what did i do to my glow plug relay?
 

so ... in my weekend of over-repairing (i love the compulsion thread, btw) ... i decided to dust off my multimeter and decided to look at my glow plugs ... AKA how diesel giant did them.

my dash glowplug light was fine ... everything worked great ... (i think), but was compelled to learn a bit more while i had an hour in the day

battery voltage: 12.5
running: 13.2

turned everything off, and checked each plug from the relay connector
everything was fine.
less than 1Ω on all 5.

plug the only plug back in ... nice and snug
go back into the house to deal with other things.

come back about 4 hours later to give it a cold-start ....
and the glow plug light on the dash no longer illuminates
i double checked the tightness of the connector
try again
nothing ....
i have not had time to look at anything else since then.
and I'm not even really sure where to look

did i fry something?
what testing should i do?

one solution finds 3 more problems.

geoff

Pull the cover off the GP Relay and check the 80 amp strip fuse.
It may look good, but could be cracked.

I keep a spare taped to the inside of the Glove Box door.


Charlie

could i have cracked it (or been the final straw) by opening the cover and pulling the connector?

i'll give that a look at lunch.

thanks
g

The odds are severely stacked in favor of the connector being the culprit.

are you saying the connector from where the GP wires come into the relay?
does that need to get R&R?

I'd pull the cover, and turn on the key to the glow position, and quickly measure voltage to ground on both screws in the fuse. if 12v on both, it's not the fuse.

Take the connector out again and look for green or whitish corrosion on the contacts. Use 600 grit wet dry paper rolled up lubed in soapy water to clean the contacts, rinse, blow dry, coat it with some Nolux (or any grease will do in a pinch) and see it that fixes it.

I got that voltage on 2 screws without turning the key...

.

As you should. The glow plug timer is "hot at all times."

Did you plug it back properly?
On the E class the connector need a good push, its a tight fit.

the drama continues...
 

Hi Everyone:
here are my findings from lunch...

• pulled the relay cover and measured the 80A fuse
  • no short, near 0Ω across the two screws
  • switched to voltage: 12.5V on each side of the fuse
• looked at the fusebox for fun ... all was good there
• took the glow plug harness connector out of the relay, cleaned each of the five female connections with sand paper ... they were not bad, but now they are shiny
• checked continuity from harness connector to glow plug in engine on all 5: all had good connections near 0Ω
• firmly re-connected the harness connector to the relay
• turned the key to 'glow' ... the buzz is still buzzing, but no glow light
• turned off the key, went back out and just looked at everything
• turned the key to 'glow' again, and went to the relay. stared at it for about 20-seconds and then heard it click.
• key off, then 'glow' again: then ran out and measured voltage from battery ground to the 'head' of each glow plug: ~10.9V on all 5 plugs ... then click and near-zero voltage at all plugs
• decided to try and start it.
• local temp says about 41°. feels warmer, but not much
• anyway, it was horrible. really horrible. lots of cranking and sputtering. on the third try, i got it to idle, barely. went inside and had some lunch while i left it running.
• came out about 10-mins later ... and there was more smoke than idles in the past. wtf?
• then turned off the car, turned key to 'glow' again ... still no light
• started up and it started quite easily.
• turned it off, put all my junk away in the garage and blasted back to 'work'

so ... there is my recap of my lunch.
is there anything i can rule out?
any more diagnostics?


why the heck did i open the thing in the first place ... it was working fine!


sorry for that scream ...

as always, any help/ideas are appreciated


geoff

If you were getting 10.5 volts at the Glow Plugs the Relay is working. You were also hearing the Relay Click off.

Could it be that the Bulb in the Glow plug Light is no good?

Also does you 81 300SD have the Glow Plug Relay that is connected to a Temp Sensor?

pre-glow system...
 

here is a diagram of the 300SD pre-glow
there appears to be a temperature sensor...
the diagram says 'as temperature increases, resistance decreases'

i also find the coincidence of the light going out at the exact same time as my messing-about to be tooooo coincidental.


anything else to check?

On your car assuming that is the correct schematic, the thermal sensor is inside the relay body.

The only external connections are the 4-pin control plug and the 5-pin (6 position body with 5 holes used) connector for the glow plugs. The control plug on your car does not have any thermal sensing wires. Later cars had external sensors so they had more external wires.

You can do a quick check by jumping pin 1 to 3 on the body side of the control harness. With the key in II you should have a glow light.

The glow relay does 2 things with the light.
First is the light turns on to help you know how long to pre-glow based on ambient. Note that the relay continues to supply current to the plugs after you crank, that is known as "after glow". That is why there are 2 power wires in the control plug. One is hot in II and III, that triggers the pre-glow, and the other hot in III only, that is the after glow trigger. The light does not illuminate during the after glow interval.

The second function is that the relay has a current sensor built in. It compares the current draw from plug #1 to the sum of the current draws from the other 4 plugs. The approximate mathematical operation it is doing with the sensing coils is (plug 1 current * approximately 3.5) - (sum of currents in plugs 2 - 5) = 0. If the result is not zero then there is a plug not drawing current. If the summing circuit is not satisfied then the glow light will not illuminate.

Unfortunately the design of this circuit had enough room for error that it is possible for 1 bad plug to get through.

Also there are documented cases on this forum of members whose glow plugs passed the ohms test but did not glow hot enough to do their job as verified by bench testing with jumper cables. (Danger- extremely high temperatures - almost 3k *F - and risk of short circuit - battery explosion likely if you short the cables - so proceed with extreme care on this if you decide to try this test.) If you had a plug that was on the way out maybe it just failed while you were testing.

If your bulb passes the jumper test I would go ahead and change out the plugs.

Be sure to use only Bosch or Beru plugs. Auto lite is bad juju. And use anti seize paste on the threads, you will thank yourself next time.

There is no "after glow" capability for the OE timer on a 1981 300SD.

i love 'more projects'!
 

okay ...
thanks for all the checking.

looks like i can add 'change glow plugs' to my order of operations...

taking them out and testing them (they are of unknown age and pedigree) is more expensive than the $53.75 the set will cost here.

i will also jumper the two wires to confirm the bulb is okay.

is there a good way to test if the relay is faulty?

With the 4 pin plugged in and the 6 pin unplugged, turn the key to pos II. Probe the 5 posts for voltage to ground. Relays can go bad but not as often as the plugs so let's try the plugs first. Also inspect your leads and make sure they are not cracked or frayed and the end terminals are securely crimped.

One more thing, be really delicate with the terminal nut on the new glow plugs. It is very easy to over torque that nut and break the heater inside the plug. Just finger tight and a gentle tweak with a box wrench or 1/4 drive ratchet is enough. There is a thread on this where it shows a glow plug sawed in half, you can see what is inside. I think whunter posted it.

thanks, jay_bob.

i basically did this test with the 6-pin connected, i think i did that test
'pos II' is the same as what i was calling 'glow' correct?
basically the farthest position that will not engage the starter.

when i did that test, i got 10.9V at the nut-connection of each glow plug before the relay shut off (maybe 30 seconds)

i will inspect the wiring harness when i have them off to replace the plugs

joy

When the Plugs are out the Manual calls for the Glowplug Holes to be reamed or otherwise cleand out. There is some alternative metnods: DIY Repair Links
DIY Links by Parts Category - PeachParts Mercedes ShopForum

PeachPartsWiki: Do It Yourself Articles - Mercedes Vehicles

So i went out in the pre-dawn here and did the test listed above. For all those who may search for something like this in the future. The glow plug light at your dash is connected to the relay harness via pins 1 and 3 bulb light. The schematic says red/black (?) and white/blue. On my 300SD these were on the same 'side' of the connector...above one another. You can identify them via the relay-side of the connector by pin 15 (12V input) and via a pin labeled 'LA'-something (my headlamp light was too harsh to see it all. But by deduction, i could see the other pin call outs (31 = cruise; 50=cranking) ... so that left 'LA...' which made sense as 'lamp'.

ANYWAY, in the dark this morning I was greeted with the warm dash light of a glowing glow plug light.
the lamp is fine ...
so i can cross that off my mystery list.

looks like some new plugs are in my future as another way to decode the mystery.
look at reamers ... that will be another thread all together ...

geoff

Glow plugs fail, it's a working one time, not the next affair.

As mentioned, pull each one and apply 12 volts, only the tip should glow.

Do this after taking ohm readings again.

I picked up a GP Reamer off E-bay a few years ago for $34.

Charlie

i can find a 7mm reamer on Mcmaster for ~30$
but that has to have some fab associated with it
Kent has a fabricated one for ~35$

i mostly pledged to myself to buy as much from Pelican as possible (because this form is so amazing) ... but the reamer here is 68$

am i missing something on why the reamers are such different prices?

For the occasional DIY reaming, a drill bit (approx diameter of GP tip) works reasonably well.

Prices are all over the place. The one I saw on eBay from a reputable tool seller (who usually has decent prices) was $97.

Prices likely vary according to the same criteria as other tools: professional/mechanic quality vs. medium duty... made in Europa vs. made in Asia, etc.

so why does the lamp not work?
 

okay ... it would be kinda weird to quote myself ... but because it is an election year, i'm making an exception.

ANYWAY...
i feel like i've got a ton of questions, and i am currently wearing my 'know just enough to be dangerous' hat. if anyone has answers to these questions, i'd love to know them:

• what circuitry or electric logic is going on that is preventing the glow lamp from illuminating on my dash?
• is there only one branch of the logic tree that explains my situation (80A fuse fine, conductivity to glow plugs fine, GP resistance fine, 10.9V to plugs during 'pre-glow', relay clicks off at about 20-30s, dash lamp does not light, starting is hard)?
• did the glow plugs suspiciously tip over to the 'bad' side to make the lamp not light?
• what does the circuit describe as bad at the glow plugs?
• if my plugs are good, is the relay bad?
• if i replace the plugs, and the light is still not on, is the relay bad?
• is the answer simply black-magic in the relay box?

i fully respect electronics ... but hate it at the same time.
i almost became an electrical engineer, but could never come to terms with not 'seeing' what was going on. i fried some $4K linear induction motors in grad school because i wired something backward with the 240V supply, and not understanding the actual information scares me to this day.

i plan on replacing the plugs this weekend, more out of diagnostic curiosity than causal certainty...

any answers to my questions are warmly welcome.

thanks
geoff

go buy a Radio Shack 22-172 if you want to see where and how many electrons are going to your glow plugs.

www.peachparts.com/shopforum/diesel-discussion/288970-best-easiest-way-test-glow-plugs-2.html

If the timer unit is operating properly, the reason for "no light" condition is a current imbalance between the glow plugs. The current monitor in the timer unit compares the current in the #1 GP circuit to the total current in the remaining 4 GP circuits. (4a=b+c+d+e). If the current monitor isn't happy with the current balance, it doesn't shine the light. The light is a totally artificial indication and has no direct affect on the actual operation of the GP's. Given that you have voltage to the GP's, there is no reason to suspect the relay. (Relay being defined as one component of the timer unit.) The "hard start" condition, coupled with "no light," would suggest a problem with one or more GP's or their associated wiring/connections.

+1.

I do not care for how the glow plug light operates and have modded my glow plug relay to manual control via a separate switch such that the glow plug light is on as long as the glow plugs are on. I only activate the glow plugs when the engine is dead cold (overnight start). On warm engine starts, they do not get used, extending glow plug life and reducing battery drain.

thank you!
 

i believe the timer unit is functional ... relay is working, voltage is going to plugs, sends electrons to the dash lamp with a jumper.

i also tested the resistance at each leg of the connector/wire chain, and i seem to have good continuity.

thank you for the explaining the theory of the unit. now it makes sense to me, i think. the lamp is only an indication of glow plug health relative to GP1.

when i replace all the GPs (for good measure since the vehicle had no good maintenance history ), i will test each old one and keep any good ones as back-ups.


one solution creates 3 more problems!

thanks
geoff

The monitor compares the current to #1 against the sum of the current to #2-5. When a significant imbalance is detected, the light does not illuminate. (Per the FSM, a single GP failure of #2-5 may not trip the monitor.)