If you have read this entire thread you know that probably at least 10% of my post are dealing with engine coolant temperatures and trying to get them down. If you haven’t read the entire thread here is a recap:
GEP 6500 engine, Banks Sidewinder turbo, 4L80E transmission, Radiators.com 2 row aluminum radiator, Leroy’s Diesel stand alone oil cooler, serpentine belt set up, Heath diesel fan, fan clutch and water pump. Plus Dakota Digital VHX gauges.
From mile 1 of driving this build the engine will hit 203°-208° on flat ground when driving 68 mph or faster with outside temps over 70°. The aluminum radiator dropped temps a degree or so and vastly quickened cooling once out of boost. The Heath set up didn’t really make a difference until I had the Dakota Digital warning system flashing 214°-217°. Then the fan clutch would lock up and that big 22 inch fan would sound like a low flying airplane on my roof. Temperatures would drop 10°-15° in less than a minute when that fan got engaged.
However, Heath advertises you should never get over 207° with their kit. I have always been 10° over that. I also called and talked to Dakota Digital several times back at the beginning about the accuracy of their temperature gauge. I could find no similar stories of Dakota Digital systems reading 5°-10° hot on the internet. I just lived with it. 36,000 miles later something happened.
I was pulling a 16 foot box trailer in relation to a Trail Life event. All was my truck usual until I stopped for lunch. 15 minutes later I fired it up on a 45° day and had EEE flashing on the display at me and a dead temperature gauge. I shut it off, checked under the hood and found nothing wrong. I ignored it and went on my way. While at an event I went through the Dakota Digital manual and discovered EEE indicates an open circuit. I wiggled the wire a few hours later before the drive home and all was well again. However while unhooking the trailer EEE came back.
I ordered a new Dakota Digital sensor and wiring harness for it from Amazon. I replaced the wiring plug and the sensor which was longer than my old one. The test drive for that had me indicating 225° when my 195° thermostat was opening on a 50° day basically just idling around. That won’t do being 30° too hot.
This time internet searches showed a lot of very similar stories. It seems all of the LS V8 conversions have created a community of people who can use the LS computer OBDII to get a coolant temp to compare against what their Dakota Digital gauge is telling them. 20°-30° greater Dakota Digital readings compared to the Chevy readings seems to be a thing. One suggestion is to cut off the end of the longer sensor at a machine mark it has grooved into it. Most find that equalizing the 2 readings.
I looked at the Amazon listing for what I bought and it said it can be cut. I did and found a bunch of white goo and cut wires inside. That sensor was destroyed. I called Dakota Digital customer service. That guy said there were reports of a bad batch of sensors and that they would send me a new one under warranty. I pointed out that I bought my gauges in 2017. He said I bought a replacement under their name that was faulty. They would send me a new one. Very pleased with the customer service.
I asked him about the reading hot trend. He said cutting the end off often does help. He also asked if I had a bushing or spacer between the sensor and the head? They only sell 1/8 NPT sensors. I thought at the time that the hole in the head of a 6.2/6.5 was 1/2 NPT. I told him I had a bushing and 1/2 NPT hole. He said forget their 1/8 NPT 2 wire sensor and get a VDO 323-060 1 wire 1/2 NPT sensor. He said with no bushing it would be cooler and how to wire it up. I went out and measured a spare head to discover our engines are 3/8 NPT. I called back to Dakota Digital and was told only their sensor with a bushing would work for our 3/8 NPT size. They sent me a replacement.
The serpentine cross over I have does not have a threaded sensor port like the J code 6.2 units do. My only option is the single port on each cylinder head.
It looked just like the one Amazon sent me and read about 20° hotter than thermostat temp. I cut the tip off successfully this time and it was around 10° hotter than thermostat temp. I did a video about this. It should post later this week. Follow the link in my signature.
Through this I began to realize my original sensor which failed open when heat soaked had been reading 5° - 10° hotter than it should have been all along. This is my only serpentine belt 6.2/6.5. It has a different coolant cross over and takes a different thermostat than the military V belt trucks. The V belt trucks have a 195° thermostat. I thought the same for the serpentine set up. Thinking my thermostat might be the issue instead of the gauge I went to order the exact same AC Delco unit I have. It is rated at 190°. That means to me that when watching the temperature gauge with a digital read out. It should rise to 190°-194° at idle and then drop rapidly to 185° or so and repeat. With a load on the engine going down the highway it should be a very steady 193° or 194°. I have never had the steady state highway temp be lower than 199° with the old sensor. Normal was 203°. Exactly the same as what the AutoMeter mechanical temperature gauge in my M715 with a 6.2 and 195° thermostat tells me.
I went through all of the Dakota Digital stuff again and found a small chart. 300 Ohms at the sensor should be 100° F. 75 Ohms should be 200°. 0 Ohms should be 300°. I configured the Dakota Digital to show me the raw Ohm reading. My sensor was indicating 67 Ohms when the thermostat would open and start flowing cool coolant. That is 8 Ohms hotter than 200° with a 190° thermostat installed. I needed a better sensor.
Dakota Digital had tried to steer me to VDO but said what I needed didn’t exist. I started searching VDO. They don’t publish Ohm ratings and part numbers together. But, they do publish part numbers with the gauge sweep and in separate charts publish gauge sweep with Ohm ratings. All of the VDO 90° sweep gauge sensors are 328 Ohms to 18 Ohms with a 300° max reading. They also only offer 1/8 NPT or 1/2 NPT sizes. I ordered a VDO 323-057 single wire sensor in 1/8 NPT.
I installed it, wired it up and warmed up the engine with the Dakota Digital set to read Ohms. 83 Ohms is when the thermostat would open. Compared to 67 Ohms with the Dakota Digital sensor. 75 Ohms is supposed to be 200°. So, the VDO was 8 Ohms cooler than the 200° line and the Dakota Digital sensor was 8 Ohms hotter.
I put it in digital temperature read out mode and went for a drive. 70° day with the ac on. Sitting at a stop sign it would get up to 191° and go back down. Accelerating with 2-3 pounds of boost and it would hit 198° But go back down once I reached a steady speed. Cruise control set at 76 mph with 1-2 psi of boost and it was rock steady at 194° never moving for the 15 miles I drove a flat highway out of town. I slowed to turn around and down to 192° it would go while coasting. 198° while getting back up to 76 mph. Steady ay 194° all the way back to town.
The original sensor would have been showing 204° or so in the same conditions. A 10° difference in read out to me is amazing. It gets the temperatures in line with everything I have read about how it should be performing. It also gets me right at the 207° max temp you will see claim by Heath Diesel.
Heath also claimed the AC Delco thermostat I have starts to open at 191° and is fully open at 202°. That tracks with the numbers I am now seeing.
The point of writing all this out? I like mechanical gauges! Getting into military vehicles it was difficult for me to accept the electric sensor gauges. I kept telling myself the “They are safer” argument. Having been subject to oil psi hoses leaking on my legs before I can understand that. This is my only complaint against the Dakota Digital unit. Everything else has been above expectations. The new units, which cost more than double what I paid 8 years ago, will Bluetooth into your phone and apparently you can calibrate them that way. Mine doesn’t have that feature. I guess the point is if all your information and seat of the pants feeling say one thing and 1 item is off. Count on that 1 item actually being off.