That is with the original cost of 2 x 100 amp batteries £1400 divide by 4000 cycles = 35pence. No need to worry about recharge/ look after worrying if they are fully charged.

Or pay £380 for 4 x 100 amp lead acid batteries ,fart about watching them all the time .then replace every 2 to 3 years . Life is to short for all that. 

Like all new technology how long do we wait till it gets cheaper. My first mobile phone was silly money compared with today’s phones but wouldn’t have waited till the price came down

[nicknorman]

5 minutes ago, mross said:

Why do think this way?  When technologies become popular, costs drop.

Not if it is based on a finite resource. https://www.ft.com/content/90d65356-4a9d-11e7-919a-1e14ce4af89b

screenshot hint if you can’t read the FT site.

Edited October 23, 2017 by nicknorman

[system 4-50]

22 minutes ago, Mike the Boilerman said:

 

Indeed, but new readers here might not know that is how they seem to be sent out from the factory these days, and checking the calibration on installation is essential before relying on a SmartGauge. The manual does not tell you this. 

How can a device that depends solely on voltage require "calibration"? Either the device is manufactured correctly, or your installation isn't giving it the true battery post voltage, or the device is totally faulty? Is "calibration" a manufacturer concept or is it something invented in the retail chain to give the customer something to believe in? What is "calibration"?

[Jon57]

2 minutes ago, nicknorman said:

Not if it is based on a finite resource. https://www.ft.com/content/90d65356-4a9d-11e7-919a-1e14ce4af89b

While I agree about the raw material cost of lithium (supply and demand) competition will have an overhaul effect on the final cost of the batteries 

[Robbo]

11 minutes ago, mross said:

Why do think this way?  When technologies become popular, costs drop.

When technology becomes popular and you can't keep up with demand costs rise.

[nicknorman]

11 minutes ago, system 4-50 said:

How can a device that depends solely on voltage require "calibration"? Either the device is manufactured correctly, or your installation isn't giving it the true battery post voltage, or the device is totally faulty? Is "calibration" a manufacturer concept or is it something invented in the retail chain to give the customer something to believe in? What is "calibration"?

Seems an odd question! Calibration means adjusting something, typically a measuring device, so that its reading matches a known good reference, or at least producing a table that shows the error and how to calculate it out.

in the case of the SG it is a matter of adjusting the internal analogue to digital conversion so that the measured voltage matches the actual voltage.

Just an aside, a digital voltmeter works by using an analogue to digital converter with a specified number of output bits, maybe 10 or 12. 12 bits is of course 2048 so if the measurement range is say 20.48v, the smallest change in the measured value is 0.01v. This is far too coarse for the Smartgauge, which is looking for very slight changes in voltage and dV/dt to work out the battery’s discharge profile and hence SoC. So the SG uses a capacitor charging method. The capacitor is charged via a circuit whose current is proportional to the voltage to be measured. When the capacitor’s voltage reaches a reference, it is discharged and the charging begins again. Whilst the charge is taking place a timer runs (16 bit (ie 65536 max count), I suspect), and stops counting when the capacitor reaches its reference. Thus there is a lot more resolution to the measurement which allows small changes to be detected.

But this type of measurement is not particularly accurate (note, big difference between accuracy and resolution) especially as manufacturing tolerances for capacitors tend to be fairly large. Thus some calibration is required at assembly, working out a mathematical compensation so that the counter’s number accurately reflects the voltage being measured. The calibration value is stored in the device’s memory and applied to each measurement cycle (I think it’s around 10 per second, can’t remember exactly)

Edited October 23, 2017 by nicknorman

[mross]

All resources are finite!  The cost of extracting and incorporating Li in batteries will likely fall, even if its commodity price rises.

[Robbo]

3 minutes ago, Jon57 said:

While I agree about the raw material cost of lithium (supply and demand) competition will have an overhaul effect on the final cost of the batteries 

The raw material is a good portion of the batteries cost in lithium batteries.

[mross]

2 minutes ago, nicknorman said:

Seems an odd question! Calibration means adjusting something, typically a measuring device, so that its reading matches a known good reference, or at least producing a table that shows the error and how to calculate it out.

in the case of the SG it is a matter of adjusting the internal analogue to digital conversion so that the measured voltage matches the actual voltage.

 

How true.  Very few instruments actually 'measure' anything; they are indicators.  A voltmeter uses the voltage to drive a small current through a device and 'deduces' what the voltage is by calculation or influencing a moving coil or moving magnet.  And, of course, by drawing a current it affects the very thing it is measuring.  

[system 4-50]

4 minutes ago, nicknorman said:

... in the case of the SG it is a matter of adjusting the internal analogue to digital conversion so that the measured voltage matches the actual voltage.

 

Are you saying this is different for different customers? Presumably not. So this "adjustment" should be the same for all devices? So it is a factory fault?

[dmr]

2 minutes ago, system 4-50 said:

Are you saying this is different for different customers? Presumably not. So this "adjustment" should be the same for all devices? So it is a factory fault?

Less common these days, but it was pretty normal for electronic equipment to need some final adjustments, done between leaving the production line and going into the box for shipping. I believe the Smartgage uses an unusual ADC (voltage measurement chip) that has high resolution but is not intrinsically very accurate so every device needs a final factory adjustment to get the required level of accuracy.

...............Dave

[mross]

8 minutes ago, Robbo said:

When technology becomes popular and you can't keep up with demand costs rise.

Give me some examples..........

[Jon57]

1 minute ago, mross said:

Give me some examples..........

Diesel

[nicknorman]

6 minutes ago, system 4-50 said:

Are you saying this is different for different customers? Presumably not. So this "adjustment" should be the same for all devices? So it is a factory fault?

Yes, no, yes.

Each device will have slight differences in component values (especially the capacitor) so the compensating value will be different for each device. Of course, as a final part of assembly, the device is connected up to a test rig and the compensating value for each device is determined and written into NV memory. It seems that this testing and calibration process has been messed up recently, resulting in devices being sold which are out of spec by quite a bit. Piss poor, but it reflects badly on the manufacturer not on the designer.

3 minutes ago, dmr said:

Less common these days, but it was pretty normal for electronic equipment to need some final adjustments, done between leaving the production line and going into the box for shipping. I believe the Smartgage uses an unusual ADC (voltage measurement chip) that has high resolution but is not intrinsically very accurate so every device needs a final factory adjustment to get the required level of accuracy.

...............Dave

It doesn’t have an ADC chip as such. It has a variable current source, a largish tantalum capacitor, a reference voltage, a comparator and uses a microprocessor register or two for the counter. The only surprising thing is that there isn’t more temperature drift, but it seems pretty good in that respect.

[dmr]

5 minutes ago, nicknorman said:

Yes, no, yes.

Each device will have slight differences in component values (especially the capacitor) so the compensating value will be different for each device. Of course, as a final part of assembly, the device is connected up to a test rig and the compensating value for each device is determined and written into NV memory. It seems that this testing and calibration process has been messed up recently, resulting in devices being sold which are out of spec by quite a bit. Piss poor, but it reflects badly on the manufacturer not on the designer.

It doesn’t have an ADC chip as such. It has a variable current source, a largish tantalum capacitor, a reference voltage, a comparator and uses a microprocessor register or two for the counter. The only surprising thing is that there isn’t more temperature drift, but it seems pretty good in that respect.

Yes, I was trying to use a wording suitable for the questioner who does not understand this sort of stuff. I did suggest that it was unusual.

 So, as you know what goes on inside it....is the "charge" capacitor really a tantalum? that's an odd choice. More interestingly, I thought "charge balancing" was the accepted method for high accuracy slow voltage measurement and this avoids the issue of capacitor tolerance, so I wonder why Gibbo chose a "capacitor charging counter" approach rather than full charge balancing??? Maybe it just saved a couple of components at the expense of needing calibration???

............Dave

[nicknorman]

Tantalum? Maybe not on reflection.

[dmr]

4 minutes ago, nicknorman said:

Tantalum? Maybe not on reflection.

I realise that as one gets older its quite common to look back to the "good old days" but I really do believe that surface mount electronics as just not as pretty as the proper old through-hole stuff. Everything is just so ..... Black.

...............Dave

[nicknorman]

4 minutes ago, dmr said:

I realise that as one gets older its quite common to look back to the "good old days" but I really do believe that surface mount electronics as just not as pretty as the proper old through-hole stuff. Everything is just so ..... Black.

...............Dave

It’s not particularly pretty but also has the look of being hand-soldered rather than screen printed with paste and cooked. Anyway, can you tell what sort of capacitor it is? Not sure I can!

As to charge balancing ADC, I don’t know much about that but perhaps it is more prone to thermal drift? 

[Dr Bob]

1 hour ago, Robbo said:

I don’t think battery technology will significantly speed up to be honest (I think it may actually slow down).  If cars use lithium batteries more and more the cost of may actually go up.   In 5 years time if you installed lithium’s they would have paid off for them selves anyhow, and if you always wait for the next big thing you’ll be always waiting.

I am usually at the bleeding edge of technology, cameras, lenses, sky boxes (!!!) etc etc but batteries, naw. Too much will change. There is a huge increase in money being invested in new ways to store renewable energy and a lot of it is not visble to the layman. This is one area I am going to wait. Yes a gamble but coming from a reasonably well informed position. Will I ever see it in my lifetime? I doubt it but have plenty of time faffing around with power etc without worrying about it....what else is there to do on a NB?

[Dr Bob]

1 hour ago, nicknorman said:

Seems an odd question! Calibration means adjusting something, typically a measuring device, so that its reading matches a known good reference, or at least producing a table that shows the error and how to calculate it out.

in the case of the SG it is a matter of adjusting the internal analogue to digital conversion so that the measured voltage matches the actual voltage.

Just an aside, a digital voltmeter works by using an analogue to digital converter with a specified number of output bits, maybe 10 or 12. 12 bits is of course 2048 so if the measurement range is say 20.48v, the smallest change in the measured value is 0.01v. This is far too coarse for the Smartgauge, which is looking for very slight changes in voltage and dV/dt to work out the battery’s discharge profile and hence SoC. So the SG uses a capacitor charging method. The capacitor is charged via a circuit whose current is proportional to the voltage to be measured. When the capacitor’s voltage reaches a reference, it is discharged and the charging begins again. Whilst the charge is taking place a timer runs (16 bit (ie 65536 max count), I suspect), and stops counting when the capacitor reaches its reference. Thus there is a lot more resolution to the measurement which allows small changes to be detected.

But this type of measurement is not particularly accurate (note, big difference between accuracy and resolution) especially as manufacturing tolerances for capacitors tend to be fairly large. Thus some calibration is required at assembly, working out a mathematical compensation so that the counter’s number accurately reflects the voltage being measured. The calibration value is stored in the device’s memory and applied to each measurement cycle (I think it’s around 10 per second, can’t remember exactly)

The bit in bold is the key to how the SG operates. This is not an easy thing to do and is accomplised by taking a set of input data viz dV/dt, V over time (not quite the same as dV/dt) etc etc and 'guessing' output data ie SG. The 'guessing' is done by a software model - likely a linear regression type (or multi linear regression) unless Gibbo was using more advanced techniques like topology or sparse data matching (which is unlikely as he would have said that to help sell it). There is always going to be some error in that model and particularly if a 'system' being monitored has different characteristics to the systems used to build the models (ie is outside the training set of the model). I know from what is said here that 'thousands' of systems were tested so the model should be quite robust BUT it will not be a 100% global model capable of being 100% correct with EVERY system. It obviously works extremely well on Nick's system. It may have difficulty tracking some 'outlier' systems. Nick you have on a few occations said that if the Voltage is correct then so will the SoC. That will only hold if the model is a 100% global model. I dont think anyone knows the answer to that (I certainly dont) apart from Gibbo. I dont think I have seen a 100% global model for inputs of the complexity we have here.

[dmr]

15 minutes ago, nicknorman said:

It’s not particularly pretty but also has the look of being hand-soldered rather than screen printed with paste and cooked. Anyway, can you tell what sort of capacitor it is? Not sure I can!

As to charge balancing ADC, I don’t know much about that but perhaps it is more prone to thermal drift? 

I've never done it myself as my interest was always fast data acquisition, but I think its the method used in digital voltmeters; charge the capacitor with the applied voltage and count up, switch to a good quality current source and discharge back to zero, again counting. As long as the current source is accurate this gives a measure of voltage independent of capacitor value.

Just been looking at these Lithium boat batteries again and I want them, the fast charge without all the slow absorption stuff would be the real bonus for me as an off grid liveaboard. The lighter weight is a problem. I think this thing about "use of full capacity" is a red herring. Trojans can be taken down to 20% and Lithium will in practice only go down to 20% as any sensible boater will always want a little safety margin. So if we compare on a 20% like for like basis the Lithiums do have about twice the cycle life but they are still just toooooo expensive.

..............Dave