MEP-831a Inverter Overhaul

[marcjs14]

Peter, just for grins I took a sample of four of my new replacement Nichicon 560uF electrolytics and measured them using my Chinese capacitor tester. Guess what they each measured? 480uF! I think this tester is junk.

I'm going to bring them into work tomorrow where we have a proper electrical department and see what I get for them as well as one of the originals. I think I've only used the tester once before a couple of years ago when verifying some loudspeaker caps for a crossover and it made sense, but those were much lower capacitance numbers than 560uF.

BTW, the unit has a calibration knob and it didn't change the value when I turned it low-to-high when testing the caps. Ugh!

-Marcel

 

[peapvp]

That’s why we say measure thrice, cut once.

 

[peapvp]

Marcel,

Btw. Resistor values come in E Series standard value. For Gold 5% it’s the E24 Series and here we have only a 62K or 68K available.
The “green” ring is either red or grey - I would say it’s grey.

Peter

 

[marcjs14]

Peter,

I just checked two of my new Nichicons on a Snap-On tester (non-ESR). They confirmed the values I got using my Chinese tester...the two new Nichicons measured 480uF and the original Panasonic came out at 500uF.

Given they are supposed to be 560uF +/-20%, the 80% value would be 448uF, so both sets meet advertised spec.

I'm not sure what to make of this. I procured the Nichicons from Arrow Electronics, which I have had good experience with in the past. Not sure if there is a date code to be found on them???

Thanks,
Marcel

 

[peapvp]

Peter,

I just checked two of my new Nichicons on a Snap-On tester (non-ESR). They confirmed the values I got using my Chinese tester...the two new Nichicons measured 480uF and the original Panasonic came out at 500uF.

Given they are supposed to be 560uF +/-20%, the 80% value would be 448uF, so both sets meet advertised spec.

I'm not sure what to make of this. I procured the Nichicons from Arrow Electronics, which I have had good experience with in the past. Not sure if there is a date code to be found on them???

Thanks,
Marcel

Is there a date code on the Nichicons?

 

[marcjs14]

Is there a date code on the Nichicons?

Peter,

Apparently this is it...B2040. So manufactured the 40th week of 2020?


Thanks,
Marcel

 

[peapvp]

Peter,

Apparently this is it...B2040. So manufactured the 40th week of 2020?

View attachment 876938
Thanks,
Marcel

Marcel,

this is pretty bad, I would verify Date Code with Newark.
if date code is correct then those caps may have COVID 19 and need to be returned.

I check all incoming parts and I would return those.

I have seen this kinda thing more frequently in the last 15 years or so, where these large e caps are off by more then 5% when they arrive here.
We require certification from vendor with shipment and that includes shelve live.

Unless you have two inaccurate meters, which I doubt, and assumed date code of 2020 week 40 is correct, then they were manufactured very poorly which raises a flag now with Nichicon products

Peter

 

[peapvp]

Marcel,
The Nichicons we use, 100uF / 50V and 470uF / 16V come from one of their Japanese facilities and those have tolerances of 1% and less when they arrive here. ESR is also 1% and below. We check those when they come in, as they are certified and we have to certify our products
Peter

 

[peapvp]

Marcel,

Just for verification, you ordered 560uf 400V caps as replacement? Not accidentally 470uF 400V?

Both, Kurt and You got 560uF / 400V installed:

 

[marcjs14]

Marcel,

Just for verification, you ordered 560uf 400V caps as replacement? Not accidentally 470uF 400V?

Peter, that is correct:

 

[peapvp]

Marcel,

Found one more piece of info on your new caps:

You have to perform a reforming after cap being stored before you can us them:

8. Storage
Some characteristics of an aluminum electrolytic capacitor are
temperature-dependent. The higher the temperature is, the
more deteriorated the capacitor will be. An increase in tem-
perature accelerates the increase in leakage current and tanδ
and the decrease in capacitance. Leaving a capacitor exposed
to high humidity for long hours may lead to discoloration of the
lead wires and terminals, and poor solderability. To store alumi-
num electrolytic capacitors, keep them at normal temperature
and humidity without exposure to direct sunlight.
Leaving them exposed to high temperatures (higher than the
normal ambient temperature) may lead to chemical reactions
between the anode oxide layer and electrolyte, which drop the
withstanding voltage and increase leakage current. If this is the
case, applying the rated voltage to the capacitor will lead to
dielectric breakdown due to the heat produced with the large
leakage current, which finally causes the pressure relief vent to
open.
Capacitors that have been stored for long periods of time
should be subjected to a voltage treatment process (see Note
1) which will reform the dielectric (AI 2 O 3 ) by electrolyte and
return the leakage current to the initial level. Leakage current
increase during storage will vary with the withstanding voltage
of a capacitor. In general, the higher the rated voltage, the
larger increased the leakage current tends to be. Also, since
storing for long period of time may shorten the lifetime of the
capacitors, consider storage conditions according to the
requirements of device life expectancy.
(Note 1) In the voltage treatment process, connecting resistor
(around 1kΩ) in series with the capacitor, applied the rated
voltage and then be kept the rated voltage for 30 to 60 minutes.

 

[peapvp]

Marcel,

The leakage current has increased from them sitting on a shelve, container for shipping etc.
That's the reason why you fluke shows the capacitance it shows.
They call reforming "Voltage Treatment Process"
The 1KOhm (at least 10W) and 375VDC for 30 min to 60 min will bring capacitance back up to nominal value.
Don't forget to bleed capacitor with one of your 68KOhm resistors until you got 0V across both terminals.
Then you can use your fluke and see what it got.

Peter

 

[marcjs14]

Sorry for the "radio silence" - I've been awaiting parts and therefore had nothing to report until now...

...the 30 amp diodes as well as the 2w 65k Vishay resistors arrived from Mouser. Both raise packaging questions:

1. I read somewhere that a metal-film vs. carbon resistor-swap leads to a one-step drop in package size, but the 2W 65k ohm metal-film looks to be less than half the size of the (presumably) 2W 68k ohm carbon resistor it is replacing. Uber-efficient on the packaging side, but will its small surface area deliver on heat-rejection?
2. I should've paid more attention to the size of the Vishay fast recovery rectifiers, since the 30A version comes in the next size up package than the (original) 15A version. Not only will their leg spacings require tweaking in order to use the 30's, but the bodies will not allow for any space between them, essentially packing the space with silicone. Maybe I'll just stick with the 15A versions that I also ordered .

I'll post some images when I get the opportunity.

 

[peapvp]

Sorry for the "radio silence" - I've been awaiting parts and therefore had nothing to report until now...

...the 30 amp diodes as well as the 2w 65k Vishay resistors arrived from Mouser. Both raise packaging questions:

1. I read somewhere that a metal-film vs. carbon resistor-swap leads to a one-step drop in package size, but the 2W 65k ohm metal-film looks to be less than half the size of the (presumably) 2W 68k ohm carbon resistor it is replacing. Uber-efficient on the packaging side, but will its small surface area deliver on heat-rejection?
2. I should've paid more attention to the size of the Vishay fast recovery rectifiers, since the 30A version comes in the next size up package than the (original) 15A version. Not only will their leg spacings require tweaking in order to use the 30's, but the bodies will not allow for any space between them, essentially packing the space with silicone. Maybe I'll just stick with the 15A versions that I also ordered .

I'll post some images when I get the opportunity.

Carbon Resistors are one of the oldest breeds of passive components. They are about 120 years old. Since then, the technologies, materials, manufacturing processes etc have advanced.
One advantage of this progress was the ability to reduce the package size to achieve the same power rating and better heat dissipation.
Hence, the metal film's are smaller. Carbon Resistors are Dinosaurs and are becoming extinct.

The DSEI 12 are TO 220 Package and the Vishay are TO 247. I had looked at the Datasheet to check the pinout, which is the same but I didn't check Package. I made an assumption that they were both the same.
Over the weekend and evenings I am mostly on my iPhone 12 mini. I do not like big phones, but it makes reading and commenting on these threads here a lot more complicated.

If you got the TO220's then I would just put the TO247's up on eBay if Newark doesn't allow returns.

Peter

 

[marcjs14]

...If you got the TO220's then I would just put the TO247's up on eBay if Newark doesn't allow returns.

Peter

Peter,

Thanks for the reassurance re. the metal films...I initially had to verify the wattage since they just didn't look appropriate based on their size!

Regarding the FREDs, the Vishay VS-E5PX3006L-N3 (30A) are TO-247 packages, so I will try sending them back and install the VS-E5TX1506-M3's (15A) which are TO-220 package.

I hope to get back to working on this a bit over the weekend.

Thanks,
Marcel

 

[peapvp]

Peter,

Thanks for the reassurance re. the metal films...I initially had to verify the wattage since they just didn't look appropriate based on their size!

Regarding the FREDs, the Vishay VS-E5PX3006L-N3 (30A) are TO-247 packages, so I will try sending them back and install the VS-E5TX1506-M3's (15A) which are TO-220 package.

I hope to get back to working on this a bit over the weekend.

Thanks,
Marcel

Hi Marcel,

I think I may have looked at this part when we talked about it:

VS-E5TX3006-M3 (To220 30A Version)

Data Sheet

I hope the Stripper is working for you. Give it plenty of time to dissolve. The other project I did took 40 hours of soaking in Stripper.
Looking forward getting it moving again

Peter