Milwaukee Midget

[fordboy628]

HEADER SIZING,  PART EINS. . . .

Midget,

Ran the exhaust pipe sizing calculations, using the following formulas:

Ac/s = cylinder volume*rpm/88,200

Where Ac/s = Area cross section of I.D.
rpm = peak torque rpm desired

AND,

Pipe I.D. = square root of (Ac/s*1.273)

So for the jewel @ 15.25 cu. in./cyl.

rpm                               pipe dia. I.D.   
5500                                            1.100
6000                                            1.149
6500                                            1.196
7000                                            1.241
7500                                            1.284
8000                                            1.326
8500                                            1.367

Since your Formula One like rod/stroke ratio is going to narrow the rpm spread between peak torque & peak bhp (from the normal 1500/1750rpm) to (???1250/1500rpm???) I think a goal of 6500/7000rpm for peak torque is reasonable.

Your existing 3 into 1 pipe at 1.25" I.D. fits the goal, although it is probably too short in primary length.    The other 1 3/8" O.D. headers from Mini Maven, with their longer primary pipes, may be the better choice.   It appears, as I suspected, that the 1 1/2" O.D. pipes will place the torque peak beyond the rpm range desired. . . . . .    although we could do one pull on the dyno to be certain.

I need some cam card specs from you to calculate the desired/optimum primary pipe length.    What I need is the EVO in degrees BBDC.     If you can't provide numbers, I can substitute a range, but I hate to guess. . . . .     What I want to do is to get both the pipe dia & length to be tuned for the torque peak, enhancing V/E.   We can then jack the torque curve with intake tract length, etc, . . . . .

For BMC mavens following along:  Differing displacements & rpm ranges dramatically change the primary pipe diameter required.   Take the time to calculate what is required for your engine, do not guess. . . . . 
 
Fordboy

[fordboy628]

It’s been a while since I can personally say I’ve made any progress on this project.  Yeah, a lot of running around, fact finding, etc., but to have actually done some physical work?  Well, that ended today.

Part of what needs to happen to dial in the engine on the dyno is creating some method of transporting the entire brain box and coil packs to the shop so any changes I make with the computer can be locked down.

To that end –



I’m fortunate to work at a place that has large quantities of audio cast-offs back in “the Junk Room” – A chassis, some old connectors – everything needed to recreate the electrical dynamics of an MG.  

Yeah, a junk room – which coincidentally also houses the lathe – and I’ve already expressed my opinion on that.

Now to dig up the throttle parts . . .

Midget,

OOOKAAAY. . . .

Nice job on the ignition box!!   Including the labeling!!    Just one question.    The red, chicken head knob, labeled "Knob".

Uhmm, I'm sort of afraid to ask what that knob does. . . . .

Does it flip the coil order from 1/4 & 3/2 to 3/2 & 1/4 ?   That might be useful. . . . . .
 
Fordboy

[Milwaukee Midget]

KNOB is an acronym for KNock sensor Output Balance, but that name is just a humorous function of the components, not what it does.

The KNOB knob is a trim control for a pair of crude piezoelectric digital/analog converters.  They take the firing output signals – two in this case – wasted spark ignition - from the ECU, and activate a pair of piezoelectric transducers, which when DC is applied, emit an 8000-9000 hertz audio signal. Yup, hf peizo buzzers, made from knock sensors out of a Ford Taurus.

That signal is passed into another pair of piezoelectric transducers (GM - less susceptible to noise) through a physical connection – essentially bolted directly face to face with a threaded pipe coupler - which creates a frequency modulated high impedance analog signal from the GM transducers mirroring that of the Ford units.
This signal is fed to a simple analog stereo amplifier which has a high pass circuit, shunting any electronic noise less than 7500 Hz, and a shelving circuit, which attenuates anything above 10k by up to 95 db. Essentially, it’s a bandpass filter.  The amp is an old Texas Instruments circuit I pirated out of a junk 386 desktop computer gathering dust in the attic.
 
I created a pair of half wave rectifiers on the output side with no capacitors, which eliminates the bottom half of the waveform.  That signal drives the amplifier until it just starts clipping, giving me essentially the bottom half of a square wave, and replicates a high frequency pulsating DC signal.  This analog synthesized pulsating DC is routed into the ECU which fools it into producing multiple sparks at the coil packs similar to a MSD setup, but at a frequency of 8000 sparks per second.

That's the signal the KNOB knob attenuates - so I don't overdrive the coil packs.

We'll see if it works.

[maguromic]

Learn something new everyday, I always thought KNOB was bourbon from Kentucky. Tony

[wisdonm]

     Is that going to be on the test? 

[Captthundarr]

I see you simplified the design by eliminating the flux capacitor, and adding bulltronium 235 may have caused unwanted radiance.....er something. Now I have a headache

[Frankie7799]

I see you simplified the design by eliminating the flux capacitor, and adding bulltronium 235 may have caused unwanted radiance.....er something. Now I have a headache

Im with ya on the headache part Capt. My heads been spinning for quite some time with all the numbers a flying. I can only read the build in small bits 

[Jack Gifford]

... pulsating DC is routed into the ECU which fools it into producing multiple sparks at the coil packs similar to a MSD setup, but at a frequency of 8000 sparks per second...

And... just how useful would these [frequent] "sparks" be, given the small coil saturation times? (A coil with that capability could run a single-coil 4-cylinder engine up to 240,000 RPM)

I doubt that you're just toyin' with us, so... what ARE you up to?

As all the manufacturers freely tell us, "MSD" (multiple spark discharge) only functions up to moderate engine speeds- usually around 3,500 RPM.

[Dr Goggles]

Hmmm, curiouser and curiouser. Is it cabin fever or inspired genius.......?   

[fordboy628]

HEADER SIZING,  PART EINS. . . .

So for the jewel @ 15.25 cu. in./cyl.    (J Baechtel formulas)

rpm                               pipe dia. I.D.   
5500                                            1.100
6000                                            1.149
6500                                            1.196
7000                                            1.241
7500                                            1.284
8000                                            1.326
8500                                            1.367

I need some cam card specs from you to calculate the desired/optimum primary pipe length.    What I need is the EVO in degrees BBDC.     If you can't provide numbers, I can substitute a range, but I hate to guess. . . . .     What I want to do is to get both the pipe dia & length to be tuned for the torque peak, enhancing V/E.   We can then jack the torque curve with intake tract length, etc, . . . . .

HEADER SIZING,  PART ZWEI. . . .

Midget,

OK, cam info for EVO is:  56 deg BBDC @ .050" lift;  73 deg BBDC @ opening.

Using the J Baechtel formula. . . . . . .

L(in) = [850 (360 - EVO) / rpm] -3                       (using 73 deg BBDC to calculate for primary length. . . . . )

rpm                       length
5500                      41.35
6000                      37.65
6500                      34.53
7000                      31.85
7500                      29.52
8000                      27.49


Using the A. G. Bell formula. . . . .

L(in) = [850 * (180 + EVO) / rpm] -3

rpm                       length
5500                      36.1
6000                      32.84
6500                      30.08
7000                      27.72
7500                      25.67
8000                      23.88

Slightly shorter than the Baechtel formula. . . . .

Using PipeMax (tm), depending on the peak bhp rpm (8000/8250) & the estimated V/E (95%/100%), the recommended primary length is between 22.3" to 25.8" which agrees more closely with the Bell numbers.

So starting at the 24" primary length typical of most of the header possibilities appears to be a good choice.

HEADER SIZING,  PART DREI. . . .

Also:   Using Bell's formula for pipe I.D. generates pipe sizes very close to the sizes from Baechtel. . . . . .

I.D. = {Square root of  [vol in cc's / (Primary length, in + 3)]} * 2.1

rpm          pipe dia I.D.       O.D.
7000          1.195             1 5/16
7500          1.237             1 3/8

These numbers are a touch smaller than the numbers generated by the Baechtel formula, but slightly larger than what PipeMax (Tm) predicts @ 1.046" to 1.208" O.D.'s   . . . . . .

My practical dyno experience with BMC XSP's used 1 1/4" & 1 3/8" primary pipes of various lengths, but the development program was in what I would term as "infancy".   As our buddy Dean (and probably Mistress Helga. . . ) would say:  "Time to put away ze calculators, shust de engineers, und shpank it!!!"
 
Fordboy

[Milwaukee Midget]

As all the manufacturers freely tell us, "MSD" (multiple spark discharge) only functions up to moderate engine speeds- usually around 3,500 RPM.

Coil saturation is the issue I hope this little experiment addresses, my fear being potential heat buildup in the coil packs - pulsating DC will still charge the coil, but might physically damage it, too.  Input frequency of the synthesized pulsating DC remains relatively constant (8000 per second, the frequency at which the coil pack is seeing charge signal), BUT the event at which the spark is released is variable – 1500-9000 RPM.  

So yes, if it works, as RPM increases (variable), the number of sparks per ignition event will decrease as a function of the shorter time the actual event occurs with respect to the pulsating DC input frequency (constant) – same as with MSD.

What I think I’m likely to see is a major initial spark, draining the collapsed coil of initial stored energy, followed by numerous smaller voltage sparks occurring at a rate of 8000 pulses per second at each sparking event until the circuit closes and the coil can actually take a full charge.

Right now, I don’t know if the coil packs can keep up.

Physical noise isolation is another issue I have yet to address for the transducers.
 
Working to my advantage is that the coils only produce a sparking event once per revolution, whereas if this were a standard ignition coil in a V-8, it would be required to create a spark event four times per revolution.

It’s all a grand experiment.

Hmmm, curiouser and curiouser. Is it cabin fever or inspired genius.......?  

It hasn’t been above freezing here since the February rains.  I'd like to think they go hand in hand . . .

[manta22]

MM;

Coil saturation is related to the points & coil (Kettering) ignition. Think of the capacitive discharge (CD) ignition coil as a pulse transformer. A properly designed system does not saturate the coil as it steps up the voltage from the capacitor discharge current ino its primary winding. What usually limits the CD performance is a limit of h w many joules the internal high voltage power supply can produce.

It's entirely possible to build a system that can melt the electrodes off the spark plugs but that would be overkill. 

[Duck-Stew]

As all the manufacturers freely tell us, "MSD" (multiple spark discharge) only functions up to moderate engine speeds- usually around 3,500 RPM.

MSD ignitions are limited to multiple sparks on a standard distributor to 3200 rpm on a V8.  4800 on a 6 & 6400 on a 4 cylinder. (IIRC- it's been a # of years since I sold speed equipment.)

[Freud]

It's a  d a m n e d wonder that the Bean Bandits were able to even get an engine started

let alone burn the ends off of the plugs unless they were too lean.

Just watch. Next year there will be a line in the rule book that keeps teams from talking that

way about pulsing DC.

That sounds criminal to me.............put your kids in the car trunk when these dudes start talking like that.

FREUD

[Milwaukee Midget]

It's a  d a m n e d wonder that the Bean Bandits were able to even get an engine started

let alone burn the ends off of the plugs unless they were too lean.

Just watch. Next year there will be a line in the rule book that keeps teams from talking that

way about pulsing DC.

That sounds criminal to me.............put your kids in the car trunk when these dudes start talking like that.

FREUD

Gee, Glenn, I thought you went to Hawaii to chill out.  

I’ll be the first to admit it’s an unlikely improvement over the system in place, but I certainly had no intention of putting myself in a position where I’d have to live more than 2000 feet from a school or walk around with a monitoring bracelet.

I’ll watch my language, Dad.