Alternative to Final Stage/Fan Controller For Engine Cooling.
#18
Nordschleife Master
Thread Starter
Thanks John.
A lot of research and that went in to this. Alan's help was invaluable. Conversations with Brian Baskin of dc control was also incredibly helpful and reassuring. He was familiar with the common problem of amplifier failures in older units. He states it was not just the FET, but the bipolar switch on the board that would heat till its demise. He said in the 'old days' they would but these two components in proximity on the boards and the heat would cause the damage/failures. He said on his units his components are of newer technology and placed in such a way on his boards so that internal overheating is not an issue. He also rates his boards to withstand 155 C/ 310 F deg ambient temps.
It's only day one, but the function when I attempted to stress the new system passed with flying colors.
A lot of research and that went in to this. Alan's help was invaluable. Conversations with Brian Baskin of dc control was also incredibly helpful and reassuring. He was familiar with the common problem of amplifier failures in older units. He states it was not just the FET, but the bipolar switch on the board that would heat till its demise. He said in the 'old days' they would but these two components in proximity on the boards and the heat would cause the damage/failures. He said on his units his components are of newer technology and placed in such a way on his boards so that internal overheating is not an issue. He also rates his boards to withstand 155 C/ 310 F deg ambient temps.
It's only day one, but the function when I attempted to stress the new system passed with flying colors.
#19
Rennlist Member
Join Date: Oct 2005
Location: Gatineau, Québec, Canada
Posts: 5,186
Received 1,335 Likes
on
506 Posts
Great job Craig and all.
It looks like a very nice alternative for our cars.
I don't have a problem with mine yet but it will be real nice to have something like this if/when needed.
+1 to what Jon said.
A good wright-up "à la Dwayne" and/or a complete kit with instructions would be great too for newbies like me!
Thanks again Craig.
It looks like a very nice alternative for our cars.
I don't have a problem with mine yet but it will be real nice to have something like this if/when needed.
+1 to what Jon said.
A good wright-up "à la Dwayne" and/or a complete kit with instructions would be great too for newbies like me!
Thanks again Craig.
Last edited by Bertrand Daoust; 08-19-2014 at 09:19 AM.
#20
Drifting
Craig -
It takes 4 weeks for delivery of the DC unit. While waiting I could use your list of materials to get all the other pieces together. If you don't mind helping me along, I'll create the DIY thread. I'll be back at work, so I'll be doing it in small stages on weekends (easy for you to guide me in small steps). I'm anticipating I'd do it over 3 or 4 weekends, taking time to take good pictures and write up each step thoroughly.
It takes 4 weeks for delivery of the DC unit. While waiting I could use your list of materials to get all the other pieces together. If you don't mind helping me along, I'll create the DIY thread. I'll be back at work, so I'll be doing it in small stages on weekends (easy for you to guide me in small steps). I'm anticipating I'd do it over 3 or 4 weekends, taking time to take good pictures and write up each step thoroughly.
#22
Rennlist Member
For those playing along, Derale licensed the DC Controls design, and it is available under model number 16795 at your performance shop of choice. Summit is stocking them.
#23
Nordschleife Master
Thread Starter
Jon,
Of course I'd lend a hand.
This job is mostly laborious, tedious work. On the scale of working on the 928, a TB/WP job ranks with changing a transmission and is a big job. Ones first and second intake removal is less - the fifth and sixth even less. This job is is even less than that, it's just a lot of small electrical work (quality connections and feeding of wires through tight spaces).
An equipment list will vary with the operator. You may like one type of connecting/crimping union, and I may like another. You may have your 'system' of joining an 18g to a 10g wire, and I may have another...and so on.
There's only two areas in this install that were thought provoking and a kind of 'pick your poison' type of thing. This was where I had considered contingencies that I spoke somewhere on this topic. The first was the crucial seating of the radiator sensor. I chose my 'A' - you may chose your 'B'. I thought my 'A' was reasonable, but you may improve on it and use different methods and materials. The other was feeding from inside the car the current line of the 87 off the added relay for the car's sensors to get it to the dc controller under the hood. I played for over an hour to try to get it through the wiring grommets above the CE Panel without success. I considered other options, but honestly I did not want to do an entire console/dash/pod dissection for this - I had another contingency already considered and was going to safely make it work. I chose to drill a 0.25in hole between the hardlines from under the car and come up strategically in the passenger footwell. A rubber grommet 9/16 OD x 1/4 ID fit nicely and my 10 awg wire fit well through it. I slathered some urethane over the opening from the underside and light would not pass through. I had about 3 ft. of some very good rubber tubing/fiberglass lined from Pegasus Racing and this is what protects the line. It fits well between the hardlines and is nylon tied in place. This was the poison that I chose - others may choose of different one.
This job is not technically or electrically over taxing, but one's poisons are highly individualized and how you achieve your goal.
Of course I'd lend a hand.
This job is mostly laborious, tedious work. On the scale of working on the 928, a TB/WP job ranks with changing a transmission and is a big job. Ones first and second intake removal is less - the fifth and sixth even less. This job is is even less than that, it's just a lot of small electrical work (quality connections and feeding of wires through tight spaces).
An equipment list will vary with the operator. You may like one type of connecting/crimping union, and I may like another. You may have your 'system' of joining an 18g to a 10g wire, and I may have another...and so on.
There's only two areas in this install that were thought provoking and a kind of 'pick your poison' type of thing. This was where I had considered contingencies that I spoke somewhere on this topic. The first was the crucial seating of the radiator sensor. I chose my 'A' - you may chose your 'B'. I thought my 'A' was reasonable, but you may improve on it and use different methods and materials. The other was feeding from inside the car the current line of the 87 off the added relay for the car's sensors to get it to the dc controller under the hood. I played for over an hour to try to get it through the wiring grommets above the CE Panel without success. I considered other options, but honestly I did not want to do an entire console/dash/pod dissection for this - I had another contingency already considered and was going to safely make it work. I chose to drill a 0.25in hole between the hardlines from under the car and come up strategically in the passenger footwell. A rubber grommet 9/16 OD x 1/4 ID fit nicely and my 10 awg wire fit well through it. I slathered some urethane over the opening from the underside and light would not pass through. I had about 3 ft. of some very good rubber tubing/fiberglass lined from Pegasus Racing and this is what protects the line. It fits well between the hardlines and is nylon tied in place. This was the poison that I chose - others may choose of different one.
This job is not technically or electrically over taxing, but one's poisons are highly individualized and how you achieve your goal.
#24
#25
I haven’t used it, but this looks pretty factory for a car of that era http://www.delcity.net/store/Fibergl...-Loom/p_795096
With regard to the dual fan reliability, the fuses aren’t going to work, the controller looks at the impedance of the load to determine whether or not to output power. If it sees a short circuit, the output current will be limited to about 200 mA until the short circuit is removed. Pins 10 and 11 provide an LED driver that’s taken from the oscillator. Any two pin LED will work and will glow green when the controller starts, will fade from green to yellow at about 25% power and from yellow to red at full power. The advantage is that full-off to full-on of the controller is only seven degrees, so if the controller is set to 180, the LED will glow bright red by about 184 degrees. If something isn’t right, it’s a lot more indicative than a gauge.
With regard to temperature, I think there is some loss in the interpretation. Reliability can be determined by the temperature of the silicon in an active device, and that temperature depends on the device. For a Field effect transistor, the max temperature is 175C (347 F), the mean-time-before-failure (MTBF) at that point is 50,000 hours. For every 10 degrees C that the operating temperature is below that, the life span doubles. The one caveat to that is that operating at this temperature is precarious, in that it will fail only a few degrees hotter than that. For a design for industrial use, a maximum operating temperature of 155 C (311 F) under the worst-case conditions is used in the design, the MTBF is then determined to be 50,000 x 2^2 = 200,000 hours, which is about 22 years of continuous usage at the worst case temperature. At that point 63% will have failed.
A bipolar transistor has a maximum operating temperature of 150C, and most of the integrated circuits have at least some bipolar circuitry. Normally, these are packaged with the FET, so they will run at the same temperature. I chose to separate the two, so that the control board, which contains the integrated circuitry, will run at a much lower temperature.
The reason for the 260A current capability is not the capability itself, but that a high current FET also has a low RDs (equivalent series resistance) and, since that acts as a voltage divider between the controller and the load, it will determine the efficiency of the controller. This particular FET has an on-resistance of .002 ohms, which is about .003 ohm hot.
The goal is to have the controller operate at the lowest possible temperature, and you can determine that temperature in an application pretty effectively by measuring the heat sink temperature. I would guess in this case that it will be about 150F. The controller will shut itself off at 250F, and cycle back on when it cools ten degrees.
The reason the fans are so quiet is that the airflow is a function of power in watts^.333, so for two fans to flow the equivalent of one fan at full power, each as to only run at 12.5% of that power
With regard to the dual fan reliability, the fuses aren’t going to work, the controller looks at the impedance of the load to determine whether or not to output power. If it sees a short circuit, the output current will be limited to about 200 mA until the short circuit is removed. Pins 10 and 11 provide an LED driver that’s taken from the oscillator. Any two pin LED will work and will glow green when the controller starts, will fade from green to yellow at about 25% power and from yellow to red at full power. The advantage is that full-off to full-on of the controller is only seven degrees, so if the controller is set to 180, the LED will glow bright red by about 184 degrees. If something isn’t right, it’s a lot more indicative than a gauge.
With regard to temperature, I think there is some loss in the interpretation. Reliability can be determined by the temperature of the silicon in an active device, and that temperature depends on the device. For a Field effect transistor, the max temperature is 175C (347 F), the mean-time-before-failure (MTBF) at that point is 50,000 hours. For every 10 degrees C that the operating temperature is below that, the life span doubles. The one caveat to that is that operating at this temperature is precarious, in that it will fail only a few degrees hotter than that. For a design for industrial use, a maximum operating temperature of 155 C (311 F) under the worst-case conditions is used in the design, the MTBF is then determined to be 50,000 x 2^2 = 200,000 hours, which is about 22 years of continuous usage at the worst case temperature. At that point 63% will have failed.
A bipolar transistor has a maximum operating temperature of 150C, and most of the integrated circuits have at least some bipolar circuitry. Normally, these are packaged with the FET, so they will run at the same temperature. I chose to separate the two, so that the control board, which contains the integrated circuitry, will run at a much lower temperature.
The reason for the 260A current capability is not the capability itself, but that a high current FET also has a low RDs (equivalent series resistance) and, since that acts as a voltage divider between the controller and the load, it will determine the efficiency of the controller. This particular FET has an on-resistance of .002 ohms, which is about .003 ohm hot.
The goal is to have the controller operate at the lowest possible temperature, and you can determine that temperature in an application pretty effectively by measuring the heat sink temperature. I would guess in this case that it will be about 150F. The controller will shut itself off at 250F, and cycle back on when it cools ten degrees.
The reason the fans are so quiet is that the airflow is a function of power in watts^.333, so for two fans to flow the equivalent of one fan at full power, each as to only run at 12.5% of that power
#26
Electron Wrangler
Lifetime Rennlist
Member
Lifetime Rennlist
Member
I haven’t used it, but this looks pretty factory for a car of that era http://www.delcity.net/store/Fibergl...-Loom/p_795096
Alan
#29
Nordschleife Master
Thread Starter
Just a tidbit I remembered which may be of import for installers. The red 3 pin connector provided by dc connects to pins 12 - 14.
This connector is for:
12 - Designed for additional input for augmented fan speed by 10% for AC - can be used for other analog voltage such as the 928's sensors that I used them for.
13 - Testing pin for voltage values given by dc when adjusting radiator temp fan activation.
14 - Pin for standard AC activation of fans to 50% speed.
These pins have very small caliber wires. These wires would likely be adequate for small amperage such as that off the hvac relay. In my case I had tapped into pin 86 (switching pin off the hvac relay) from my secondary hvac relay. Pin 87 has jump post voltage and amperage and might have been to 'overpowering' for these very small wires built into to this connector.
This connector is for:
12 - Designed for additional input for augmented fan speed by 10% for AC - can be used for other analog voltage such as the 928's sensors that I used them for.
13 - Testing pin for voltage values given by dc when adjusting radiator temp fan activation.
14 - Pin for standard AC activation of fans to 50% speed.
These pins have very small caliber wires. These wires would likely be adequate for small amperage such as that off the hvac relay. In my case I had tapped into pin 86 (switching pin off the hvac relay) from my secondary hvac relay. Pin 87 has jump post voltage and amperage and might have been to 'overpowering' for these very small wires built into to this connector.
#30
Electron Wrangler
Lifetime Rennlist
Member
Lifetime Rennlist
Member
These pins have very small caliber wires. These wires would likely be adequate for small amperage such as that off the hvac relay. In my case I had tapped into pin 86 (switching pin off the hvac relay) from my secondary hvac relay. Pin 87 has jump post voltage and amperage and might have been too 'overpowering' for these very small wires built into to this connector.
As I said - you could have reused the wires between the inside fan controller and the final stage for this purpose - this is what makes most sense here - those wires are redundant otherwise anyway.
Alan