Connect with us

Capacitor Substitution

Discussion in 'Power Electronics' started by H2814D, Aug 24, 2019.

  1. H2814D

    H2814D

    75
    9
    Nov 4, 2017
    Hi all. I have tried to find a definitive answer to my question on my own, however, I am not confident in what I have found. This was my last, but obviously, most trusted attempt.

    I realize substituting capacitors with higher voltage and temperature ratings is acceptable, if you cannot find the exact spec replacements, but I am trying to replace a cap that I cannot find the same uF rating for. I've tried the common online supply stores (Digikey, Mouser, Arrow) and ebay and amazon. No luck.

    It is a 950uF/200V snap in cap. 2" long, 1" diameter, 10mm lead spacing. There are tons of 1000uF's with the correct dimensions, and those would be very close, considering there is often a 10-20% tolerance with these things. So my question is, would I be able to get away with a 1000uF (or other suggested option) cap in place of this, or does that alter the eventual DC output voltage of the unit? There are two of them on the board and the application is for an RV power supply from 110vac to 12vdc, so it both powers the 12 volt electronics in the RV and acts as a house battery charger when plugged into AC. Note that the circuit board does list the cap values under the removed caps. 950uF/200V.

    Pics of the board and the cap are attached. If it matters, the mains are on the right side of that board and the DC output on the left (white and blue wires).

    Thanks in advance for what I'm sure will be another interesting education. :)
     

    Attached Files:

  2. 73's de Edd

    73's de Edd

    2,728
    1,116
    Aug 21, 2015
    Sir H2814D . . . .

    It says . . . .
    Note that the circuit board does list the cap values under the removed caps. 950uF/200V.
    Then DO ALSO NOTE +- 20% so, you are just fine, just be sure of the +105 degree temp range due to its potential rough temperature environs that it may be subjected to.
    PLUS I would be changing the pair simultaneously. . . . but only IF ?

    In " reading " the board, I am seeing AC power coming in the right bottom corner and then getting an X capacitor, then a contrawound line inductor pair and yet another X cap and a line fuse and then over to the right to a heat sink mounted FULL WAVE RECTIFIER block. Its rectified DC comes back to those two black capacitors for filtering. They DO operate at a VERY non taxing S...L...O...W up and down sinusoidal 50/60 cycles.
    So, typically, they are QUITE trouble free, as being compared to the other caps that handle the decades or hundreds of kilohertz of switching cycles of HARD pulses / square waves operating frequencies .
    That with gradual time, just heat up and eventually dehydrate those caps electrolyte . . . in an exponentiall manner.

    Now, relevant to your passing judgement, do you visually see any physical fault or electrolytic leakage being associated with those two big boys ?

    HOWEVER ! . . . . NOW . . .do look back at the fans direction , and don't you see an intermediate size E-cap that . . . looks like to me . . .that it has a domed cap pooching upwards at its triangular imbedded stress striations.
    If so, that might be your units REAL problem . . .? ? ?

    Aside observation / comment . . . .

    Located betwixt the cooling fan and the central MAIN power transformer . . . . . lies one HELL(x100) of a ferrite based inductor !
    Also, as you have already found, its dang nigh impossible to find that S.H.I. ( t . t . y . ) brand of capacitors.

    73's de Edd . . . . .

    Every morning is the dawn of a new error.



     
    Last edited: Aug 24, 2019
    HellasTechn likes this.
  3. dave9

    dave9

    807
    188
    Mar 5, 2017
    Yes 1000uF, 200V or higher rated would be fine. Why are you replacing them? These are at AC line frequency and generally have an easy life, rarely fail unless there was a huge surge that came in, and a surge that big would probably take out the switching transistor(s) if not the bridge rectifier too, before the caps.

    I agree with 73's suggestion that the other E-cap may be a problem. Depending on what value it is, I might replace it with one of these or similar:

    https://www.digikey.com/products/en...tity=&ColumnSort=0&page=1&stock=1&pageSize=25
     
  4. H2814D

    H2814D

    75
    9
    Nov 4, 2017
    Sir 73's de Edd. Thank you for your reply. I am confident, with your reply, that I can substitute the 1000uF for the 950uF. Replacing both of them at the same time was my intention. One of them had a slight bulge and was leaking.

    Your observation about that other cap is absolutely correct and it was the first one I noticed and the first one I pulled off the board. A 2700uF 25V. In addition to the obvious bulge, the covering appears to have melted downward toward the base/leads. I've never seen that happen before, but the heat in Las Vegas can do things you wouldn't believe to electronics. I attached a pic.

    Once again, thank you for your response.
     

    Attached Files:

    davenn likes this.
  5. H2814D

    H2814D

    75
    9
    Nov 4, 2017
    Thank you, too, dave9. I'll be ordering shortly. :)
     
  6. 73's de Edd

    73's de Edd

    2,728
    1,116
    Aug 21, 2015
    LAS WEGAS . . . . you're in LAS WEGAS . . . .thats what my nephew there, thought when he encountered this . . . .
    upload_2019-8-24_1-47-23.png
     
  7. H2814D

    H2814D

    75
    9
    Nov 4, 2017
    Well I thought I had this fixed after replacing the caps, but in real time use, it is still not working correctly, so I'm back for help.

    This is a Parallax Series 7100, Model 7155, 55-amp RV Power Converter. It is very widely used in the RV industry. It will power the 12 volt systems and will charge the RV battery when connected to 120vac, if it is working properly.

    Here is my issue: Upon replacing the caps as indicated above, I powered up the converter on the bench and was reading 13.8vdc out of the blue and white DC side wires, so I thought I had fixed the problem. I then re-installed it in the RV. I had purchased a new battery and it was charged up, so after powering up the converter in the RV to check it there, I figured the battery voltage level wasn't low enough to cause the charger portion to kick in. I then used the RV for a few days on battery power and noticed a voltage drain on the battery. I decided to hook up the generator to recharge the battery. After doing so, I went to test the charging system. I had brought a voltmeter with me and noticed that the converter was not charging the battery at what I expected to be around the 13.8vdc charge level. It wasn't charging it at all.

    So after all this and returning home, I removed the converter from the RV and put it back on the bench. On the bench, and after supplying 120vac to the mains, the converter again produced a no-load 13.8vdc at the outputs. I decided to hook up a 12 vdc fishing LED light to the dc side and the voltage dropped down to 8.7vdc and caused the LEDs to slightly flicker and not produce the wattage normally seen with the 12 volt battery. So under load, the converter is dropping the voltage down way too far to be effective.

    So that is why I am here now. What components on the board should I look at first? The transformer maybe, but I would think if I'm getting the rated voltage without a load, that wouldn't be the problem? I do not see any obvious signs of failed components, however, there are a couple of cement vertical resistors on the mains side of the board that have obviously gotten very hot. The PCB is discolored. You can see that in the pic. I haven't removed them (the resistors) to check them yet. I did remove the large heat sink, so I could look closely at the components attached to and using it.

    I will wait on your suggestions.

    Thanks in advance.
     

    Attached Files:

    Last edited: Oct 6, 2019
  8. dave9

    dave9

    807
    188
    Mar 5, 2017
    I'd check for bad solder joints, particularly anywhere discolored from heat like those resistors you mentioned (and the transistors on the heatsink), reflow the solder just to rule it out. I'd also look very closely at the copper traces there to see if they are lifting off the PCB... your pic of the back of the board is a bit dark and low resolution but it does look a bit crusty?

    Maybe it's just the flux, you might try scraping that off, if a soak with alcohol on a paper towel won't soften it up. Gget the area clean to inspect and then reflow the solder. If the copper is oxidized you may need more flux, separate or in new solder.
     
    Last edited: Oct 6, 2019
  9. **Rafael**

    **Rafael**

    16
    0
    Apr 27, 2015
    Hi there H2814D,

    You must have already checked for the obvious, I presume. Check the resistors you mentioned, cold solder joints, open traces, open resistors, etc. PWM chip may not be getting output voltage feedback due to any of the above reasons.
     
  10. HellasTechn

    HellasTechn

    1,434
    205
    Apr 14, 2013
    That Cap has definetly kicked the bucket !
    Oops just realised you had already replaced it.

    So true !
     
  11. H2814D

    H2814D

    75
    9
    Nov 4, 2017
    OK, so I have checked every solder connection and made sure there are no cold solder joints anywhere on the board. I removed and tested the mosfets (IRFP450) on the right side of the board against the heat sink and have done the same with the diodes (MPR4060PT) on the left side of the board. The mosfets and diodes tested ok. I found a 220uF 25V cap near the mains that tested inconsistently, so I replaced that one. The vertical cement resistors all checked ok. All of the other caps had been tested or replaced.

    Thinking that last cap replacement and the other solder joint re-flows had done the trick, I powered up the converter. This time, I started smelling that nasty burning electronics smell and could see a bit of light arcing at the drain (middle post) of the right most mosfet through the PCB. This arcing was actually visible through the thickness of the board and not on top of it. I immediately shut it down. I then looked it over to verify the solder connection was ok. It was, but I decided to remove and test the component again just to make sure it was still ok. It was. I then soldered it back on the board and powered it back up. This time, both mosfets were doing the same light arcing at the drain. At this point I stopped messing with it, since it became close to being thrown through a window. :)

    The only real difference in getting the low voltage during the load the other day and the results of today are the replaced cap I mentioned in paragraph 1. Other than that was just removal and replacement of components. And before you mention it, all of them were replaced back in the same positions as they were removed.

    Any ideas, guys? And thanks again for your help.
     
  12. Hunter64

    Hunter64

    9
    4
    Nov 20, 2018
    Do the integrated circuit have a number (type), and what is the TO-220?
     
  13. H2814D

    H2814D

    75
    9
    Nov 4, 2017
    Hunter64, I have attached photos of each component that could be an "IC." And a photo for location reference. I think you are looking for the KA3846 or MC34072P,
    though. I don't understand your question about "what is the TO-220?" Can you explain what you are asking for please?
     

    Attached Files:

  14. WHONOES

    WHONOES

    730
    150
    May 20, 2017
    Sounds like you have a problem with a current limiting circuit cutting in to early. This could be due to a sense resistor getting a bit frazzled and its resistance increasing a bit or it might be the active element be it transistor or IC having a problem. I would check for a low value resistor that looks as though it has got very hot. An indication of this might be carbonisation of the PCB or the part itself or, it may even have created its own dry joint by getting hot enough to melt its solder joints.
     
  15. Bluejets

    Bluejets

    3,517
    713
    Oct 5, 2014
    TO92 case device to the left of the MC34072P looks like it has the top chipped off it.
     
  16. davenn

    davenn Moderator

    13,325
    1,769
    Sep 5, 2009

    TO220 is a case type ... commonly used for power transistors and 3 terminal regulators

    upload_2019-10-8_11-33-56.jpeg
     
  17. davenn

    davenn Moderator

    13,325
    1,769
    Sep 5, 2009
    looks more like hot glue etc
     
  18. H2814D

    H2814D

    75
    9
    Nov 4, 2017
    It is some type of substance on the device, a hard clear glue like stuff. It is not chipped. Like davenn said.
     
  19. H2814D

    H2814D

    75
    9
    Nov 4, 2017
    There are two gray parallel resistors (.1 Ohm) just directly above the two large caps in the photo above. They both test out at .1 Ohm, however, the one on the left seems to rise up to .4 Ohm and then back down to the correct value. That is quite a percentage-wise fluctuation. Are those the ones you are talking about? They do not appear as if they have been overheated, on either side of the board.
     
    Last edited: Oct 8, 2019
  20. H2814D

    H2814D

    75
    9
    Nov 4, 2017
    OK. There are pictures of the mosfets above. Is he asking for more than that? Forgive me if I sound like I'm an amateur, but I am. :)
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Similar Threads
There are no similar threads yet.
Loading...
Electronics Point Logo
Continue to site
Quote of the day

-