Yet another Climate control Fix

[quack]

Just to add. I fixed my climate control based on the plentiful prime posts for $5.07 cents, including some new solderwick.

Symptoms were like most. a year ago the degrees numbers went away and it only worked on high. two days ago it stopped working all together no lights no controls. I got some caps and went to work. Most important besides the caps was fixing traces. I had to fix 5 in total using either jumpers or the component leads. The cap goop didn't eat the traces as it turned them to mush so my exacto just melted the traces around the caps. I got tired after two hours so i only fixed all the traces and repaired mostly just the leaked caps. 5 looked in really good shape. Unit worked happy the first time!
parts source in sanjose was hsc electronics, they had a sweet used server I may go back for but that's another story.

Only Problem is I still have two fastening screws left over ..not kidding on that.

 

[Nick675]

Congrats! Sorry about the two extra screws. Just double check the console. It only took me <10 min to pull apart.

 

[BryanZublin]

Replacing just the leaking capacitors is risky because all 13 capacitors will leak eventually. I have repaired two boards so far that previously had some caps replaced and the original caps that were not replaced leaked at some later date.

 

[passline]

Just to add. I fixed my climate control based on the plentiful prime posts for $5.07 cents, including some new solderwick.

Symptoms were like most. a year ago the degrees numbers went away and it only worked on high. two days ago it stopped working all together no lights no controls. I got some caps and went to work. Most important besides the caps was fixing traces. I had to fix 5 in total using either jumpers or the component leads. The cap goop didn't eat the traces as it turned them to mush so my exacto just melted the traces around the caps. I got tired after two hours so i only fixed all the traces and repaired mostly just the leaked caps. 5 looked in really good shape. Unit worked happy the first time!
parts source in sanjose was hsc electronics, they had a sweet used server I may go back for but that's another story.

Only Problem is I still have two fastening screws left over ..not kidding on that.

Did you remember the 2 behind the clock?

 

[Larry Bastanza]

Did you remember the 2 behind the clock?

And the two behind the ashtray

Regards,
LarryB

 

[Briank]

Replacing just the leaking capacitors is risky because all 13 capacitors will leak eventually. I have repaired two boards so far that previously had some caps replaced and the original caps that were not replaced leaked at some later date.

I see boards that have been previously repaired by so called professional shops that only replace some of the caps and use 85C rated caps when they should be 105C.

 

[ftuhy]

Brian,

Is the leaky capacitor phenomena isolated to just the first few years of production, or does it eventually hit all NSX's? Ie., one would think Acura or the OEM eventually discovered the problem and selected better capacitors.

The CCU in my NSX looked good about a year ago when I had it apart for other reasons. But, I'm wondering if preventative maintenance makes sense for my '96.

 

[Briank]

Brian,

Is the leaky capacitor phenomena isolated to just the first few years of production, or does it eventually hit all NSX's? Ie., one would think Acura or the OEM eventually discovered the problem and selected better capacitors.

The CCU in my NSX looked good about a year ago when I had it apart for other reasons. But, I'm wondering if preventative maintenance makes sense for my '96.

It will happen to all of them, time and heat are the factors that determin how soon.

 

[BryanZublin]

I have repaired units from 1991 through 2000 model years and I suspect that there are later years that have the same issue. As BrianK stated, time and heat are the enemies for electrolytic capacitors. The circuit design appears to be identical for all model years so it seems that there were no improvements made over time. Contrast that to consumer electronics devices which get refreshed nearly every year and you can get an idea of the different strategies between consumer and automotive electronics. The NSX is further cursed by it's long model life and low production volumes so there was probably little incentive for Honda to make changes.

The OEM caps are rated for 105 deg C but I don't know the life rating. For my repairs I use 105 deg C capacitors rated at 5000 hours. The life at lower temperatures is exponentially higher (roughly doubling of life for every 10 deg C decrease in temperature).

 

[nis350]

The OEM caps are rated for 105 deg C but I don't know the life rating. For my repairs I use 105 deg C capacitors rated at 5000 hours. The life at lower temperatures is exponentially higher (roughly doubling of life for every 10 deg C decrease in temperature).

So the life is based on usage or age?

 

[BryanZublin]

For a new capacitor, the actual life in the application will depend primarily on its operating temperature. The temperature of the CC board will be higher than the cabin temperature of the NSX because it is enclosed in a plastic box and the power dissipation of the board and display is more than 3 watts.

As an example, if the NSX cabin is at 20 deg C, assume the CC board is 25 deg C hotter or 45 deg C. A capacitor rated at 5000 hours / 105 deg C would have a minimum life of (2^((105-45)/10))*5000 hours = 64 * 5000 hours = 320000 hours = 36 years. This is a very simple approximation.

As another example, assume that the OEM capacitor is rated at 1000 hours / 105 deg C. Then the same calculation above would decrease from 36 years to 7.2 years.

The actual life of the capacitor should be higher than the calculations above since you are not driving the NSX 24 hours / day. Since we know that the OEM capacitors fail within 10 years, there must be some other failure mechanism at play other than operating time.

 

[nsxtasy]

As BrianK stated, time and heat are the enemies for electrolytic capacitors. The circuit design appears to be identical for all model years so it seems that there were no improvements made over time. Contrast that to consumer electronics devices which get refreshed nearly every year and you can get an idea of the different strategies between consumer and automotive electronics. The NSX is further cursed by it's long model life and low production volumes so there was probably little incentive for Honda to make changes.

Do most new cars nowadays no longer contain electrolytic capacitors for the climate control or sound systems?

 

[Briank]

Do most new cars nowadays no longer contain electrolytic capacitors for the climate control or sound systems?

They are still being used and a 10 to 15 year life span is plenty long for most cars. :wink:

 

[Hksdrager7]

could you post the link you use to fix the ccu please.

Just to add. I fixed my climate control based on the plentiful prime posts for $5.07 cents, including some new solderwick.

Symptoms were like most. a year ago the degrees numbers went away and it only worked on high. two days ago it stopped working all together no lights no controls. I got some caps and went to work. Most important besides the caps was fixing traces. I had to fix 5 in total using either jumpers or the component leads. The cap goop didn't eat the traces as it turned them to mush so my exacto just melted the traces around the caps. I got tired after two hours so i only fixed all the traces and repaired mostly just the leaked caps. 5 looked in really good shape. Unit worked happy the first time!
parts source in sanjose was hsc electronics, they had a sweet used server I may go back for but that's another story.

Only Problem is I still have two fastening screws left over ..not kidding on that.

 

[ftuhy]

The actual life of the capacitor should be higher than the calculations above since you are not driving the NSX 24 hours / day. Since we know that the OEM capacitors fail within 10 years, there must be some other failure mechanism at play other than operating time.
__________________
Zublin Engineering
http://www.zublin.com

I agree. I just did a rebuild on my '96 CCU. C44 failed the most spectacularly, burning the surface coatings off the board and etching some of the copper. C11 also failed, but only blew out the +lead without damaging the board. BrianK noted that C32 is also a frequent failure because it is next to a semiconductor (IC7) that get warm.

If over a large number of CCUs, all capacitors fail at approximately the same rate, its a "cheap capacitor" problem. If only certain capacitors fail, its a circuit design or application problem. And, certain capacitors don't simply fail, they vent hot gases out the bottom and blow off the + lead. That again points to an application problem. (In college, when we got overtired in the lab, we would put fins on the electrolytic capacitor can, then hook the capacitor up to 100Vac. They would blast off like a rocket and stick in the ceiling tiles. Great entertainment and stress relief!:tongue: )

I believe this is a result of some design deficiencies. Per numerous posts on CCU failures, C44 is probably the most frequent capacitor to fail. Briank indicated it appears to have a square wave going into it. Read the industrial applications literature from most any aluminum elctrolytic capacitor manufacturer, and they will note that high frequency content is not good for electrolytics. Well, even a low frequency square wave has significant high frequency components. In my design days, we put a small regular capacitor in parallel with electrolytics to help filter out the high freqencies.

While my C32 had not failed, I leaned the replacement capacitor away from IC7 to reduce the localized heating.

One thing I thought of after the fact is I should have put a small disc of telflon under C44 to prevent it from damaging the printed board any further the next time it blows. That might be a good thing to do any any capacitor that blew and vented out the bottom of the can.

I'm now retired and so no longer have easy access to an oscilloscope, much less to Briank's nice test set-up :biggrin:. It would be interesting to observe the waveforms at the capacitors that fail most frequently to see if some simple modifications could be made to greatly extend capacitor life.

As a final note, my '84 Corvette with 102K mi has never had a failure in the electronics of the "Atari dashboard" or its other electonics. I might be wrong, but I doubt Chevrolet's OEM spent more $$ than Honda's OEM in chosing electronic components.

 

[warrenw]

Frank, don't you know these things run on smoke.
When the smoke escapes, they stop working! :biggrin:

I sent mine off to Brian K only because I can't buy the dam capacitors cheep enough, in small enough quantities.
There isn't an electronics supplier within 1000 miles of here. If I have to pay for shipping, I may as well ship the CCU to Brian.

I have an Electronics Engineering Diploma, but I am also lazier as I get older. I did have the smarts to get it fixed before any caps leaked out all the crap and destroyed the PCB. More preventative than reactive.

 

[ftuhy]

warrenw,

I tried to capture some of that smoke to put it back into the capacitors but alas the smoke, but not the odor, had already escaped.:biggrin: So, rats, I had to buy all new capacitors!:frown:

You did the right thing going to Briank. He's definitely the go-to guy with tons of experience when it comes to NSX CCUs, etc.

 

[Ryuji]

Hi,

I have the common problem with my CCU (fan only works
on high) and I wanted to repair it...
I managed to repair almost all (3) bad capacitors but one
is my problem...

I don't have a 330µf-10V new capacitor, can I use a
330µf-35V new capacitor instead ?
I know I can use a capacitor with more voltage but 35 instead
of 10, is it acceptable ?

 

[Briank]

Hi,

I have the common problem with my CCU (fan only works
on high) and I wanted to repair it...
I managed to repair almost all (3) bad capacitors but one
is my problem...

I don't have a 330µf-10V new capacitor, can I use a
330µf-35V new capacitor instead ?
I know I can use a capacitor with more voltage but 35 instead
of 10, is it acceptable ?

10v or more is fine, it is only filtering 5 volts.
35v is overkill but not a problem.

 

[Ryuji]

10v or more is fine, it is only filtering 5 volts.
35v is overkill but not a problem.

Thank you it works just fine now

 

[rsevo6]

Mu CCU is also acting up, most of the time not starting up for several minutes after taking off and often only working on lowest temp setting and/or highest blower setting.
Rediculous that such a quality car as the NSX is let down by some bad quality components

Is there a partslist available on which/ how many parts to replace (capacitor values etc)?
It would be easier to search and get the parts before I take the unit apart to keep repair downtime as short as possible.
Don't want to take things apart and then find out parts are not readily available :frown:

 

[Ryuji]

PICS of A/C fix do it! :biggrin:

 

[rsevo6]

PICS of A/C fix do it! :biggrin:

Thanks, but in the partslist in that topic, the caps are all different (low) voltages.
I understood from previous postings by BrianK these were supposed to be 105V or if not available, as close to that as possible to give them longer life??

 

[Ryuji]

That's not 105V but 105°C (same caps are also available in
85°C with shorter lifetime).

Honda original caps are 105°C...

 

[rsevo6]

I stand corrected, didn't think enough before putting up the question:redface:, thanks :wink: