The Fan Tome Gourmet

The FanTome Gourmet

(Too subtle? Too foodie? Too weird? Too much time on my hands?)

Finding Replacement Fan Blades for the Fan in the Heater Box

(Better?)

When I took apart my heater box, ostensibly to deal with the fact that the foam had dissolved off the flaps, allowing huge amounts of outside air in (a detriment to my freshly-resurrected air conditioning system), it made sense to address anything the heater box needed while I was in there. I rebuilt the heater box as per the excellent instructions here on the FAQ. The heater fan motor worked without squealing, but when spinning the fan blades with my finger, I could feel a rough spot in the bearing. I tried dripping sewing machine oil onto the bearing, but it made little difference. I’m usually not a slippery slope guy, but the idea of the fan dying and having to pull out the evaporator console and heater box and crack the heater box open a second time gave me pause.

So I looked at replacing the fan motor. I was stunned to see the nearly $300 cost for a new motor with the fan blades attached. I looked here on the FAQ and found posts relating to use of a nearly-identical Porsche electric motor, Porsche part number 0130007081, costing about $60 through several online sources. However, there were three issues:

1. The Porsche part came without fan blades.
2. The Porsche part spun in the opposite direction.
3. This part number is only valid if you have a later-style fan motor (I’ll come back to this).

The second issue can easily be handled by crimping on new connectors to the wires and reversing the leads to the fan. But the business of it coming without fan blades is problematic.

Note that the original fan blades can be reused; Mike Self reports:

If your old fan motor is dead, dead, dead and can't be revived, simply saw through the armature between the fan and motor housing, the tap the shaft through the fan hub. Place the hub/shaft on top of a 3/8 drive deepwell socket and use a hammer & punch to knock the shaft out. Be sure and note exactly where the fan is situated on the shaft so when installed it won't hit the heater housing.

However, my original motor wasn’t dead, and I didn’t want to risk damaging the blades. I thought “how hard can it be to source, if not the original fan blades, a set that are close enough?” The answer is – harder than you’d think.

Initially I thought that the problem was the size of the shaft on the fan motor, but that’s only one issue. I measure the shaft as about 0.24”, but others report that it’s metric, exactly 6mm (0.236”). My web-sleuthing Kung Fu is pretty good (we all have things we can get obsessive-compulsive over), but, try as I might, I could not find a source for a fan with a metric bore. However, there are several sources for small aluminum fans with fractional bore sizes, as these are widely used in the HVAC industry. The closest bore sizes to 6mm are 3/16” (0.1875”, substantially smaller) and ¼” (0.25”, slightly larger).

I decided to take one for the team, order a whole bunch of these fans, and try them out. Fortunately they’re cheap. In the end, it would’ve been far easier to simply drop the $300 and buy the new fan with the blades on it, but I’ve done this. As your Hack Mechanic. I did this ridiculous thing. For you. (Any “My Cousin Vinny” fans in the audience?)

Seriously, in the part of my life that generates income, I’m an engineer, and there’s a part of me that, when it wants to know the answer to something, just won’t let go. I pursued this thing way past the point of reason.

Let’s start with the original fan blades. What’s pictured is out of my early ’72 tii.

This is a five-bladed fan, about 5 ½” in diameter, with a 6mm bore. As you can see, the blades are curved (quite beautifully, actually) for counter-clockwise (CCW) rotation as you’re looking at the front of the fan. The blades are held in place with top-mounted hub with a press-on nylon bushing. In contrast, every replacement fan I found (listed below) has a hub with a little Allen set screw instead.

Note how the curve of the fan blades faces front. That’s the discharge side of the fan, the side that blows into the car. Most fans have a preferential direction; it just happens that the original curved blades make the preferential direction very easy to see. If you put fan blades on upside down, it is not going to blow in the opposite direction (just try it). For straight-bladed fans you can simply reverse the direction of the motor, but for curved-bladed fans, the curved faces of the blades must face in the direction you want the air to go, and you must spin the motor so the base of the curve scoops the air. For the original fan, that’s counter-clockwise. As people have noted, the Porsche replacement fan motor (and, according to Jim Gerock, even the BMW replacement motor) spins in the opposite direction (clockwise versus counterclockwise). So if you’re using the original fan blades on a motor that spins clockwise, you must reverse the wires at the fan motor to spin it counter-clockwise. However, we’re trying to find a replacement fan, and since many fan blades are available in both CW and CCW orientations, all other factors being equal, perhaps we can find a fan that spins CW (the natural direction of the new motor) and leave the wires in place. The problem is, as I explain below, that all other factors aren’t equal; it’s challenging to find one that fits.

Since the application for most of these small replacement fans is HVAC, some of the fans have intake and exhaust variants. That is, some of the fans have the hub on the intake (back) side, and some have the hub on the discharge (front side). If you imagine a fan shoehorned inside a piece of machinery or behind a refrigerator, the location of the hub becomes important. If you’re sliding a fan down onto a shaft, you can’t reach the Allen set screw on the hub to tighten it if the hub is on the “wrong” side. Note that, because the curved faces of the blades must face forward, the hub orientation cannot be changed by simply turning the fan over or reversing the direction of the motor.

Let’s look at how the stock fan fits in the housing. As you can see from the picture below, the tops of the fan blades are just about flush with the top of the well in the housing. If they come up any higher, the fan hits the heater flap when it is opened. This is a fundamental constraint you have to live within.

The way the fan sits down in the well is possible because the hub is on the front (the discharge side). It’s nestled inside the pitch of the fan blades. It doesn’t take up any extra space. If the hub were on the back (the intake side), it would act like a big spacer, pushing the fan blades forward and causing a clearance problem. This turns out to be a crucial limitation on which fans you can use and which you can’t.

Just to confuse things, note that Jim Gerock reports that the fan on a very early fan motor of his doesn’t have a plastic press-on hub on the discharge side – it has a hub on the intake side with an Allen set screw, raising the possibility that this fan can be easily removed and reused. Then again, Jim reports not being able to pull these fan blades off after loosening the set screw. Jim supplied the photo below. The fan blades look shallower and less curved than the fan on my ’72 and other fans I’ve seen pictures of.

The total front-to-back height of the fan blades is also crucial. On the original fan from my car, it’s 1.25”, as shown below.

So what we want is a fan:

• 5 ½” in diameter
• Whose bore is 6mm
• With the hub on the front (discharge) side
• Whose front-to-back depth is no greater than 1.25”, but as close to it as possible to maximize air movement

Put another way, the original fan is perfect and should be reused if possible.

But, damn it, I’m in this up to my hips. I’m going to see this through to the end and give a recommendation.

Let’s step through my candidate replacement fans. Rather than jump to the answer, I’m going to go through these in the order I bought and tested them. Note that not all spec sheets list the cubic feet per minute (CFM) a fan moves, or list it at a consistent RPM, so a rigorous comparison of CFM difficult.

First is the cheapie 10-blade fan from McMaster-Carr:
 

(I believe this is the same as Grainger AIR-DRIVE AD5CW 3/16

Other posters have made reference to this same fan here on the FAQ. McMaster-Carr (www.mcmaster.com) is quite simply the best industrial website in the world, sort-able by virtually any database tag you can imagine. Their web site is instantly addictive. Just try it. Type in “rubber.” You want sheets of rubber? Balls of rubber? Blocks of rubber? What hardness? What thickness? What temperature rating? You’ll still be looking 20 minutes later. They ship quickly and inexpensively and are rarely out of stock.

McMaster has a good assortment of fans, but only one this small. 5” is the closest McMaster comes; they have no 5 ½” fan.

The McMaster catalog lists the direction for this fan as CW, but it doesn’t say which side the hub is on. If you look at the picture above, you’ll see a slight curving of the fan blades. The curves need to face front. Note that this puts the hub on the bottom – on the intake side, not the discharge side. This is the opposite orientation from the stock fan. At $3.41, this thing is dirt-cheap, but you get what you pay for – as you can see from the pics, it is primitive, with unfinished corners and edges that are wicked sharp. Also, the aluminum is somewhat thin, making the blades easy to bend. I ordered several of these, and those that arrived in a bubble-wrap envelope (as opposed to a box) had bent blades.

But I thought, man, drill out the bore to 6mm and for $3.41, I’m done. Unfortunately, that proved more challenging than I ever would’ve thought. Even doing my best to clamp the fan so it was flat and on-center, drilling did not prove successful for me. The enlarged hole I drilled is noticeably off-center (see below), and when installed on the fan motor shaft, the blades wobble slightly. Perhaps others have more skill than I do, but if I have to take a three dollar fan to a machinist to enlarge the bore, the cost-effectiveness vanishes.

So I began looking at fans with a larger ¼” bore. The first one appears for all the world to be the same 10-blade fan, just with a ¼” bore. Surprisingly, McMaster doesn’t list it in their catalog, but Grainger does.

As with the McMaster version with the 3/16” bore, you have to flip it upside down and put the hub at the bottom to spin it so the curved blade faces point up and push the air. With the hub on the bottom, does this fan fit? Below I show that it does; even with the hub on the wrong site, because the fan blades are short in the front-to-back dimension, clearance is not an issue. So, yes, this fan will work, but it’s pretty cheesy. I imagine an acorn finding its way in there and bending the fan blades. We can do better.

Next is a four-bladed fan.

I thought I hit the jackpot when I found PackardOnLine, and to a certain extent, I did – they had a wider selection of small fans in one place than anywhere else I found. Unfortunately Packard doesn’t sell direct to the public, only through distributors. You can call them and they’ll tell you a distributor in your area. I found I could order Packard fans through AppliancePartsPros.com, since these small fans are used in HVAC.

This is a 5 ½” four-bladed fan with the hub on the intake side that looks not unlike the very early fan in Jim Gerock’s picture. Below are pics of this fan in the housing, showing that, even with the hub on the wrong side, it fits easily. So this is a definite candidate. It is sturdier than the 10-blade fan, but the blade pitch seems shallow. Of course that helps in terms of clearance, but I can’t help but think that it leaves a certain amount of air flow on the table.

Next is a five-bladed fan that I thought would be “it.”

I list two part numbers for the ¼” bore fan in a choice of CW and CCW rotations (Packard also shows similar fans in 3/16” bore configurations). The fan shown on the Packard web site looks very similar in shape to the original fan, but the fan that arrived (as pictured) has a less finished more angular look. Like every fan I’ve shown thus far, the hub is on the intake side.

Although the front-to-back of the fan blades is 1.3” – basically the same as on the original fan – there’s a problem. Below are pics of the fan in the housing, showing that with the hub location on the bottom (intake) acting like a big fat spacer pushing the fan blades up, the blades clearly protrude above the well – it doesn’t fit.

Note how important this hub location is. Just as a test, if we flip this fan over and put the hub discharge side, the fan sits lower in the well and it fits, but now the curve of the blades is pointing the wrong way. The Packard web site doesn’t show an option for this fan with the hub on the discharge side.

Below is the Lau fan. These are listed on the Packard site, but more detail is shown in the Lau catalog. Like the Packard fans, the Lau fans can be ordered through plumbing and HVAC supply houses, including AppliancePartsPros.com:

I was excited about the Lau fan because it was the only fan I found where the CW and CCW variants had the hubs on different sides – that is, the CW fan (60834601) has the hub on the intake side, but the CCW fan (60834701), like the original fan, has the hub on the discharge side. Note that although the CFM is not listed, I thought the fact that this fan has six blades and a 45 degree fan pitch should yield increased air flow as compared to the four-blade fan above. The blades on this fan are more robust and have a much more finished appearance than the 10-blade fan shown above.

Unfortunately, even though the hub is on the correct (discharge) side, the physical size of the fan blades do it in. The photos below show that the fan doesn’t even come close to clearing – the blades hit on the bottom of the well and protrude through the top.

With this hard-won knowledge, I thought I’d give one more try at finding a fan with the hub on the discharge side and a less severe fan pitch than the Lau fan. More web searching lead me to GlobalIndustrial.com. They only sell fans in quantities of 10 or more, but they list the Packard, Lau, and other fans in one place, and some new ones, including Alltemp (the actual manufacturer appears to be Rotom). Searching for the part number, Alltemp fans appear to only be available through Canadian supply houses, so it was a pain to get and took about a week to ship. The fan is about $10, with Fedex shipping from Canada about $18, making it pricey compared with the others, but, hell, in for a penny, in for twenty-eight more bucks.

Like the Lau fan, the Alltemp fan comes in two flavors with the hub on different sides depending if rotation is CW or CCW. The CW fan has the hub on the discharge side. This means that, if it fits, the leads on the new motor (which runs backwards) don’t even need to be swapped. Hitting on all cylinders.

When the Alltemp fan arrived, I was surprised to see it looked almost identical to the 5-bladed Packard fan, but with the hub on the other side.

A careful measurement of the Alltemp fan shows that the front-to-back of the blades is about 1.4” – slightly more than the original fan, but less than the Lau fan. I test-fit the fan into the housing, sliding it down onto the shaft until the backs of the blades hit the cage. As you can see, the blades seemed to stick up a bit too high.

I fit the fan housing onto the heater box and spun the fan blades by hand while I opened up the heater control flap. To my profound disappointment, the tip of the fan blades began to hit the flap. I thought, well, that’s it. I’m done. The winner is the four-bladed fan. I put it on the shaft and began to assemble the heater box.

Then, I did what many of us do. I stopped and did it again.

Upon closer examination, only two of the five fan blades were hitting the flap. This was because the blades weren’t level. I tried to level the blades inside the fan housing, sliding the fan down as far as it would go until the blades hit the cage, backing it up slightly, then eyeballing how close each blade came to the cage and bending them until they were within 1/16” of each other, but ultimately I pulled it out and did it on a table. This allowed me to level both the top and bottom surfaces very accurately by adjusting it so the tips of all blades touch the table, bending whole blades to level the bottom, then twisting them slightly to level the top. Once this was done, there was enough clearance both top and bottom.

However, when I tightened everything down and prepared to button it up and tested the motor one more time, I heard some buzzing. Small changes in alignment were causing the fan to hit. I looked at the fan cage carefully. There is a series of small notches, shown in the picture below by my holding an Allen key, intended to give a little more clearance to the original fan blades. Taking a cue from that, I used the Dremel tool with a grinding attachment and cut notches for the new fan (second pic). This gave the fan some breathing room in terms of clearance. I reassembled everything and, with the cutouts, the clearances were much more comfortable.

Note that there’s gobs of clearance at the top (the front) of the fan until you open the heater flap. When you test for clearance, it is imperative to move the heater flap through its full range of motion. It is possible for the fan to clear the flap when the flap is fully open but hit when the flap is about 2/3 open, as that is the flap’s closest approach to the fan blades.

So the Alltemp fan fits. It’s tight, but so is the original.

But Which One Blows Hardest?

Since not all the fans list the CFM at which they blow, or list it at inconsistent RPMs, I wanted to do a test that was less subjective than simply sticking my hand in front of them but less expensive than buying an anemometer. I took a microphone stand, found a piece of cardboard from an industry-standard 12 pack of Magic Hat Number Nine Not Quite Pale Ale, hung the cardboard off the mike stand so it was free to swing, and taped on a screwdriver at an angle of about 30 degrees to the vertical. Then I placed the mike stand with the cardboard directly in front of the blowing fans and moved the entire stand fore or aft until the cardboard swung back so it was in line with the screwdriver. I compared the 30 degree swingback distance for the different sets of fan blades.

To my surprise, the little four-bladed Packard fan (A61473) showed significantly more air flow in this test than any other fan (that is, it pushed the cardboard back at the greatest distance). More than the Alltemp fan. Even the Lau fan (which didn’t fit), with its six 45 degree blades, didn’t even come close. Most unexpected. I repeated the test. It’s not a fluke. Data is data.

But something didn’t seem right. More blades, at a more steeply pitched angle, should move more air. And the 5 and 6-bladed fans sounded like they were blowing more air. I realized that my test was only measuring the air flow directly in front of the fan, not the total air flow. I needed to test the fans installed in the heater box. So I repeated the test with the Packard 4-blade fan and the Alltemp 5-blade fan each mounted on the motor, the fan cage attached onto the top of the heater box, all the flaps closed except the defroster flaps, and the industry-standard Magic Hat measuring device positioned in front of one of the round openings in the heater box where the defroster vent air exists.

In this test, the five-bladed Alltemp fan spit out measurably more air than the 4-bladed Packard fan, though I’m not sure I could tell the difference just using my hand.

Unfortunately, I could not collect data with the original blades because, for some reason, the original fan motor, which was not dead when I removed it, is dead now. Yes, I am aware that, now that it is dead, I can use Mike’s technique for removing the fan blades, but I don’t see how to get a saw in there.

I’d love to complete this test by putting the original fan on the new motor and measuring its air flow, but I do have a life you know.

So, Damn It, Rob, Which Fan Should I Use?

I don’t see any reason to use the 10-blade fan; the Packard 4-blade fan is more sturdy and moves more air. So the choice is between the Packard 4-blade fan (part number A61473) and the Alltemp 5-blade fan (part number 9364587).

The Packard 4-blade fan is readily available through a number of domestic channels. If you don’t need to use the heater fan that much, are concerned about the clearance issue – if you can’t button up the heater box without paranoid fear that you’ll hit a pothole, jar something, and start hearing tick-tick-tick evolve into scrape-scrape-scrape – and if you don’t want to grind some notches into the fan cage, use the Packard 4-blade fan. You’d have to install it incredibly carelessly to have it rub.

But if you’re in a climate where you need the heat and defroster to blow as hard as possible, use the Alltemp fan, level the blades, grid out some notches for insurance, install the fan carefully, and I think you’ll be fine. It’s a pain to get, though, because it appears to only be available through Canadian supply houses.

I’ve installed the Alltemp.

Giving The Fan The Shaft

Now that you have selected a fan, you have to make it fit on the motor’s shaft. If you want to use a 3/16” fan and enlarge the bore, go for it, but I was not successful doing that. I decided to use a fan with a ¼” bore and shim it.

If the shaft is 6mm, that’s 0.236”. Subtracting that from ¼” (0.25”) yields a difference of 0.01378”. Since we want the shaft centered inside the bore, we want to take half that – 0.0069”.

I went to – you guessed it – McMaster-Carr – and typed in “shims.” Do so and you’re met with a dizzying array of shimming material. I settled on 18-8 stainless slotted shim tabs. I bought part numbers 97235K129, 97235K134, and 97235K138 (0.003”, 0.004”, and 0.005”, respectively), for about $1 each, cut them into small pieces about the size of a thumbnail that I could wrap around the motor shaft. If life were perfect, you could use a .003” and a .004” shim to come as close as possible to 0.007”, but I couldn’t make those fit. I settled on a single 0.005” shim.

When everything was done, I test-fit the blade, with shim, on the motor, fit the assembly in the heater box, applied power, moved all flaps, made sure nothing hit or buzzed, then took the motor back out, unscrewed the Allen set screw, put a dab of Locktite Red on it, and tightened the living snot out of it.

And The Motor Itself

In this thread:

http://www.bmw2002faq.com/component/option,com_forum/Itemid,50/page,viewtopic/t,382412/

I noted that there were previous posts saying that Porsche part 0130007081 was touted as being a nearly identical fan motor. Turns out, it’s wrong. Well, not wrong, just incomplete. The early fan motors have a housing with a 2.3” diameter, and for that you need Porsche Part 0130007002 with a plastic sleeve or “girdle” around the outside. The best price I found on this was $73 from BMAParts.com. The later fan motors appear to be smaller, with a 2” outer diameter. That’s the one that Porsche part 0130007081 replaces. Below I show, left to right, my original early-style fan motor, the correct replacement (0130007002) with the sleeve, and the replacement for the smaller later-style 2” motor without the sleeve (0130007081). The original motor looks slightly bigger because it’s sitting up higher.

When the correct motor (with the sleeve) arrived, I test-fit it into the heater box cage, and swore it didn’t fit. There are four raised ridges on the inside of the fan housing. It made sense that these would slide into the four slots on the outside of the sleeve, one of which you can see on the right side of the sleeve. But they don’t line up. I was mystified.

Then I looked carefully at how the original fan motor fit. It has no slots running down the outside. The mating is achieved by a notch on the front of the motor’s case, beneath the blades. It mates with the cut-out in the housing, shown at the 5:00 position in the photo above. As you see from the picture above, there are other cut-outs that neatly receive other areas of the original motor, but when fitting the replacement motor, what’s important is that this cutout at 5:00 receive the notch in the plastic sleeve. If you orient the electrical tabs toward the hole where the wire comes through, you’ll see how it needs to go. The correct fitment is shown below, with the notch and cut-out rotated into the 12:00 position.

Before I held it in place with the original clips, the fit of the replacement fan in the cage felt a little sloppy – not the tight fit of the original fan – but once the clips were in place, it didn’t feel like it was going anywhere.

Motor Direction

Jim Gerock’s notes in the heater box FAQ say:

“1) I have confirmed that the new replacement Behr heater fan with Bosch motor (made in Spain) has reverse polarity as compared to the original Bosch German motors. Looking at the motor with the (2) terminals facing you, the left-hand (Male spade) is negative (-) and the right-hand (Female) is positive (+). While "bench testing", I connected my new fan up to the original wires (red + from left-side of the resistor board and brown - from the right-side of the resistor board) and it turned backwards, blowing air out of the box.”

Other posts say that the Porsche fans labeled “Made in Spain” also spin backwards.

So I assumed if I ordered a new Porsche fan, it would spin backwards.

But it doesn’t.

At least not the two brand new Bosch motors I have (0130007081 and 0130007002). Both say “Made in Hungary” on the boxes. Both, when wired like the original fan with the male spade hooked up to negative and the female spade hooked up to positive, spin counter-clockwise – the same direction as the original motor.

So my idea that, in ordering the Alltemp fan that spins clockwise and putting it on a new motor that spun clockwise, I didn’t need to reverse the wiring? Wrong. (Obviously it’s trivial to flip the wires, but it took me a while to be certain of what I was seeing.)

Verifying the polarity of the fan is essential. Don’t believe me. Don’t believe Jim. Verify it yourself.

As Groucho Marx said, “who are you going to believe? Me or your own eyes?”

One Final Note on the Words “Early” and “Late”

I casually throw around the words “early” and “late” to connote the larger and smaller diameter motors, respectively. Jim Gerock’s photo shows what is probably an “early early” motor, and Jim wonders if the 6V cars have something even earlier. In this write-up I make no claims whatsoever with regard to representing a complete archaeology of the differences in heater boxes across the 2002’s eight year span; I was just trying to replace a freaking fan.

So… twelve pages later, we reach the end of The Fan Tome Gourmet. A fan has been selected. It’s time to button this puppy up, test yet again, reinstall the heater box, make the car drivable, and get back to what I should be doing – finishing the a/c resurrection and getting the car sorted out to take to The Vintage.

And if the thing starts buzzing, or the shim on the shaft fails, I won’t bother posting about it. You’ll know because wherever you are in this world, you’ll hear my screams.

Join the Discussion

Enjoy what you read? Have a comment or question for Rob Siegel about this "Interesting Hack"?

Click the button below to come join The Hack Mechanic in discussing this article in his very own technical discussion forum hosted by Bentley Publishers.

Related