jkayed

subscribed!

Howard Coleman

iTrader: (6)

as you may know i am in the late stages of wiring in my ViPEC V88 ECU.

one of the reasons i opted for the V88 instead of the V44 was that it contains a highly sophisticated knock system. actually two discrete systems, one for the front and one for the rear rotor. after receiving knock data i can slice and dice it, and ultimately define knock. once defined i will be able to program an immediate response w re to fuel and/or timing.

i was talking w Cam Worth sunday night and we were talking knock sensors. he suggested i look at the RX8 sensor making the point our sensors were probably designed in the late 80s.

i will say that i think they work well based on my Power FC Datalogit experience. the FD sensor is a "Broadband" sensor. it has one wire, is tuned to the rotary knock frequency (i seem to remember 4200 Hz) plus or minus 1000.

the newer sensors are two wire "Flat Response" sensors that cover 5K to 18K Hz and the computer sorts it out. the RX8 sensors, and most other contemporary sensors, i believe are of this type.

is anybody out there running an RX8 sensor?

howard

Mouse07

subscribed!

Howard Coleman

iTrader: (6)

as i wait for my initial manifold to be finished i came up w something interesting. as you may know i will be doing my AI preturbo and i really wanted to get some temp numbers.

most intake air temp senders have a 5+ second reaction time. i guess that's why i always see the same IAT number on my dyno runs.

how about Milliseconds.

i am going to position a type K thermocouple (similar to an EGT sensor) before and after my intercooler. of course we don't need Ms but we do need to find out what's going on before it has gone.

Rick at EGT Technology is making me a couple of T-couples that are designed to do the job. a different wire, sealed w epoxy so the water/meth will not be an issue. accurate to .083 of one percent from 32 to 3200 degrees F.

i plan to retain the two 1000 cc injectors in my elbow and hook (only) one of them up to my WM50 system.

i will then be able to switch between pre and post turbo w a keystroke or two. both injectors are similar in delivery and totally/separately tuneable.

hc

scotty305

iTrader: (2)

Howard, the OEM IAT sensors may be slow but I would be surprised if an open-element IAT sensor (for instance, GM IAT, Triumph IAT) would take 5 seconds to respond.

This should be pretty easy to test, just borrow your wife's hairdryer and hold both the IAT sensor and the EGT sensor about a foot away.

rxspeed7

iTrader: (5)

Open elements react pretty fast. But no where near as fast as a thermocouple.

Howard, as a cheaper(money not quality) alternative to buying "automotive" tc's check out omega. They are a lab grade company that specialize in thermocouples. They are all the same thing, people just slap egt in from of the word thermocouple and charge 6x the amount for them.

RXtacy

iTrader: (7)

If you're just putting the thermocouples in the intercooler piping you don't really need one with that large of an operating range.

rxspeed7

iTrader: (5)

Very tru, a j-type tc would prob give you a more precise measurement for your range but it's really a moot point to measure it before and after. Unless your just wanting to see the intercooler do it's job.

Howard Coleman

iTrader: (6)

that is a nice site for thermocouples... and the J looks good. since i have dealt w EGT Tech for a number of years and they always provide good stuff custom to my needs, they sell a really nice compact amp and Rick will be using his pick for my specific needs (water and meth) i will just roll w him. and yes i do want to have a dynamic accurate measurement as to my IC. i have close to unlimited logging inputs so i will take advantage of it.

hc

Grant M

Having the sensors before and after the IC will help identify if it actually working once the pre turbo injection kicks in and if the IC is actually reducing the temps any further.

blue87

iTrader: (2)

Just FYI J type TCs use iron as one of the metals so they rust. Use K type with modern amps you won't see any significant difference in accuracy for what we are looking at.

Howard Coleman

iTrader: (6)

update:


things are moving ahead on all fronts. my fresh motor is together, the initial manifold is a few days away from completion, the trick AI pre-turbo fuel injector fixture is complete, the ViPEC V88 is almost completely wired and the BW turbo is all set w a new CHRA. lots of detail work remains but the finish line is in sight.

here are some details...

i continue to be impressed w Full Function Engineering's products. as mentioned previously, we are lucky to have good stuff still being created for our aging but awsome FDs.

here's a few snaps of the beautifully engineered crank pickup. while the FD has a wonderful OE crank pickup things get a little drifty around 8-9000 rpm and a 32 tooth resolution greatly improves the stock 12 tooth.



FFE gives you the choice of the stock type pickup (reluctor) or a Hall Effect. i chose the Hall sensor for a number of reasons.




here's a pic w the toothed wheel installed. a perfect fit.



FFE has recently released new design Primary and Secondary fuel rails. a very rational clean sheet of paper design. i am going to plumb them w the kevlar wrapped lines.



the Bosch EV14 fuel injector comes in short, standard and long lengths. i found my 1000 cc primary injectors were the long variety so i will have to swap them for standards. i am set w the 2200 secondaries as they are standard.



i love my double disc 8.5 inch Quartermaster cerametallic unsprung. it is fairly civilized and yet will easily handle the 600 (max ) hp i am looking for. 3 years and looking good w re to the frictional surfaces. being a larger (18% more frictional area) double disc than the Exedy, i am able to run less spring pressure (2000) so it is easy on the leg. of course the reason i run it is my Tremec T56 6 speed. Exedy would not make me T56 friendly discs. if you are wondering, that is a FD automatic flywheel (flexplate)... weight 3 pounds.



the motor should go in tomorrow. the manifold should be done next week. lots of finish wiring to do. i will be using two RX8 (Bosch) knock sensors. four IGN-1A inductive coils....

i will be running braided line from my AI pump in the rear thru a "Y" which will go to one EV 14 1000 CC in the elbow and one EV 14 1000 CC pre turbo. each will be commanded separately so we be able to compare pre turbo and post ic & turbo. since each has it's own tuning map there will be no problem w drowning the motor in spool. injectant will be WM50. i do not anticipate running both injectors at the same time but it is a remote possibility.

first up will be the previously tested BW S300/63, then the Garrett 3574/P trim back end. (we will be testing two rear housings thanks to Elliot White/Turblown). then it is my GT4094 and after that probably a GTX35.

oh, another item will be on board. two K thermocouples. one just after the turbo and one just after the intercooler. since they react immediately and are accurate we will learn a bit about charge air temperature dynamics.

the manifold is taking a bit longer as we are setting up to build an initial run of ten.

lots of good things are happening at Beyond Redline, Luke continues to work his magic, so things are lining up.

howard

FullFunctionEng

iTrader: (13)

Looks good Howard!

Be careful with the ev14 (SS ball/seat) without any lubricants (i.e. straight water). They can/will rust.

shaun.sheldrake

Great thread Howard - thanks for all the insights, keep it up.

Down here (NZ) there are a number of guys running the TO4Z, I'm planning to retire my trusty stock twins (non-seq) at the end of the year, so your timing is perfect.

Did you consider the TO4Z? Thoughts?

Thanks
Shaun

Howard Coleman

iTrader: (6)

"Did you consider the TO4Z? Thoughts?"

i group our turbos in the small, medium and large categories.

GT35 in it's various iterations is small

TO4Z, GT500 (A-Spec), PT67, BW S300 are medium

GT42. T51R are large

grouping is based on average compressor wheel area. (for all turbos see my sticky thread "Turbo Comparison.")

GT35 comp wheel is 6.38 sq inches

TO4Z, et al, is 7.0 sq inches

GT42, et al, is 8.38 sq inches

and sure there are turbos that fall between the groupings as well as newer billet comp wheels that offer additional flow V area.

one overlooked turbo aspect is the actual exterior dimensions... for example one of my favorite turbos, the one i own, is a GT4094. it has 97.5% of the area of the GT4294 yet is significantly more compact. because of their size, GT42s have to locate in front of the engine rather than alongside like the medium and small class turbos. there are loads of reasons why this is disadvantageous.

granted, the GT42 has a 7.5 sq inch turbine while the GT4094 has a 6.4. we will soon see if this is important.

GT4094r.... big output, medium size.

the medium turbos are interesting. they are compact, fit alongside the motor and when everything is right can make very close to 600 SAE rwhp. for most of us that is enough in the power department.

the current medium turbo in my test is the BW S300/63. the turbo is an outlier within the medium turbos because of two reasons...

it has a 52 trim!

all others have a 63 trim. lower trim equals more midrange... in theory. theories are what this thread is about. let's test the 52 against the 63 and see if there is a difference. i have just a few pulls on the 63 and have found it likes top end. it was very happy at 8600 and still making good power.

EGT at 8600 was 1550! which brings me to the second point re the BW...

it has a 6.31 sq inch turbine. most of the other medium turbos have a P trim or smaller. a P trim wheel is 5.89.

in my book whether the trim or hotside differences result in meaningful output/reliability advantages is yet to be established.

certainly RiceRacing has gotten more out of the TO4Z than most... running over 30 psi boost for a year and maintaining a healthy engine is amazing. most of the key is in the setup rather than the turbo.

his pre turbo AI along w the proper balance of water/meth cools the in-turbo compression extending the turbo oxygen capacity while keeping the motor cool.

we will soon see how this works.

howard

duval7

iTrader: (1)

I may be late asking this, but I have always been told that a pre turbo AI setup will corrode the compressor wheel? Whats the advantage of injecting pre turbo as opposed to just before the throttle body?

Howard Coleman

iTrader: (6)

"Whats the advantage of injecting pre turbo"

"pre turbo AI setup will corrode the compressor wheel?"

i thought i would wait for others to take a swing at this and since no takers..

first off, i have no first hand knowledge whether pre or post turbo AI is best. i do expect to gain that insight the next time i am on the dyno.

there are numerous people on this board, as well as preturbo military aircraft installations, that provide a good indication if done right there may be serious advantages to pre turbo.

all compressors reach a max point where they cannot produce any more output. the limit is created by the heat of compression.

cool the heat and you move the max point.

it is easily possible to calculate backwards from power output that a compressor is putting out more than the compressor map displays.

maybe 10%.

as to the fragile cast aluminum compressor wheel... i must admit that was always the reason i discounted preturbo.

the key is proper atomisation of the injectant and it is not a difficult problem to solve. i will be running an approx 50/50 mix of meth and water. meth corrodes aluminum but when mixed w water, actually diluted since meth and water mix, corrosion is not an issue.

others w actual experience are welcome to chime in.

hc

tom94RX-7

iTrader: (5)

My car made some more peak power when I switched the nozzle from the post ic location to the pre turbo location while on the dyno.

Shortly after I first tried pre turbo water injection last year I was having a weird problem with my AEM progressive controller, it would spray water for a second or two every time I turned the key on or off. Then I found damage to the outside edges of the compressor wheel, I think because that puddle of water in the intake pipe was getting pulled through the turbo and damaged it. I didn't have any damage before the problem with the progressive controller, and no extra damage after fixing that problem.

So its very important that everything works properly, for many reasons, and good atomization

Howard Coleman

iTrader: (6)

what's weird in this picture?

congratulations, but what the heck is the oil injector doing in my knock sensor boss? of course if you have read my stuff you know the last place would be the stock location. i really don't want dirty carbon filled 4 cycle oil on my apex seals.


todays job was to mount the two RX8 new generation Bosch knock sensors. they are sort of tricky as the hole in the block is 10 mm and the hole in the sensor is 8 or thereabouts. i thought i would build a stud that fit the block and then slimmed down to get thru the knock sensor body and add a nut.

so i was looking for a 10 mm bolt and needed to make it about .6 inches long. there sitting in front of me happened to be a discarded oil injector... perfect width, perfect length after a quick chop job. all my 10 mm bolts are breathing easier.


here are the pieces to get it done... an oil injector, the modded chopped off threaded portion to be used ready for welding and an additional thinner threaded extension along w the knock sensor.




finally, the Mazda oil injector doing what it was supposed to do. thanks Mazda




howard

1989fc3s

suscribed for final BW numbers. If i can find a good shell I'm planning a 13b-RE w/ s366 .91 a/r. On pump only for now

shaun.sheldrake

Thanks Howard - as always, plenty of great insights ... will keep watching with interest to see what the dyno sessions reveal.

Howard Coleman

iTrader: (6)

given the fact i expect to be running around 30 psi w my new AI system i thought it was time for a few system upgrades. out w the plastic line and in w the dash 3 serious stuff. i run 115 psi AI line pressure and didn't want to bet the motor.

i also decided not to retain my FJO AI 40 micron filter. bigger is better w filters. i found an Aeroquip 40 micron, stainless steel insert that is good for 200 psi. most fliters of this type come w humongous ports and the AQ 12335 is no exception. the good news is AQ sells a nice O ringed reducer to dash 6. i have a 6 to 4 and a 4 to 3 coming from Brucespeed.com.



the CAD pre-turbo injector fixture should be complete friday along w the manifold. i have a few of the fuel lines to finish (waiting on a hose end which is in the mail). finish some ECU wiring, install the motor and measure for K thermocouple lines (intake air temps post turbo and post IC).

enjoying the journey i think.

howard

lOOkatme

iTrader: (19)

I am not sure if this is a true statement. Our fastest reacting sensors in aerospace were the open type platinum elements. They responded in like 2 Ms.....they were used in liquid oxygen for pratt whitney engines and the wires were completely immersed into the liquid and open. you cannot get any faster than that. Thermocouples cannot beat this.

Thermocouples are faster in responding in the normal configuration when the wire is wrapped around a mandrel/substrate. But they are inferior in terms of accuracy..and stability. And super short thermocouples seem to have more error.

Thermocouples rule in rugged environments and heat...like inside jet engines....otherwise they use RTD's most everywhere else.

rxspeed7

iTrader: (5)

You are very correct, but in relevance to what were using them for and what the average consumer has access to, thermocouples are faster,cheaper, and more robust than most open element sensors.

Jobro

iTrader: (2)

Thermo-couples are terrible for noise immunity. Put one in an environment with large electric fields and they are next to useless, basically 0 dependability on their measurement.

IR is no good for accuracy I found I needed to put black electrical tape on anything remotely shiny to still get an inaccurate reading.

Many semiconductor based temperature sensors are also crap they get offset noise that makes them always read low on average when they are in noisy environments.

GM spec open element air temp sensors are just about as good as it gets for this application....