1KZ-TE Glow system questions (1 Viewer)

[bloc]

Hi all,

I'm in the process of swapping my 1KZ-TE injection pump to a zexel mechanical unit to ditch most of the electronics, and add more reliability, predictability, and parts availability. Will be writing up the required modifications, but that's a little later.

I was trying to figure out how to control the glow plugs and out of curiosity checked the voltage from the ECU to the glow plug control relay and noticed some unexpected behavior. The voltage of the control circuit doesn't turn on then off. It starts high and gradually lowers. The relay appears to only turn on then off, with 10v being supplied to the plugs when they are on, and it turning to 0v abruptly. This is all in relatively warm conditions (about 90F), I'm in the process of trying to figure out how to trick the ecu into thinking things are cold to check for appropriate data then.

Basically I was wondering if anyone had knowledge of this system and why it is set up in the way it is. Why would Toyota design it like this? Is this similar to the "Super Glow" system used on cruisers? or the same thing?

I will most likely end up setting up the factory Glow Plug relay to be controlled manually, but I want to make sure things will work like that.

Also, is some kind of resistor in the relay responsible for stepping the voltage to the plugs down to 10v? Will I burn them up if I use a different (more robust) relay that will supply them 12v?

Thanks for any help,

Justin

 

[toddm]

I just checked mine. 10.6V at the Glow plugs, same as the voltage into the GP relay. Relay has .3ohms resistance, so its either the wire between the 80A fusible link or there is an actual resistor in the fuse block, which I doubt. The wire supplying power to the GP relay is actually smaller than the wire that goes on to the Glow Plugs. Its 104F here and the Glow plugs stayed on for about 5 seconds. I think the only thing that changes with colder temps is the time the relay stays on.

If I were you I would add a momentary switch for the glow plugs, and hit it when you need it. You may not even need them till November.

Todd

 

[toddm]

I wanted to measure the voltage at the Fuse box end of the wire, but its too hot out!

 

[bloc]

I had the fuse box completely apart when I combined the JDM and USDM wiring, and don't remember a resistor at all. I did notice that the positive voltage supply wire going to the glow plug relay was pretty hot after cycling them a few times in the apartment garage trying to check voltages and stuff. It must be the wire itself that provides the resistance.

According to the 1KZ-TE manual, it should run the plugs for 6 seconds up until the water temp is about 130F. It lists this as "after glow time", which I would take to mean "after the engine is running", but says it should do it even when you just turn on the ignition (ie, engine not running).

Also, I remember something about the plugs staying partially on for a couple minutes once the engine is started? I need to look into this now that I vaguely remember that detail.

It clearly says not to apply more than 11v to the plugs in the manual, so that answers one of my above questions. I just wonder why they decided to use a long run of (hot) wire to get that reduced voltage?

And, why does it slowly lower the voltage from the ECU on the GP control circuit, instead of on then off?

 

[toddm]

I saw the voltage actually rise on the glow plugs, from 10.4 to 10.6 over a few seconds. I'm not sure why the ECU output voltage would drop, unless the battery voltage was dropping? I did my tests with the engine off and saw Batt. voltage go from 12.57 to 12.00. I have both Batteries. That indicates quite a load on the system. Since the ECU is controlling a relay I doubt it has any fancy controls for it other than ON or OFF.

You are probably right, the wire from the fuse box could be the resistor.

According to the graph in the engine manual (page ST-2) the "light lighting time" (I assume this is the dash Glow plug lamp) could be as long as 5 seconds if the coolant temp is somewhere around 0.0 F and only a second after about 40ish degrees F. This would allow the Glow Plugs to get up to temp even when its really cold out. Then the "afterglow time" is up to 2 minutes until the coolant temp is above 122F, dropping rapidly in time to about 1 second at coolant temps above 131F. Afterglow time probably starts when the engine does.

I think it also stated that these times are calculated at the time the ECU is powered, meaning that if if the coolant temp rises above 130 before the 2 minutes are up the ECU will Keep the GP's on till the timer runs out.

The Graph on page DI-55 is slightly contradictory to the Graph on page ST-2. Its also alot harder to read.

Todd

 

[bloc]

Got a chance to mess with it some more today.

First off, the ECU definitely steps down the control voltage gradually. Beginning at about 95F ambient temp, with the engine not warmed up, I removed the GP relay and turned the ignition on with a multimeter tapped into the control signal for the relay. I saw about 20 seconds of full 12.6v, then it gradually lowered to about 1v over the next minute or so. The voltage dropped quickly at first, then more slowly. If you were to graph it it would have an elliptical curve. I got impatient before I waited to see just how low the voltage would go.

Interestingly, when I started the truck, it did full battery voltage (~13.6v now that the alternator was charging) for just less than 2 minutes, then when the relay would normally turn off, the control voltage from the ecu dropped off slowly, just as before. It's almost like the ECU, under all conditions, cannot just turn the GP circuit OFF. it must do so gradually.

I didn't have the right resistors, so i tried unhooking the water temp sensor for the ECU to simulate cold conditions, but it just detected it as a bad sensor and turned the GP on then off nearly immediately under both ON and START conditions.

The supply voltage to the GP relay was measuring battery voltage (12.6-13.6 depending on when the engine was running) and the glow plugs showing 10v when on, so the wire must indeed be the resistor.

I was leaning toward simply running the regular GP relay off a momentary switch (as you suggested), now that I've figured out the GP control doesn't have any fancy resistors or voltage control, That's almost certainly what I'll be doing.

I'd like to figure out the relationship between the 2 minute GP cutoff and lots of 1KZ-TE engines immediately starting to run like crap at altitude/cold.. but then thats a big part of why I'm ditching the electronic pump and ECU in the first place.

Can't wait to get rid of all of those unnecessary wires.

 

[toddm]

Got a chance to mess with it some more today.

First off, the ECU definitely steps down the control voltage gradually. Starting at about 90F, with the engine not warmed up, I removed the GP relay and turned the ignition on with a multimeter tapped into the control signal for the relay. I saw about 20 seconds of full 12.6v, then it gradually lowered to about 1v over the next minute or so. The voltage dropped quickly at first, then more slowly. If you were to graph it it would have an elliptical curve. I got impatient before I waited to see just how low the voltage would go.

Interestingly, when I started the truck, it did full battery voltage (~13.6v now that the alternator was charging) for just less than 2 minutes, then when the relay would normally turn off, the control voltage from the ecu dropped off slowly, just as before. It's almost like the ECU, under all conditions, cannot just turn the GP circuit OFF. it must do so gradually.

I didn't have the right resistors, so i tried unhooking the water temp sensor for the ECU to simulate cold conditions, but it just detected it as a bad sensor and turned the GP on then off nearly immediately under both ON and START conditions.

The supply voltage to the GP relay was measuring battery voltage (12.6-13.6 depending on when the engine was running) and the glow plugs showing 10v when on, so the wire must indeed be the resistor.

I was leaning toward simply running the regular GP relay off a momentary switch (as you suggested), now that I've figured out the GP control doesn't have any fancy resistors or voltage control, That's almost certainly what I'll be doing.

I'd like to figure out the relationship between the 2 minute GP cutoff and lots of 1KZ-TE engines immediately starting to run like crap at altitude/cold.. but then thats a big part of why I'm ditching the electronic pump and ECU in the first place.

Can't wait to get rid of all of those unnecessary wires.

Could you try the test on the control wire again, but with the relay plugged in? I suspect the voltage was decaying slowly because the multimeter was not discharging the circuit very quickly or at all after the transistor in the ECU switched off. I don't think the ECU would try to modulate the output since its driving a relay, which will only open or close the connection to the GP's. I suppose it could use a PWM signal, but a relay is not really designed for that kind of duty. Also it says in the engine manual that the GP's are ceramic so no current control is performed.

Todd

 

[bloc]

I could, but will need to find a way to do it without hacking up the waterproofed connections or wiring.

Would the type of multimeter effect the results of said decay? This is a high-dollar fluke true RMS meter?
would breaking the connection of the multimeter (without the relay) then hooking it back up during the early part of the voltage decline "zero" it or tell you anything?

 

[toddm]

I don't think the multimeter matters too much. But I don't really know. I'm sure they are designed to be as transparent to the circuit you are testing as possible.

Ill test mine tomorrow. I have the ECU out and open so I can clip onto the S-REL pin internally without jacking up the sealed connectors.

Todd

 

[toddm]

I tested the S-REL pin with the Glow Plug relay in place, truck and ECU fully functional. I got 11.8VDC on the output for 17.5 seconds according to the logic analyzer. The multimeter also showed 0.0VDC right when I heard the GP relay click off. Voltages measured with a Fluke. Negative on ground, positive on S-REL. The coolant was at ambient temperature as the truck had not been started before the test.

It was 108F here today, Probably only 98-99F when I did the testing. Based on the graph on the glow time on page ED-55, my coolant temperature was about 100F. based on the graph on page ST-5 My coolant temp was almost 130F. I think I'll use the graph on ED-55!

Todd

 

[bloc]

So a voltmeter uses so little current that it takes over a minute for the voltage to bleed off when there is no other use of the current?

great.

 

[toddm]

How did you have the multimeter hooked up?

 

[bloc]

pulled the relay out, stuck the + dmm lead into the + relay control terminal, and - dmm lead into the ground terminal.

 

[NorCalBorn]

J, off-topic..

How's life!! You have a pm.