Hertz Multiplier

Has any one tried a frequency multiplier on the MAF? I think this might be a good solution to the 1g count misses. Just thougt I'd ask before I waste my time trying it:D

-Dan

It seems the output is a 5v amplitude square wave, where more air increases the cycles/sec. A multiplier will allow a severly hacked mas to read correctly; provided you know the ration of the areas: metered vs. un-metered. Just some stuff I've never read about.

-Dan

Will this allow my O2's to cycle at idle and part throttle cruising with my hacked 1G MAS? Even with my S-AFC at +50% and 660cc injectors, my O2's don't want to cycle. My car idles at 24 Hz, and anything above 30 Hz and the O2's will cycle. Do I need to switch to a 2G MAS or will this frequency multiplier fix this? Where do I find one and how much $?

I have a 2g MAF and don't recommend one. The multiplier will make your ECU think its seeing more Hz. So yes it could solve your problem. I haven't made it yet, and have never heard of any one else going it before. It should allow a hacked 1g MAF to run like stock, or at least much better that just putting a resistor in to richen it up. I'm hoping someone will respond who has tried this before I begin to work on it;)

-dan

I wondered how will this multiplier circuit help MAF overrun since the problem lies with the vortex sensor? I am not an expert in this sensor but a multiplier circuit will help those unmetered air to be added (by some factor) post sensor to MAS but it still does not help MAF overrun problem.

The key is having a MAF hacked enough so that the max Hz count is low. Then, multiply the (lower) output frequency by the ratio of the area hacked + total area. So the MAF will count only what is was designed to, say 1500Hz, but the multiplier will send out 2200Hz to the ecu; provided you have figured the Hacked area correctly. I think this will work better than just tricking it into thinking its cooler outside.

-Dan

If you have a small turbo. I can see how that will help. But if you have a big enough turbo pushing 20+psi, no matter how much you hack the MAF, you will still have chance of MAF overrun. The way around this problem is to allow more unmetered air like we agreed. How? Using bypass pipe like the one Dejon Tool is selling.
Using a multiplier may help. However, I do not see the advantage if you already have a AFC.

How tough to build the circuit? Can it be adjusted using a potentiometer? If you have something, I am game enough to try it.

You can get a microprocessor and use a look up table for different factors of multiplication, depending on the area hacked. This would work well with any turbo combo that needs more air that the MAF is designed to flow correctly. As for over run at 25 psi with a 20g (my situation) and a 1g MAF I will need another opening, other than the MAF. I plan on using a stock MAF with another outlet about the same size as the MAF's. This will give me double the flow, and regular Hz. The ECU will think it is stock. I don't know what processor, nor have I designed a circuit. I am just throwing the idea out to see in anyone has tried it, if not then I will. I will do it for free, not try to sell it or some crap like that, just parts cost. If it works the way I want it to it will be like an automatic afc, but with nearly infinite +/- fuel control.

-Dan

Originally posted by tanner261
Will this allow my O2's to cycle at idle and part throttle cruising with my hacked 1G MAS? Even with my S-AFC at +50% and 660cc injectors, my O2's don't want to cycle. My car idles at 24 Hz, and anything above 30 Hz and the O2's will cycle. Do I need to switch to a 2G MAS or will this frequency multiplier fix this? Where do I find one and how much $?

What are your o2 trims at idle? You seem to be dumping waaaaaaay too much fuel at idle even with a fully hacked mas.

There is another thing that came to my mind. We actually want to fool the ECU into seeing less air to get better timing table. If you compensate the airflow with the multiplier, then you are not using the advantage of larger injectors.

I have a 2g mas on my 1g and love it!!!

Compared to my 1g mas... my car idles a 1000x better and runs alot better too.

My max HZ reading on the 1G mas was over 2000

Now with the 2g mas and stock intake pipe (dejon tool CIP on the way) max HZ is at 1860 ... although that may be high... I may need to get the dejon intake pipe that has the bypass tube to help out???

Originally posted by jw
We actually want to fool the ECU into seeing less air to get better timing table. If you compensate the airflow with the multiplier, then you are not using the advantage of larger injectors.

My goal was to flow more air, but read it correctly. If I have a 2in diameter hole, unmetered, No injectors will compensate. The processor will be programmable; to you can let the ECU see what ever you want, if you have bigger injectors you just won't need to compensate as much( i.e. multiply by a smaller #).

-Dan

Strikes me that this is in effect, what the apexi and hks fuel trim computers do... lie about the airflow (hz) sent by the MAF.

I'm working on the same thing, using a microcontroller (dragging the microcontroller/breadboard/laptop programming environment with me to work, bars, friends, restaurants. i have no life).

have you started on your project? i've yet to put the signal from the maf on a scope to verify the pulse train coming out tho.

one should be able to change the frequency based on rpm, map, temp, and anything else one takes a fancy to, just depends how much work one wants to get into, and how many a2d channels you want to mess with. my sense is that you'd have to either use a pretty fast cpu, or use external freq2volt facilities.

anyone have any suggestions for a clean pulse train (5 v square wave) souce for determining RPM?

michael

Originally posted by shabby


What are your o2 trims at idle? You seem to be dumping waaaaaaay too much fuel at idle even with a fully hacked mas.

My fuel trims are near 140% <-too lean. My O2s won't cycle at idle because there isn't enough Karman counts (Hz) going to the ECU. If I barely blip the throttle my O2s go way rich (.96). And when I blip the throttle, my Hz count goes over 30 Hz.

So, if I put a flapper in front of the unmetered section, I know my O2s will cycle. But I don't want to put a piece of metal in front of my compressor wheel. Another way around this would be to install a 2G MAS. But that costs at least $100, unless you can find a 2G MAS for really cheap and build an adapter plate for the exhisting K+N. Which I may end up doing. The other option would be this Hz multiplier which is the same thing as an AFC.

I guess the more I think about it, I really need a 2G MAS to cure my driveability issues. I have no problem with WOT, I am getting 28-30 degrees of advance with no knock and a hacked 1G mas and 660cc injectors.

Also, I think my problem has something to do with the injectors I'm running. They are out of a fuel injected Bombardier watercraft engine. They are made by Bosch and are designed to run at 4 Bar. I'm only running 43 psi as a base fuel pressure. I don't want to crank my fuel pressure up too much because my AFC is already at -15% or so across the board for WOT. If I go leaner than that, by Hz count goes below 1400 Hz on the logger, and suposedly (from Todd's site) the ECU won't advance timing as much if the ECU is only seeing 1400 Hz or less. The 1G MAS has different timing tables for different Hz readings.

Another idea for high hp aplications would be to just get two 1g MAFs. This way I could just get the two square waves, convert them into a numeric value (Hz), then generate an equivalent summed square wave out to the ecu. BTW what do 1g UN-hacked MAFs run?

Dan

Now, I am confused how 2 MAF can work together. Do you convolute the signal or something? I am a little rusty on Electrical 101 :D

I do not think it is difficult to build a multiplier or divider circuitry. The fact that our vortex sensor overrun cannot be solved. Allowing unmetered air is just a temporary fix to the problem, when more unmetered is allowed, it cause a vacuum leak effect and hence bad idle problems come into play. I do not know if the unmetered air portion gets a mulitplying factor over the metered portion will do the trick. It's no harm to try anyway.
If you can come out with a circuit real quick to do a test. I am game. I am just too lazy to do the calculations and stuff :)

However, if you think about it, isnt that what an AFC is doing?

I would like to know why you feel that a 2G MAS is not good and what you wouldn't recomend it?? I have installed MANY 2G MAS's in 1G and not 1 with a bad effect on the car and all of them run fine after proper tuning. I see more things that are based on lackl of tune then on the hard parts used. Many cases it was something skipped or over looked but in most cases it's a lack of proper tuning. But I am open ears to what you feel is the reason.

BM-

Originally posted by Rdy2race
I would like to know why you feel that a 2G MAS is not good and what you wouldn't recomend it?? I have installed MANY 2G MAS's in 1G and not 1 with a bad effect on the car and all of them run fine after proper tuning. I see more things that are based on lackl of tune then on the hard parts used. Many cases it was something skipped or over looked but in most cases it's a lack of proper tuning. But I am open ears to what you feel is the reason.

BM-
Is this directed at me? I did not mentioned that 2G MAS is not good or anything but my post is directing at two 1G MAS running in parallel that captain4g63 suggested.

I've been getting inconsistent readings rich/lean around 2k with the 2g meter. It might be a vacuum leak, anyway, I never had to fiddle with an extremely hacked 1g sensor after I tuned it. I've had to much school work to mess with it right now. If anyone has any ideas...it would be helpful.

I'm still going to explore my multiplier idea. I think it will work the best with 2 1g MAFs. Reading will be as accurate as stock. I'm hopping this will eliminated the need for the AFC/tuning.

-Dan

JW, no that was to capt. I think it was him that said it. The 2G's seems to work nice but yes do need more messing with on a daily but a over all better running car. Im a 1G hacked mas it is more stable as in when its tuned in it's good but the quility of idle and low end responce or state of tune might be better said (like when 5 out of 10 tries you do a half stap of the gas in 3rd and getting different reading and WAY different. ) the inaccurate count and turblient air make for a harsh tune if it is for all out power or some power is given up for sake of driveability.

Don't get me wrong many have 1G hacked MAS and run very well but this is the group that puts the time and effort to make it work. the 2G MAS seems to be more "user friendly" it's easier to tune the car and it will idle well and still be able to make the same power. this is all that I ment.

I didn't see what capt was getting to with the twin idea at first (must have missed that I thought it was just playing with the mas signal. )

BM-

Been here, done this. I was plaving around with an EVO MAF for awhile. It looks alot like a 2G MAF, but it has a 3" round oulet opening like a 1G MAF and a hex 2G inlet. It tends to read about 50% more air than a 1G MAF. I took the output of the MAF (5v square wave) as an input to a Frequency-to-voltage convertor. Voltage went into an A->D convertor. A->D went to a divide-by-X circuit. I now have a digital representation of a root frequency directly from the MAF without having used any microcontrollers (no boot code, no unexpectrd resets under WOT, just pure logic). This base frequency goes to multiplier with a DIP switch for multiplication factor. Now, much like the MASC Hotwire MAF, you can adjust the final base Hz for whatever compensation factor you need applied for your chosen injectors to get you back near zero on the AFC in closed loop. This gives you the base correction you need for better adjustability at WOT.
If you are willing to do all this work, toss the MAF entirely. Get a 3-Bar GM MAP sensor (5v P-P output). Go MAP->a/d->microcontroller. Add an IAT sensor->a/d, a Baro sensor -> a/d. Plug it all into your uC, send a uC output to a d/a, the d/a to a voltage to frequency convertor, add code and stir. Voila! you just built your own VPC without the need to "chip" and you can toss the MAF out the window.
Or you just buy a VPC for $800.00 and call it a day.

I would still like to try and get a 400hp car to run reliably with 1g MAF for under $100 that doesn't need much tuning; hopfully no afc. I've dropped the multiplying idea. My only concern is not knowing how high of a frequency the ecu can understand. I think TMO has this figured out though. The highest Hz ive seen is 2.2k, I hope the ecu knows what to do with 3K.

Actually, the ECU stops caring what the incoming MAF count is at 1606 Hz, from what I understand. No, wait a minute...now that I think about it, the 1606 Hz limit is a limitation of the logger's ability to READ the MAF count, not necessarily the ECU's ability to use it. So nevermind that last bit. It's early and I'm in need of more caffeine.
Your goal is not completely unreasonable, but understand that the ECU's maps won't change. Hacking the MAF and reducing the count while simultaneously increasing the engine's actual flow (bigger Turbo, more boost, porting stuff) helps to prevent fuel cut and missed counts at high flow rates. Al Blaha understood this, (as do you from your post here...) and opened up his MAFs for big flow at the top and put a flapper door to close that gaping hole for a better idle. It is easier to get there with a Dremel tool and some smarts than with a soldering iron and a handful of MAXIM's IC's. I promise. ;)

I'll get back in a few weeks when finals are over. I'll either have fallen flat on my face or figured somthing cheap out. I just don't have the time to test things out right not. thanks for the info