Jump to content

Kodak MKSIII and DC input protection help needed.


Recommended Posts

Hi, awesome forum this .

Disclaimer : absolute beginner.

I am busy test benching 5kw inverter and 2 x  5 x 465w panels. Panels are on ground but I have cloned my roof tilt angle and direction.

I have single requirement, that is to augment solar energy with mains in order to reduce Eskom consumption. No batteries. Loads are limited to geyser and swimming pool pump for now.

When mains is not available, and little sun like today here in Germiston, the inverter will limp with the small amount of energy that comes from solar, it actually adjust AC output voltage to xfer as much input power to output. My load is old oven plates for testing, but I don't like this lower AC voltage for say swimming pool pump. Can this be disabled? Or should I detect this and cut it out with  external device. Inverter is in SUB mode.

 

Also, my absolute max open circuited DC voltage is 250 volt DC . At the moment I have worked a MOV arrangement on the fuse to start clamping at 428v . That would allow me not to invite the inverter to clamp.

But commercial offerings is in the order of 600v and 1000v DC, is this fine? Also they pretty expensive, so I want buy right.

 

Thanking you.

 

Link to comment
Share on other sites

On 2021/08/13 at 12:54 PM, 87 Dream said:

Hi Dropkick

Disclaimer: The most knowledgeable of people on this forum all started off as absolute beginners at some point🤣 Kudos for starting!!

I am more familiar with other Axpert models namely the MKSll & the larger Axpert MAX. Can you please post us the sticker on the side of the inverter to assist you with the Solar Voltage questions you have. This sticker has all the technical specs for setting up the inverter.

Very smart of you to run the inverter in this hybrid way without batteries. It's like running a grid tied system with AC coupling. There is on other models a way to set the output Voltage & options range between 220,30,40. There is also another menu setting that enables you to choose between UPS mode or APL or appliance mode.

I unfortunately do not have access to a Axpert MKSIII manual. However, the MKS range seems to be standard across the range. 

The Menu options are found in:

Setting 3: APL or UPS

Setting 8: Output voltage 

Maybe also found under setting 10??

Not so sure. But the options will be as above 220, 230, 240V.

Enjoy the new setup & the adventure 👏

 

Hi 87, thank you so much for inspiring words from a guru. Have been following you in coloumb for a while and got some really good insight.

I am trying to following the manual as much I  can , learning curve still. But I will try those inputs of yours soon as I can.  Please find below the requested image of inverter.

 

Regards dropkick

 

16290148511401516213331093512495.jpg

Link to comment
Share on other sites

Yesterday I have performed a load test of a different kind. That is 800w pool pump over plus minus 25m cable.

 

Very surprised and happy with results. Added an Ellies 8/20 13500 amp surge protector to the load side, and then I threw the kitchen sink at the inverter,including , inter alias , abrupt interruption of load, abrupt interruption of incoming mains. Very happy.

 

Also , because if lack of sun yesterday , I observed power blending at work. At times 380w available from solar , and the inverter imported the balance from Eskom. Exactly what I needed. Magic.

Link to comment
Share on other sites

41 minutes ago, dropkick said:

Sadly , I do not see incoming mains current as a monitoring parameter on serial or WiFi watchtower.

I think you mean WatchPower. Yes, this is a limitation of these inverter-chargers; there are sensors only on AC-out (load) and inverter out. Nothing on AC-in or the battery. So you need a current shunt to measure battery current, then estimate AC-in power from conservation of power and using an estimate for the inverter (DC-AC converter) efficiency. One more current transformer would have been very welcome.

Edited by Coulomb
Link to comment
Share on other sites

2 hours ago, Coulomb said:

I think you mean WatchPower. Yes, this is a limitation of these inverter-chargers; there are sensors only on AC-out (load) and inverter out. Nothing on AC-in or the battery. So you need a current shunt to measure battery current, then estimate AC-in power from conservation of power and using an estimate for the inverter (DC-AC converter) efficiency. One more current transformer would have been very welcome.

Thank you , it confirms that I will not unneccessary  duplicate resources if I employ current xformer and RMS converter, and to some extent , as a semi-retired light-current engineer, I feel worthy again! 

For my configuration it is important , as I need a power blending percentage monitor , which gives instantaneous measure of what energy I am importing from Eskom , in relation to load. Or in other words , to what degree is solar helping to pay the bill . 

Link to comment
Share on other sites

2 hours ago, 87 Dream said:

🤣🤣🤣This had me in stitches laughing 🤣

Glad you reached a happy place with your setup. Can you explain the use of a MOV in your Solar panel setup? I get that you want to get the panels to within the MPPT range & don't want to exceed the 430V but how can this MOV help you given your current PV capacity?

I have zero experience using a MOV & actually want to learn about how it works from people whom incorporate them.

87

Yes, my Vmp is 170vdc  and max open circuit is 250V DC. Because the axpert range is 450V , I reckon they clamping at 500v upwards. So knowing that I will not exceed 250v , I want my protection to be ideally 300v upwards , so I am going to suck up the garbage surges before it even enters the surge protection stage of the machine itself.   To ensure that ,  I need to clamp between 300v and 450v .

Unfortunately , as good as MOV s are, they possess a sloppy characteristic in terms of where they will start clamping in the real world.  For instance a 280v MOV may not be suitable for my 250v , as the tolerance of the MOV is too big. Currently I am using two back to back epcos 20mm 115v movs , which combined will clamp at 328v . And of course they are placed between the DC fuses and the inverter, to ensure the fuse will protect the MOV  and inverter when there is sustained surge energy.

Of course a number of assumptions is  made by me, experienced users must please correct me where I am wrong:

1 the PV output is regarded as floating, ie. No earthing on any PV leg, only PV frames are earthed. Therefore no protection required to earth. The movs employed only serves to protect the differential surges between plus and minus feed conductors.  

2. The leakage current and capacitance disadvantage of MOV is not applicable here , so no need for gas arresting assistance to block leakages to earth. Hence the need for a simple MOV only arrangement.

 

Kind regards

 

Link to comment
Share on other sites

21 minutes ago, 87 Dream said:

this is the equivalent of the Axpert MKSll. 

Actually, with the "a" on the end (OGX 5.48a), it's equivalent to an Axpert MKS III, which is still not documented on the Voltronic Power website, even though the Axpert MKS IV is. The MKS III is basically an MKS II with a removable display, like an Axpert King or VM III, but not the round front like an MKS IV.

I updated the Voltronic Zoo post with the extra info just today.

Edit: The Axpert MKS III is also known as a PIP-5048MGX, and of course many other names.

Edited by Coulomb
Link to comment
Share on other sites

6 minutes ago, Coulomb said:

Actually, with the "a" on the end (OGX 5.48a), it's equivalent to an Axpert MKS III, which is still not documented on the Voltronic Power website, even though the Axpert MKS IV is. The MKS III is basically an MKS II with a removable display, like an Axpert King or VM III, but not the round front like an MKS IV.

I updated the Voltronic Zoo post with the extra info just today.

Edit: The Axpert MKS III is also known as a PIP-5048MGX, and of course many other names.

Very valuable and interesting info.

So if I read correctly, then if one is ok with no parraleling option, as well as lower charging current, then the Kodak  vmiii is just as good as mksiii. And if so then vmiii is a lot of  bang for your buck. I see here on the forum shop it's going for like 7700 rand.

Link to comment
Share on other sites

36 minutes ago, 87 Dream said:

Took a quick look at your model number under Kodak & this is the equivalent of the Axpert MKSll. 

One of the first Axpert Models to be released with the higher 450V volt Solar input voltage.

To answer your question above about grounding: The racking/ mounting systems are usually made out of an electrical conductive material like Aluminium. The entire frame of the Solar cells themselves are also made out of aluminium. So yes, grounding is mandatory required by the SANS code in South Africa. On the PV side fuses & surge protection are also mandatory under the code for obvious reasons. This is my limited understanding as an installer at an installation site. The protections are there as required. Now, how the magic will work when the Paw Paw hits the fan is maybe for the guys in white long coats & spectacles and insurance companies with their actuaries😂🤓

I really cannot see anyone else using a MOV so honestly you are the first to mention it. 

87

 

 

 

 

Thank you, but can you confirm that while the Aluminum  frame needs to grounded, the DC feeds are not grounded at any end, ie. Floating?

On the note of MOV , I would dearly like to think that the DC surge arrestors available commercially also employ movs . If not can anyone confirm what technology is used.?

 

Kind regards

Link to comment
Share on other sites

7 minutes ago, dropkick said:

can you confirm that while the Aluminum  frame needs to grounded, the DC feeds are not grounded at any end, ie. Floating?

The PV input terminals are far from floating, they have lethal voltage to earth even if the panel voltage is low. It's basically half wave rectified mains with the array DC voltage added to one side. So you must not ground them or touch them, but they are not floating. 

Link to comment
Share on other sites

1 minute ago, 87 Dream said:

In other words is the panel mounts grounding part of this earthing process ?

No, the panel frames make essentially no connection to the PV terminals.

The weird half rectified mains comes from the fact that the SCC in these high voltage models is an unisolated boost converter connected to the DC bus. The DC bus is used to generate the 230 V output, via a full bridge arrangement. The neutral output finds itself at close to earth potential. The negative end of the DC bus is connected variously to the active or neutral output depending on the phase of the 50Hz output. There is also PWM happening with smoothing, so that's where the half mains sine waves come from.

When it rains and there is leakage to earth from some types of panels, all sorts of problems can occur, e.g. the DC bus voltage can become too high, resulting in a fault code. There could also be tripping of earth leakage devices. 

Link to comment
Share on other sites

@Coulomb @87 Dream 

 

Curious Case of Serious Energy Waste , or is it?

Hi guys , sorry for so many questions , but  I have noticed a rather peculiar phenomena whilst testing. As you will recall , I have attached ammeter to the inverter grid input on my test jig.

1. Before the inverter is turned on , there a small quiescent current  , which is normal, id say.

2. When I switch the inverter on , with zero load , the ammeter reads 1.9A . Huh? That's 427W of power ! . Is it coming in or going out? What is it?

3.  That unaccounted 1.9Amps   energy diminishes and vanishes as I introduce and increase load. pump , 2kw kettle , 2kw  kettle and pump etc.

4. Remove grid , solar takes on all 800wats of pump power. pretty efficient transfer of energy from solar to load  = 95.2%.

 

This is not random behavior , pretty systematic / deterministic .  I am battling with readout sequences on the kodak screen , so I am going to introduce a logger by next week in order to provide detailed logs.

PS :

I have stumbled upon this Spanish thread  where the MAX is reviewed (I think this info also applicable to the VMIII and MKSIII). It would appear that energy are released onto grid when the load cannot handle it . This stuff is really above my paygrade.

https://www.solarweb.net/forosolar/fotovoltaica-sistemas-aislados-la-red/48226-voltronic-axpert-max-review-profundidad-2.html

In SUB, that is, when it is coupled with the network to add with it towards the loads, it also turns back (discharges) a lot of active power. We are not talking about a few watts .... In the following graph you will see it clearly: The discharge measurements are made with an EASTRON SDM230 network monitor connected to the AC INPUT of the MAX. The other active power measurement is with another EASTRON SDM230 connected to the AC OUTPUT of the MAX. Solar power measurements come from the MAX. It seems to me that it pours a lot into SUB. Too. In SBU it does not pour anything since it does not connect with the network.

[Voltronic] AXPERT MAX in-depth review-vertidomaxsub.jpg

 

Is this what I fear it is , grid tied issues?

 

Kind Regards

dropkick

 

 

 

Link to comment
Share on other sites

At the same time realized yesterday a simple characteristic of AC , that is ,  because of alternating current flow , we cannot detect the direction of power flow  without calculating the vector relationship between  current and voltage. Damn , this solar thing is getting complicated . 

 

But to see the sun operating my pool pump , whilst my middle finger is pointing towards Escom HQ , makes all effort worth it !!!

Link to comment
Share on other sites

On 2021/08/16 at 1:03 AM, dropkick said:

So if I read correctly, then if one is ok with no parraleling option, as well as lower charging current, then the Kodak  vmiii is just as good as mksiii. And if so then vmiii is a lot of  bang for your buck.

There are a few shortcuts with the VM models, in particular the fan control is crude, and several owners have complained about the noise.

I thought that they maybe don't have the plug-in control card, putting all that logic on the main board. But lately I see evidence that seems to contradict that.

It's also one of the high voltage SCC models, but they all are apart from the MKS and King models.

So given those limitations, it is quite good value, as it was designed to be.

Link to comment
Share on other sites

6 hours ago, dropkick said:

When I switch the inverter on , with zero load , the ammeter reads 1.9A . Huh? That's 427W of power ! . Is it coming in or going out?

No, that's 427VA of apparent power. As well as going in or out, it could be averaging to zero, if the current and voltage are 90° apart.

But I think I worked out that the capacitor in the LC filter is likely supplying only about 50VA, so most of this is likely real power, possibly charging the battery.

Per 87 Dream above, Tejota is the expert on this model, and it seems that firmware updates will fix it. Hopefully, the fix is rolled into the latest non-experimental firmware.

Edited by Coulomb
Link to comment
Share on other sites

1 hour ago, Coulomb said:

No, that's 427VA of apparent power. As well as going in or out, it could be averaging to zero, if the current and voltage are 90° apart.

But I think I worked out that the capacitor in the LC filter is likely supplying only about 50VA, so most of this is likely real power, possibly charging the battery.

Thank you for your valuable response, Coloumb .

I have no battery connected, so only grid,solar and load.

I'll tell what I will do Tommorow . I will connect both voltage and current( via CT) and measure the phase relationships on two channel DSO. That will  give me a measure of (a) how apparent the power is , and (b) direction of power transfer . 

 

 

Edited by dropkick
Spelling
Link to comment
Share on other sites

4 hours ago, 87 Dream said:

Hi Dropkick

I was just going to point that out to you that you cannot say which direction the current is going with AC 🤣

Watched a very interesting video in Andy's off grid garage on a very similar topic.

I am familiar with the Spanish forum & I did look a lot into this before choosing the MAX as a viable option. All I can say is that I have had to upgrade the firmware to an experimental version which is 90.19 & I have never seen the machine pump energy into the grid. 

I did not even bother with the stock firmware of 45.?? Because after the advice of these Spanish mafia in which they also have an Axpert "Sensei" like Coulomb, Tejota, under his advice after the upgraded firmware it fixes this issue. I have been using Solar Assistant in all my Installations & cannot say that I have logged anything that resembles grid feed. Also they have prepaid meters & these would go into fault mode should a back feed occur. However, I have also taken precautions as to the sizing of the PV array etc. 

However, you are quite correct in say the VMIII & MKSlll will share these traits as the tech & hardware are similar in architecture. 

Maybe @Coulomb will have more light to shed especially on the systems guts & internals. However, if I remember a while back he mentioned he hasn't had test subject to dissect. 

Anyhow thats the whole point. Eskom doesn't get more of your hardworking $$$.

87

 

 

 

Thank you 87 , for valuable response. I have been so far avoiding the inevitable , ie. Flashing new firmware. But I suppose that going forward, I'll have to learn to do that sooner than later.

Link to comment
Share on other sites

When it rains and there is leakage to earth from some types of panels, all sorts of problems can occur, e.g. the DC bus voltage can become too high, resulting in a fault code. There could also be tripping of earth leakage devices. 

 

I had this problem and did ask for help with this, but got now reply. I have installed 2x 25A blocking diodes in between the panels and the fuses and the problem is gone. 

I was wondering if the MOV's are not the same as installing surge protectors on the DC line, I think the surge protectors has MOV's in them.??

Link to comment
Share on other sites

3 hours ago, Krokkedil said:

When it rains and there is leakage to earth from some types of panels, all sorts of problems can occur, e.g. the DC bus voltage can become too high, resulting in a fault code. There could also be tripping of earth leakage devices. 

 

I had this problem and did ask for help with this, but got now reply. I have installed 2x 25A blocking diodes in between the panels and the fuses and the problem is gone. 

I was wondering if the MOV's are not the same as installing surge protectors on the DC line, I think the surge protectors has MOV's in them.??

Good to know that thank you, I think it will not hurt to insert them anyways, so I put them on my installation BOM. 

 

Yes , unless someone correct me there will be movs in DC surge devices.  Maybe gas arrestors as well , but they are used invariably to block leakage to earth  from the movs.    Maybe a bit of inductors as well to smooth rate of current change. But I cannot fathom why a good DC arrestor must cost in the order of 1000 rands. Because my take on them is to replace them every 2 years (consumables) , in order to reduce the probabilty of a faulty open circuit MOV ( bad undetected failure mode )  not being able to provide defence in the next round. We have learned these lessons in the railway , where qualative measures needs to prevail over quantative measures ( where you cannot measure a failure).

If COC is going to require me to put on a commercial arrestor, then fine , but I will put my own devices also in place .

 

Oh and I think your diode steering has another advantage , that is blocking reverse voltages in panels that are not exactly same output. Though I do think there is some block in the panel moulds themselves. But I like the idea , plus you can measure individual panel voltages at the combiner.

 

Link to comment
Share on other sites

@Coulomb, @87 Dream

 

Ok so today setup a  a DSO to check vector relations between current and voltage on the grid incoming. By setting it up I probably violated 0.6 COC rules per square inch of layout.😁 Note to myself : listen to people on this forum, be carefull, be safe, specially ⚠️⚠️that blikemse High tension solar DC.!!!

 

Added a pure resistive load , kettle, drawing 8.5 amps and checked voltage and current  nearly 0 degrees difference . Then removed load and checked the infamous 1.9 amp flow. Interesting the phase shift nearly approaches 90 degrees , which infers 100 % virtual power ( powder puff!) . Cannot really tell if current was lagging or advancing. But I suppose it was capacitive if anything else. So coulomb , you were spot on , near pure apparent power . So I am stepping down on this one.

 

lessons learned:  one thing I want in my future solar toolbox is an instantaneous real and apparent power monitor.  Yes please, let's leave nothing to the imagination. Maybe the problem with us engineers is we pedantic, and as a result we don't sleep well. ! But hey that's how we roll.

so no it's not power exporting. I have another discussion regarding blending , but will leave it for another day.

 

 

Link to comment
Share on other sites

1 hour ago, dropkick said:

Oh and I think your diode steering has another advantage , that is blocking reverse voltages in panels that are not exactly same output. Though I do think there is some block in the panel moulds themselves.

No, these days the diodes in the panels are to prevent each section (often a third or a half of the string of cells in one panel) from reverse power flow. They are in parallel with the string of cells. In normal operation, these diodes would be reverse biased (not conducting), and only start conducting in the case of shading causing reverse power flow into that string of cells. I think that @Krokkedil's diodes will be in series with the whole string (correct me if I'm wrong, please). For reverse cell protection, you also need 2 or 3 (possibly more) diodes per panel, not just two.

Link to comment
Share on other sites

1 minute ago, dropkick said:

Another note to self: that power monitor must indicate flow direction as well, thanks.

Was always looking for serious application for the raspberry Pi pico embedded controller. The moment has cometh. With a nice OLED readout. Alas , pity there is no built in WiFi on the pico. But again that's just for yuppies.

Link to comment
Share on other sites

22 hours ago, Coulomb said:

But I think I worked out that the capacitor in the LC filter is likely supplying only about 50VA

I was wrong!  🥺

The capacitor on 5 kVA models is 20 μF, so a 7.2 kVA model would be expected to use about 7.2/5 x 20 = 28.8 μF, closest preferred value is 27 μF. The capacitive reactance at 50 Hz is 1/(2πfC) = 1/(2·3.14·50·27x10⁻⁶) = 1/0.0085 = 118 Ω reactive. So the expected current at 225 V would be 225 / 118 = 1.9 A. Whaddayaknow 😯

This is good, because it also avoids a hefty fine for breaking the law of conservation of energy (and power), since unbeknown to me you don't have a battery or anywhere else for some 427 W to go 👮‍♂️!

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...