Jump to content

Can a simple Axpert inverter assembly mistake be the reason for Error 08, 09, 51 etc. ?


Recommended Posts

Posted

After my Axpert inverter progressed from intermitted Code error 08s to nearly constant code 51s (regardless of weather, time of day or whether or not Solar and/or grid was supplied), I started to study the many related posts on this site. I wish to thank the following contributors for their valuable contributions, in no particular order,: @charlez @Maximus777 @Coulomb @APV @Chris Knipe @jaco de jongh (not sure how to tag him correctly).

What stopped me short of starting to replace mosfets and caps, were the amount of boards to remove just to get to them! I started weighing the costs of getting a new unit and developing my assembly skills when I made a discovery, quite by accident, that have allowed the unit to run today for 12 hours straight(!), through 2 load shedding sessions, intermittend cloud cover, without as much as a single beep.

I can't find who mentioned something in one of the many posts about reversing the fans, which I did not pay much attention to then and also because no one really reacted on it but I remembered it when I looked at the uncovered monster before me (still mounted and connected on the wall) and me very reluctant with a set of screw drivers in hand.

With only the front panel removed, I switched the unit on one more time and when the fans started, felt the wind blowing down, out of the unit, not up and in over the heatsinks.

Now I have never designed or sized heatsinks or any electronics cooling systems before, but I know hot air moves up and these fans are trying(?) to sucking the air down. That just seems counter productive.

I wanted to remove and turn the fans around but again, their screws are only accessible if you dissasemble many other parts first, so I sighed again.

Being Easter monday and not wanting to spend more time on the inverter, I left it at that, and then this morning I had an idea: Let me test this: I openend the front panel again, openend the plastic sheet to allow air in, and placed a floor standing fan at full speed blowing at it!

Wolla, the inverter still works without any error while it would have been tripping 20 times by now in the same time frame.

Can it simply be that the inverters are assembled incorrectly, with their fans blowing out (down) and not in (up), to cool the heatsinks?

 

My experimental setup (no, I don't have small children in the house anymore!):

image.thumb.jpeg.61d506775827b56fac35223542c9fdc9.jpeg

 

image.thumb.jpeg.4aecd77d98920bcd5070bec2e9c3e51f.jpeg

And yes, I have a list of error codes on the wall. I should have ticked all the ones I have encountered already...

So, this seems like a succesfull hack and worthwhile to get those fans out and reversed.

Posted
16 hours ago, PaulNaude01 said:

Can it simply be that the inverters are assembled incorrectly, with their fans blowing out (down) and not in (up), to cool the heatsinks?

In my Synrgj/Axpert MKS5 inverters the fans are blowing downwards too. No error message ever in almost 2 years of service. Try to leave just the terminals cover open to ease the air flow.

Posted

Hi @Beat

My inverter ran without any issue for about 4 years now and in some cases like in this post,

the issues also only started many years later. 

I don't really want to re-design the cooling for the supplier (they should do that) but there is clearly a problem and changing the cooling fan direction can only help in my opinion. The only downside I can spot is more dust that can accumulate over time, so maybe the inverter must be installed upside down 🤣

 

 

Posted

I note that this appears to be a clone; I can see the "Warning" sticker above the non-genuine-looking barcode label. So it's quite possible that they are using cheap fans that don't have enough air flow to adequately cool the inverter.

Inverting the fans should help some; hopefully it will help enough.

BTW, it's possible to invert the fans without opening the case: https://forums.aeva.asn.au/viewtopic.php?p=60086#p60086 should get you started. The clue is removing the comms board; it's a relatively easy job.

Of course, this is on a genuine Axpert.

Posted
19 hours ago, PaulNaude01 said:

  I've had my Axpert for the better part of 7, or 8 years now (when was the Eskom cockup?).

the issues also only started many years later. 

I don't really want to re-design the cooling for the supplier (they should do that) but there is clearly a problem and changing the cooling fan direction can only help in my opinion. The only downside I can spot is more dust that can accumulate over time, so maybe the inverter must be installed upside down

Perhaps the heat sink of critical components got covered up with dirt in all those years. Try to clean them, also the input air filters, if any.

Posted

I'm a bit doubtful that reversing the fan direction would have a profound effect on cooling efficiency. If using that extra floor standing fan has stopped the error codes it might be reducing the symptoms of another issue such as deteriorated thermal paste on a component or even a component that's producing extra heat because it's in the process of failing.

 

A thermal image camera can be a valuable tool to help resolve these types of issues.

  • 1 month later...
Posted

So I managed to borrow a thermal camera and here is a semi-overlay showing the hottest elements at the top of the inverter. Considering that the factory default fans are installed at the bottom, and blowing down, it makes sense that these components won't cool down. 

image.png.98861c3661c7e8bd86bb9f701fe2274e.png

Posted (edited)

When taking the fans out to reverse them, I discovered that the one fan did not rotate smoothly anymore so I decided to replace both of them with new fans (which also happen to have a higher RPM and flowrate).

Then something very strange jumped out of the wood work. After installing the new (also 4-wire fans), the new fans ran at top speed all the time and the noise was just too loud to ignore. Despite the noise, it bothered me that although the load on the inverter changed as normal, that the fans did not ever slow down.

image.thumb.png.f0c7dd56b7a6a36315cf514b944cecd7.png

I searched the internet for datasheets of both fans and could not find anything useful on either regarding the extra wires (speed control and speed measurement) to double check that these wires are indeed doing the same thing.

Since the new fans had a higher RPM spec, I though that maybe the controller on the inverter are simply tuned to the old fans and therefor let the new fans run too fast but I could not find anything to work with.

Now comes the scary bit.

After searching some more, I found a site explaining the functions of the blue and yellow wires of IBM PC motherboard fans and this correlated with what I could test on the new fans: If the blue wire is free (or have a high resistance to ground), the fan runs at full speed and when it is connected to ground (or has a low resistance to ground), the fan runs at a slow (minimum) speed. I measured the resistance range and found that at 2.5kOhm or higher, the new fans runs at maximum speed.

I tested this on the old fans and they didn't even run with 12 volt supplied, except for a quick jerk of blade movement when switched on.

So I set of to test what the inverter supplies to the blue wire (I had my suspicions) and indeed, the blue had a voltage (of 0.35V), NOT a resistance supplied! I fitted the old fans and indeed, without the blue wire, they didn't run and with it connected, they ran at a high (possibly max) speed.

So now my question: Why on earth would fans be sold without this information freely available, especially when they function totally different and void of clearly any standard?

And then more importantly, what would be the easiest way to convert a voltage to a resistance, to use the new fans? 

 

 

Edited by PaulNaude01
spelling and grammar mistakes
Posted
7 hours ago, PaulNaude01 said:

So I managed to borrow a thermal camera and here is a semi-overlay showing the hottest elements at the top of the inverter. Considering that the factory default fans are installed at the bottom, and blowing down, it makes sense that these components won't cool down

Hi Paul

I think you should keep in mind that with factory fans blowing down out of the panel it will create a low presure inside when the door is closed and suck colder air in from the side vents right next to the thermal hot spots you are showing on the camera.So the hot spots with the door open might not be the problem area.

I might be wrong thou.

Posted
On 2022/06/17 at 7:40 AM, Piper said:

suck colder air in from the side vents

You are correct as far as my experience is concerned. Something I learned from bitcoin mining in 2015 is that these types of PC fans are designed to pull air and not push it, they are actually quite useless at pushing air. The best configuration we found was a push-pull relationship where if you had one fan pushing air in then you needed another fan pulling air out the other side (this reduces pressure against the pushing fan) or chip temps just increased higher than they were with 1 fan pulling.

I would suggest against changing the fan configuration without adding another to pull out the top.

Posted
On 2022/06/17 at 12:53 AM, PaulNaude01 said:

blue wire

That's the pwm cable for speed control, different fans use different cable colours (I know it's crazy) and some don't have pwm, I don't remember all the details it was a long time ago that I dealt with fans, you may need to switch some of the wires by pulling out the connectors and switching them around. You might find some better info here: https://forums.aeva.asn.au/viewtopic.php?p=54866#p54866

Posted
On 2022/06/22 at 11:09 PM, jumper said:

You are correct as far as my experience is concerned. Something I learned from bitcoin mining in 2015 is that these types of PC fans are designed to pull air and not push it, they are actually quite useless at pushing air. The best configuration we found was a push-pull relationship where if you had one fan pushing air in then you needed another fan pulling air out the other side (this reduces pressure against the pushing fan) or chip temps just increased higher than they were with 1 fan pulling.

I would suggest against changing the fan configuration without adding another to pull out the top.

Thanks @jumper. I would like to look into the fan designs in more detail. My limited fluid flow knowledge says the mass flow before and after the fan must be the same. What I would like to investigate however is what happens further away before and after regarding the flow patterns which could make a difference in terms of the airflow patterns through the heat source area. In a closed horizontal pipe system, the air flow has to be identical on both sides otherwise mass conservation does not hold. What I do think makes a difference is gravity and with that, dust. In the pull down configuration, the heated air is less dense and wants to rise while the fan pulls it down. In the push configuration the air wants to rise and the fan is helping it on. Regarding dust, in the pull scenario the dust will happily enter the inverter and collect on anything in the way. In the push configuration, the only way dust can enter is if it goes up, against gravity and with less ability to collect inside. I wouldn't mind being proven wrong since I like to learn.

On 2022/06/22 at 11:20 PM, jumper said:

That's the pwm cable for speed control, different fans use different cable colours (I know it's crazy) and some don't have pwm, I don't remember all the details it was a long time ago that I dealt with fans, you may need to switch some of the wires by pulling out the connectors and switching them around. You might find some better info here: https://forums.aeva.asn.au/viewtopic.php?p=54866#p54866

Yes this was a really weird discovery. Apparently the correct terminology is that the one fan needs a pull up resistance and the other a pull down resistance. Fortunately, it was a relatively easy fix with an Arduino laying around to monitor the Inverter signal and provide the opposite to the fan. It was only a few lines of code. My inverter now runs silently (my laptop fan is now more audible!) and the inverter hasn't had a single error 51 since. 

In all honesty, I think the higher RPM (and by implication the CFM rating) on the new fans is far more important than the direction of air flow, but every bit should help.

Posted
On 2022/06/30 at 10:37 AM, PaulNaude01 said:

In all honesty, I think the higher RPM (and by implication the CFM rating) on the new fans is far more important than the direction of air flow, but every bit should help.

I must agree with you there, the increased CFM will make the most difference to your temps and those sunon fans are nice and quiet compared to most cheap oem fans, glad everything is running smoothly since.

As for dust, I think the amount of dust is more a function of the amount of air moved through the machine (and of course environment etc, which remains constant here) than gravity (as the air is moving), so you might be getting more dust with increased cfm, but not sure if dust is a problem for you, it is for me (on a farm), you'll see in the pic below.

As for flow direction, I would stick to the original design for a couple of reasons. I think there is a temp sensor under the big transformer (TX2?) between the 2 big heat sinks and it is located to measure the temp after the air moves over the transformer and changing air flow direction will alter that reading... pretty sure I read this on the Aus blog. Of course if you have increased air flow and keep the temps lower then this probably won't be a problem for you.

I've also had my machine apart recently as it has given up completely and when taking out the main board I noticed some weather tape in the back of the casing I thought was for support, but later realised it is used to try to force air from the air holes at the top, behind the unit, against the wall (you might not have these) and bring it around the back of the board over the back of the upper components and then around the side of the board in the middle and to the bottom out through the fans. Altering the flow direction would probably mess wit this. As mentioned you might not have this on yours as I think I have a newer version with larger SCC, but also from TSP.

The other thing was, the only obvious damage I can find on my machine is bulging caps at TX9 which is right at the top of the main board which would logically be the hottest part due to heat rising, so maybe they want the air to move in the top first and over these parts and then over the parts with massive heat sinks, changing direction means you will move already heated air up to the top, possibly causing hotter temps up there. Logic also says that if it is sucking in from the top it is also sucking in the hotter air from the top of the room rather than colder air from below, so there's that too.

Some pics for interest sake:

Bulging caps also showing the holes that pull air in behind the board:

image.jpeg.5d9adf1bfcc2d7fafbe945fa5aa3944a.jpeg

Dust pulled in behind the mainboard on to resistors:

image.jpeg.6d16b634df8c07b954293cdd5dae67f4.jpeg

Anyway glad all is well with your machine, not so much for mine.

Posted

Thanks @jumper

You make some very valid points. I didn't think about the location of hot components and the temp sensor. That could surely make an important difference.

I have also managed to take a few thermal images (the overlay has a slight vertical offset)

image.jpeg.2f5549821dbbd165000054b60160babb.jpeg

Apart from the coil, the other hot (visible) components are these:

image.thumb.png.de01470463fc18c8a607f942f7dc669a.png

Regards

Posted

Thanks Paul, those images make things a lot clearer why the air is sucked in at the top, with that big coil sitting right at the vent and that very hot transformer up there. I also think the clear plastic 'wind tunnel' is quite important to pull air in over the big transformer in the middle and the igbts under the big heatsinks. I'm pretty sure the manufacturers would have left it out if they could have, seeing as how they cut costs on the oem fans.

I reckon the best mod is upgraded fans and the second is what seems the silliest at first: mounting it upside down 🤣🤔

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.

×
×
  • Create New...