The ebag link is for a new style fan like Andy and I run. Look for the metal heat sink. If it's there, it's an old style fan.

I did some water testing on the black box.
in both the horizontal and vertical positions it did allow water into the box.

In the horzontal, IE, the screws are visible looking straight down at the box, there was enough water pooled at the bottom, that I would be concerned. Very short test. Used a veg sprayer in the sink to hit it for about 30 seconds. About like if you hit it with a power washer washing your car.

Mounted vertical, with the long axis being hit by water from above, it was much better, leads me to believe the screw holes are the weak link.
A few drops of water pooled at the bottom.

I will test out the white box in a few days, but if you chose the black one, mount it inside would be my sugguestion.
OR, tape up the edges and screw holes.

Thanks for the test results

Another thing to keep in mind, is that a hole still needs to be drilled for the wires to pass through. If mounted on a wall, I recommend the hole to be on the bottom edge, so most water should drain itself.

Although maybe it's just best to stick it inside somewhere

OR, a tiny bead of RTV around the edge and screw holes, since you shouldn't have to open it often once you have it set up. This is likely what I'll end up doing.

I was thinking something less permenent.
Tape should be enough.
the screw holes were the worst leakers. Sides not as much.
but a strip of tape around the sides and over each screw hole would be enough to "seal" it.

I bought a Mercedes fan about 6 months ago for my 67 Coronet and it looks like I'm going to need one of your controllers as it has the metal heat sink. Looking at your website I'm finding that you are going to look at controlling the fan speed. My car won't be on the road until next year so I was thinking about waiting until you figure it out as I would like the ablity to have the fan run more quietly and draw less amps on a cooler day. I really apprieciate you taking the time to make these controllers up.

got the ociliscope hooked up. Hopefully the readings will help get some variable speeds for these fans.

Got busted by the wife, so I might not be able to get other readings. lol.

With the info I got from Andrew earlier today, we now have variable speeds

I still need to do a bit more testing, but it appears that I will need to change the hardware a bit, and the controller will need both small wires from the fan going to it. I'll know more after some more testing.

I realized there might be some issues with variable speed.
The mercedes has a few extra sensors to tell it when it is okay to switch to low speed. VSS and tps.
you only have temp.

I sugguest a bit of extra logic for the temp. Start the fan at high speed at the start temp. After a 4 degree drop, switch to med speed, another 2 degree drop go to low and shut off at the 10 degree total drop.
IF at any point temp increases, kick back to high speed. Not sure if you need to change the logic at this point to keep it at high, or if the switch up is enough.

I believe the 2 speeds I sent you plus high should be safe, I don't think you should put in any other inbetween speeds just because that would require more testing of the older fan to be sure it doesn't hurt it at other variable speeds.

I think this will need some drivablilty testing, in that does any of those speeds make it worse and if the up temp logic works.
I would prefer to verify this at warmer temps, but regardless it needs to be checked out.

Finally, if you have the time I would like to check out some stock mopar temp sensors to see if you can use those for temp. Now more than ever you need a temp sensor for this to work.

I'm thinking something a bit different. How about running 50% or 60% speed for most cooling needs and fire up to full speed with high temps or on demand (A/C kicking in)?

And I was thinking a third idea

With a temp sensor installed, I can use the feedback from it to maintain the set temperature (ie, 205*). If the temp continues to rise, fan speed would increase to full speed in 5 or 10% increments. Likewise at idle, it may only need 10% to keep 205*, so why have the extra noise and current draw.

Or, I make the minimum speed 30-40%, so at idle it will still be able to lower the temp and kick the fan off for a little bit, while minimizing noise & current draw.

For AC it would still kick to full speed.

In theory if the feedback loop is correct, if you hit the highway, it should cool off from the increased airflow, with the controller responding with a lower fan speed. Continue until the temp drops enough to turn the fan off.

I may have to make a video of the fan at low speed. It's almost silent

As for the hardware changes needed... I basically need to start over The current board I sent Andrew can't get a clock speed low enough that can generate the speed of duty cycle needed. The basic design stays the same, it will just have a different microcontroller controlling it, which is all transparent to the end user.

Now with more video! (Do we all sound weird when we hear our own voices, or is it just me? )

http://www.youtube.com/watch?v=8aGq7f0SSYk

In the video the fan starts at 90%, then I drop it to 70, 50, 30, and 10%, where it finally turns off, then it turn it back on to 15%.

That's friggin awesome!!!

I'm glad you were able to knock it out so quickly.

I think your idea is great assuming the user can select the starting point. Some folks like 160 degrees, some like 190 degrees.
You need to keep the lowest speed fast enough to cool the engine below the thermostat. That way the fan will actually shut off.


Don't worry. We all think your voice sounds funny in the video.

Correct, they'd still be able to choose the temperature using the knob. Does anyone want a temp of less than 160*?

I should post a picture of the board I used in the video... it's ugly I still need to design the new board and have it made, yadda yadda... add a couple weeks + more testing.

I can add some more functionality to this test board and at least do some testing in the truck now. I'll also let the fan run at 15% for a while to make sure it doesn't die.

Quote:

---

Don't worry. We all think your voice sounds funny in the video.

---

I have a couple of issues or changes.

First off, I think I know the reason for the way the merc fans work, and perhaps you need the same logic.

So when the cars are warm and just start, you have a heat soak condition. Most times the temp will spike up to around 230 ish.
So they must have some sort of over temp limiter to where the fans kick high due to the temp.
I agree with this procedure for a couple of reasons. If you are moving faster than the 40 percent speed of the fan, you are now impeeding air flow.
I think if you add logic to set speed based on how much over set temp you are just in case, because if you hit that temp while running you don't want to step up, you want to be at max cooling right now, then your idea of starting at 10 percent, and moving up if you don't maintain is fine.

Second, a/c really doesn't need max cooling. I don't see the point of hitting max speed for a/c. The only times I have needed the fan for a/c is when I am dead stopped. if I have any movment at all, it is enough to keep it from kicking the fan for a/c in 100 degree weather.
Here I would stick to the 10 percent with cooling taking precendent for speed.
That being said, you should have a timer on it. If it gives the signal for longer than 30 seconds or so for a/c, kick it to max cooling.

I noticed the merc fans do stay running all the time once the engine is warmed up. And I think I prefer that too them kicking on and off all the time. So my vote is for min duty to maintain say a 5 degree dead band, vs the 10 degree shut off when moving.

I've seen the heat soak in action with my controller (the controller in the video has a display I can put the measured values on). The heatsoak doesn't show up on the gauge in the dash, but on a warm start the temp will read 215*, then immediately drop to the 190s.

(For the following assume a set temperature of 205*)

So you're saying, if the temp is greater than 220*, don't bother slowly stepping up, just go to full speed? That makes sense. On startup I can also add a 5 or 10 second delay before the fan comes on to give the water some time to circulate. After that on a warm start, it would go back to the standard logic of varying the speed to maintain 205*.

Otherwise, on a cold startup, the fan will be off until 205* is reached, then it can start at 50%, and vary its speed from there.

I was wondering if people would prefer to have the fan always running at 15%, or run at 20-25%, and turn off occasionally. I almost think I would prefer the always running


For AC, when I was looking at the trinary switch stuff, most comments stated they needed full fan to move air over the condenser. But then again, half speed on this fan is full speed for others I could see going to 50% initially, then going to 90% if the AC signal is still there after 30 seconds. Although that would negate some of the effectiveness of putting a switch on the AC signal as an override.


I need to install the fan in the truck, then initially I'll make it so I can vary the speed with the knob. Then driving down the road I can go from full speed to idle and see how the temp reacts. After that I'd have a lot better idea of what to expect.

that is it pretty much.
at 205 start and then step down to maintain.
at anything over 218 220 kick to full speed, then say at 205 try to maintain again.

not sure what you mean about the a/c switch. I can assure you it does not take that much to maintain that off. I would hope that the coolant need for cooling will prevent it from ever coming on, and I agree, I would rather see it on at 10 percent all the time rather than kick on and off while idling around town or in stop and go traffic.

If it will be a while for you to get it in your truck, send the prototype to me or the variable speed and I will rig it to my car.

Infact, I was wondering if you can reprogram the board you gave me to do just that instead?
ignore the other feeds, just power, ground, both leads to the fan.
Change the temp pot to speed?
I can verify temps seperatly and then see how it affects it by speed of the car.

Unfortunately, the boards I sent you can't produce a clock signal slow enough to generate the correct signal to the fan. Its PWM hardware is woefully inadequate compared to the 8051 I'll use in the new design. I just tried to control the fan with the clock speed I can get on those boards, and it just turns the fan on or off.

I should be able to get the fan mounted in the truck this week. In semi-related news, it just started snowing again

Snow shouldn't be a problem.

looks like this is going to work out great ... Thanks to all of you who have donated your time and $$ to this project .. can't wait for the final version so I can put one on my coronet