bmwm3coupe
08-05-2008, 06:57 PM
If you have a Zionsville Radiator with the electronic fan options (single speed or dual speed), here is something that could/should extend the life of your cooling system.
When researching the radiator failure in our E39's it became obvious that failures of the radiator, expansion tank, and/or hoses happened more often AFTER you turn off the car. This was something that happen to me and to the prior owner of the car as well (both times using the OEM piece-of-crap radiator).
When talking to the folks at Zionsville about this, they explained to me that what happens is that once you turn off the engine and remove the key, the engine is still pretty much as hot as it was while running, but that now the radiator does not have any forced air to keep it cool and control the water pressure in the cooling system. Therefore, with no "active" cooling, the water temp rises and the pressure rises since the engine stays warm for a relatively LONG time. Most of the time, the system can cope with this extra higher pressure, but as parts get older, they fail: this is why most often the radiator/reservoir/hoses will fail after you turn off the engine.
The problem does not go away with the outstanding Zionsville radiator, since as soon as you turn off the engine and remove the key, the electric fan is no longer enabled. However, I noticed that when I got home, when I turned off the engine, but left the key in the ignition, that the electric fan will either stay running, or that it would come ON by itself, sometimes on HIGH and then LOW, or sometimes on LOW. What was amazing, is that this cycling of the fan turning ON by itself (with the key in the ignition and the engine off) keep going on for a while.
So over the course of several months I experimented to see how often this turning ON/OFF would take place, and found that it seems to eventually stay off after 15-20 minutes or so. Of course, I can't simply leave the key in the ignition every day since this will drain the battery after 4-5+ hours since lots of circuits are active while the key is in the ignition, even with the AC FAN and Radio OFF (don't ask me how I found this out - many times!).
So the ideal situation would be to have a circuit that would "latch" the ignition system signal so that even if you remove the key the circuit would keep the fans enabled for them to keep the cooling system from overheating - basically allowing the fan to slowly cool the radiator and preventing the water pressure from increasing too much > the ultimate goal of course being to extend the life of the cooling system without draining the battery. Basically we need a FAN DELAY CIRCUIT.
I started playing with some paper designs:
http://m3coupe.com/E39/Radiator/img_5867.jpg
until I came with the final circuit (I used a circuit emulator in Linux to verify operation before building the actual circuit):
http://m3coupe.com/E39/Radiator/img_5868.jpg
If you are curious about the theory of operation, read on. If not, just skip it. Although in principle a simple RC is all that is needed to latch a signal, in practical terms is not feasible to use a simple RC to do this - the problem is current. You need current to keep things like relays enabled, specially since a relay could use anywhere from 100-****A to stay "ON" (my relay needed 135mA). The solution is simple - isolate the RC from the actual relay by using a transistor. Of course, it can't be just your generic NPN or PNP transistor, since those also need current through the base/emiter junction to stay "ON". I therefore used an FET transistor that needs a voltage differential only - no current to stay ON - this allows the RC to be completely isolated. So now, looking at the circuit, I have a pull-down resistor to make sure the input to the circuit does not "float" once the ignition key is removed. I also have a simple plain diode to make sure I can't send/discharge any current into the car's computer (that could be bad!). This diode then has (not shown in the paper design) an inline resistor to (along with the CAP) provide the charging time (I selected something like 5 seconds or so - this is how long the key has to be in the ignition for the cap to fully charge). Then the RC for the discharge - values for about 15-20 minutes provide the voltage necessary to keep the transistor ON. While the transistor is ON, the output relay is activated, which provides 12V+ to the Zionsville circuit - this basically mimics what the Zionsville circuit is expecting to see from the car - a 12V+ signal when the key is in the ignition. Lastly, I added a simple LED on the output side so that I could have a visual signal that the circuit is latched. Not the most elegant and simplest circuit, but it works great.
I then tested the circuit on a breadboard:
http://m3coupe.com/E39/Radiator/img_5869.jpg
and tested the operation with a digital storage scope with shorter values - the goal was about 15-20 minutes but I used a fraction of this to test values on the bench as I did not wanted to have to wait 15-20 min after every change!. The yellow trace is the "ignition signal" the blue trace is the Voltage of the storage Capacitor, and the pink signal is the output which activates the relay:
http://m3coupe.com/E39/Radiator/img_5872.jpg
I then found a suitable box and built the actual circuit:
http://m3coupe.com/E39/Radiator/img_5873.jpg
This sliding switch on the corner is the by-pass. When active, the ignition signal is directly connected to the original Zionsville circuit - nothing changes. When the switch is on "active", the circuit latches the ignition signal, and provides this signal to the Zionsville circuit for about 15-20 minutes). The green LED stays lit to indicate the circuit is latched:
http://m3coupe.com/E39/Radiator/img_5874.jpg
I then inserted the new circuit in the car (of course everything is clearly labeled):
http://m3coupe.com/E39/Radiator/img_6060.jpg
http://m3coupe.com/E39/Radiator/img_6061.jpg
I have been testing the circuit for a couple of weeks now and it works perfectly - exactly as I envisioned. Now, when I turn off the car and remove the key, the fan will have 15-20 minutes to bring the cooling system's temp to a "normal" level, and after 15-20 minutes the circuit disables itself, preventing any possible drainage on the battery.
That's it. I hope some of you will find this useful :-)
When researching the radiator failure in our E39's it became obvious that failures of the radiator, expansion tank, and/or hoses happened more often AFTER you turn off the car. This was something that happen to me and to the prior owner of the car as well (both times using the OEM piece-of-crap radiator).
When talking to the folks at Zionsville about this, they explained to me that what happens is that once you turn off the engine and remove the key, the engine is still pretty much as hot as it was while running, but that now the radiator does not have any forced air to keep it cool and control the water pressure in the cooling system. Therefore, with no "active" cooling, the water temp rises and the pressure rises since the engine stays warm for a relatively LONG time. Most of the time, the system can cope with this extra higher pressure, but as parts get older, they fail: this is why most often the radiator/reservoir/hoses will fail after you turn off the engine.
The problem does not go away with the outstanding Zionsville radiator, since as soon as you turn off the engine and remove the key, the electric fan is no longer enabled. However, I noticed that when I got home, when I turned off the engine, but left the key in the ignition, that the electric fan will either stay running, or that it would come ON by itself, sometimes on HIGH and then LOW, or sometimes on LOW. What was amazing, is that this cycling of the fan turning ON by itself (with the key in the ignition and the engine off) keep going on for a while.
So over the course of several months I experimented to see how often this turning ON/OFF would take place, and found that it seems to eventually stay off after 15-20 minutes or so. Of course, I can't simply leave the key in the ignition every day since this will drain the battery after 4-5+ hours since lots of circuits are active while the key is in the ignition, even with the AC FAN and Radio OFF (don't ask me how I found this out - many times!).
So the ideal situation would be to have a circuit that would "latch" the ignition system signal so that even if you remove the key the circuit would keep the fans enabled for them to keep the cooling system from overheating - basically allowing the fan to slowly cool the radiator and preventing the water pressure from increasing too much > the ultimate goal of course being to extend the life of the cooling system without draining the battery. Basically we need a FAN DELAY CIRCUIT.
I started playing with some paper designs:
http://m3coupe.com/E39/Radiator/img_5867.jpg
until I came with the final circuit (I used a circuit emulator in Linux to verify operation before building the actual circuit):
http://m3coupe.com/E39/Radiator/img_5868.jpg
If you are curious about the theory of operation, read on. If not, just skip it. Although in principle a simple RC is all that is needed to latch a signal, in practical terms is not feasible to use a simple RC to do this - the problem is current. You need current to keep things like relays enabled, specially since a relay could use anywhere from 100-****A to stay "ON" (my relay needed 135mA). The solution is simple - isolate the RC from the actual relay by using a transistor. Of course, it can't be just your generic NPN or PNP transistor, since those also need current through the base/emiter junction to stay "ON". I therefore used an FET transistor that needs a voltage differential only - no current to stay ON - this allows the RC to be completely isolated. So now, looking at the circuit, I have a pull-down resistor to make sure the input to the circuit does not "float" once the ignition key is removed. I also have a simple plain diode to make sure I can't send/discharge any current into the car's computer (that could be bad!). This diode then has (not shown in the paper design) an inline resistor to (along with the CAP) provide the charging time (I selected something like 5 seconds or so - this is how long the key has to be in the ignition for the cap to fully charge). Then the RC for the discharge - values for about 15-20 minutes provide the voltage necessary to keep the transistor ON. While the transistor is ON, the output relay is activated, which provides 12V+ to the Zionsville circuit - this basically mimics what the Zionsville circuit is expecting to see from the car - a 12V+ signal when the key is in the ignition. Lastly, I added a simple LED on the output side so that I could have a visual signal that the circuit is latched. Not the most elegant and simplest circuit, but it works great.
I then tested the circuit on a breadboard:
http://m3coupe.com/E39/Radiator/img_5869.jpg
and tested the operation with a digital storage scope with shorter values - the goal was about 15-20 minutes but I used a fraction of this to test values on the bench as I did not wanted to have to wait 15-20 min after every change!. The yellow trace is the "ignition signal" the blue trace is the Voltage of the storage Capacitor, and the pink signal is the output which activates the relay:
http://m3coupe.com/E39/Radiator/img_5872.jpg
I then found a suitable box and built the actual circuit:
http://m3coupe.com/E39/Radiator/img_5873.jpg
This sliding switch on the corner is the by-pass. When active, the ignition signal is directly connected to the original Zionsville circuit - nothing changes. When the switch is on "active", the circuit latches the ignition signal, and provides this signal to the Zionsville circuit for about 15-20 minutes). The green LED stays lit to indicate the circuit is latched:
http://m3coupe.com/E39/Radiator/img_5874.jpg
I then inserted the new circuit in the car (of course everything is clearly labeled):
http://m3coupe.com/E39/Radiator/img_6060.jpg
http://m3coupe.com/E39/Radiator/img_6061.jpg
I have been testing the circuit for a couple of weeks now and it works perfectly - exactly as I envisioned. Now, when I turn off the car and remove the key, the fan will have 15-20 minutes to bring the cooling system's temp to a "normal" level, and after 15-20 minutes the circuit disables itself, preventing any possible drainage on the battery.
That's it. I hope some of you will find this useful :-)
