P0674 Glow plugs

[a5150nut]

I know I posted about this about September 2018 but I can't find it.

Back then I replaced #4 glow plug but still kept getting SEL with new plug. After changing the glow plug twice I still got P0674. So I determined it was probably the GP Controller. Changed it still same problem. Not so bad when its warm enough they don't cycle so now SEL.

Now recently I started getting P0678. Today I decided to fix it once and for all. Did some research and figured I could figure things out. Pulled #4 get 0.6 ohms on it. Same on #8. Well lets try #2 its working and its the same.

Key off I have 0.9 volts going to all 3 leads when disconnected. Key on 2.4 volts. Start and when plugs cycle 11.4 volts then drops back to 2.4 volts.

Started it with #2 disconnected and got P0672, P0674, and P0678. Cleared again P0672 is gone but P0674 and P0678 keep coming back and setting SEL.
Tomorrow I plan on disconnecting both batteries and grounding positive leads for an hour. Did that after I replaced the GP Controller just hoping it will work this time.

Any other suggestions?

[More Power]

Ohms and/or Voltage can each provide a good indication for correct operation, but current flow (amps) is the best test for a glow plug. Insert your ammeter in series with the glow plug wiring at the glow plug. Measuring the amps at a couple "known good" plugs during a key cycle test will help to evaluate any suspect plugs.

For example... 1- a glow plug controller could appear to function correctly, when applying battery power to the 8 gp circuits, but one or more circuits could fail to pass full current when asked to. 2- GP systems usually employ fusible link wiring to help protect the GP controller system. Fusible links keep the gp wiring from melting if a glow plug were to become shorted (for example). Fusible links can fail in a way that might present as intermittent.

[a5150nut]

OK that makes sense. Digital meters can read trace voltage. What kind of amperage should I expect so I dont burn up my meter? And where are these fusible links? I know the harness is part of of the main harness and $1k.

[a5150nut]

Got a second code for number 8 not working. Tested them with a test probe and 4 and 8 no glow. Replaced both now for a couple weeks all has been good. That would be the the third GP in 4 and second on 8. Nice to not have to clear codes after first start of the day.

[DmaxMaverick]

If the same hole is repeatedly frying (otherwise) good plugs, something is wrong (sorry for stating the obvious). It's either glowing too long, or that hole is too hot, for whatever reason. Traditionally, that spells bad injector, which comes with other liabilities. What do the fried plugs look like, compared to others? Excessive soot, white metal, blisters, deformation?

[a5150nut]

Nothing different about No 4 looks same as the others.
I'm at 128k miles now and very little has had to be done.
Just thermostats and two glow plugs and a GP controller trying to solve GP problem

[CoyleJR]

I know this is a bit off point but I see where GM is putting ceramic glow plugs in the Duramax 3.0. I have a 2005 Jeep Liberty with a VM Motori 2.8 Diesel and it came from the factory with ceramic glow plugs. When the ceramic gets old it breaks and falls into the combustion chamber and you have big troubles. DON'T LET ANYONE SELL YOU NEW AND IMPROVED CERAMIC GLOW PLUGS FOR YOUR 6.6 DMAX.

GM quote
"Ceramic glow plugs used in the Duramax 3.0L heat up more quickly and hotter than conventional metal-based glow plugs, helping the engine start and heat up more quickly in cold weather."

[More Power]

OK that makes sense. Digital meters can read trace voltage. What kind of amperage should I expect so I dont burn up my meter? And where are these fusible links? I know the harness is part of of the main harness and $1k.

You'll find a bad fusible link by load testing each GP at its location. Probe the wire that connects to it in series with the other meter probe on the electrical terminal of that GP. If you're measuring a low (lower than the others) amperage, go directly to a battery for power instead of through the GP controller as a test. You only need 3 or so seconds of power application for a good test. The early LB7s used GPs where each bank of four were ganged together. The California versions of the LB7 and all later Duramax engines used individual wires for each glow plug. To see whether the GP controller or the GP wiring could be at fault (assuming you measure a low amperage at one glow plug) you may need to swap wires with another known good glow plug circuit. Who knows, the controller that feeds 8 individual glow plug circuits could be the cause of a single plug not working correctly.

In the 6.5, for example, the fusible links are built into the harness, either near the GPC or near the GPs. The links are usually just a short piece of light blue and somewhat more flexible wiring - it feels different... 6-10" long or so that are spliced into the wiring. I bought a length of this fusible link at an auto parts store years ago. As I recall, the fusible link is wire sized by cross sectional area of the wire... not by its gauge. So, a 0.5mm2 would be an example. If someone has a bad fusible link, there should be writing on the wire that specifies what size it is.

The GPs in a 6.5 can each pull about 11 amps. Most digital multimeters I'm familiar with can handle up to 20 amps, which should be more than adequate for a single Duramax GP. Other than an ohm check, I haven't load tested the GPs used in a Duramax. I do know that they are a small diameter "quick heat" variety, so they shouldn't draw more than those used in a 6.5. Let us know what develops... Jim

[a5150nut]

Thanks for the follow up MP I have them working now. Replaced 4 and 8 and its been 3 weeks now.

[More Power]

Thanks for the followup.

I suspect the LB7 in my 2001 GMC has a bad glow plug. The 4 plugs in each bank are connected together using a buss bar. That bar will have to be removed to load test each plug... Wish I'd done that before putting the wheel well liners back in last summer. :(

[DmaxMaverick]

If you suspect bad plug(s), the LB7 fed emissions may not set the SES and a code on only one or a couple (CA emissions will fault on one weak plug). If you can read the cycle voltage (simple with a Predator), it will show how many are failed, or open for any reason, but won't show which one. I don't recall them offhand, but I posted the voltage scale some years ago in another thread.