quality inverter?

[Justin Laureltec]

What breed is that? Pos and neg terminals are a bit close! They need both to be moved to their alternative bolts.

This is a Victron MultiPlus 12/3000/120-50. As for the terminals, well, that air gap is Victron-tested to 500V and if my battery ever throws more than that out toward the inverter, both my MRBF fuse at the battery and my ANL fuse before the MultiPlus are going to blow well before any arcing occurs... I think I'll be alright

 

[b.james]

If your spanner rotates the lug towards the other side you will get a chance to test those fuses out for quality won't you ?.

 

[Justin Laureltec]

If your spanner rotates the lug towards the other side you will get a chance to test those fuses out for quality won't you ?.

Hahaha yep... I literally trust Bussman with my life.

 

[Justin Laureltec]

If your spanner rotates the lug towards the other side you will get a chance to test those fuses out for quality won't you ?.

In all seriousness, though, you're assuming that I don't know the first three safety rules of any electrical work:
1. Never work on an energized circuit
2. Use insulated tools
3. Pay attention to what you're doing

Besides that, the MultiPlus has connections for 2 banks, thus why those posts are there to begin with. I could move my current connections, but then when I go to connect my second bank I'd be right back where I started anyway, so... I prefer to just pay attention to what I'm doing and connect things correctly from the start and that way I'll never *need* to test those Bussman components.
I appreciate your concern, and can understand your reasoning, but this is a manufacturer-approved connection that is only dangerous if someone doesn't know exactly what they're doing... As is the case with most electrical connections. Electrical work is only dangerous if you're sloppy.

 

[Samueru]

Probably won't be popular saying it but it comes down to the model. There are teardowns of Samlex units on youtube that show they aren't any necessarily anything flash. That goes right down to using what ever random FETs they could get their hands on at the time. Closely matched FETs matter, and I'd guarantee that they aren't.

This video right?

 

[gnubie]

That's one of them. Even if you ignore 99% of the video you can see for yourself that they have used different brand components. That by itself isn't necessarily an indicator that anything is wrong as they could well have had someone (or some automation) testing each FET to ensure it was closely enough matched with its partners to ensure good load sharing, but in a market where low end products are manufactured in China and every cent is tracked to maintain margins, do you think it's likely this degree of testing has been done? I sure don't.

If the models I'm making sweeping statements about were quality they would at least go to the expense of using parts from the same manufacturer and same production batch. That's no guarantee of device matching when it comes to cheaply made components, but it's a long step up from using random grab bag parts.

 

[blueyedevil54]

Probably won't be popular saying it but it comes down to the model. There are teardowns of Samlex units on youtube that show they aren't any necessarily anything flash. That goes right down to using what ever random FETs they could get their hands on at the time. Closely matched FETs matter, and I'd guarantee that they aren't.

Respectfully, Yikes! I think you are saying something that has no basis in reality.
Offshore manufacturing plants can be faulted with many things, but to categorize them as "using whatever random FETS that they can get their hands on ..." is just wrong. It would cost manufacturing more to have "random parts" as their discount comes from using the same part multiple times.
Additionally, if FETs are used in parallel and they are all the same part number, there is no benefit of "closely matching" them.
They are an on/off switch with a documented resistance (loss) when on.
Closely matching transistors harkins back to the dark ages of bipolar transistors.
John

 

[gnubie]

It is extremely common for cheap Chinese products to use salvaged parts. That scope covers using salvaged wall warts all the way down to salvaged capacitors. There are actual workshops full of people pulling components off boards, binning them, and taping them. It happens.

The on resistance of FETs varies. It's a simple fact of life. The easy way to get closely matched on resistances is to use devices from the same production batch. When dealing with high currents it is important that the on resistance is closely matched otherwise one FET will carry considerably more of the current than the other, potentially exceeding its ratings leading to failure, especially considering the first point.

Cheapie manufacturer simply don't care if their product goes bang after a month.

It's not just AC inverters that demonstrate a random mix of components. Random mismatched FETs are usually found in the boost converters in high power car audio amplifiers too, and yes, they go bang for that very same reason.

 

[Steve_S]

Just some input from a Working Solar Installations:

My current primary solar system uses a SAMLEX EVO-4024 Inverter/Charger. MADE IN TAIWAN ! (not China or USA). REF: https://www.samlexamerica.com/products/ProductDetail.aspx?pid=575 The documentation is overdone and confusing because too much info, BUT very unlike Chinese documented stuff which is so thin you can see through it. Their specs are bang on and accurate and performs very well. Had a few minor issues and their Tech Support jumped right in and perfect ! IT MUST BE NOTED that their Industrial Inverters and others are NOT all Pure Sine which will affect whatever testing you are doing., Apples to Apples !

My Secondary system uses a "Cheapo Chinese" Inverter/Charger (paid like $550 USD - 2015) made by Yiyen APC-3024 Combi REF: https://www.yiyen.com/product/apc-series-pure-sine-wave-inverter-charger/ While not as efficient or feature rich (really minimal) the unit has worked flawlessly for 3 years now and no real problems of any sort (only issue is remote only works up to 30') BTW: AIMS Inverters are YIYEN Inverters packaged for AIMS USA... compare the price and be seated when you do. Some other "Rebrands" come from Sigineer (China) which is a higher end OEM. REF: https://www.sigineer.com/

BTW a Point of Interest. Are you folks aware that DeWalt, Black & Decker, Porter Cable are all owned by the same International Holding Company ! Did you know that the same engineers design all the equipment ? These products are all made in China by "Sub-Contractors" who make the products per the specs of the designers & engineers (who are American BTW) and yet B&D is now Crap Homeowner toss away tools, Porter Cable is good mid range and DeWalt has consumer grade & high commercial grades. `The Subcontractor is not at fault for a poor design, they only bid on the spec provided to produce whatever. If you have a crappy spec with junk parts, that's what you get and the price reflects such.

 

[Steve_S]

View attachment 887

IF anyone thinks that is a snug & tight assembly, look inside a Midnite Classic Charge controller ! An ant can't squeeze in there !

 

[Justin Laureltec]

Respectfully, Yikes! I think you are saying something that has no basis in reality.
Offshore manufacturing plants can be faulted with many things, but to categorize them as "using whatever random FETS that they can get their hands on ..." is just wrong. It would cost manufacturing more to have "random parts" as their discount comes from using the same part multiple times.
Additionally, if FETs are used in parallel and they are all the same part number, there is no benefit of "closely matching" them.
They are an on/off switch with a documented resistance (loss) when on.
Closely matching transistors harkins back to the dark ages of bipolar transistors.
John

I think perhaps a bit of hyperbole can be excused in the point @gnubie was making; "random mismatched parts" doesn't necessarily mean that random parts will be in a single device, because indeed, most parts purchased by a manufacturer will be in million-count packaging... but a very important consideration here, particularly when discussing device reliability, is matching components across devices of the same make/model. Yes, perhaps all the FETs inside device A match, but if device B, of the same make and model as device A, does not have the same FETs, then the reliability -not to mention the real-world tech specs- of both devices become highly questionable. A prime example of this is Renogy... Renogy is not a manufacturer, it's a branding company that buys up random warehouse overstock and discontinued devices from solar panels to controllers to inverters to batteries, and slaps its name on them and sells them. And hey, some of those components it purchased and rebranded are great.... and some are utter crap. The problem is, there's no way to tell which is which until something fails, because there's no guarantee of -or even convincing attempt at- consistency.
You can see an example of this in Renogy's battery specs and charge recommendations... try to get a straight answer out of them from one day to the next of what their max current rating is, or proper charging voltage/current, and you'll never be able to... because they themselves have no idea what the specs are because they didn't design or manufacture the cells or, more importantly, the BMS.
For our customers, for whom proper operation of their equipment can literally be a mater of life or death, consistency and reliability is absolutely critical, and that's why we would never distribute Renogy, or Xantrex, or Aims, or any of the truly off-brand components like the knockoff Blue Sea breakers that I see being touted everywhere because oh, they're half the cost!! Yes... but you have no idea how -or if- it will perform, and considering that a malfunctioning component can easily cause a fire, or electrocution, or strand you in the middle of nowhere... you have to ask yourself whether that little bit of money you saved was really worth the ever-present doubt that you now live with.

 

[blueyedevil54]

This video right?

Nope,
As far as matched parts, it's likely there are FETs and Diodes mounted to the same heatsink. They have different part numbers.
As far as his statement that the FETs cascade fail is wrong too. Once a FET fails it takes the voltage rail down since it fails shorted.
The other FETs don't see voltage if the blown FET shorted out the supply. Then the fuse opens and... where's the cascading failure?
This fellow presents information as if he knows what he's talking about, but all I hear is BS.
John

 

[gnubie]

I'm guessing you've never seen a device with multiple popped paralleled FETs. It's a quite common mode of failure in this cheap stuff. As far as pulling down the rail goes, take a look at the design of these things. The fuses, if they are even present, are on the DC input, before some nice beefy reservoir capacitors. Those capacitors can, and do, provide nice high current flows. It's not at all unusual to see FETs blown completely open.

I'm going to leave it there because I think it's fairly clear that there will be no convincing you of anything different, and that's fine. This forum welcomes all views.

 

[blueyedevil54]

Likewise!

 

[Justin Laureltec]

Nope,
As far as matched parts, it's likely there are FETs and Diodes mounted to the same heatsink. They have different part numbers.
As far as his statement that the FETs cascade fail is wrong too. Once a FET fails it takes the voltage rail down since it fails shorted.
The other FETs don't see voltage if the blown FET shorted out the supply. Then the fuse opens and... where's the cascading failure?
This fellow presents information as if he knows what he's talking about, but all I hear is BS.
John

For the record, FETs don't necessarily fail open; they also fail closed and and half-closed as well (I realize that this is using mechanical-relay language, but the principle remains the same).
As evidence, look at all the people -on this forum, even- who have incorrectly connected their Victron BatteryProtects in between inverters or inverter/chargers and now have BPs that don't disconnect their loads (because the FETs failed closed), don't connect their loads (because the FETs failed open), or only provide a percentage of their voltage/current throughput (because the FETs failed half-closed and are now turning that shedding current into heat and indiscriminately melting things around them).

 

[blueyedevil54]

Actually they fail shorted, or not able to turn off.
What happens next is variable. The failure plasma can blow out enough material that there is not enough conduction to keep the short.
Or there can be a big carbon crater with smoke. And after the failure when power gets turned back on there can be the resistance of the debris.
The FET works if it turns on and turns off.
John

 

[Justin Laureltec]

Actually they fail shorted, or not able to turn off.
What happens next is variable. The failure plasma can blow out enough material that there is not enough conduction to keep the short.
Or there can be a big carbon crater with smoke. And after the failure when power gets turned back on there can be the resistance of the debris.
The FET works if it turns on and turns off.
John

If you're "actually" familiar with the way a FET works, then surely you're aware that the gate can weld open, closed, or half-open. I'd say you're entitled to your opinion about how they work except that it's not a matter of opinion. Source/gate/drain, that's how a FET works. If the gate welds open, the FET passes no current. If the gate welds closed, the FET won't stop passing current. If the gate welds half-open, the FET becomes a current-restricting heat generation device. There are also several other ways a FET can fail, depending on whether it's doped to produce an n-type or p-type semiconductor, let alone which method of insulation is used, but since pretty much everything we're talking about is a MOSFET-based device, I'll keep it to that for now.
Yes, your description above can certainly happen, no argument there. All I'm pointing out is that there are more ways in this heaven and earth for a FET to fail than you dream, Horatio.

 

[svetz]

If you're "actually" familiar ...
I'd say you're entitled to your opinion ...
...there are more ways in this heaven and earth for a FET to fail than you dream, Horatio.

These are not justifications for arguments, they are attacks on a member. It's great to argue issues, but please let's not attack other members.

Now obviously this is a really minor case and I'm only bringing it up to stop it cold not just here, but as a reminder everywhere. Play nice! Because the next thing you know B feels slighted and says something back. Then A does something witty that escalates it further. Next thing you know, there's actual drama and one or both end up getting warnings and bans. It also discourages members from asking what they believe to be dumb questions because they don't want to be put down. Will has built something special for us, let's all do what we can to protect it.

Remember we're all here to learn. If you feel you're not being believed then link a reference. I don't know the first thing about MOSFETs, but for example A could say:

Mosfets usually fail short, but they can fail open, ref. It doesn't sound like they can fail half-way, at least not for ever as the heat would cause them to short open (ref).

Then B could respond,

That article is for high temperatures. It you look at this paper...

Even then, some people are just never going to agree with you. And that's okay too!
You can lead a horse to water, you can't make it drink. Sometimes though, you just have to let things go.

Next my apologies for taking the thread off topic. Feel free to gripe about me, have your two cents on the issue, or anything else... but please do it in the "complaint" forum rather than take this thread even further off topic.

Okay, off my soap box.

 

[Steve_S]

No such thing as a "Stupid Question" with the exception of the one Not Asked ! Not everyone knows everything and to ask when you do not know something is a sign of intelligence, to attack someone for not knowing something is.... well I think everyone get's it.

 

[Justin Laureltec]

Point taken, and I apologize for inappropriately escalating the argument!