UPDATE: check the Ideas & Products page for a better way of supplying brake fluid... all those chronically leaking low pressure tubes have now been eliminated.

Preparatory to starting the cowl I have taken a preliminary step that I want to recommend. I got this tip from Scott McDaniels at Van's. He says that with all the new small alternators (mine's a B&C) you can cut the bracket off by over an inch at the inboard end, re-drill the hole, and use a shorter belt. This will keep the alternator in tighter to the engine and eliminate any interference with the cowl as the engine moves around. My Sam James cowl is quite a bit tighter in this area than the stock cowl so I needed a small blister anyway. Look at how far down most alternators hang and then look at this pic. BTW, finding a shorter belt was no problem - the local auto parts store has Goodyear Gatorback belts in .5" increments. I cut my bracket off 1.5" and note where I am in the adjustment slot. I'm using a 35.5" belt

(Niagara cooler on left, Stewart Warner on right)

Since flying my RV-8 (four months ago) my oil temps were a bit higher than I'd like on days where the ambient temp is 80 or above. When it is above 90 they were definitely too high (like 218 at low altitude low power cruise). On one 90+ degree day in Central Oregon departing from a 5,000 elevation airport with the engine already hot, and while following another plane's climb profile, I saw 245... yikes! Since then I've learned to manage oil temps and have never seen over 230 regardless of the ambient temp. Still, I did not feel like I had adequate headroom.

Several RV List contributors had spoken with Pacific Oil Coolers lately and been advised that Stewart Warner coolers were the most efficient, so I called them also to see if a swap with my Niagara would lower my temps. They told me "absolutely" and that I "could possibly see a 20% reduction". Of course they wouldn't guarantee anything. Frankly I simply didn't believe the 20%, but I'd be happy with 10-15 degrees (5-8%).

When doing my final FWF installation I had originally bought a Posi Tech. Solely on the experience of several RV-Listers I sold it and bought a Niagara from Van's. At that time I thought the Niagara was the identical product to a SW... ignorance is bliss. Again, the Niagara has worked well for 120 hours for all but very high ambient conditions, which is only a problem here in the Great Northwest for 2-3 months a year. If you lived in Arizona or Florida it wouldn't have been unacceptable.

Curiosity got the better of me so I ordered one and swapped the SW for the Niagara while doing an oil change. The flanges and bolt hole dimensions are identical, but the core is recessed about 1/4" deeper than Niagara. This required a very minor modification on my mount... no problem. I ground ran it as usual when doing an oil change to look for leaks, none noted, and buttoned it up to be flown the next morning. I have now flown 7.5 hours over the last two days in ambient conditions of around 80 degrees on the ground. The comparisons I made are not good scientific method, rather my own feel for what temps would have been before versus what I was seeing now. After flying the plane 120 hours in the last four months though I feel like I have a good handle on what was normal for a given flight condition. Sorry, proper data would have been tough to get (identical ambient conditions combined with identical climb/power profiles) and way too much work.

So here's what I found: when doing an extended climb before at 120-130 mph on days like this I would see peaks of around 220-225. This weekend I saw peaks of 205-215. Cruise at 7,500-8,500 feet, with the power settings I typically run, before would typically be 189-202, this weekend they were 182-185. During long high-speed descents temp would never drop below 180 with either cooler so the vernatherm must be working. When assessing this information it should be noted that my plane is non-standard and may make the oil cooler's job a bit tougher. I have a Sam James closed cooling plenum. This forces you to mount the oil cooler lower behind cylinder #4 than with Van's standard baffles, hence half the cooler inlet is cylinder fins. This probably impedes airflow and also may pre-warm the air, not good for cooling.

Hmm, what conclusions can we draw from this? My own sense is that the SW does indeed cool *just a bit* more effectively than the Niagara. In fact my own belief is that I'm seeing an 8-10 degree reduction. The Niagara is $241 from Van's, the SW was $397 outright from Pacific Oil Coolers. Was the swap worth it? In the long run, probably. Was it dramatic? No. Will I keep the SW installed? Yes, that 8-10 degrees is important on those hot days. So there you have it, as close a comparison as I can give you.

Many builder's have asked if Randy Griffin would make more of these for sale. Randy says that it's just not economically feasible. He has however provided the part drawing as a .dxf file. You could take this file to your local machine shop, and with CAD equipment they should be able to reproduce it.

(right click to download)
gsbrkt.dxf

Turns out the E.I. FP-5 fuel flow gauge does NOT properly filter out the pulses from the aux fuel pump — once flying I was getting erroneously high flow readings with the aux pump on. I was concerned because this would in turn lead to incorrect "fuel consumed" readings. E.I. admitted I was given bad information on that issue originally, and the source of this information is no longer with the company. Both in their installation manual and on the phone they insist that the flow sender needs to be AFTER all fuel pumps to avoid a vacuum on the sender at any time. There's only two ways to do this, at least in a carburetor installation. One, mount it inline between the engine driven fuel pump and the carb, or two, mount it on the firewall and run long lengths of flex line from the fuel pump to the sender, then from the sender to the carb. While at a local aircraft part store I found a special AN fitting allowing me to hard-mount the sender to the carb. This installation yields even larger errors. At the same K-factor setting the flow readings are probably 40% too high. Then when I turn on the aux pump they actually go down. Go figure. 

After flying with it this way for a couple of weeks I felt it was more accurate in the original configuration. I next relocated it back behind the firewall, but used my special hard coupling to mount it directly to the back side of the firewall thereby leaving an even longer straight tube from the Facet pump to the flow sensor. I've flown many tanks of fuel since then and even though the flow does indeed jump up 2-3 gph when the aux pump is on, the accuracy is impressive. After a few tanks to get the k-factor adjusted properly the all-important actual "fuel remaining" calculation now yields an error of .2-.5 gallons when I fill up... good enough for me. Frankly I don't understand how it can be so accurate when the boost pump throws it off so much, but there is no denying the accuracy. If you spent a whole tank in the pattern with the pump on I suspect it would throw it off, but for normal RV flying I have come to trust this computation as the true reflection of how much fuel I have left. The regular fuel gauge works pretty well also, but of course is completely inaccurate when the plane is on the ground in the three-point attitude. If I want to know how much fuel is in it I just flip on the master and the FP-5 defaults to the "fuel remaining" screen... voila, I know instantly how much fuel is left and can plan accordingly.

Additional factoids: E.I. uses a Floscan 201 flow transducer which seems to be the standard of the industry. I understand Vision Microsystems uses the same sensor as does Matronics and JPI. The 201 is for low wing aircraft and has a smaller orifice for the fuel to flow through, but it has greater resolution in that it breaks a gallon of fuel into something like 8,500 pulses. They also offer another sensor with a larger orifice (205?), but it has lower resolution and may be slightly less accurate. The reason all of these instrument manufacturers instruct you to mount the sensor downstream of ALL pumps is to avoid a vacuum condition. Apparently it is possible for avgas to be separated into a "benzene state", and they want to avoid that under all conditions. Frankly, I think that in the low pressure carburetor installations this is just about impossible. Anyway, these instructions come from Floscan and I'm sure are driven a bit by liability concerns.

To the their credit, Dave Campbell at E.I. has been very helpful and attentive through this process. I continue to have confidence in both my E.I. instruments and the company.