Miller MAF. How this started


4 months ago a group of SpecE30 racers asked me to look into the Miller MAF kit as an alternative to our AFM. Their points were:

  1. There is no source of AFMs that are “known good”. As a result, in order to end up with a good AFM, the routine had become to purchase 3, at $300-800 each, and take their car to a dyno to test them. After choosing the AFM that seemed to work best, they sent the remaining 2x AFMs back, sucking up any restocking fee. Dyno cost and hassle being worse then the cost of the new AFM.

  2. AFMs are the primary engine management failure point. This has cost racers a lot of headaches and money as they struggled to figure out what the problem was, that often could only be tested at the track.

  3. AFMs are tough to rigorously test. Any other part, you can buy one off of Ebay and, one way or another, figure out in a couple minutes whether or not it works ok. Buy a used ICV and if it vibrates it’s probably ok. But, this side of BavRest’s airflow and V ratio test, there’s no way to even come close to checking AFM function. The test procedure in the Bentley is wrong because it’s the eta test. If you figure out the right test protocol for our AFM, the V ratio result will tell you whether the AFM is functioning, but not about it’s calibration or linearity. The only decent way to test the AFM is to get on a dyno, preferably with AFR being sniffed. Testing on the track will give you rough info, but few people are going to be able to sense the brief loss of 3hp at 5k-5500 rpm because the ebay/refurb AFM is making your car a little lean.

  4. Relative to the other parts we buy new/used for our cars, AFMs are unique in that they are “tunable”. There is a tradition of folks turning the tension spring. Even to this day there is disagreement as to the impact of moving it CW and CCW. Therefore any AFM you might get might have been screwed up by tinkering.

Prior to that conversation, I had no idea that there were concerns re. not having a source of refurb AFMs that could be relied upon. Their points struck a chord with me because I’d gone thru AFM hell in 2010 trying to understand why my car ran like crap at 4500-5500 at WOT. 6 events were hosed as I tried to figure out what the problem was. I was hamstring by the fact that I could only test at the track. I replaced every goddamned thing that had anything to do with engine management, to include the AFM. As the hosed events came and went, I got more desperate, I found info that said that V drops could be the problem. So I replaced my battery, wired in a direct (switched) line from battery to DME. To exclude the possibility of power cut-outs, I wired in switched 12V from the battery to the fuel pumps and a master relay bypass. Every month I would try harder to exclude possibilities.

I drove 14hrs (one-way) to MidO for an event and it too was hosed by the engine cutting out. That was the final straw. When I got home I pulled apart every power and ground connection still in use in the car, every damned one. Then I spent 2wks abrading the mating surfaces of everything, to include every male/female pin, and re-assembled with conductive grease. Several connectors that lore said were problematic, I cut out and hard-wired. And I still didn’t beat the problem.

So I started replacing shit all over again. The problem went away when I replaced the AFM a second time. So it was all about 2 bad AFMs. The 3rd AFM is the one in my car right now, 8yrs later.

Remembering that AFM horror show, I told the guys that I’d look into their Miller MAF idea.

I created a couple facebook threads to collect information from the racers. The question was…“Do we have a source of AFMs that we can trust? That is to say, no anecdotes re. bad AFMs from that source.” The threads were barraged with all sorts of tales of woe. To my surprise, it was clear that every source of AFMs had multiple stories of bad ones purchased.

I contacted BavRest. They graciously spend a bunch of time with me discussing how they refurb AFMs. I think that it’s likely that they are the only source of AFMs that is genuinely trying to recalibrate them. My guess is that the other sources of refurb AFMs just pretty them up and do a cursory test of “seems to work” or “inop”. Unfortunately, there are as many complaints about BavRest refurb AFMs as any other source.

I contacted the Aussie SpecE30 racers because they’d been using a Miller kit for years. They had gone to it for the same reason…too hard to find reliable AFMs. The Aussies were very happy with the Miller kit. I checked with Shawn and then contacted Miller and asked them to send a kit for testing. The cost to us would be $300.

Charms of the Miller MAF kit kit.

  1. No moving parts, so the probability of component failure is low. Replaces a part that is notorious for being a problem.

  2. Cheap. Lower cost than an AFM. A helova lot lower then buying several and then taking your car to a dyno to decide which AFM to keep.

  3. Fixes chipping. With the Miller kit the temptation to chip one’s car goes away. It’s one thing to seek the significant advantage available in the performance difference between an OEM chip and an a aftermarket chip. But that “significant advantage” is no longer available if you already have chip from an industry leading aftermarket chip tuner. There is temptation to chip the car if there is a 5hp delta. But if it’s only a 0.5hp delta, the temptation for the other guy to chip his car has gone down 90%.

There is a problem developing behind the scenes that could cause SpecE30 some damage. It’s the uber motor that has had it’s intake blocked in some way to reduce peak hp. Chipping is part of the uber motor problem. The MAF kit taking away most of the temptation to chip would go a long way to mitigate this developing problem.

  1. Miller says that the most significant feedback they get about their kit from E30 types is the improved throttle response. It apparently goes from “clumsy and laggy” to “instant”. It remains to be seen how much this resonates with folks. Note however, that “More fun to drive” is not a good justification for a rule change.

The dyno standard. If the community is going to seriously consider the Miller kit, we need to know what it’s going to do to dyno results. Not just “how much more power?”. The community needs to see if the shapes of the curves become more standard from car to car, that is to say “fewer unusual exceptions”. The community needs to see if anything interesting happens witih our problems with dyno accuracy and repeatability. Consider just the issue of the jagged dyno line where someone becomes at risk for DQ just because their jagged line hits the max hp.

So, drivers that are interested in checking out the Miller MAF kit are going to get a chance to do so. We’ll get a variety of cars doing before/after dyno runs and see if we find anything interesting. Then lets see what people think.


Scott, again, thank you for the transparency, and now the “backstory”.

Anecdotal story in line with yours. I’ve been involved in two similar “WTF is going on with the car under power” situations. One on my son’s old DD (non-track car) and another with an “to be unnamed SE30 racer”. On my son’s, after much reading and testing and poking around, we installed a Miller setup he bought on R3v and bam! Instant fix. His problems happened all over the place in terms of throttle position. Some of the odd behaviors (like hard or no start, shutting off when lifting, etc) all gone. Completely new car to drive and I got zero calls or texts about the car leaving him stranded. Which hurt him as he was in high school at the time and couldn’t use the car as an excuse to miss curfew.

The SE30 example was similar, although different because it is a race car and at full throttle. But this guy was getting mad because on long straights, other SE30s would pull away from him and he knew all of the other variables between the cars in question were nearly identical (weight, etc) as the other guys were his friends and shared information. He also knew it wasn’t them cheating, he had worked on their cars several times with them. It was only after he started looking closely at the data and their’s to figure out that they were leaving the corner before the straight at nearly identical speeds, and accelerating away at the same rate… for awhile… then there was a definite time the other cars were accelerating at a higher rate and his rate dropped, then kinda came back only to repeat again after the shift. It was so minimal he couldn’t feel it while driving, but was there in the data. I believe he purchased upwards of 6 AFMs and spent an afternoon at the dyno, picked the two (for a spare) that were the best (and not in peak HP, that wasn’t the primary requirement based on the data). He has since run that same track with those same guys and is no longer frustrated with them pulling away. When I pointed out to him over beers that $$$ was not the goal of SE30, he said he agreed and that he didn’t do it to be national champ or the winner at all races, he was just getting frustrated with the experience of it all and was about to build the car into something else, so it was worth the dollars (to him at least) to have a better experience.

Now, I haven’t gotten that far in my build yet, I hope to have the car running this summer, but I’ve been concerned the entire time about the AFM because everything else on the car is refreshed/rebuilt (WAY high mileage donor with a destroyed head) and I know I would flip my lid if things didn’t run right because of something like an impossible to find correctly functioning AFM.


The attached doc was put out by the Aussies in attempt to help people with the problems they had with the MAF. The bottom line is that the AFM and the OEM chip is more tolerant of engine management problems. AussieMAFProbs.pdf (483.9 KB)