Replacing Liquid Electrolytic Capacitors With Solid Capacitors

Discussion of hardware and software development for Super NES and Super Famicom. See the SNESdev wiki for more information.

Moderator: Moderators

Forum rules
  • For making cartridges of your Super NES games, see Reproduction.
Post Reply
Jagasian
Posts: 421
Joined: Wed Feb 09, 2005 9:31 am

Replacing Liquid Electrolytic Capacitors With Solid Capacitors

Post by Jagasian »

The liquid electrolytic capacitors in electronic devices such as the SNES start to dry out after 15 years roughly. Some last longer than others before needing to be re-capped. Dry caps may also accelerate the aging of other components in the SNES, degrade audio and video quality, as well as cause other issues.

Since the release of the SNES, there have been advancements in capacitor technology, and so there are solid polymer and solid ceramic capacitors with high enough capacity to replace electrolytic capacitors in some situations.

There are two main benefits of these solid capacitors: they do not dry out and they have very excellent electrical noise filtering. So replacing the liquid caps with solid capacitors means you will never have to replace the capacitors ever again.

So I gave it a shot and it works great! Every liquid cap is replaced with a solid cap and it looks great when measures on my oscilloscope and my CRT TVs.

Below is a capacitor parts list and other pics showing the finished job. I used high quality capacitor brands (Panasonic and Kemet are in the same league as Nichicon and a small list of other top tier brands). The solid capacitors are rated for use up to 125 Celsius versus 105 Celsius used by the high grade liquid electrolytic caps that most people use when replacing caps.

Is this overkill? Yes, but if it means never having to replace caps again, it is worth it when you have dozens of consoles, many arcade PCBs, dozens of CRTs, and other old electronics. Replace with solid caps and never worry about it again.

One last note. Capacitor C67 that Nintendo used is rated for 25 volts, but the SNES power supply itself only uses a 16 volt rated capacitor and using my oscilloscope I confirmed that it is safe to use a 16 volt rated capacitor for C67 as the voltage at C67 never goes above 10 volts.
Parts list
Parts list
Ceramic capacitor replacement for C62
Ceramic capacitor replacement for C62
Solid polymer capacitors made by Panasonic
Solid polymer capacitors made by Panasonic
Voltage ripple on C62 before replacing caps
Voltage ripple on C62 before replacing caps
Ripple on C62 after replacing caps, notice less dip in the voltage
Ripple on C62 after replacing caps, notice less dip in the voltage
Jagasian
Posts: 421
Joined: Wed Feb 09, 2005 9:31 am

Re: Replacing Liquid Electrolytic Capacitors With Solid Capacitors

Post by Jagasian »

Voltage regulator output before upgrades
Voltage regulator output before upgrades
Voltage regulator output after upgrades
Voltage regulator output after upgrades
calima
Posts: 1745
Joined: Tue Oct 06, 2015 10:16 am

Re: Replacing Liquid Electrolytic Capacitors With Solid Capacitors

Post by calima »

Why recap the old voltage regulator instead of replace it entirely? I hear the modern replacements are quite a bit better.
lidnariq
Posts: 11432
Joined: Sun Apr 13, 2008 11:12 am

Re: Replacing Liquid Electrolytic Capacitors With Solid Capacitors

Post by lidnariq »

Modern switching regulators may not be a fantastic plan¹, and linear regulators haven't gotten much better in the past 30 years, and even an ideal regulator can't help with power demand that's too physically far away from the regulator output (because the feedback loop is at the regulator output).

¹ Krzysiobal has had some recent bad experiences with switching power supplies failing and taking out the irreplaceable parts of the Famicom at the same time.

For efficiency, a two-stage design (switcher for efficiency, followed by LDO for safety) might make sense. Maybe. Also, the SNES still uses the unregulated 10V all over the audio path.
Jagasian
Posts: 421
Joined: Wed Feb 09, 2005 9:31 am

Re: Replacing Liquid Electrolytic Capacitors With Solid Capacitors

Post by Jagasian »

calima wrote: Thu Jun 25, 2020 11:39 pm Why recap the old voltage regulator instead of replace it entirely? I hear the modern replacements are quite a bit better.
Sorry if I wasn’t clear, but I replaced the voltage regulator with a modern higher rated regulator, in addition to upgrading the caps and adding the cap on the output of the new regulator.
NewSchoolBoxer
Posts: 2
Joined: Wed Jul 13, 2022 8:47 am

Re: Replacing Liquid Electrolytic Capacitors With Solid Capacitors

Post by NewSchoolBoxer »

Long time lurker, first time poster. Replacing electrolytic with solid polymer or ceramic is genius and you go out of your way to show the improvement under scope. I've linked this thread on Shmups forum and Reddit. I like the same brands you do. Would add Vishay ceramic to the elite group. However, the part of replacing the 25V max capacitor with a 16V one bothered me.
Jagasian wrote: Wed Jun 24, 2020 8:41 pm One last note. Capacitor C67 that Nintendo used is rated for 25 volts, but the SNES power supply itself only uses a 16 volt rated capacitor and using my oscilloscope I confirmed that it is safe to use a 16 volt rated capacitor for C67 as the voltage at C67 never goes above 10 volts.
From SNES diagram:

Image

Why would Nintendo use 25-50V caps that cost more when PSU isn't sending more than 12VDC? They produce less leakage current at the low voltage ratio. This technical video about creating an ultra low noise voltage reference mentions the greater the range you can place between DC input and the cap rating the better. For 7V and 10V references, 50V and 63V electrolytic capacitors worked best without becoming too huge or expensive.

I believe Nintendo's 25V on a pricey 1000uF cap was a compromise value to keep costs down on a console sold at a loss. Your 1000uF capacitor for $2.50 has a leakage current of 3.2mA. Kemet's cheaper one at 16V has 470uA, so I'd roll with that.

Leakage current is in part a function of capacitance * applied voltage. Lower leakage at lower capacitance but we don't want the higher resonance frequency. Other strat is feeding lower DC. 7.5VDC supply is better with a modern 7805 that can handle 7V.

Sorry if I sound pedantic here but solid caps don't last forever. They have upsides of lasting several times longer, not getting worse from age alone and not leaking. Costing several times more isn't exactly a negative if we can extent a fragile console's lifespan.
User avatar
dink
Posts: 157
Joined: Sun Jan 12, 2020 8:42 pm

Re: Replacing Liquid Electrolytic Capacitors With Solid Capacitors

Post by dink »

Jagasian wrote: Fri Jun 26, 2020 1:42 pm
calima wrote: Thu Jun 25, 2020 11:39 pm Why recap the old voltage regulator instead of replace it entirely? I hear the modern replacements are quite a bit better.
Sorry if I wasn’t clear, but I replaced the voltage regulator with a modern higher rated regulator, in addition to upgrading the caps and adding the cap on the output of the new regulator.
I was going to say - there was an aweful lot of hifreq noise on the regulator output after recapping :? That explains it :)
Garth
Posts: 246
Joined: Wed Nov 30, 2016 4:45 pm
Location: Southern California
Contact:

Re: Replacing Liquid Electrolytic Capacitors With Solid Capacitors

Post by Garth »

I worked at TEAC (semi-pro tape recorders) in 1982-83 as a repair tech. Although the cassettes hadn't been out many years, the open-reel machines had, and the only times I had to replace an electrolytic capacitor were where they skimped on the voltage to save money in the large power-supply capacitors, putting for example a 25V capacitor in a 21V application or 16V capacitor in a 12V application. Other than that, I never saw any electrolytics fail. I also have a cheap tape recorder from about 1962 with the original electrolytic capacitors in it, and they're still fine. I'm not saying they never fail, but I think that if proper derating is applied, it is very rare.

Monolithic ceramic capacitors will generally have less error tolerance than electrolytics do; but do stay with something like X7R. Cheaper ones like Z5U will have a much greater temperature coefficient, and even voltage coefficient. The voltage coefficient is something we really have to pay attention to in audio circuits, because the chip capacitors have a piezo effect where the voltage on them makes them skew to one side, like bending a saw more sharply to get a higher note when they play the saw, musically, with a violin bow. I cut an article about it out of one of the industry magazines. I don't know if it would be frowned upon as a copyright violation to post scans of it here. In my design work for my job, basically I keep the WVDC on X7R capacitors at least 5x as high as the charge on the capacitor will ever be in the circuit. The curve is pretty flat in the ±20% range (like up to 10V on a 50V WVDC capacitor), and when you get outside of that, the capacitance starts falling off rapidly. For Z5U, Y5V, and others, it's pathetic—really, really bad, and it doesn't wait 'til you get to the 20% mark! (That's ok for something like power-supply bypass capacitors, but not for audio circuits where they'll cause a lot of distortion.)

There's a lecture on capacitors from a Kemet engineer at https://www.youtube.com/watch?v=ZAbOHFYRFGg .
http://WilsonMinesCo.com/ lots of 6502 resources
Post Reply