73StreetRace

My work is almost finished. I should end the final details next Friday.
Here are some new pictures :

The electronic box ( fans controller ). It will be mounted on the frame, in front of the radiator :
This is the true color of the box ( Red anodized look varnish ) :


The electronic box ( inside view ) :



The home made fans shroud ( before wiring ) :



The same fans shroud after wiring : The color, as it appears on the picture, is false.
Actually, it has the same color as the electronic box ( Red anodized look varnish ) :



Excuse me the the bad quality of pictures, but the autofocus of my digital camera seems to work bad since a few days

For people who didn't see my previous post last week :

Aluminum fans shroud before painting :





The electronics :





I will post new pictures of the engine bay with all this inside next week
I just hope it will work as well as it looks : great

Theiskell

Nice work

mrvette

Looks super nice, whos' electronics??

gkull

Very nice fan shroud. How much would you sell me one for?

I have the so so Dewits upper and lower sheet metal pieces

78Vette-SA

Your Schema Du Circuit Rocks! I don't understand it, but it rocks.

Definately some talent and skills there. Looks great!

73StreetRace

Electronics by myself
I'm working on a new controller for these two fans, PWM this time instead of direct on/off operation. But I will wait to see if the first controller works correctly or not. If anybody is interested in building a PWM fan controller, here's a diagram :

mrvette

Damn man, U do ok, even IF your names for stuff are a bit funny....

I never was convinced any SS PWM or other controller than a relay was necessary for cooling fans....but long as it works well, that is the rub.....question, are you running it off the battery directly, or off the alt side of any wiring...??

for instance, I wire my fan heavy currents through a short fuse link to the one relay doing both Spals....

they howl like a hurry caine, but WTF, it's FLORIDA, right at home....

73StreetRace

Here is a picture to show you how i'm running :





Now, here are some explanations about the PWM controller circuit. First of all, I have to say that I didn’t build yet this circuit. It is still a work in progress. All what I can say is that it should work as it is , but it could be necessary to make some slight adjustments to make it work reliably.
Note that this circuit can of course work with a single fan, twice powerful.
You can easily find all the characteristics of the electronic components of my diagram on the net, with the help of google. Now how it works :
1/. The power output circuit :
I decided to use 2 MOSFET ( HEXFET ) power transistors : the IRFZ48N can hold 55V ( peak ), 64 Amp ( continuous ) and 400A peak current ( during a few nanoseconds ! ) The internal impedance is only 0.014 Ω. It uses the common TO220 case. Two of them should be large enough to drive two electrical DC fans, even the biggest ones available. We also need an external Schottky power diode to absorb the peak voltage generated by the commutation of the charge ( 2 DC fan motors can be a really inductive load ! ). I chose a MBR1545CT. There are 2 diodes in the same case ( TO220 ). We only need one, so we will use both in parallel. These transistors will act as switches. Each of them will work 50% of the time, and each of them will be off when the other one is on. I used this principle because each transistor will only “see” 50% of the maximum current and it will give them more time to evacuate the heat. Another factor was the use of the SG3525 for generating the PWM. This chip is not dedicated to DC motor control ( actually it is designed for DC power supplies ) but it is often used in robotics ( fighting robots) : low power consumption, almost 0% to 100% of duty cycle, very reliable and it can be used with a single power supply ( most DC motor controller chips use a symmetrical power supply ). A 3300µF ( or more ) condenser, placed near the power MOSFETs will avoid the peak voltage generated when the inductive load is switched off by the transistors.
2/. The MOSFET drivers :
I used the EL7212C chip to drive the 2 MOSFETS, because the SG3525 can only give 200mA on each totem pole. This current would not be large enough to drive the MOSFETS with a decent slew rate.
The EL7212C has a peak current of 4 Amp and a very high slew rate. Note that it is an inverter buffer. We will soon see why…
3/. The PWM generator (SG3525) :
The main advantage of this chip is that it can generate a duty cycle of about 49% on each output. As the power MOSFET are used in parallel, this will give us a duty cycle from 0% to 2x49 = 98%. A good number but as we would prefer a maximum duty cycle of 100%, we will use an inverter ( EL7212C ) : this chip will give us the opposite signal to drive the MOSFET. So now the duty cycle can reach 100% ( = 100 - 0 ) ( maximum speed for the fans ) and still can come down to 2% ( = 100 - 98 ). Now we only have to choose a switching frequency : this is really easy, only one condenser and one resistor. The condenser must have a value as little as possible, because it will determine the maximum duty cycle of the SG3525. This maximum is reached with a 1nF condenser. Now we choose a switching frequency of about 20kHz. At such a high frequency, the inductive load of the DC motors will only “see” a DC source-like, and not a rectangular waveform any more. A higher frequency would not give us any benefits, but a slower one, between 400 and 1000Hz for instance, could generate an audible and unpleasant noise… This frequency is obtained with a 68kΩ resistor. The 2 inputs ( pins 1 and 2) of the SG 3525 drive an operational amplifier ( AOP ). We will use it as a voltage follower by connecting pin 1 to pin 9. So we now have pin 2 only. When the input on pin 2 is O volts, the outputs of the SG3525 are low ( O volts ). The inverter will give us two continuous high output, therefore we will have 100% power to the MOSFETs. When the input on pin 2 is about 5 volts the output of the SG3525 reaches the maximum duty cycle of the chip, about 98%. The inverter will therefore give us 2 rectangular waveforms on each output, with a 2% duty cycle ( that means a very low average voltage on each output of the EL7212, about 2% x 13Volts = 0,26 V : the fans will certainly not operate with such a low voltage, so it’s low enough ).
The built-in FLIP/FLOP and the NOR gates of the SG3525 generate a rectangular signal with the appropriate duty cycle on each output alternately.
All we have to do now is to generate the appropriate DC 0 to 5 volts signal according to the temperature sender…
4/. The temp sender circuit :
I used a TS-71 ( GM ) temp sending unit, mounted on the passenger side head, for my first design, and it worked very well. So I used the same one for this PWM controller. Any NTC resistor ( negative temperature coefficient ) can be used, but you will have to calibrate it first with an accurate digital thermometer. I used a µA723 ( or LM723 ) chip to create a reference power supply of about 7,25V ( can vary from 7,10 to 7,60V from one chip to another ). This voltage will not vary with the ambient temperature or with the battery voltage ( from 9 volts to 40 volts ), therefore giving us an accurate measurement. The maximum output current of the 723 is only 150mA, but it is enough for our purpose.
These are the values I got with a 330Ω – 1% serial resistor and my sending unit ( measured at pin “+CAPT” on the diagram ) and the 7,25 V ( LM723 ) power supply :
200°F --> 2,93 V
195°F --> 3,07 V
190°F --> 3,24 V
185°F --> 3,36 V
180°F --> 3,52 V
175°F --> 3,69 V
We will use the CA3130 operational amplifier, because it works with a single power source and can almost reach 0 volts and ((V+) – 0,5V) on its output. Once again, you can find the characteristics easily on the net.
The first CA3130 is a voltage subtractor. Now we have to choose two temperatures for our controller : let’s say we want 100% speed for the fans at 190°F and 0% speed at 175°F. We now know that the temp sending unit will give us 3,24 V and 3,69 V respectively at these temperatures : a low voltage at pin “+CAPT” means a high temp, and of course a higher voltage means a lower temp. We need to adjust the first trimmer on the board ( 1kΩ linear ) to obtain 3,24V on pin 3 of the first CA3130 ( the one which is on the left of the diagram ). We can adjust this value with a simple multimeter set in DC voltage mode ( All voltage values are ground referenced ).
At the output of this CA3130, we now get the voltage difference between the temperature sender signal and 3,24 V. It means that for any temperature higher than 190°F the output of this CA3130 will be 0 V. For any temperature lower than 190°F, the output will be : (Temp sender Voltage ) – 3,24 V = positive value.
We must now modify the gain of this output to get the second point : we want 0% speed = 5 volts output at 175°F. At this temperature, the sender gives us 3,69 V, so the first CA3130 will give 3,69 -3,24 = 0,45 V. And we need about 5 Volts. This will be the task of the second CA3130. This amplifier has an adjustable gain and is non inverter. In this example, the gain must be adjusted to 5,00 / 0,45 = 11.1. This gain is adjusted with the second trimmer ( 2kΩ linear ), and is equal to ( Pot + 100 ) /100 = 11.1. The result of this equation is P = 1010 Ω, easily achieved with the 2kΩ trimmer .
Now, we must limit the maximum output of this second CA3130 to about 5 volts. This is the task of the Zener diode ( 5.1V) and the 1kΩ serial resistor. We now have a linear output ( 5V – 0V ) between 175°F and 190°F respectively, with 5V for any temp lower than 175°F and 0V for any temp higher than 190°F : just what we wanted !

Don’t forget that the power output components are not present on the main board.
You must choose a box with a large enough heatsink for these components.
The two gates of the mosfet transistors ( pins 1 on the TO220 case ) are driven by the 2 pins named G1 & G2 on the board.

Note : The printed board circuit picture above has just been updated ( upgrade )