Pre-Order Discussions

dinkle is distributed out of Florida in the USA.

From  Venice  available Feb 25  on  Ebay .   

1 hour ago, JIT said:

So, basically you are doing it for space reasons?  Otherwise, you should able to use multiple caps to get what you need.  Multiple caps has the benefits of lower ESR/ESL as well.  It just seems like adding supply chain risk with a single sourced custom caps.

Using multiple smaller caps actually INCREASEs the ESR...we're limited by space limitations, airflow, and PCB routing constraints from putting more than 6 smaller caps on the board.  6 smaller caps have more than double the ESR apiece, resulting in a notably higher ESR total.

These aren't intrinsically "custom" caps from the ground up--like I've said before, 22,000uF, 80v is a "standard" size.  It just comes down to finding legitimate and good sources that don't have a 46+ week lead time.  And if it has to be customized from a manufacturer, that's better than unknown quality from a trading company ("reseller").

case in point: here's 2 capacitors from the same manufacturer, and the same product line (apples to apples): United Chemi-Con SMH-series:

ESMH800VQT223MB80U: 22000uF 80v -> ESR of 11mOhm, ripple current of 11.62A.  (Lead time: 46 weeks, 0 in U.S. stock)

ESMH800VSN103MA50S: 10000uF 80v -> ESR of 25mOhm, ripple current of 6.63A.  (Lead time: 50 weeks, 0 in U.S. stock)

If we used the 10,000uF cap, we'd have to use NINE of them to get the same ESR as we'd get with four of the 22,000uF caps.  (And there's only room on the PCB for six of the smaller ones.) 

Oh and by the way, while looking, I came across this absolute gem on a 10,000uF 80v Nichicon (LLS1K103MELC)...

<fileStore.core_Attachment>/monthly_2023_02/image.png.d76cde095efab7a8e4b2328eb092b10b.png

I'm sorry, but if this is the future of American manufacturing, we're all in trouble.  THAT'S ALMOST TWO YEARS!  (1.83 years to be close.) 

Yes, I know our lead time on this order of inverters has been terrible--but like I've posted before, this sort of problem won't happen again: because we aren't going to use the previous factory again!

In the case I referred to it was FAA approved equipment and we couldn't substitute the parts

It's likely worth it to order the ones you have been using and find an alternative for the current order 

9 hours ago, Sid Genetry Solar said:

Using multiple smaller caps actually INCREASEs the ESR...we're limited by space limitations, airflow, and PCB routing constraints from putting more than 6 smaller caps on the board.  6 smaller caps have more than double the ESR apiece, resulting in a notably higher ESR total.

If smaller caps with appropriate ESR are really not available then the total ESR could end up higher but custom/special caps still sounds extreme.  I doubt your manufacturer for custom caps is in the same league as Nichicon.  It's hard to imagine standard smaller caps with appropriate specs are not available from other similar quality manufacturers.

That makes a lot more sense...yeah, when there's regulatory requirements involved, non-approved substitutions are not allowed!

So the ones I have been using, like I said, are obsolete and no longer produced--and they aren't even on the Mouser website anymore.  Not even available for backorder.  The "substitute equivalents" from the same supplier as we've been using (ESMH800VQT223MB80U, linked above) have 0 pieces in stock in the entire U.S., with a 45-week lead time...which is simply not acceptable.  We've already had more than enough delays as it is.

The physically smaller the capacitor, the higher the ESR has to be by simple mechanical limitations.  There is no magic bullet around getting a lower ESR...apart from perhaps using solid polymer caps (MUCH more expensive).

So with the math, if the large caps (40x80mm) have 11mOhm ESR / 4pcs = 2.75mOhm overall. 
The "smaller" caps are 35x50mm (only 5mm smaller in diameter), and a maximum of 6pcs will fit on the board.  25mOhm ESR / 6pcs = 4.16mOhm, almost double the total ESR.

February 20, 2023:

• Transformer order is paid: actual production of spec GS inverter transformers has started.  I am expecting "sample verification" photos later this week--after confirming them, the whole order will be produced.
• still building an order list from the fastener supplier + throwing a lot of connectors on that order 😉.

February 22, 2023:

• oh the joys of working with 5-6 different suppliers at once!
• I've gotten production quotes on busbars (right about the expected price) and heatsinks.  After sorting out shipping logistics, they will be ready to order.
• working with the fan company for some slight modifications (latching connector ends, laser engraving the part number, and seeing if we can reduce no-load current)...
• still working with the capacitor company.  I appreciate their honesty about capabilities...but after some test results last night, I'm pretty sure they can produce what we need.
• I keep tacking items onto the fastener supplier's order, I'm sure I'm wearing them out 🤪
• need to figure out how to consolidate all these shipments to one freight shipment...otherwise the multiple express shipment costs really add up quickly.

On 2/18/2023 at 11:08 AM, Sid Genetry Solar said:

Worth noting that these filter caps do not "improve surge ability" in the inverters, as commonly assumed.  They're there to try to absorb the inductive "kicks" from the battery cables (resulting from the SPWM switching methodology.)  They don't store enough energy to maintain the battery voltage for even a fraction of a single AC wave at full load.

the size of the transformer and keeping all of the electronics cool gives the surge ability....the electrolytic capacitors filter the noise and reduce ripple correct???

I was looking at some electrolytic capacitors and they have a 2000 hour life span rating....how long are yours going to be rated by the new source???? and of course the hotter the electrolytic capacitors get the shorter their expected life span is also it seems...

the larger toroidal transformer with thicker enameled coated aluminum primary and secondary transformer windings gives it more ability,, 

so ambient temperature will play a big role.....also

1 hour ago, pilgrimvalley said:

I was looking at some electrolytic capacitors and they have a 2000 hour life span rating

This is extremely typical--indeed the expensive "U.S. caps" linked above only have a 2,000-hour lifetime rating.  In the larger caps we're looking at, this is about the longest lifetime you can find.  (Much smaller caps sometimes can be found with 5,000 and 10,000-hour rated lifetimes.)

However, it is worth noting that this paltry not-quite-3-month lifetime is at full rated voltage (i.e. 80v), full rated temperature (i.e. 105C), and full rated ripple current load (whatever that is; the datasheet doesn't list this size).

In normal operation, however, none of these 3 limits are encountered.  Which is why capacitors can last decades.

1 hour ago, pilgrimvalley said:

the electrolytic capacitors filter the noise and reduce ripple correct???

Yes, they do this; however, the bulk of their purpose is to "dampen" the inductive kicks resulting from the SPWM switching of the FETs.  The hard on/off switching (well, not-so-hard with a weakly driven inverter...) results in strong inductive kicks from the battery cables and everything else between the FETs and the power source.  If these "kicks" aren't dampened by the caps, the FETs will fail almost instantly.

Which is why if these main filter caps go bad, the FETs tend to also go *poof*: failure caused by brute overvoltage.

the 2000 hour rated ones ((2000-hour electrolytic capacitors )) I was seeing were advertised on Alibaba made in China for the larger 80-volt caps,,,, but thanks for the explanation / reply as I learn a little more about how things function....I was looking at some similar to what you are seeking (they are so many different suppliers on Alibaba to weed your way through)>>>>((( kind of like finding a needle in a haystack)))

hope your new electrolytic capacitor supplier works out...😎

An inverter whose functionality depends upon the pulse width modulation technology is referred to as PWM inverters. These are capable of maintaining the output voltages as the rated voltages depending on the country irrespective of the type of load connected.

the inverter  LF driver board is where you think the improvement comes about in the design (and a few other parts improvements) ....to improve the DC to AC PWM (pulse width modulation) change to a SPWM (sinusoidal pulse width modulation) for the quasi AC pure sine wave (PSW) to the toroidal transformer....

I can vision where the sudden change in loads from the DC battery voltage cause the inverter parts to need compensation or dampening to maintain the constant DV input voltage.... would a bigger battery bank at the preferred voltage also help this design issue???

the toroidal transformer functions on AC input (SPWM) and steps it up to the desired higher AC output voltage...

"Victron" is said to over engineer and under rate their inverters so the end result is a better inverter sold at a high price with lower claimed output so the end user is not as disappointed (((( vs the PowerJack over rated sales advertisements....)))) "Victron" uses a toroidal transformer in their LF inverters also....

so beefier better rated quality controlled parts should help to keep your inverter from blowing up all the mosfets as easily....

please forgive >>>  my simple but still learning about how the small electronic pieces and parts function....🤔

thanks, 😎

17 hours ago, pilgrimvalley said:

the size of the transformer and keeping all of the electronics cool gives the surge ability....

And perhaps more importantly the battery system including all wiring and connectors must support the surge demand.  Many people incorrectly think LF inverters can magically provide 3X surge power when their battery system only supports 1X rated power.

27 minutes ago, JIT said:

And perhaps more importantly the battery system including all wiring and connectors must support the surge demand.  Many people incorrectly think LF inverters can magically provide 3X surge power when their battery system only supports 1X rated power.

Thanks JIT,

I was kind of thinking along the same lines >>>> to get the voltage out and keep all functioning good>>> the battery bank needs to be able to not sag in voltage much upon demand loads >>>> hence the idea of a larger battery bank...

as I have also noted that if the battery bank goes low in voltage the inverter voltage output often starts to get lower until the low voltage protection of either the inverter or the BMS kicks in to protect the expensive battery from permanent damage....

so the need for a big battery and adequately sized solar PV array should be sought 1st as the system is in its growing phases...IMO

😎

6 hours ago, pilgrimvalley said:

as I have also noted that if the battery bank goes low in voltage the inverter voltage output often starts to get lower

this on LF Inverters is due to the transformer spec not having enough headroom.

6 hours ago, pilgrimvalley said:

or the BMS kicks in to protect the expensive battery from permanent damage....

Admittedly, these BMS units tripping and "cutting the inverter loose" often end up damaging the inverter due to severe DC overvoltage as it's sent flying by the BMS.

February 24, 2023:

• waiting on a phone call around lunchtime from the local shipping company regarding logistics for the different manufacturers in China.  Looking to see if we can consolidate shipping from the 4-5 smaller suppliers, instead of sending the boxes express air (DHL/UPS/FedEx/SF Express, etc.) at $200-$600 a pop.
  • once this is sorted out, I can get several orders placed (heatsinks, busbars and fasteners are all ready).  And like I've mentioned, the transformer order is already placed.
• Still working with the capacitor company...but I think we have a deal.  Caps will have a slightly higher ESR of about 15mOhm (vs 11mOhm listed for the U.S. ones).  But that's going to be our best bet--I would much rather know what we have, than hope we happen to end up with something better.
  • It's surprisingly hard to find true capacitor manufacturers--as the bulk of the companies are "trading companies" and can deal with genuine parts just as easily as counterfeit parts.  And you would never know the difference until there's an inexplicable unit failure in the field.
  • I checked with a PCB supplier, and they gave me 2 quotes for said caps:
    • First option, and I quote: "refurbished (There are quality problems between 3% and 4%, and it is recommended that you consider reserve part)，Imported brand."  Wow, that gives me so much confidence.........NOT!
    • Second option: "Made in China product. Note that the size is slightly different. No datasheet."  No datasheet?  How am I supposed to know what it's supposed to be?  Could be a ball of wax for all I know.