I have two issues that are weird. I installed Marshall gauges - temp, boost, oil pressure. The temp gauge takes a long time to come alive after starting then constantly fluctuates - from 0 degrees to about 220/240 degrees. Today it fluctuated all day and finally came to rest at 0 while driving. The second issue is the boost gauge. It never goes above 0. Does this indicate there is no boost? It shows vacuum. Is there a way to calibrate the vacuum/boost relationship?

 

Please draw a diagram of how these gauges are connected to the car --- placement of sensors, power connections, etc.

 

I'll have to do it simpler as I included an electronic part that I didn't install. Maybe tomorrow. BTW, do you know of any electrical connectors that are weather-proof and always connect? The only thing I could find are those rubber trailer [plugs from NAPA.

 

"~ included an electrical part ~"? Other than the gauges and sensors, what else is needed to monitor coolant temp and manifold vacuum / pressure?

For weather-proof electrical connectors, take a look at RealOEM's Vehicle Electrical System, there's all kinds of connectors shown. Or, shop your local auto parts store, they'll have splicing connectors in addition to mating connectors. Are you looking for mating connectors, so a component can be removed, or a splicing connector to "hard-wire" a new component in place?

I'm not familiar with Marshall gauges but most gauge suppliers have kits with all needed electrical components. I needed to order my coolant sensor separate from the gauge, and the manifold adapter separate from the boost gauge. Both are ProSport gauges and "hard-wired" in place.

 

If the boost gauge never reads positive boost, I would suspect it was either plumbed to the wrong spot or there is a leak in the boost line, letting out boost but still holding vacuum.

The temp gauge sounds like bad wiring...

 

The answer... I took the turbosmart gauge out and replaced it with an oem diverter so I could still drive the car. The TS was broken. It was the electrical one with the Plumb-back design. The OEM valve gave me 2-9 psi of boost. At least it showed up. I replaced the original diverter with a revised valve and now I get about 10psi. Now, I am going to try a TS vacuum operated valve. If it doesn’t work properly I’m out $200+. Or a return fee. I am hoping it will work as promised and hold a bit more boost for longer.

I will report back after I install the part.

Best, DR

 

I still don’t have the drawing for you. The gauge works sometimes. I doesn’t report the temp quickly until it decides to start working. I checked all the wiring
-Common ground to all devices
-unswitched power to the correct connections + its fused separately
-switched power to the relay - I wired in a BMW relay to turn on all the electronics I added.
-signals all delivering signal to each gauge

Now here is my thinking at this point. The sender works and the gauge works. The sender is mounted to a metal connector in the top radiator hose which replaced the plastic connector for this two-piece hose. Is it possible that the thermostat opens and closes so many times that the sender is too slow to respond and the gauge reading takes a dive? After discussing this with techs at the dealer I think that the reading from the radiator hose is incorrect. The gauge needs to read the temp from the block. I have installed in my car an N18 filter body that comes with two ports. One is oil for sure as I use it to connect to my oil pressure sender. If in fact the other port is coolant and I can find the correct fittings to connect the sender I may have the problem solved.

What do you think?

 

Here's a pic of my coolant temp sensor installation. Both sensor wires go directly to the gauge while the sensor housing is wired to the chassis. Note the actual sensing element is in direct contact with coolant and is completely independent of any OEM items other than power. Also, this hose is the t'stat output to the radiator. The gauge readings follow OBD readings within 5 deg so accuracy isn't a concern. If I wanted more accuracy, the sensor can be moved closer to the t'stat, it would just be harder to access.

From the above description, it sounds like your sensor is monitoring conducted heat thru a piece of metal. Mounting a coolant sensor on the block will probably not monitor flowing coolant. Most coolant passages are too narrow to accept traditional sensors without partially blocking the already narrow passage, and I wouldn't want to measure coolant that's stuck in a pocket without flowing. It would be great for pressure readings, just not typical temp. As for the added relay, it might offer some piece-of-mind for safety / convenience / battery life, etc, but personal opinion --- it's one more piece of equipment to someday cause grief. I'm old-school and firmly believe in the KISS principle.

Maybe I misunderstand your description of how the coolant sensor is connected. A pic or two would help.

 

I think my gauge is hooked up the same as yours. However i didn't ground the metal coupler to the chassis. Isn't the black wire from the sender the ground? Here is a simplified explanation of the wiring.
Three gauges with led lighting
Tunable RPM switch - for the oil pump solenoid
Magnetic pick-up - to pick up the firing of one plug and send the information to the RPM switch
Relay - switched power turns the relay on and straight battery power runs the gauges, rpm switch, and pickup
Every item is grounded to the cigarette lighter. Switched power is from the cigarette lighter.
I'll try to re-draw the electrical plan. It changed a bit.

I don't think the sender or the gauge is malfunctioning. If the reservoir is full I am assuming the system is full but I could be wrong. I have an N18 oil filter housing on my car. It has two extra ports. One I've used to supply the oil pressure sender. I'm assuming the other is coolant. That may be my solution. I cannot understand why I cannot get a clear and consistent reading from the gauge.

 

I took the brass fittings off and screwed the sender directly into the coupling. Don't be confused...

 

Black wire from sender is definitely NOT a ground, at least on my system. It is a signal return which may or may not be connected to ground thru the gauge, dependent on the type of "bridge" circuit used by ProSport. The only "Ground" connection is for gauge power and coolant sensor housing.

How many pins are on your sensor's connector and does the cable have a shield? Or is the cable for coolant, not electrical? Hard to determine --- seeing all the plumbers tape implies there's liquid in this connection. Definitely doesn't appear as straight-forward as my ProSport system. Is this the way Marshall expects the sensor to be configured --- with a 90 deg elbow? Then, "I took the brass fittings off and screwed the sender directly into the coupling." Any chance the electrical circuit was damaged during this "modification"?

I just looked up Marshall Instruments and found this --- http://www.marshallinstruments.com/products/7474.cfm. If this is the gauge you're working with, it's significantly different from mine. They also list two different temp sensors. Are you sure you have the correct one?

 

OK, after evaluating what I found on the Marshall site, it looks like you need to connect both sides of the appropriate sensor to the gauge. From their pics, I see only one wire from the sensor. This means that for it to function correctly, the other side (the return side) of the sensor needs to be connected to the gauge. The main signal wire connection should be obvious. I believe your problem is getting the return side connected and your plumbers tape is causing an electrical connection problem. Marshall says connect the gauge signal return to ground. This will work if and only if the sensor has its return connected to the same ground. There's different methods to do this:
1. Connect a wire to the sensor body via a lug washer --- wire terminal that fits around the sensor like a washer, then ground this washer to a chassis ground. Gauge sensor connector must use the same chassis ground, not a signal ground. (Preferred method)
2. Connect a wire to the sensor body via a lug washer --- wire terminal that fits around the sensor like a washer, then connect this wire directly to the gauge sensor signal return.
3. Use the current connections WITHOUT the plumbers tape, if and only if the cable is shielded. This shield will complete the return signal when it's connected to the gauge sensor return.
There are probably other methods to complete the sensor return circuit, I've just listed the more obvious ones. I didn't learn what was "in the box" you bought from Marshall, but it appears you modified the installation with the 90deg elbow, possibly the cable, and definitely the plumbers tape. Grounding the sensor housing (the part inside the hose) is not necessary but won't hurt either. It's the actual sensor that needs to be electrically connected to the gauge, both sides. Plumbers tape is great for protecting against liquids. It also insulates electrical connections. Yes, new bolt threads will cut thru the tape, but not enough to make the good electrical connection needed for this sensitive circuit. LOSE THE TAPE everywhere except where the sensor mounts into it's housing. This housing doesn't NEED to be grounded.

Test the sensor wiring by measuring the sensor resistance with an ohmmeter at the gauge with the gauge disconnected (don't want to put an ohmmeter on the gauge circuit, just the sensor). Ohmmeter should read a value just below the 100deg sensor value. How much below depends on your ambient temperature during this test. I gotta emphasize how important it is that the sensor MUST be the one designed for the gauge. You haven't told us which part numbers you bought.

If it helps build confidence in my suggestions, I'm retired from a career in electronics manufacturing as a test and quality engineer, with a BS in Applied Physics.

 

Apologies. I'm getting back to you late. We had a tragedy in our family.
Yes it's the same SCX gauge with the sender located in the upper radiator hose. One wire is for the signal and the other goes to common ground.The gauge works, it just doesn't show the water temp in the engine. It shows temp only when the thermostat opens up and sends water to the radiator.

The boost gauge only tops out a 10 PSI. Contrary to my scanner which shows various boost levels as I drive from .7 to 1.5 Bar.

 

When the sender is located in the upper radiator hose, that's what it's going to measure --- coolant flowing from the t'stat. You want to measure coolant temp someplace else, find the fittings and relocate the sensor.

Can't comment on your scanner boost readings. My AccessPort has two different "boost" indications and they are significantly different. If your scanner isn't reading a vacuum at idle, you're monitoring the wrong sensor. The MAP sensor location on the intake manifold is the correct one to use with an external gauge. Typically an adapter is mounted under the OEM sensor and the external gauge is connected to the adapter, via the appropriate sensor.