Need help with fault code "21" - VTEC solenoid "spool" valve

[YellowHaze]

Hello NSX techies! During my last track event on August 8, my CEL came on. Although it went off after stopping, it came back on during my next run. Oil was fine, etc. At high revs (>7500 rpm), my car began to hesitate. I have a Prospeed stage III chip set at 8200 rpm so my car happily reaches that rev limit. The car ran fine otherwise, but the CEL then stayed on.

After recently learning how to pull the OBD codes, three codes (2, 21, and 36) were read. I already knew about my TCS (disabled due to being faulty). On the track, my guess about needing new O2 sensors was correct for code 2. I have Prime (and a keen interest to engage in DIY projects on my car) to thank for my growing knowledge base. Last Saturday, I spent a few hours changing the O2 sensors and installing a new used TCS unit. After resetting the system by pulling the clock fuse, my CEL still persisted.

Today, I pulled the one remaining code 21. After doing some research, I now know what and where the VTEC "spool" solenoid valve is. My question is, does the valve just need to be removed, cleaned and reinstalled or replaced? Or, does code 21 point in some other direction? Your help is appreciated!

Steve

 

[Daedalus]

Troubleshooting the error code begins on 6-15 of the '91 manual.

There are 2 segments of the system: the spool valve (#9) and the feedback sensor (#14). The valve itself is a coil, and thus needs to show the correct resistance (at the ECU). Under the correct conditions, the ECU sends a signal to the valve to open, which feeds oil pressure to the VTEC circuit. There is a pressure sensor in the circuit that senses the oil pressure and confirms to the ECU that the valve is open. A lot of suspects, but troubleshooting is fairly logical.
* Bad harness or connection
* Bad spool valve (shorted and/or doesn't open enough, or always remains open)
* Bad sensor (always closed or always open)
* Insufficient oil pressure

 

[YellowHaze]

Troubleshooting the error code begins on 6-15 of the '91 manual.

There are 2 segments of the system: the spool valve (#9) and the feedback sensor (#14). The valve itself is a coil, and thus needs to show the correct resistance (at the ECU). Under the correct conditions, the ECU sends a signal to the valve to open, which feeds oil pressure to the VTEC circuit. There is a pressure sensor in the circuit that senses the oil pressure and confirms to the ECU that the valve is open. A lot of suspects, but troubleshooting is fairly logical.
* Bad harness or connection
* Bad spool valve (shorted and/or doesn't open enough, or always remains open)
* Bad sensor (always closed or always open)
* Insufficient oil pressure

Thanks for your input, Daedalus. Unfortunately, troubleshooting electrical components is not my strenght. What do you specifically suggest?

 

[Daedalus]

Is it too much to ask you to take an ohmmeter and check resistances of the valve and pressure sensor on the front bank with the engine off? Just a WAG, I think you'd have around a 50% chance of finding the problem just by doing this.

At the very least, wiggle the connectors and make sure they're on and making good contact, if you haven't done so already.

 

[EdinKali]

If it isn't electrical, bad spool valve gaskets might be responsible for low oil pressure at the valve.

 

[C-speed]

Just a note.. low oil or low oil pressure will NOT cause a code 21. Code 21 is SPECIFICALLY the VTEC solenoid itself. The solenoid functions independently of oil pressure, that's what the VTEC oil pressure sensor is for (on non-JDM vehicle's).

One quick way to check if the VTEC solenoid is working, is to simply give it 12-volts. If it works, you'll hear it click. Its only 1-wire, you can't screw it up.

Test it from the solenoid side of the connector, then re-test it from the harness side of the connector. Generally, it'll be a bad connection.

If the solenoid works when you test it, then you have a wiring issue - break in the wiring somewhere from the solenoid to the ECU. If it doesn't click at all, then you have a bad solenoid - replace it.

 

[YellowHaze]

Regarding your second comment, Daedalus, I do not own electrical diagnosis equipment nor know how to interpret that kind of data. Double checking the connectors and making sure everything is fitted in properly is something I will do. Upon a quick inspection, all looked in good order with no fluids showing. I keep a tidy engine compartment. Thanks for your follow-up.

Thanks for your comment, EdinKali. I've read a couple of threads about how the earlier engine configuration has an oil movement weakness when driving at high speeds and pulling some good g-forces. Something about oil collecting at one end of the upper valve cover, pooling around the spool valve, and causing an ECL. Replacing the gaskets was one possible corrective measure. I'd hate to do that and not have it be the culprit, though.

 

[Larry Bastanza]

Steve,

Find a buddy with an ohm meter. This is a quite simple test and you only need to check the two items, as mentioned above. After that it would be a wiring issue, but that is doubtful, unless the harness has been disturbed for some reason. Did you move around a lot of wiring for your 02 sensor change?

Regards,
LarryB

 

[Daedalus]

Just a note.. low oil or low oil pressure will NOT cause a code 21. Code 21 is SPECIFICALLY the VTEC solenoid itself. The solenoid functions independently of oil pressure, that's what the VTEC oil pressure sensor is for (on non-JDM vehicle's).

It will if too low to close the feedback pressure switch (how else would the ECU know the spool valve is bad?). The manual doesn't say how much pressure will close the switch, but it does say you should be able to read at least 57psi at the switch tap at 5000 RPM with the spool valve engaged.

 

[YellowHaze]

Great help as usual. Thanks guys. My oil pressure is perfect and has never had a low reading since I've owned my car. So, hopefully it's simply a matter of a bad solenoid. I'll post my findings once I get my hands on an ohm meter.

PS. Hi Larry! Thanks for sharing your expertise. Did you get my last email? Haven't heard back from you in a few days. Glad to see you perusing Prime. :biggrin:

 

[YellowHaze]

At the risk of getting ahead of myself, looks like the only option is to replace the entire spool valve assembly, part no. 15810-PR7-A00. The discount price at AcuraOEMparts.com is $287. D'oh! Anyone know of a better (less expensive) source?

 

[Larry Bastanza]

You can buy this:

37250-PR3-003 SWITCH ASSY., VALVE TIMING OILPRESSURE (#14 in the drawing) $48.83

and this:

36171-PR7-J00 SOLENOID ASSY. (#9 in the drawing) $168.87

Regards,
LarryB

 

[C-speed]

It will if too low to close the feedback pressure switch (how else would the ECU know the spool valve is bad?). The manual doesn't say how much pressure will close the switch, but it does say you should be able to read at least 57psi at the switch tap at 5000 RPM with the spool valve engaged.

If there wasn't enough oil pressure to activate the VTEC solenoid, then VTEC simply will not kick in. You may or may not get a CEL. If the oil pressure were low enough, it will only trigger the low oil pressure light on the gauge cluster.

If the VTEC oil pressure sensor was bad/unplugged, you would get a code 22.

Neither of which will throw a code 21, which is again, SPECIFICALLY the VTEC solenoid itself.

So either A) the ECU tried to engage VTEC and the solenoid did not work or B) there is a short somewhere from the ECU to the VTEC solenoid. Its only a single wire.

Actually, now that I think about it, unless there is a short in the VTEC solenoid wiring or internally, you will not get a code 21. The solenoid could simply be stuck for example, and the ECU wouldn't know it and you will not get a code 21.

So basically, he's got a short in the wire from the ECU to the solenoid, or internally in the solenoid. I already told him how to check it very quickly and easily by applying 12v in my previous post. Get a length of wire and jump to the 12v in the fuse box. You'll know in 10secs if the solenoid is bad or if its bad wiring.

 

[nikey22]

somebody please get this guy a $12 ohm meter and lets get this solved...... I'm betting its a bad coil (solenoid)...

 

[YellowHaze]

Again, I really appreciate the support guys. As I already mentioned, I can figure out how to do many mechanical DIY jobs (brakes, O2 sensors, spark plugs, etc.). When it comes to electrical fault checking, I have little knowledge on what to do and how to do it.

So, I will go to Home Depot and buy an ohm meter. I will Google 'ohm meter' and hopefully learn what it does and how to interpret the reading. Noone gave me any metrics, so I presume this is a go, no-go type of reading using an ohm meter to check the solenoid, correct?

Can someone tell me where specifically I need to hook this up to check for proper functioning of the spool (VTEC) solenoid valve? Comments like "apply 12 volts" and "test from solenoid side" are not helpful as I don't know where to connect what. I suppose I need a little "hand holding" so I know exactly what I need to do. Referencing the schematic by identifying the part number and what location of the part to apply 12v power or hook up an aligator clip, etc. will be helpful. Thanks again!

 

[Daedalus]

If there wasn't enough oil pressure to activate the VTEC solenoid, then VTEC simply will not kick in. You may or may not get a CEL. If the oil pressure were low enough, it will only trigger the low oil pressure light on the gauge cluster.

If the VTEC oil pressure sensor was bad/unplugged, you would get a code 22.

Neither of which will throw a code 21, which is again, SPECIFICALLY the VTEC solenoid itself.

So either A) the ECU tried to engage VTEC and the solenoid did not work or B) there is a short somewhere from the ECU to the VTEC solenoid. Its only a single wire.

Actually, now that I think about it, unless there is a short in the VTEC solenoid wiring or internally, you will not get a code 21. The solenoid could simply be stuck for example, and the ECU wouldn't know it and you will not get a code 21.

So basically, he's got a short in the wire from the ECU to the solenoid, or internally in the solenoid. I already told him how to check it very quickly and easily by applying 12v in my previous post. Get a length of wire and jump to the 12v in the fuse box. You'll know in 10secs if the solenoid is bad or if its bad wiring.

My apologies. The switch troubleshooting section follows right after the solenoid troubleshooting, and I read through both, thinking they both applied to code 21. After realizing this I now agree with most of your statements. However, VTEC solenoid activation does not depend on oil pressure (though physical cam operation does). As long as coolant temp, engine speed and (I think) throttle position meet certain parameters, the valve will open. Hypothetically if oil pressure was too low, the switch wouldn't open, and you would certainly get a code 21.

I'll post some pics later if no one else does. I'm all in on the solenoid itself or its wiring, so checking it won't be hard.

 

[C-speed]

However, VTEC solenoid activation does not depend on oil pressure (though physical cam operation does). As long as coolant temp, engine speed and (I think) throttle position meet certain parameters, the valve will open. Hypothetically if oil pressure was too low, the switch wouldn't open, and you would certainly get a code 21.

Incorrect, unless I'm misunderstanding you. The VTEC operation DOES depend on the VTEC oil pressure switch 100% (on USDM models).

If the VTEC oil pressure switch does not see enough oil pressure (on/off switch), then the ECU will NOT activate the VTEC solenoid. It is a safety feature build in to the system - not sure why the VTEC oil pressure switch is not found on JDM vehicles though.

Thus, the VTEC solenoid is soley dependant and will only activate on sufficient oil pressure, coolant temp, throttle position (manifold pressure actually) and RPM.

 

[Daedalus]

The pressure sensor is downstream of the solenoid and thus sees no pressure until the solenoid opens. One of the checks in the FSM is to ensure that there is less than 7psi at 5000 RPMs at the sensor tap with the solenoid off (low enough to keep the switch closed), and at least 57 psi with the solenoid activated (high enough to open the switch).

 

[NSXclear]

somebody please get this guy a $12 ohm meter and lets get this solved...... I'm betting its a bad coil (solenoid)...

It's only $1.99 most of the time at Harbor Freight. I'm sure there is one in OC.

I guess Ohm meters are intimidating for some people. Not really hard to use.

 

[Daedalus]

 

[Daedalus]

Make sure you don't bend the pin when hooking the alligator clip to it. When you're done just hook the harness up the way it came apart. I make a habit of using dielectric grease in each pin terminal before fastening connectors together.

I won't lie--the connectors on the front bank are in a pretty tight spot. But I was able to do the entire job in under 10 minutes, including taking pictures. Long-nosed pliers helped a little.

 

[YellowHaze]

Make sure you don't bend the pin when hooking the alligator clip to it. When you're done just hook the harness up the way it came apart. I make a habit of using dielectric grease in each pin terminal before fastening connectors together.

I won't lie--the connectors on the front bank are in a pretty tight spot. But I was able to do the entire job in under 10 minutes, including taking pictures. Long-nosed pliers helped a little.

Awesome Daedalus! Hopefully, this was something you already had done and posted from your computer files. If you went thru all the trouble to do this tonight and post it for my benefit, I'll need to take you out to lunch and buy you a drink as a gesture of my appreciation. A big thank-u!

 

[YellowHaze]

OK, I purchased a multi-meter and just followed the photo instructions posted by Daedalus for checking resistance of the solenoid connector. I registered a reading of 19 ohms, so that appears to be OK. C-speed suggested testing the solenoid valve by "giving it 12v". If C-speed or someone else could provide a little instruction on how to do that (where to connect what), I will do it.

My solenoid, connectors, and all adjoining components appear to be in good order- clean and properly connected. At this point, looks like my solenoid is faulty. I'd sure like to verify that before buying a new one or just the two components Larry suggested. Still, over $200! Thanks again to all for chiming in with your help and comments.

 

[YellowHaze]

Interesting. After reconnecting the solenoid connector, I decided to start the car just in case the CEL decided to go off. It did. I had a funny hunch about that. Like I said, the connector connection was fine and was 'clicked in' before I removed it to check the resistance. Am I out of the woods or might the solenoid valve act up again upon doing some hard driving?

 

[Daedalus]

That's good news. I was going to ask whether you tried firing it up yet, and was going to post after taking a few more pix. Since you didn't play with anything mechanical in the spool valve I'm going to guess it was purely a connection issue. You won't see "try wiggling the connectors" in any service manual, but it's a quick thing to do and it hits every now and then. Next time you hit VTEC on the road see if the light comes back.

BTW Larry B., I had someone engage the solenoid while I sat in the car revving the engine. As always you're right--there's no way I can hear it from inside the cabin. Must be the fuel pump solenoid as you said, but it's clicking on about 1000RPMs higher than I expect it to, right where VTEC is supposed to.