bad caps of the 1990's played havoc with car ECU's

[theOracle]

the Nichicon PF (M) series from the late 1980's through most of the 1990's had leakage issues, these caps played havoc with many automotive ECU's, especially high end Japanese cars. (i.e. Lexus LS400, Toyota Supra, etc.)

The problem was the use of quaternary ammonium salts.

the following capacitors from this time period had issues:

Nichicon: RZ, RT, VZ, ET, PR, PY, PF, PL, PQ, PG, MX, WX, WP, MP
Nippon Chemicon: LXF, TXF, SXE, SXG, MVK, MFK, MVF, MFF
Elna: RSH, RSG, RSE, RC2S, RC3S

The production of capacitors that contain quaternary ammonium salt compounds was ended in September 1998. The final one was Nichicon PF series. The PF series was kind of special version with expensive materials both rubbers and lead wires used. These had extended lifespan over the other caps above, but still it was shorter than conventional ones.

[japlytic]

I've seen Nichicon low-ESR units from around that era go bad (and found to be leaking). Also, I have seen certain series of Rubycon electrolytics from around that era go bad too.

[Wester547]

^ Do you remember what series those Rubycon capacitors were (if you don't mind me asking, just out of curiosity)?

[japlytic]

Originally Posted by Wester547

^ Do you remember what series those Rubycon capacitors were (if you don't mind me asking, just out of curiosity)?

One of them included SSP series.

[ratdude747]

I remember finding a blown Elna listed in a 1997 Pioneer headunit... sadly, that wasn't why it was fried (PO overdrew it's power)

[dmill89]

Although I'm not sure they it so much an issue of "bad caps" as decent caps that have exceeded their lifespan and failed. An engine compartment is a rough place for a capacitor with massive temperature extremes, and the vehicles in question are now 15+ years old.

[PaulMc]

I understood the OP to mean that the caps failed during the '80s and '90s. Not that they were now failing. I remember a lot of ECU problems during those decades. Did I misunderstand? At the time I was working in the HVAC field and solid state boards were just becoming common in a/c and heating equipment. Now I wonder if some of the many failures were due to bad caps.

[RJARRRPCGP]

But, most of them at least lasted longer than many caps of the early 2000s.

[momaka]

Nichicon PR definitely seem to have that problem. I pulled a small bunch from an old monitor/TV board and almost all of them have started to leak from the bottom, especially in the last 3-4 years I've had them in storage.

[lti]

I have a dead socket 7 motherboard from 1997 with Nichicon PL caps on it. They haven't started leaking yet.

[theOracle]

Originally Posted by RJARRRPCGP

But, most of them at least lasted longer than many caps of the early 2000s.

yes, this is exactly the point

the best of these bad caps I listed, the Nichicon PF (M) series was the most premium cap money could buy in the mid-90's and generally lasted at least 10 years before leaking and failing.

they were used in some high end luxury cars such as Lexus

Although I'm not sure they it so much an issue of "bad caps" as decent caps that have exceeded their lifespan and failed. An engine compartment is a rough place for a capacitor with massive temperature extremes, and the vehicles in question are now 15+ years old.

actually they were bad caps in the sense that their lifespan was prematurely short, but still generally 8-10 years or more. The error of your thinking is that automotive ECU's live in the harsh and high temperature engine compartment, they don't, they are located in the climate-controlled passenger compartment, typically in the passenger side kickpanel, behind the glovebox, under the passenger seat, etc. - The auto manufacturers wised up very early to the fact that the main computer for the vehicle needed to live in a gentle environment, as it is so critical and may control not only the engine, but the transmission and climate control as well.

[PeteS in CA]

momaka, Nichicon stoped making the PR series in 2005 or 2006, IIRC. So at the least, thse PRs are 7 years old, and thePR series was a 3000 hour series (at rated max temp and ripple current), IIRC. These things have a finite life, even in storage.

[dmill89]

Originally Posted by theOracle

The error of your thinking is that automotive ECU's live in the harsh and high temperature engine compartment, they don't, they are located in the climate-controlled passenger compartment, typically in the passenger side kickpanel, behind the glovebox, under the passenger seat, etc. - The auto manufacturers wised up very early to the fact that the main computer for the vehicle needed to live in a gentle environment, as it is so critical and may control not only the engine, but the transmission and climate control as well.

That depends on the vehicle. I'm not too familiar with these 80-90s high end Japanese cars so that may very well be the case with them but it is certainly not the case with all vehicles. Both of my vehicles have the PCM/ECM under the hood. My 95 Jeep Cherokee (4.0L I6) has the PCM mounted on driver side inner fender, on the same side of the engine as the exhaust manifold to be specific. Even many late model vehicles still have the PCM/ECM under the hood, my 07 Pontiac G6 (3.5L V6) has the ECM mounted in front of the battery, right behind the driver side headlight, this is significantly farther away from heat sources and offers much better ventilation than the PCM location on the Jeep but is still under the hood. While the Pontiac G6 may be discontinued (since Pontiac no longer exists) it is based on the GM epsilon platform and virtually all current GM midsized vehicles use the same design including the Chevy Malibu, Buick Regal, and Opel/Vauxhall Insignia.

[severach]

The entire car varies from -30*F to 150*F. The engine isn't substantially worse than the interior. In the winter the interior varies more than the exterior. The newer the car the more modules it has and they place them randomly inside or outside the passenger compartment.

[PeteS in CA]

The entire car ... hmmmmm ... I lived for several years in the Phoenix, AZ area. I could easily believe the engine area getting up to 150F and the passenger area up to 125F or 130F, but I doubt the passenger are got much hotter than that. OTOH, I could easily believe temps in Nome AK or Anchorage, AK getting lower than -30F.

But the basic point is true. There's no part of the car that doesn't get exposed - possibly for short durations - to significant temperature extremes. In more practical terms, Part 1, 130F is >55C, and when the electronics dissipate heat that will raise the ambient for those electronics close to their maximum ambient temp. In more practical terms, Part 2, -30F is getting close to the minimum rated temp for low impedance types like Nichicon's HE series, and at that temp the impedance will be up to 3X-8X the 20C impedance (both per Nichicon's datasheet).

[Topcat]

I've recapped quite a few Mitsubishi 3SI ECM's (Stealth and 3000GT). Back in the 80s and early~mid 90's, the bad cap issue wasn't even on the radar...but looking at it today, does it really surprise anyone that 20+ year old electronics are failing? It shouldn't....hell, some of these cars, the manufacturers never intended to still be around that many years later.

[larrymoencurly]

Originally Posted by PeteS in CA

The entire car ... hmmmmm ... I lived for several years in the Phoenix, AZ area. I could easily believe the engine area getting up to 150F and the passenger area up to 125F or 130F, but I doubt the passenger are got much hotter than that.

Engineers at the GM test track east of Phoenix, in Mesa, measured over 160F in the passenger compartment.

I have a TV digital converter box that needed most of its capacitors replaced in the first 2 years. In comparison the vintage 1976 TV connected to it still works and has needed just 3 of its original white Sanyos replaced. I credit the protective powers of the TV cabinet's fake wood grain,

[theOracle]

maybe inside ECU placement is a Japanese thing

I have owned

1984 Nissan 300ZX
1986 Nissan 300ZX
1992 Honda Accord EX
1996 Lexus LS400
1996 Acura RL
1997 Toyota Camry
2000 Toyota Avalon
2004 Lexus LS430

ECU location on all of these cars was inside car (usually passenger side area)

[momaka]

Originally Posted by PeteS in CA

momaka, Nichicon stoped making the PR series in 2005 or 2006, IIRC. So at the least, thse PRs are 7 years old, and thePR series was a 3000 hour series (at rated max temp and ripple current), IIRC. These things have a finite life, even in storage.

Well, mine are actually from the early-mid 90's - so yeah, they are quite old indeed. I've seen (and have) older caps that still work fine, though. The fact that all of my PRs leaked from the bottom (I have like 10 or so) just tells me that perhaps they did have some problems (though probably nowhere near as bad to call them "bad" or "unreliable").

[PeteS in CA]

True. I remember seeing some Sprague axial lytics (30D series, IIRC) that had to be replaced in 2007 ... the date code on the parts was 1977.