Air filtration - K&N Air cleaner test
In the past few years (well, must be nearly 6 years by now!) since I got involved with writing for the specialist Mini Magazines I have been continually fighting with the problem of limited space. This has limited what I have been able to cover, both in subjects and in details. Up until recently (say the past year or so) I have been covering fairly broad topics in a fairly broad terms and detail. Basically trying to cover perhaps the most popular areas dabbled in by the Mini owning masses. Having covered pretty-much all that I think can be covered in this fashion, I decided time had come for a more detailed look at specific items, supported and underlined by the frequent number of questions concerning certain popular subjects on the message board and in the 'Knowledge' section of Mini magazine. Hence the articles on tuning specific engines in specific ways, what the deal is when fitting 13-in wheels, etc. Air filters being one of those popular and specific subjects.
In the forefront of Mini tuning, improving the over-all breathing capability of the very asthmatic A-series the air filter set-up is one that sees priority attention. All the widely available (and widely varying) stage one tuning kits include at the very least a replacement - and hopefully - high-flow air filter element to fit into the standard plastic case, others going for the complete replacement cone/pancake type. The oft-asked question associated with this is 'which is best - element in standard case or cone/pancake type?' To satisfy some folk's desire, herewith is a test I did on a small-bore unit.
Despite a fair increase in the number of after-market air filter manufacturers over recent years, one still holds sway in the Mini market.
All manner of materials have been tried and tested to establish suitability for air cleaner duties - namely to filter as much engine-damaging crud out of the air as possible whilst not restricting airflow at all if it can be helped. Even when seriously clogged up. Two materials seem to have been settled on, and are widely used by a number of manufacturers. Two camps basically exist - the 'cotton gauze' mob and the 'foam' mob. Comparisons have been made, extensive tests carried out, verbal and literary battles fought, and many a personal experience thrown in in an attempt to establish one or the other as being 'the champion'. The 'cotton gauze' mob usually carry the day, and the one that was mentioned above as holding sway in the Mini market despite all-comer's attempts to oust it is the original (and many acclaim the best) cotton gauze filter manufactured by the long established company K&N. The range they produce is massive, quality outstanding - mirrored in what is available from them for the Mini/SU market.
Testing the whole range for the Mini/SU would have been brilliant - but time and finances are short - and we really wanted to look at specific applications. Considering I'd just covered tuning the small-bore engine types, and this perhaps being the most prolific engine size at the moment to which these stage one kits are fitted, it was obvious where the starting point should be. Added to this the filters tested should be those that are the popular, true direct fit in/bolt on replacements with no other jiggery-pokery needed when fitting, this shortened the candidates to two. The replacement element that fits inside the standard plastic case in place of the standard paper element and the cone type that replaces the standard filter casing as a whole.
Now, K&N generally reckon that fitting said cone type filter would find another 6-7bhp on the A-series engine over the standard set-up. There's no meaningful figure given for the element, but 'a marked improvement' should be seen. So what did I see…
And the first contender is…as this is a straight fight between the K&N replacement element and the cone type I didn't bother with trying the standard paper element. It's known to be seriously restrictive as far as maximum power is concerned, and gets horribly inefficient when dirty. So we have the standard plastic element with the hot air intake removed, leaving an open nozzle. This ensures the coolest intake temperature from the standard set-up. It's also the set-up I've been recommending over the cone type for years, if for no other reason than it's a great deal quieter. I have also had my reservations over the years as to the 6-7bhp claims from K&N - all my testing hadn't shown this. My original tests conducted some fourteen years ago went AWOL along with all my other stuff in my pilot's case alleviated from my car by some low-life years ago, so I didn't have hard proof. This test was going to hopefully set that right…
To add a little flavour to this 'shoot-out' I decided to throw something else into the fight - a variation on the standard casing. Giving credence to the rumours that the standard filter case is restrictive as far as breathing potential goes - moons ago I carefully calculated the cross sectional area of the 'nozzle' versus that of the opening that sits up against the elbow on the back of the carb (bottom of picture); the nozzle proving slightly smaller. The filter case itself is more than big enough, and the filter surface area is unquestionably big enough when you consider the cone type filter has sufficient filter area to cope with 110bhp! So to eradicate the nozzle area as a possible restriction I have bored holes in the plastic casing in the past, with seemingly good results. So I decided to try it here. I had tried holes in both the position you see here and also around the front vertical 'wall' on the casing. The ones shown here seemed to give the same results as those in the front 'wall' of the case but without the fierce increase in induction roar that accompanied the others. The six three-quarter-inch diameter holes very nearly the same cross sectional area as the oval elbow hole - so the original nozzle is a bonus.
Those paying attention in the first picture will have noticed the black duct tape around the edge of the casing. To save time at the rolling road, I pre-drilled these holes then simply covered them over - thus shortening the test time. Note that the holes are drilled around the front edge of the filter so they are between the filter and the outside world. Pointless drilling where the air won't be filtered!
Beautifully made and aesthetically pleasing - the cone type K&N. Stainless steel backing plates, perfectly formed cotton gauze filter with moulded sealing edges, and very simple fitment. The promise of an extra six horses for such a simple swap that looks this good proves beyond temptation for thousands of folk. But then all that glitters ain't necessarily gold. Long term the investment for any K&N filter is worth it, as you'll never need to replace it ever again. A million-mile warranty will see to that providing you wash it and re-oil it carefully - following the manufacturer's instructions to the letter of course. Main draw back with this type of total replacement filter is the row. The induction roar when putting the engine under load (accelerating and when cruising at speed) is ear-bleed inducing. But then again - there's those that love it.
Bonus of using the replacement element is it keeps it looking very standard under the bonnet. Can be very useful in certain instances, a-hem…
It's a neat and tidy fit too - more than can be said for the state of the under-bonnet/engine area of my poor old nail. 120,000mile and still just hanging in there! The mechanicals are all OK, but the bodywork - yoy! Andrew at GRV finds it highly amusing. He's concerned that if I get it to go any faster, the front is just going to disappear in a cloud of rust particles! As you can see, the holes drilled through underneath are invisible with the bonnet up.
Oh the shame - that beautifully crafted shiny new filter being forced to dwell in that messy, dirty engine bay!
It's a really easy fit job though, and why it was chosen for this test. Fitted in minutes by even the most unaccomplished DIY-er. The only draw back is tuning it. This style of filter creates such a massive increase in fresh airflow available to the carb; it will cause fuelling calibration problems. There's absolutely no way the standard needle will be anything like adequate. And this is a continual bone of contention for me. I'd love a pound for every time I've been asked what needle would be needed for such-and-such an engine with one of these filters on it, or even why a person's engine isn't running very well since they installed one. Every shop that sells these should alert the purchaser to the problem and be able to supply a needle that may not be spot on for their engine - but at least have a fighting chance of getting the fuelling somewhere near so the car can be driven to a rolling road for proper attention.
That's an extra expense vendors often neglect to mention too - a rolling road session is absolutely vital to get the best out of either of these filter conversions. Not only is the fuelling affected - but ignition timing too.
Fine gentleman that Andrew at GRV (unfortunately now not currently working there since they have moved to a newer, smaller premises with no room for the rollers! EEK!). Doesn't mind mucking in to help things go smoothly/quickly. Once we'd done the two test on the standard filter case, he was in there like a shot to whisk the old filter case assembly and elbow off the back of the carb and stick the shiny new cone on. Mere minutes later we were ready for test three.
Before it was done though - I asked Andrew what he thought the result would be. He said 'seen plenty of K&N equipped cars on 'ere of all sorts. 1300 Minis seem to benefit from them most, but I reckon the cone'll win by a couple of extra horses'. Now that's what I had generally heard and had been advising for some years - as in 'it's not worth going to the pancake/cone filter unless you want to be deafened, there's only a couple of extra brake at the top end and fuelling is a pain to sort'. Sound familiar to you? Turning the tables, Andrew asked me for my forecast on the fight. Having driven the car with the standard case, then with holes in it, I was sure there wasn't going to be anything in it. Andrew just smiled.
Another thraping on the rolling road for this super-sonic 1100 engine. I have to tell you I am enjoying driving this engine around immensely. It is so tractable, pulls so strongly, and is such fun, I'm starting to look for reasons to 'pop down the road' in it. I always come back with a silly grin on me mush! Anyway, back to the plot. Or plots in this case.
Graph A shows the results of the standard case with element (KCFILT1 trace), and the standard case with the holes un-taped (KCFILT2). A quick buzz up on the standard case to get a base line (KCFILT1) - and it gave the same figures as when I had it tuned up last month except we picked up an extra 2 lb ft of torque (now 61 as opposed to 59 last time). Possibly because the engine now had 1500 miles on it instead of 300. Consistent that. Even that bloomin' vertical 'hiccup' is still there!
The second run was preceded by a quick run up to check the fuelling wasn't going badly wrong. Last time on the rollers, the mixture was slightly rich, but was left alone as finding a needle to provided spot-on fuelling wasn't likely and 'tweaking' (filing) was out of the question as it would be impossible to reproduce it accurately, and definitely not available from any shop. I wanted to be able to use the results as recommendations for future inquiries. It proved to be spot-on with extra air supplied by the drilled holes everywhere except at maximum load/peak power where it had leaned out slightly. To rectify this, we changed the needle from the 'AAA' to an 'AAM'.
The result is trace KCFILT2. A further 2bhp was gained, and another 3lb ft of torque. Serious stuff for such a mildly modified small-bore engine. These peak figures don't tell the whole story though - the graph clearly shows the gains were wide spread.
Now the 'big one'. Graph 'B' shows trace KCFILT2 with KCFILT3 overlaid - the result of fitting the cone type filter. From the K&N 'blurb' and the prospect of an extra few horses - we were expecting a leaning out of the mixture everywhere - but particularly at the top-end. Boy were we surprised! The CO leaped from 3.8% with the standard case with holes to 5.5% with the cone!! Wild. And as the trace for KCFILT3 shows - a drop off in power and torque - a max power of 54bhp and 63lb ft. The CO wasn't out of the ballpark, and certainly not enough to cause the power loss. And again, the graphs show these differences weren't restricted to peak power.
Andrew was more surprised than I was. My prediction seemed proven, but better than I expected. Andrew was amazed the cone filter actually made less power. But why?
Experience and a great deal of testing over the years suggest the answer is in the way the standard filter case is designed. The A-series engine with its Siamese 5-port design causes some very powerful shockwaves within the induction system. I believe the elbow on the back of the carb and the volume of the filter case dissipates these very effectively. The cone/pancake type filters reflect these shockwaves back into the induction system, causing induction pulse problems.
The unexpected increase in CO reading at peak power would also suggest that the cone filter is causing a constriction at the carb mouth, reducing airflow. A reduction in airflow will cause power loss through depleted volumetric efficiency.
That famous phrase 'all that glitters isn't gold' is perfectly illustrated here. Something that looks and sounds like it should produce optimum results doesn't. Without a doubt, the optimum set-up on a small-bore engine is the standard plastic filter case with half-a-dozen three-quarter-inch holes drilled in it and using a K&N element.
A bit of good news for those 'noise' junkies - those extra holes give the induction sound of a Weber side-draught when under load, but are as quiet as a mouse when on part throttle. The noise IS addictive though!
I would also state here that this is not necessarily the case with the large bore engine units. I have had reports of gains of up to 6bhp when using a pancake (not cone) type K&N over the standard filter case set-up from a reliable source. Unfortunately I do not know if that is the standard filter case with standard filter, or with K&N element and other suitable mods. That'll be another test then…