Do you have a SmartGauge?

[rusty69]

1 minute ago, nicknorman said:

Yes there is. I would say that the best option is to return the SG to the supplier for a replacement that is within spec. The user calibration function has a couple of problems, one is that the user cal is lost after a factory reset. Although it can always be done again I suppose. More importantly, what are you going to calibrate it against? Cheap multimeters look accurate because they have lots of decimal places, but in reality often aren't. And multimeters whose batteries are close to expiring can be ludicrously inaccurate.

The best option I have is the voltmeter on my Xantrex link pro. 

[nicknorman]

1 minute ago, rusty69 said:

The best option I have is the voltmeter on my Xantrex link pro. 

That quotes +- 0.3% for voltage. So if we can believe that, it should allow you to get within the +-0.5% required for the SG. What is the current difference between the Xantrex and the SG voltage readings?

[blackrose]

50 minutes ago, WotEver said:

Great that yours is spot on. I've not tested any myself (other than the one I possess which is pretty old and currently unused) so I have no personal experience of any being out of spec. Hence my post said 'check it if you can' and 'if it's out of spec...'. 

I'm truly pleased to see you state that yours is spot on - that's as it should be.

But I should also have said I bought mine 8 or 9 years ago.

The thing is that unless you can check a SG (or any other meter) against another meter of known accuracy and calibration then the whole exercise is meaningless - as are anecdotes of accuracy tests which don't include details of the meter being used for the test.

Edited April 10, 2017 by blackrose

[rusty69]

6 minutes ago, nicknorman said:

That quotes +- 0.3% for voltage. So if we can believe that, it should allow you to get within the +-0.5% required for the SG. What is the current difference between the Xantrex and the SG voltage readings?

Xantrex=12.53

SG=12.65

[WotEver]

10 minutes ago, blackrose said:

The thing is that unless you can check a SG (or any other meter) against another meter of known accuracy and calibration then the whole exercise is meaningless - as are anecdotes of accuracy tests which don't include details of the meter being used for the test.

Absolutely. Which is why I wrote in the OP "if you have a way of confirming the voltage display accuracy...". Just saying "it's different to the DVM I bought in Aldi" isn't confirming anything. 

[nicknorman]

13 minutes ago, rusty69 said:

Xantrex=12.53

SG=12.65

I sent you a PM.

[cuthound]

When I was working we used this company to calibrate meters, as all equipment used for commissioning had to hold a current calibration certificate,  even torque wrenches.

http://www.calibrate.co.uk

Their prices start from £25.

As Blackrose says, unless you know that the meter is calibrated, you really have no idea of its accuracy.

[IanD]

On 4/10/2017 at 17:30, nicknorman said:

The SG's circuitry is pretty stable over time and has virtually zero temperature co-efficient. However it needs to be set up correctly in the first place (by means of storing values in the microprocessor's memory, rather than twiddling trimpots) and for some inexplicable reason, this currently seems beyond the wit of the manufacturers.

If all the calibration data is digitally stored then it ought to be almost impossible to get it wrong -- normally you'd have the SG and an accurate DVM connected up to a (programmable?) reference voltage and a PC, run the calibration program, store calibration data in SG, job done. It takes real incompetence to get something this simple wrong...

[nicknorman]

11 minutes ago, IanD said:

If all the calibration data is digitally stored then it ought to be almost impossible to get it wrong -- normally you'd have the SG and an accurate DVM connected up to a (programmable?) reference voltage and a PC, run the calibration program, store calibration data in SG, job done. It takes real incompetence to get something this simple wrong...

Agreed. But you are missing one point, it also requires someone to be vaguely interested in getting it right, as opposed to someone who is paid according to how many devices they can pass through their hands on to despatch.

[IanD]

1 minute ago, nicknorman said:

Agreed. But you are missing one point, it also requires someone to be vaguely interested in getting it right, as opposed to someone who is paid according to how many devices they can pass through their hands on to despatch.

Still don't see how being in a hurry can cause such big errors, the procedure should be idiot-proof. Unless you're in so much of a hurry that you don't bother calibrating the SG at all, that could explain it...

[WotEver]

2 minutes ago, IanD said:

Unless you're in so much of a hurry that you don't bother calibrating the SG at all, that could explain it...

Now I think you're getting close

[MtB]

1 hour ago, IanD said:

Still don't see how being in a hurry can cause such big errors, the procedure should be idiot-proof.

 

As most development engineers know, if you make something idiot-proof the universe will respond by building a better idiot.

Maybe this is what has happened.

[jddevel]

My answer to all these queries and problems is "made in China"- for those as old as me that has replaced "made in Hong Kong" for quality control or lack of it.

[IanD]

Quality control in China is just as good or as bad as what you pay for it. A lot of high-quality electronics is made there -- most of the high-end gear you see in pretty much all applications -- as well as a lot of low-quality rubbish.

[bizzard]

19 minutes ago, IanD said:

Quality control in China is just as good or as bad as what you pay for it. A lot of high-quality electronics is made there -- most of the high-end gear you see in pretty much all applications -- as well as a lot of low-quality rubbish.

Agleed.

[jddevel]

So how do you know? Buy it and hope I suppose.

[rusty69]

Just now, jddevel said:

So how do you know? Buy it and hope I suppose.

Or don't 

[jddevel]

Well that doesn`t seem to answer the SG indicated problem. Reading the topic as a layman I`m not sure it fills me with confidence to invest in one when there is doubt in its reliability. Regarding the price factor. I agree you get what you pay for. The problem I feel , that in this competitive world we live in, is that as accountants rule even high end quality products MAY obtain components AT TIMES from sources with poorer quality control. All of this is particularly prevalent in my opinion in fast moving advancements in technology.   

[rusty69]

Just now, jddevel said:

Well that doesn`t seem to answer the SG indicated problem. Reading the topic as a layman I`m not sure it fills me with confidence to invest in one when there is doubt in its reliability. Regarding the price factor. I agree you get what you pay for. The problem I feel , that in this competitive world we live in, is that as accountants rule even high end quality products MAY obtain components AT TIMES from sources with poorer quality control. All of this is particularly prevalent in my opinion in fast moving advancements in technology.   

The trouble is, as I previously noted,unless you are in a position to check it yourself you could  be none the wiser if you do buy one. 

 

 

. 

Edited April 13, 2017 by rusty69

[Sea Dog]

Just the look of the Smartgauge, particularly the 70s calculator digits, suggests to me a company not investing in their product. I'm not knocking the SG concept, but you'd expect a refresh to update the look and probably some of the components now and again. If they aren't interested or able enough to do that, maybe that explains the apparent QC issues?  I rather wonder if the current manufacturer isn't 'rusting' on Gibbo's laurels.

[Tom and Bex]

Just an update about my smartgauge, I've contacted Merlin directly who have asked I send the unit back to them for repair/replacement. They said technically it should be returned via Cactus, but happy for me to return direct. Not even a request for proof of purchase, both of which indicate they are aware of potential problems.

I also raised the issue of quality control, adding how can such a large error get through, and their response was:

"I would also like to assure you that we do have strict quality control procedures in place here, so I will raise your concern with our Engineering Manager accordingly."

Will let you know how accurate the replacement is when it arrives!

To answer the question about how I know it's inaccurate without a calibrated meter, I have 2 digital multi meters (one cheapish once and one quality one), both from different manufactures. I also have a cheap Chinese panel voltmeter, and a tracer solar controller display panel, and all 4 agree with each other within 0.5% or less. The smartgauge however reads consistently 0.5v higher than these. I would be gobsmacked if all of them were so far out and the smartgauge was accurate! All of them combined cost less than the smartgauge, so how they can get away with selling it for so much when the accuracy is so poor I don't know. 

[WotEver]

Just now, Tom and Bex said:

Not even a request for proof of purchase, both of which indicate they are aware of potential problems.

They are certainly aware. 

[MtB]

Just now, Tom and Bex said:

Just an update about my smartgauge, I've contacted Merlin directly who have asked I send the unit back to them for repair/replacement. They said technically it should be returned via Cactus, but happy for me to return direct. Not even a request for proof of purchase, both of which indicate they are aware of potential problems.

I also raised the issue of quality control, adding how can such a large error get through, and their response was:

"I would also like to assure you that we do have strict quality control procedures in place here, so I will raise your concern with our Engineering Manager accordingly."

Will let you know how accurate the replacement is when it arrives!

To answer the question about how I know it's inaccurate without a calibrated meter, I have 2 digital multi meters (one cheapish once and one quality one), both from different manufactures. I also have a cheap Chinese panel voltmeter, and a tracer solar controller display panel, and all 4 agree with each other within 0.5% or less. The smartgauge however reads consistently 0.5v higher than these. I would be gobsmacked if all of them were so far out and the smartgauge was accurate! All of them combined cost less than the smartgauge, so how they can get away with selling it for so much when the accuracy is so poor I don't know. 

 

I bet when you get it back they say they tested it and found nothing wrong with it so are returning it to you. But miraculously it will now be reading voltage accurately.

This tends to be how it pans out in the world of boiler spares. Many years ago I returned a brand new cast iron heat exchanger with a 3mm hole in it. Potterton sent it back to me denying there was a hole, and when I unpacked it, they were right. Quelle surprise!

[Naughty Cal]

Just now, Mike the Boilerman said:

 

I bet when you get it back they say they tested it and found nothing wrong with it so are returning it to you. But miraculously it will now be reading voltage accurately.

This tends to be how it pans out in the world of boiler spares. Many years ago I returned a brand new cast iron heat exchanger with a 3mm hole in it. Potterton sent it back to me denying there was a hole, and when I unpacked it, they were right. Quelle surprise!

Either way Tom and Bex will have a working Smart Gauge which is surely the outcome which is required.

It matters not whether it is the same unit so long as it is accurate.

[nicknorman]

Just now, Naughty Cal said:

Either way Tom and Bex will have a working Smart Gauge which is surely the outcome which is required.

It matters not whether it is the same unit so long as it is accurate.

Although it will of course be unfortunate if they send another one that is equally inaccurate, maybe the other way!