13B 2 Rotor Engine

This animation of a rotary engine is courtesy of

Real World Solutions

Rotary Engine Animation

(Information from Kevin McCauley)

Engine

 

Year
Name
Horsepower
Torque
1993-1995
All Models
255@6500
217@5000
1996-1997
Japan Only
265@6500
222@5000

 

Performance

 

Year
Model
0-60
1/4 Mile
Top Speed
1993-1995
5 Speed
5.2
13.9@99.9
161
1993-1995
4 Speed Auto
6.0
14.5@96.3
158
1996-1997
5 Speed Japan
4.7
13.9@102.5
164

 

 

Engine Type

Rotary

Displacement cm{cu in}

654 x 2 {40.0 x 2}

No. of cylinders and arrangement

2 Rotors, longitudinal

Combustion chamber type

Bathtub

Compression Ratio

9.0 : 1

Air induction

4-Port induction

Turbo charger

System type

Sequential twin turbo charged

Cooling method

Water + Engine oil

Boost control actuator

Turbo pre-control +Wastegate control

Boost control method

Solenoid valve {duty-controlled} x 2

Idle Speed

700 – 750

Throttle Body

Horizontal draft {2 stage – 3 barrel}

Intercooler

Air cooled

Fuel Tank liters{US gal, Imp gal}

76 {20.1, 16.7}

Fuel pump kPa{kgf/cm,psi}

490 – 740 {5.0 – 7.5, 71.1 – 106.7}

Injector

Side-feeding

Injector volume primary {cm,psi}

550{550}

Injector volume secondary {cm,psi}

850{850}

 

20B 3 Rotor Engine

Mazda have made several prototype and experimental engines over the years ranging from the 6A (a single rotor version of the 12A, intended for use in Japanese “Kei” cars), the 2002 “4 rotor 10A” to the massive R-II 21A with 2x1046cc rotors (You can see the 2002 and 21A elsewhere at this site).Several 3 rotor prototype and racing engines have been produced over the years. Generally speaking these engines were constructed using the modular nature of the rotary engine (that is to say, the parts can be “stacked” together). This allowed comparatively cheap production items (rotors, rotor housings, side housings, seals etc) to be used in conjunction with some specially made (expensive) parts (eccentric shaft and centre housing with rotor phasing gear/bearing) There has really only been one mass produced passenger car 3 rotor engine – the 20B, which was a twin turbo fuel injected engine. Note, NSU also had a 3 rotor engine though I’m not sure if this was mass produced.Generally speaking, the following 3 rotor engines were “available” from the early 1980s: 13G Racing engine (1984-1989), 13G “Production Prototype” (1984/1985), 20B Production engine (1990-1996), 20B Race engine (1990-now)

20B – Production engine

The 20B engine was only sold in the Japanese market. It was also only available with twin turbochargers. It was fitted to the “Eunos Cosmo” (Model type JCES) sold from March 1990 until January 1996. (There was a minor model change, with mainly cosmetic changes about 1993. The first of the cosmos have an engine with plain alloy intake manifold etc. The “updated” model has these parts anodized dark grey. All other mechanical specifications remained the same) The Cosmo was a high specification, high performance luxury coupe nearly comparable to the Lexus. All Cosmos were 4 speed automatics. (See my “EUNOS COSMO” Page for more detail about the car).
It seems that the 20B engine is based on the 13B-REW engine of the “3rd generation” (FD3S) RX7. (Despite the FD3S being released in August 1991, well over 18 months after the Cosmo with it’s 20B in March 1990 – However the Cosmo ALSO was available with a 13B turbo, with very similar specifications to the RX7′s 13B-REW, except power was down a little (about 20-25 hp less). Perhaps the Cosmo’s engine was used to test how reliable the new design was.. Especially with the upcoming volume selling RX7.)

This engine appears to use many standard 13B components, such as rotors /rotor housings, seals etc. As with the “prototype production 13G” the only obvious custom components are: *Special second centre housing with gear and bearing *Eccentric shaft and counterweight assembly *Intake and exhaust manifolds/turbos *Ignition system (Distributorless, but uses FC3S/FD3S RX7 type pickup unit)

NOTE: 20B-REW is the actual correct designation for this engine. This means “20B-Rotary Engine DOUBLE TURBO” according to what I have read (it is easier for a Japanese speaker to say “W” than “DOUBLE”)

20B-REW “Production” Engine specifications

Origin Closely based on 3rd generation (FD3S) RX7′s 13B-REW Special 3 rotor parts based on 13G design
Capacity 654cc x 3 rotors = 1962cc
Compression ratio 9.0:1
Turbo Boost 0.7 Bar (=10.29 PSI)
Induction Turbocharged (twin sequential)/intercooled (intercooler mounted near car’s radiator), electronic fuel injection (2 injectors/rotor)
Exhaust Peripheral Exhaust Port
Ignition Distributorless electronic ignition (2 plugs/rotor)
Power/RPM 280ps@6500rpm
Torque/RPM 41kgm@3000rpm
Max RPM Cosmo Tacho redline at 7000rpm (scale ends at 8,000)
Dimensions Length 672mm Width 549mm Height 520mm (with accessories)
Weight 350kg With all accessories & manifolds
Special Features Ignition knock control uses one sensor per rotor (on rotor housing above trailing spark plug)

 

20B – Racing engine

(Picture from 1996/1997 Mazdaspeed catalog Page 104)

This engine is currently (1997) sold by both Mazdaspeed in Japan and Mazda Motorsports in the USA, both of which are officially part of Mazda. At first I presumed this was a racing engine based on the production 20B. However on comparing the photos of the race 20B and the race 13G, plus the Mazda catalogs list this engine as using rotor housings with grooves for the water seals, I think it is still a 13G they are selling. Also the power output and torque figures are identical (same RPM), and the weight/dimensions are almost the same. (The grooves for the water seals were moved from the rotor housings to the side housings when the “2nd generation” RX7 (FC3S) was released in 1986)

20B Racing Engine specifications

Origin Possibly same as 13G race engine, but not certain.
Capacity 654cc x 3 rotors = 1962cc
Compression ratio Unknown (But 13G is 9.4:1)
Induction Naturally aspirated peripheral port, electronic fuel injection (2 injectors per rotor)
Exhaust Peripheral Exhaust Port
Ignition Distributorless CDI ignition (2 plugs/rotor)
Power/RPM 450ps@8500rpm
Torque/RPM 40kgm@8000rpm
Max RPM Unknown (But 13G is 9500 RPM)
Dimensions Length 675mm Width 549mm Height 520mm
Weight 143kg
Special Features Dry sump

The Complete Mazdaspeed 20B parts list

These pictures have been scanned from the 1996/1997 Mazdaspeed catalogue, Pages 105 to 121 inclusive. I could see no copyright notices in the catalogue, so these are here as “free advertising” for Mazdaspeed.
Page 106 shows the rotor housings, and you can see that the groove for the water O-ring is in the rotor housings (as were all pre-1986 engines), HENCE I think this is ACTUALLY still a 13G engine. Also, check out the trick throttle assembly in page 116!
By the way, the prices are in Yen, and 1 US Dollar = 123.43 Japanese Yen (March 1997)

[PAGE 105] Rotor diagram and parts list [PAGE 106] Rotor housing diagram [PAGE 107] Rotor housing parts list [PAGE 108] Side housing diagram [PAGE 109] Side housing parts list [PAGE 110] Eccentric shaft diagram [PAGE 111] Eccentric shaft parts list [PAGE 112] Manifold diagram and parts list [PAGE 113] EFI computer diagram and parts list [PAGE 114] Fuel system diagram and parts list [PAGE 115] Water Pump diagram and parts list [PAGE 116] Throttle plate diagram [PAGE 117] Throttle plate parts list [PAGE 118] Oil system diagram [PAGE 119] Oil system parts list [PAGE 120] Ignition/Alt/Starter diagram [PAGE 121] Ignition/Alt/Starter parts list

Note: All pictures are 768x1024x4 GIF files. This was found to be the best compromise between file size and picture quality, which is about the same as a fax (to get better quality meant much bigger files).

Performance applications of the 20B engine

A second hand production 20B engine is the only multi rotor engine worth considering. The 13G engines are extremely rare (and expensive). The new price of a “20B race engine” as listed above is $US 20,000+ for the core engine alone. Many more ‘production’ 20B engines were made as they were fitted to a production car rather than being a special racing item.  However 20Bs are far rarer than their 13B cousins; so they are difficult to find and will be expensive. (I would estimate that less than 100 of these have been imported to Australia).

In Australia, the price for these units started at $Aus 12000 in 1991, dropping to a low of about $Aus 4000 in 1997. In March 2001 I was told by someone who had just bought an engine that the going rate is $Aus 4000 to 9000: *Adelaide Jap Dismantlers $Aus 6550 for front cut (Whole front half of car) *Ichiban for $Aus 2975 (some damage to engine), $Aus 3995 or $Aus 4995 Refer to my “parts” page for contact information. If you have any updates of prices or supply sources please let me know).

So, assuming a 20B can use standard RX7 13B components (rotor housings, seals etc) a rebuild on a 20B could probably be done for under $Aus 3000.

20B engines used in Australia have been fitted to many cars in the Mazda range – RX2, RX3, RX4, RX5, RX7s of all vintages, 929 sedans and Coupes, I have even seen a mid 80′s BMW 5 series with a 20B installed. (See elsewhere on this site for both a 2nd Generation RX7 with a 20B conversion and the Rod Millen 20B rally car).

Because the 20B probably has one of the most complex engine management requirements in the world and the factory computers have complex wiring, these engines generally have the twin turbos replaced by a single large turbo (e.g. TO4) and are run on a aftermarket fuel computer. Expensive engine management systems from Motec, Haltech and Autronic and others may be capable of distributorless ignition and/or twin turbo operation however in the past the ignition was typically done via an early RX7 distributor, with a Racing Beat conversion kit comprising of a big clear Perspex distributor cap and a new ignition triggering vane.

A standard Mazda transmission will bolt onto the 20B without modification, another popular choice are 5 speed Toyota Supra gearboxes. Some drag cars are using ancient 2 speed Powerglide transmissions (why they would use one of these rather than a modern 4 speed is beyond me…) The differential used is almost always a Ford 9 inch.

20Bs have been fitted to later model RX7s (FC3S/FD3S) without too much problem. (I suspect the RX7 engine bays MAY have been originally designed to hold a 3 rotor engine, particularly in the FD3S). These conversions seem to retain more of the original hardware.
Potential power outputs: *The standard power output of a 20B turbo is 280HP. *With turbos removed (naturally aspirated), these engines produce 250hp stock, 320hp if ported. *Apparently with the mere addition of a boost controller the power jumps to around 400HP. *Rod Millen’s Pike’s Peak race car made about 500HP with a nearly stock engine. *Japanese modifiers get about 700HP without too much trouble (I suspect this would be at a level where some engine porting and severe turbos would be required). *Racing beat’s Bonneville racer makes 900+ hp with a triple turbo peripheral port 13G *The above figures were all mid-late 1990s. In 2002 some drag racers are claiming 1200hp.

 

Various Rotary Engines

The Different Rotary Engines

Rotary Engine Statistics

Type

Year

Art

Displacement

Power

40A

1961

Experimental

1 x 386 ccm

L8A

1962

Experimental

1 x 399 ccm

L8A/0353

1963

Experimental

2 x 399 ccm

L8A/3804

1963

Experimental

3 x 399 ccm

L8A/3805

1963

Experimental

4 x 399 ccm

160 PS/6000

10A/3820

1964

Prototype

2 x 491 ccm

12A/3830

1966

Experimental

2 x 573 ccm

12A/3830

1966

Prototype

2 x 573 ccm

400/3867

1967

Experimental

2 x 395 ccm

400/3893

1967

Experimental

2 x 495 ccm

10A/0810

1967

Cosmo L10A

2 x 491 ccm

110PS/7000 135Nm/3500

10A/0813

1968

Cosmo L10A

2 x 491 ccm

128PS/7000 142 Nm/5000

10A/0820

1968

Cosmo

2 x 491 ccm

100PS/7000

10A/3883

1968

Singapore GPrennen

2 x 491 ccm

204 PS – 170kW/230PS

12A/3872

1968

Prototype

2 x 573 ccm

10A/8020

1968-72

R100 / Presto

2 x 491 ccm

100PS/7000 137Nm/3500

10A/3877

1969-72

R100(US)

2 x 491 ccm

100PS/7000 129Nm/4000

10B

1968-69

Cosmo

2 x 491 ccm

128PS

3912

1970

Experimental

1 x 356 ccm

35 PS

X002

1970

Experimental

1 x 360 ccm

3915

1970

Experimental

1 x 360 ccm

13A/0823  9,1:1

1970-72

R130

2 x 655 ccm

126PS/6000 175Nm/3500

6A

70er

Experimental

1 x 573 ccm

halber 12A

7A

70er

Experimental

1 x 654 ccm

halber 13B

2002

71

Experimental

4 x 491 ccm

180 PS/6000

12A  9,4:1

1970-71

R100

2 x 573 ccm

100PS/7000 129Nm/4000

12A  9,4:1

1970-71

RX2

2 x 573 ccm

120PS/7000

10A/0866

1971-75

RX3 (JP)

2 x 491 ccm

105PS/7000

12A  9,4:1

1972

R100

2 x 573 ccm

100PS/7000 112Nm/4000

12A/3905  9,4:1

1972

RX2 (US)

2 x 573 ccm

102PS/6800 137Nm/4000

21A

1972

Experimental

2 x 1064 ccm

185PS 280Nm

12A/R612  9,4:1

1972-75

RX3 (US)

2 x 573 ccm

102PS/6800 134Nm/4000

12A  9,4:1

1973

Luce GR (JP)

2 x 573 ccm

120PS/6500 162Nm/3500

12A  9,4:1

1973

Luce GR AP (JP)

2 x 573 ccm

115PS/6500 159Nm/3500

12A  9,4:1

1973

Luce GR II( JP)

2 x 573 ccm

130PS/7000 168Nm/4000

12A  9,4:1

1973

Luce GR IIAP (JP)

2 x 573 ccm

125PS/7000 164Nm/4000

12A  9,4:1

1973

RX-2

2 x 573 ccm

97PS/6500 134Nm/4000

12A

1973

Le Mans

2 x 573 ccm

250PS/8000

12B

1973-75

K�uflicher Rennmotor

2 x 573 ccm

250PS+/9500

13B   9,1:1

1973

R130 (Jp)

2 x 654 ccm

125PS/6000 175Nm/3500

15A

1973

Experimental

2 x 737ccm

135PS/5750 203Nm/3500

12A SIP  9,4:1

1974

RX2

2 x 573 ccm

97PS/6500 134Nm/4000

13B   9,2:1

1974-75

RX4

2 x 654 ccm

110PS/6000 164Nm/3500

12A  9,4:1

1976

RX3 Nikki 2B1

2 x 573 ccm

95PS/6000 143Nm/4000

13B   9,2:1

1976-78

RX5 Cosmo

2 x 654 ccm

110PS/6000 168Nm/4000

13B   9,2:1

1976-78

RX4

2 x 654 ccm

110PS/6000 168Nm/4000

12A  9,4:1

1977-78

RX3SP

2 x 573 ccm

95PS/6000 143Nm/4000

12A Sport  9,4:1

1977

Racing Kit

2 x 573 ccm

250PS+/9000

13B Racing  9,4:1

1977

Rennen

2 x 654 ccm

290PS+/9000

12A  9,4:1

1979-80

RX7

2 x 573 ccm

100PS/6000 147Nm/4000

13B Racing

1979

Rennen Le Mans

2 x 654 ccm

285PS/9000

13B Racing

1980

Le Mans Qualification

2 x 654 ccm

300PS/9000

13B Racing

1980

Le Mans Rennen

2 x 654 ccm

290PS/8500

12A Magermotor

1981-85

RX7 US

2 x 573 ccm

100PS/6000 147Nm/4000

12A 6PI

1982-85

Luce,Cosmo (Jp),RX7

2 x 573 ccm

12A/Turbo

1983-8

Luce,Cosmo,RX7

2 x 573 ccm

160PS/6000 231Nm/4000

12A/Turbo  8,5:1

1984-85

Luce,Cosmo,RX7

2 x 573 ccm

165PS/6000 231Nm/4000

12A  9,4:1

1984-85

RX-7 (S,GS,GSL)

2 x 573 ccm

101PS/6000 149Nm/4000

13B DEI

1983+

Luce,Cosmo,RX7

2 x 654 ccm

13G Racing

1983-84

Rennen 757

3 x 654 ccm

450PS/8500

13B Racing

1984

Le mans

2 x 654 ccm

Sprint 330PS, Ausdauer 310PS

13B/Turbo  7,5:1

1984

Racing

2 x 654 ccm

500PS/7800 456Nm/7500

20B

1984-

Production Rennmotor

3 x 654 ccm

500PS/9000

13B DEI  9,4:1

1984-85

RX7 (GSL-SE)

2 x 654 ccm

135PS/6000 186Nm/2750

13B DEI  9,4:1

1986-89

RX7 2.Generation

2 x 654 ccm

146PS/6500 193Nm/3500

13B/Turbo  8,5:1

1986-89

RX7 Turbo II

2 x 654 ccm

182PS/6500 256Nm/3500

13J-M Racing

1988

Rennen 767

4 x 654 ccm

500PS/8500

13B

1989

Luce (Jp, 929)

2 x 654 ccm

177PS/6500 253Nm/3500

13J-MM Racing

1989

Rennen 767

4 x 654 ccm

630PS/9000 2Schrittsaugrohr

20B

1989

Eunos Cosmo

3 x 654 ccm

280PS/6500 402Nm/3000

13B VDEI  9,7:1

1989-92

Leichtere L�ufer

2 x 654 ccm

160PS/7000 196Nm/4000

13B/Turbo  9,1:1

1989-91

Leichtere L�ufer

2 x 654 ccm

200PS/6500 274Nm/3500

13B/Twin Turbo 9,0:1

1992-

RX7 3.Generation

2 x 654 ccm

255PS/6500 304Nm/5000

13B MSPRE

1995

Exp. RX01

2 x 654 ccm

220PS/8500 223Nm/6500

26B

1991

Rennen 787B Le Mans

4 x 654 ccm

700PS/9000  607Nm/6500 var.Saugr.

HR-X

1991

Exp. Wasserstoff

2 x 499 ccm

100PS 127Nm

HRX-2

1993

Exp. Wasserstoff

13B Renesis

1999

RX-Evolv

2 x 654 ccm

280PS/9000

13B Rensis

2000

Prototype RX-8

2 x 654 ccm

250PS/8500

13B Rensis

2003

Series RX-8

2 x 654 ccm

192PS-250PS (D 192PS-231PS)

Wideband Dataloggit

Wideband O2/Datalogit Tuning

Stoichiometric:

Stoichiometric or Theoretical Combustion is the ideal combustion process during which a fuel is burned completely. A complete combustion is a process which burns all the carbon (C) to (CO2), all hydrogen (H) to (H2O) and all sulfur (S) to (SO2). If there are unburned components in the exhaust gas such as C, H2, CO the combustion process is uncompleted (taftan)  Approximately 14.7 parts Air to 1 part Fuel (14.7:1) for the perfect complete combustion.

Pressure Conversions:

Remember: when the term 1 bar is used to describe pressure what most forget to state is that it is assumed they mean “above atmospheric pressure.”  Naturally the atmosphere has pressure (1 BAR approx 14.7 PSI) but this is assumed constant.  Stating a turbo “kicks out” 14 PSI that is actually 14 PSI more than the atmospheric hence 14.7 PSI is technically 2 BAR

101.325 kPa = 1 ATM

100 kPa = 1 bar

200 kPa = 2 bar = 14.7 PSI   (referred to 1 bar assuming it is understood to be 1 bar above atmospheric pressure)

Goals of a tuner:

Fuel related:

 

For any given moment have the A/F ratio, Load, and RPM to make calculations and then output the corrected amount of fuel for the desired A/F ratio in those same conditions .

Typical A/F ratios:

  • high 14′s up to “0 PSI” AKA atmospheric pressure (1 BAR)
  • mid 12′s up to 6 PSI (1.4 – 1.5 BAR)
  • mid 11′s up to 14.5 PSI (2 BAR) and above.

Fuel related links:

Detailed information on both Gasoline/Ignition

Ignition related:

In terms of ignition, proper timing is needed to ignite the compressed air/fuel mixture at the most opportune time avoiding preigniting (detonation) or retarded ignition

What is DETONATION?

Typical timing advances:

  • 30 to 35 deg advance to 90kpa (0.9 BAR)
  • 20 odd up to 120kpa (1.2 BAR, 3 PSI)
  • as low as 10 to 17 total advance up to 200kpa (2 BAR, 14.7 PSI)

If you stick to a total engine advance of no more than 15 deg, assuming your static timing (pick up) is set correctly, then you are much less likely to suffer engine failures due to timing

EGTs:

Exhaust gas temperatures is one way to measure timing.  If the timing is too retarded (combustion happens too late) then the fuel is still burning in the manifold causing higher EGTs

Effects of porting on timing:

*1 BAR boost assuming constant ambient temp of 15 C and charge temp of 30 C*

  • STOCK PORT                  20-23 deg advance …lowest effective “dynamic” compression at max power.
  • STREET PORT                 15-17 deg advance…midrange effective “dynamic” compression at max power.
  • PERIPHERAL PORT        10-12 deg advance…highest effective “dynamic” compression at max power.

 

Before you tune you must ask yourself:
1) What ECU are you using? 2) What amount of performance/reliability do you want from your car?. 3) How modified is your car? 4) How much boost do you plan on running? 5) What octane gas do you use? (better be more than 92 octane) 6) How easy is it to access a dyno, wideband O2 sensor and data logging kits for all vital engine information? 7) Where is your EGT sender located?

Wide Band O2 Sensors

 

Why do I need a wide band 5-wire sensor on my air fuel ratio meter?

1)  The output of a sensor with less than 5-wires is binary.  That means it’ll only tell you if you are rich or lean.  It cannot tell you how rich or lean you are.   Once the mixture goes a little bit fat, the meter will pin to the rich side.  Likewise if the meter goes lean, the meter will pin to the lean side.  Very often, a meter maker using these cheap sensors will write on their faceplates that the meter goes from 10.0 to 20.0 AFR.  This is just paint on the panel.  It is simple dishonesty.  It is more likely 13.5 to 15.0 in reality.  They might as well use three LED’s to tell you the AFR – LOW, MED and HIGH!

2)  There are only two wide band sensors widely available today.  One is made by Bosch and the other by NTK/NGK.  They have five wires.  These sensors require sophisticated closed loop controllers to determine the air fuel ratio rather than a simple switch type OEM sensor that requires only a voltmeter.  All sensors can only determine the air fuel ratio at the switch point where the mixture goes lean to rich.  This is any sensor’s zone of accuracy.  The difference between the OEM sensors and the 5-wire widebands is that this zone can be moved around to different air fuel ratios in the widebands.  In other words, 5-wire sensors have an electronically controllable air fuel ratio in their reaction cavities.

What this means is that you can make the sensor operate in it’s zone of sensitivity at air fuel ratios ranging from 8 to 24 or higher air fuel ratio.  The problem is that this requires some clever circuitry to keep the sensor in this delicate balance.  (lamdaboy)

 

Many RX-7 tuners actually wire their wideband directly into the O2 sensor location for the aftermarket ECU.  As long as the Wideband O2 sensor has a 1-4V output it can replace the stock O2 Sensor.  Many do this so that, like in the PowerFC, one unit logs O2 and all other engine parameters.

The standard scale is:  4V is amount of oxygen in open air                                  2.5V is combusted fuel at a 14.7 A/F ratio
This is the Tech Edge Wideband O2 sensor kitwhich is used in proper tuning of a RX-7

 

This is the FJO Wideband O2 sensor kit which is used as well.

 

Lambda-BoyStandard Air Fuel Ratio Meter

 

Innovate Motorsports LM-1: Digital Air/Fuel Ratio Meter

 

Data Logging ability with excellent software

 

All of these kits have a data logging feature which is crucial for tuning because it retains past information so tuning does not have to be “on the spot”  Data is logged into either a laptop or a PDA and reviewed as the fuel map is corrected to get as close as possible to the desired Air to Fuel Ratio.  Stoichiometric combustion is always dreamed about, but is unattainable due to the detonation in turbocharged rotary engines.  Detonation is more likely with the air is hot and pressurized.  It is a debate as to what the “correct” AF ratio is for a FD.  I have personally seen as low as 11.5 to as high as 13 used in setups.  (this information is used as education resources only)

This website chronicles an installation of a wide band o2 sensor on a 3rd Gen RX-7. 
 

Knock Sensors

Use primarily in tuning the ignition maps for the RX-7, the J&S Knock Sensor is not only a monitor, but a safety net as well.  The unit is capable of detecting preignition and displaying that to the driver.  The unit adjusts timing as needed to avoid early timing.  If the timing is too soon preignition/detonation will occur.  If the timing is too late, the timing is considered retarded and ignites what is already expanding because the rotor is in motion causing a change in pressure and volume.  Perfect timing is the goal.

This document has several articles related to knock sensors and detonation.
 
 

Datalogit for PowerFC

Data logging unit that “splices” the signal for the commander and sends it to a laptop via serial port.  The laptop must have correct software installed to read signals and log data from ECU.  Most useful features include logging Load vs. RPM graphs and correction maps.