CS 370 1/2 cut panels, Tigo TS4 and Shading - Solis 25Kw Inverter

[curto]

Guys,

We have 23Kw of Solar on our roof in Australia - two Solis Inverters in 3 Phase - a 25Kw unit and a 5Kw - we need the Strings because of weird roof layouts.

On the 25Kw unit - one of the strings is 13 panels long.

We put this set of panels in last year in April - so this is the first year that we have experienced Solar on these panels in January/February.

These panels are laid out as a single string in a L shape - 7 panels face North and the other 6 NNW

The 6 panels on the NNW face suffer shade until mid morning in our WInter period - May through to September so we fitted them out with TS4 optimizers - the idea being that these would be isolated and the strong would operate on the 7 North panels until the others came out of shade.

All of the panels are mounted in landscape orientation in relation to the roof (so long sides face the gutter and the ridge line)

What we are finding is that 3 of the 6 panels are being shaded along their length - so across the two halves of the panels by the eaves of the house until approx 12:30 during January and February and they are dragging the whole string down so we are seeing nearly the full voltage on the sting but the amps are less than 1 until they come out of the shade and then the whole string roars to life.

I would have thought that the Tigos would have isolated these 3 panels out of the string and allowed it operate - however this is not the case - we have done testing and bypassed the shaded panels and the rest of the string does come to life.

It is almost as if the 1/2 cut nature of the panels, the triple diodes and the TIGO are all fighting each other and not achieving what we want.

Has anyone done any real world testing with Tigos, 1/2 cut panels and shading ?

My thought is that i will split the string into one string of 7 panels and the other of 6 and parallel them - the TIGOs should handle the string length mismatch by virtue of their voltage boost capability.

Do people think this would work or would we still see the shaded string bringing down the other string ?

Craig

 

[Strousy]

Guys,

We have 23Kw of Solar on our roof in Australia - two Solis Inverters in 3 Phase - a 25Kw unit and a 5Kw - we need the Strings because of weird roof layouts.

On the 25Kw unit - one of the strings is 13 panels long.

We put this set of panels in last year in April - so this is the first year that we have experienced Solar on these panels in January/February.

These panels are laid out as a single string in a L shape - 7 panels face North and the other 6 NNW

The 6 panels on the NNW face suffer shade until mid morning in our WInter period - May through to September so we fitted them out with TS4 optimizers - the idea being that these would be isolated and the strong would operate on the 7 North panels until the others came out of shade.

All of the panels are mounted in landscape orientation in relation to the roof (so long sides face the gutter and the ridge line)

What we are finding is that 3 of the 6 panels are being shaded along their length - so across the two halves of the panels by the eaves of the house until approx 12:30 during January and February and they are dragging the whole string down so we are seeing nearly the full voltage on the sting but the amps are less than 1 until they come out of the shade and then the whole string roars to life.

I would have thought that the Tigos would have isolated these 3 panels out of the string and allowed it operate - however this is not the case - we have done testing and bypassed the shaded panels and the rest of the string does come to life.

It is almost as if the 1/2 cut nature of the panels, the triple diodes and the TIGO are all fighting each other and not achieving what we want.

Has anyone done any real world testing with Tigos, 1/2 cut panels and shading ?

My thought is that i will split the string into one string of 7 panels and the other of 6 and parallel them - the TIGOs should handle the string length mismatch by virtue of their voltage boost capability.

Do people think this would work or would we still see the shaded string bringing down the other string ?

Craig

Hi Craig,

I have a related issue. We have 2 arrays (strings) next to each other. One of 12 and one of 6 panels. In winter, the same tree shade slides accross them so I installed 6 Tigos on the string of 6 panels to see what happens.

The observation is that with only one panel in the shade of the Tigo array, we virtually lose most of the power which is indeed not the Tigo promise. I haven't by passed shaded panels yet to see if the string inverter can handle 4 panels well without the shade. In other words, can my string inverter handle running on only 4 units?

you might be interested in a possible software issue: https://www.mcelectrical.com.au/tigo-optimiser-recall/
and sales@tigoenergy.com

Peter

 

[curto]

Hi Craig,

I have a related issue. We have 2 arrays (strings) next to each other. One of 12 and one of 6 panels. In winter, the same tree shade slides accross them so I installed 6 Tigos on the string of 6 panels to see what happens.

The observation is that with only one panel in the shade of the Tigo array, we virtually lose most of the power which is indeed not the Tigo promise. I haven't by passed shaded panels yet to see if the string inverter can handle 4 panels well without the shade. In other words, can my string inverter handle running on only 4 units?

you might be interested in a possible software issue: https://www.mcelectrical.com.au/tigo-optimiser-recall/
and sales@tigoenergy.com

Peter

Yes please keep me updated on your progress as i am not very happy with how these TIGOs perform with the 1/2 cut panels in Landscape

 

[Strousy]

Yes please keep me updated on your progress as i am not very happy with how these TIGOs perform with the 1/2 cut panels in Landscape

Rethinking it over, your story sounds like the story in the link I provided. A software issue. I was contacted by Tigo Australia (Jeff) who offered help after stating that Tigo never did a software recall in Australia. After I sent a detailed diagnotic story, Jeff asked for a convenient time to phone. This phone call has not happened yet (the agreed time was missed) but stating that "Tigo never did a software recall in Australia" did not address the initial query of the need for a software recall. To me this feels like Mark Cavanagh is right in his claim that Tigos don't work properly unless you connect them up to their monitoring system which then takes care of a software update. It feels like their are trying to hide a huge scandal, which of course, makes the final verdict even more dangerous. I'm in no rush for Tigo to sort my problems, I'm more keen to understand what is happening and pinpoint where the issues exactly are. With the constant cloud cover in Adelaide currently this is a bit hard, but over the next two weeks I should find the opportunity sometime. I am a retired technician myself and I need to understand. The $500,00 on Tigos is neither here nor there, I'm after the truth first and I have not much sympathy for companies hiding the truth in order to make a buck. My direct email (Peter) theidealdiet (AT) tpg.com.au

 

[Strousy]

Yes please keep me updated on your progress as i am not very happy with how these TIGOs perform with the 1/2 cut panels in Landscape

An update after sunshine and time allowed for further Tigo testing yesterday:

The 6 panels with Tigos installed have been checked with a EY800W Solar Panel Multimeter and all performed well.
The Solax inverter has demonstrated to be able to run on 4 panels after bypassing 2 panels.

Conclusion: If the Solax inverter can run on 4 panels but does not run properly on 6 panels with Tigos installed, while 2 panels are partially shaded, that does mean that the Tigos do not live up to their promise of impedance matching and removing the effect shade on 2 panels has on the whole string.

My conclusion: The $442.42 I spend on Tigos have been wasted. I wish I had read Mark Cavanagh's blog before deciding on giving Tigo a go. (https://www.mcelectrical.com.au/tigo-optimiser-recall/)

 

[ncsolarelectric]

Did you put Tigos on all 6 panels, or only on 2 panels? I've never used Tigo optimizers, but my understanding is to make them work for impedance matching, they have to be installed on every panel in the string, not just the shaded panels. That's the only way they can optimize the power output of the whole string.

When a panel is shaded, the output current is lower. To impedance match, all of the OTHER panels in the string must be boosted in voltage to lower the current to be the same as the shaded panel, and thereby maximize the output power of the string. If you only put the Tigos on the 2 shaded panels, it does the opposite and just makes it worse. I don't see how it could do anything useful that way.

It also seems interesting that Tigo doesn't have a maximum voltage output limit spec. So how high can it boost the string voltage to impedance match?

 

[Strousy]

Tigos were installed on all 6 panels that make up the array. And no, Tigo claims one could only put their units on only the panels affected by shade, they do not need to be on all panels.

You need to understand how this works: Panels in the sun have a low impedance producing energy potential (volts). Panels in the shade have a high impedance while not producing less energy potential. Because panels in an array are IN SERIES with each other, the high impedance of the shaded panels blocks the energy potential from the panels in the sun. It is like standing on a water hose with your foot. The rest of the hose has great potential to deliver water but little water is flowing because of the restriction under your foot. The restriction = shaded panels in an array.

Proper Tigo impedance matching would allow the potential energy from the sunny panels in the form of current (aka water in the hose) to flow through the restriction caused by the shaded panels. Well, nice claim and sales pitch but that does not work. Other panels are not boosted in Voltage, that concept is not correct. But anyway, every single panel in my failing array has Tigos mounted, hence even if your argument were correct, it does not change the conclusion.

When I impedance match the "shaded panels" with a by-pass cable (which is low impedance, low resistance), the inverter handles 4 panels in the array quite well (only ever swap parts with the array disabled! DANGER!!! both high voltages and serious arcing). Of course the voltage drops by two panels but the current remains similar to what one could expect. Hence the V x I (i.e., power) produced per panel is fine. The 4 remaining panels work well and the inverter can handle this lower voltage, confirming that Tigos are the issue.

As an example: With 6 panels I might get 165V and 4.5A in the string. With 4 panels I might get 125V and 4.2A in the string. The voltage drops, the energy potential drops, because of less panels. The current remains similar because the energy per panel that arrives at the inverter remains similar. Energy being V x I.

I've been trained and worked in electronics for 30 years. Perhaps I do have some understanding here. I worked on systems much more complex than solar energy systems.

See also the attached description of the Tigo testing.

 

[ncsolarelectric]

Tigos were installed on all 6 panels that make up the array. And no, Tigo claims one could only put their units on only the panels affected by shade, they do not need to be on all panels.

You need to understand how this works: Panels in the sun have a low impedance producing energy potential (volts). Panels in the shade have a high impedance while not producing less energy potential. Because panels in an array are IN SERIES with each other, the high impedance of the shaded panels blocks the energy potential from the panels in the sun. It is like standing on a water hose with your foot. The rest of the hose has great potential to deliver water but little water is flowing because of the restriction under your foot. The restriction = shaded panels in an array.

Proper Tigo impedance matching would allow the potential energy from the sunny panels in the form of current (aka water in the hose) to flow through the restriction caused by the shaded panels. Well, nice claim and sales pitch but that does not work. Other panels are not boosted in Voltage, that concept is not correct. But anyway, every single panel in my failing array has Tigos mounted, hence even if your argument were correct, it does not change the conclusion.

When I impedance match the "shaded panels" with a by-pass cable (which is low impedance, low resistance), the inverter handles 4 panels in the array quite well (only ever swap parts with the array disabled! DANGER!!! both high voltages and serious arcing). Of course the voltage drops by two panels but the current remains similar to what one could expect. Hence the V x I (i.e., power) produced per panel is fine. The 4 remaining panels work well and the inverter can handle this lower voltage, confirming that Tigos are the issue.

As an example: With 6 panels I might get 165V and 4.5A in the string. With 4 panels I might get 125V and 4.2A in the string. The voltage drops, the energy potential drops, because of less panels. The current remains similar because the energy per panel that arrives at the inverter remains similar. Energy being V x I.

I've been trained and worked in electronics for 30 years. Perhaps I do have some understanding here. I worked on systems much more complex than solar energy systems.

See also the attached description of the Tigo testing.

We have similar backgrounds. I have expertise in power supply and transformer design. I've never used the Tigo optimizers, and I agree, they need to boost or bypass the current to overcome the "pinched hose".
Okay, I see now. It's the opposite of what I was thinking. Rather than boost the voltage of all the rest, it has to buck the voltage on the shaded panel, lower voltage + higher current, to match the current that would normally flow. Maybe the panel is so badly shaded there is not enough energy to match the current even at 0V?

 

[Solar Guppy]

Maybe the panel is so badly shaded there is not enough energy to match the current even at 0V?

Correct, you have to have some production for the optimizer to function. From the description in the initial posts, it sound like there is enough shading that none of the strings ( within a single solar panel ) are unshaded.

Also the Tigo's work in full sun as well, matching current of all panels in the string. I personally wouldn't use them as putting panels in shade is a waste of money and for the cost, adding solar panels is a better return on the investment than having electronics to match the current.

 

[Strousy]

Re: you have to have some production for the optimizer to function
True, the Tigos are not externally powered which will result in design limitations........ But if they cannot work on the little power available in the shade, they cannot work, which is my experience. However, others have found that hooking the system up to the internet initiates a software update which makes them work properly. It doesn't appear to be a power issue but a software issue that Tigo appears to deny exists.

Re: putting panels in shade is a waste of money and for the cost
This is "shady" thinking..... "black and white" thinking. Shade comes and goes. The idea behind the Tigos is great: A little shade on some of the panels some of the time should not affect the total power production capacity while concurrently, these panels still contribute to achieving maximum power production. Imagine having one skinny tree casting a shadow moving from West to East over your array, only ever covering 10% of the array but this shade cutting back 90% of your power production because of the "kink in the hose" effect. Not everybody has the luxury of having 100% sunshine 100% of the time.

The design limitations come from trying to avoid the effects of the raised DC voltages one finds in solar arrays. Externally powering would move their design towards the Enphase units and make them more complex and in principle more fault prone.

 

[ncsolarelectric]

Given the choice between optimizers and microinverters, I'll choose microinverters every time. IMO, optimizers are just an added complexity when an inverter needs to be installed anyway. Microinverters are plug-n-play if all that is needed is grid-tied, without batteries. The current reliability of microinverters is extremely good.

 

[curto]

Re: you have to have some production for the optimizer to function
True, the Tigos are not externally powered which will result in design limitations........ But if they cannot work on the little power available in the shade, they cannot work, which is my experience. However, others have found that hooking the system up to the internet initiates a software update which makes them work properly. It doesn't appear to be a power issue but a software issue that Tigo appears to deny exists.

Re: putting panels in shade is a waste of money and for the cost
This is "shady" thinking..... "black and white" thinking. Shade comes and goes. The idea behind the Tigos is great: A little shade on some of the panels some of the time should not affect the total power production capacity while concurrently, these panels still contribute to achieving maximum power production. Imagine having one skinny tree casting a shadow moving from West to East over your array, only ever covering 10% of the array but this shade cutting back 90% of your power production because of the "kink in the hose" effect. Not everybody has the luxury of having 100% sunshine 100% of the time.

The design limitations come from trying to avoid the effects of the raised DC voltages one finds in solar arrays. Externally powering would move their design towards the Enphase units and make them more complex and in principle more fault prone.

Just wondering if you got anywhere on this one ? I got a loan kit from Jeff which was the TIGO TAP and comms module but we never hooked it up as Jeff and I could never get our schedules right. The one concern i have is that once the TIGOs are hooked up to a TAP they then enter rapid power down mode if they can not contact a TAP - apparently TIGO themselves have a way of updating their database to force an alternatve firmware that stops this action once you remove the TAP. Jeff was pretty puzzled as to why my panels exhibit the issues they do and swore it is not a TIGO problem. I am almost at a point of purchasing a TAP just to see if the panels will magically roar back to life.

Craig

 

[Hedges]

When a normal panel gets shaded, you can get full current at reduced voltage by enabling bypass diode.
When a half-cut panel gets shaded, depending on which of its 6 sections (wired 2p3s), you can get full voltage, half current.

Tigo may or may not process that optimally, selecting the lower voltage of 2p2s for 2/3 the wattage, vs. 3s voltage. At the higher voltage, all other panels in the string operate at half current (and slightly boosted voltage.)

It would be up to the inverter to drive voltage lower until it finds a higher local maxima at lower voltage, twice the current.

Need to see how Tigo behaves in this case. Did half-cut panels even exist when Tigo was born?

(I've bought some RSD, haven't used yet, never used optimizers. I'm a firm believer in series strings, bypass diodes, and good MPPT algorithms.)

 

[curto]

Yes thanks i am aware of the theory - but the previous poster in this thread and i are not experiencing anything like that - when the 1/2 cut panels are placed in Horizontal mode and shade creeps along the length of the panel then the TIGOs seem to do nothing - nor do the panels activate their bypass didoes which brings the whole string down to extremely low amps - the TIGO should attempt to boost the amps (not the voltage) of the affected panel to bring it in line with the whole string and enable maximum output - failing being able to do that it should then bypass the panels - it should not give up and leave whole string at a very low amperage. Similarly the panels by themselves do not appear to handle this very well either and do not attempt to bypass when both 1/2 cut sides are partially shaded.

I am looking at a panel now - 3 of them in a line with approx 1/2 the panels shaded from the top - and the whole string of 13 panels is putting out approx 500w - the other 10 panels are all in full sunlight.

When all panels are unshaded the string puts out 11amps at approx 420V

Craig

 

[zanydroid]

Have you tried bench testing a single panel to measure it’s voltage-power curve when partly shaded?

And based on the string length it looks like these half cut panels are internally arranged parallel rather than in series (former is more common but the latter is also an available config). There are also asymmetrically arranged parallel wired half cut modules (eg not all bypass sections have the same number of series cells)

Activation of the bypass diodes is a function of the MPPT and not the module. If you symmetrically shade half cut along long edge I don’t think the MPPT would be able to tell that it’s half-cut vs full cut.

I don’t know that the behavior of TIGO has been published to the extent that we know how it picks the operating point nor how it interacts with the MPPT scan

 

[curto]

thanks for the feedback - i had a long dicusssion with the head of the australian operation for TIGO who is a tech and he explained their behaviour is "meant" to be as i described above. I.e. they sense the voltage and amps of the string and if they can not raise the amps of the panel to meet that behaviour (so they do not bring down the whole string) then they are mean to bypass the panel entirely until they can.

This is not what we are seeing happen in real life and he could not understand why.

Craig

 

[JJT211]

thanks for the feedback - i had a long dicusssion with the head of the australian operation for TIGO who is a tech and he explained their behaviour is "meant" to be as i described above. I.e. they sense the voltage and amps of the string and if they can not raise the amps of the panel to meet that behaviour (so they do not bring down the whole string) then they are mean to bypass the panel entirely until they can.

This is not what we are seeing happen in real life and he could not understand why.

Craig

Well did you update the software?

Ive seen people complain about the TS4 having issues, and I get its a bummer having to buy the gateway for it to update the software. But are you saying after updating software its still not working correctly?

 

[AZ Solar Junkie]

Well did you update the software?

Ive seen people complain about the TS4 having issues, and I get its a bummer having to buy the gateway for it to update the software. But are you saying after updating software its still not working correctly?

Looks like @curto hasn’t been on the site for a few months now. Has anyone gotten to the bottom of these questions?

 

[zanydroid]

Looks like @curto hasn’t been on the site for a few months now. Has anyone gotten to the bottom of these questions?

I don't remember what TIGO thread this is, but I was on some more recent threads about standalone mode on TIGO specifically and on optimizers in general. Sadly there isn't the most streamlined benchtesting culture on the forum for one-off testing with high startup costs. The startup cost here being a CCA and TAP, so like $500, that also adds no value if you're not invested personally in using it. The MLPE itself is easy enough to get, as is a stand-in half cut panel.