Generic bluetooth PV generation monitor

thenudeone

Interesting idea to use the flashing LED to calculate generation.

Just waiting to see how long it will take someone to manufacture one to automatically switch appliances (eg washing machine) or a set load (eg an immersion heater) on at a set level.

http://www.mediafire.com/?tthrywk76e7v6vh

EDIT: LARGE PICTURE REMOVED

Martyn1981

I've got an Optismart, paired with an ENViR monitor I bought on E-bay (one of the many free Eon monitors people are flogging off). Much cheaper, but you need to check that the meter is suitable. Go to Current Costs web-site. If the TGM displays a solid red light at any time, then the Optismart may get a 'little' confused!

Mart.

Kernel_Sanders

thenudeone wrote: »

Interesting idea to use the flashing LED to calculate generation.

I've thought about this before, and concluded that the viability of such a device depends on how accurately it can measure the frequency of the flashes. The higher the number of watts being generated, the higher any inaccuracies will be magnified.

Martyn1981

Kernel_Sanders wrote: »

I've thought about this before, and concluded that the viability of such a device depends on how accurately it can measure the frequency of the flashes. The higher the number of watts being generated, the higher any inaccuracies will be magnified.

Hi kernel, I'm probably gonna regret this statement, but in my experience the system is 100% accurate to the TGM. I compared the monitor to the inverters, tricky as there are two, so I need steady generation, and some quick maths, but the monitor over-reads the inverters by 1.7%, which is exactly the difference showing on the system recently when I hit 1MWh. TGM 1,000 units inverters total 983 units.

Obviously need a slightly larger sample than 'me', but looking pretty good so far.

Mart.

PS system has just switched on at 141W, at 6.56am, nice.

Kernel_Sanders

thenudeone wrote: »

Interesting idea to use the flashing LED to calculate generation.
Just waiting to see how long it will take someone to manufacture one to automatically switch appliances (eg washing machine) or a set load (eg an immersion heater) on at a set level.

It would seem you only had to wait 42 minutes
https://forums.moneysavingexpert.com/discussion/comment/51865899#Comment_51865899

OK, he's not using the flashing of the light, but I thought it might amuse someone by posting this (if only myself!)

GaryMo_2

I've been using a Geo Solo PV, as posted by the OP for the past 6 or 7 weeks. It's a really good piece of kit and pretty accurate. I have it sat beside a general energy monitor which I use to show the live power consumption so it's easy to compare what's being used against what's being generated at any particular time.

Kernel_Sanders

Martyn1981 wrote: »

the monitor over-reads the inverters by 1.7%,

Mine overreads by precisely 1%. I'm presuming the Elster is far more accurate than the Sunnyboy.

Martyn1981

Kernel_Sanders wrote: »

Mine overreads by precisely 1%. I'm presuming the Elster is far more accurate than the Sunnyboy.

Yep, the inverters normally state that they are only estimates and not for invoicing purposes. I think they can vary under or over by 5% or so, but most people seem to find them 'similar' to the TGM.

The TGM, like all 'proper' meters should be pretty accurate, can't find it now, but I think the tolerances may be within 2% under-read, and 1% over-read (since they were designed to monitor import originally).

Not sure if all LED flash counters work perfectly, or if I just got lucky, but I seem to have 100% accuracy to my TGM (however accurate the TGM is!).

Mart.

Kernel_Sanders

Martyn1981 wrote: »

Not sure if all LED flash counters work perfectly, or if I just got lucky, but I seem to have 100% accuracy to my TGM (however accurate the TGM is!)
Mart.

It may be because it was a low output period when you did the testing. If the timer is one tenth of a second out when measuring generation at 400W, then the difference would be <5W. However, the same margin of error at 3600W would mean a grossly disproportionate discrepancy of between 327 and 400W.

Martyn1981

Kernel_Sanders wrote: »

It may be because it was a low output period when you did the testing. If the timer is one tenth of a second out when measuring generation at 400W, then the difference would be <5W. However, the same margin of error at 3600W would mean a grossly disproportionate discrepancy of between 327 and 400W.

Testing period is now nearly 7 months long!

Optismart signal is picked up by my ENViR monitor, whose cumulative total matches the increase in the TGM since I bought and fitted the Opti.

As far as I can tell, the Opti is supplying a 100% accurate report (to my monitors (also got an ENVi model upstairs)) of what the TGM is reporting (flashing).

Again, can't claim that all optical readers are 100% accurate, only have a sample size of 1, but looking good, till somebody spoils the party with a negative report.

Mart.

PS Sorry if this sounds like an advert for the Opti, or the monitors. Just saying what I see. M.

Kernel_Sanders

Martyn1981 wrote: »

Testing period is now nearly 7 months long!
Optismart signal is picked up by my ENViR monitor, whose cumulative total....

OK Martyn, it's just that it wasn't apparent to me that your monitor stored cumulative totals.