Fullscreen HTML5Med Res (??MB)HTML4

All the parts for the Smart Mixture Display are mounted on a small PC board. This prototype uses rectangular LEDs for the 10-LED mixture display but you can use round LEDs if you prefer – see text. A Smart Mixture Display For Your Car Track your car’s fuel mixtures in real time, see the operating modes of the ECU and be warned if a catastrophic high-load “lean out” occurs. This Smart Mixture Display monitors your car’s oxygen sensor and airflow meter outputs and gives an audible warning if mixtures go dangerously lean. T By JULIAN EDGAR & JOHN CLARKE HE SILICON CHIP Mixture Meter – first presented in 1995 – is one of the most popular performance car electronic kits ever produced. Literally thousands have been built, each showing by means of 10 coloured LEDs whether the air/ fuel ratio is rich or lean. While such a design – which works from the car’s standard oxygen sensor – won’t give you an absolutely accurate readout of the mixture strength, it’s far better than having no indication at all as to whether the car is running rich, lean or at stoichiometric (the latter means an air/fuel ratio of 14.7:1). As a bonus, it also clearly shows if the car is in closed or open loop mode The new Smart Mixture Display pre60  Silicon Chip sented here still displays the mixture strength by means of 10 LEDs – red for lean (red is for danger!), green for midrange mixtures and yellow for rich. However, we’ve added three important extra features with this new design: (1). Better protection of the electronics (in some cars, the old design was prone to blowing its chip); (2). An automatic dimming function for night driving; and (3). An audible lean-out alarm. Lean-out alarm The lean-out alarm is a great idea. It monitors both the air/fuel ratio and the engine load, sounding a buzzer if the air/fuel ratio is ever lean at the same time as the engine is developing lots of power. So why is this important? Well, if the engine – especially one with a turbo – goes lean under high loads, it’s almost certain that you’ll instantly do damage. One Impreza WRX that we know of lost part of an exhaust valve this way. What could cause this sudden and catastrophic condition? Lots of things – from a dying fuel pump to fuel starvation during cornering. Even a couple of blocked injectors could cause a lean condition. It’s not the complete answer – there are some conditions that the meter won’t register. However, in most situations, it will act as an important warning that things aren’t right. The lean alarm works by also monitoring the voltage signal coming from www.siliconchip.com.au Fig.1: the circuit is based on an LM3914 dot/bar display driver IC. This accepts the signal from the oxygen sensor and directly drives a 10-LED display. Op amps IC2a & IC2b and their associated components (including Q2 and the piezo buzzer) provide the “lean-out” alarm feature. the load sensor – usually the airflow meter. Most airflow meters have an analog output voltage that rises with engine load, being around 1V under light loads (eg, at idle) and close to 5V under high loads. If the output voltage from the airflow meter is high, the meter knows that the engine load must also be high. LED indicators But what about the main section of the Smart Mixture Meter – the LEDs? www.siliconchip.com.au How do they work? In broad terms, the oxygen sensors in most cars have an output voltage that varies between 0-1V, with higher voltages indicating richer mixtures. The meter lights one LED for each tenth of a volt coming from the sensor, so at 0.1V the far end red LED will be on, at 0.2V the next red LED will light up and so on. This doesn’t give a precise indication of air/fuel ratio (see the “Air/Fuel Ratio Measurement and Oxygen Sensors” panel for the reasons) but in practice, it’s still very useful. So the oxygen sensor voltage is constantly displayed by means of the LEDs and if the oxygen sensor output voltage is low (ie, there is a lean mixture) at the same time as the airflow meter output is high (ie, a high engine load), the onboard piezo buzzer sounds. However, most of the time (we hope all of the time!), you won’t have to worry about alarms sounding – instead you’ll be able to glance at the dancing LED as you drive along. Dancing? April 2004  61 One of the most common causes of turbo engine damage (along with detonation) is a high load lean-out. That’s what happened to this Impreza WRX motor – and in just a moment part of an exhaust valve was gone. [Michael Knowling] The exhaust gas oxygen sensor delivers a mixture strength signal than can be monitored by the 10-LED Smart Mixture Meter. All cars made in at least the last 15 years use an oxygen sensor. [Bosch] Won’t the illuminated LED stay constant if the air/fuel ratio isn’t changing? One of the beauties of the meter is that it will show when the ECU is in closed loop operation, with the mixtures hovering around 14.7:1. This air/ fuel ratio – called stoichiometric – allows the catalytic converter to work best, so at idle and in constant-speed cruise, the air/fuel ratio will be held around this figure. To achieve this, the ECU monitors the oxygen sensor output. If the mixtures are a bit richer than 14.7:1, it leans them out a little. Conversely, if the mixtures are a bit leaner than 14.7:1, it makes them slightly richer. This constant cycling of mixtures around the 14.7:1 point is called “closed loop” and will cause the lit LED to dance back and forth across the meter – as much as two or three LEDs either side of centre. When some people see the LEDs flashing back and forth in closed 62  Silicon Chip loop operation, they quickly decide that the meter is useless. After all, the indication is “all over the place”! However, it’s showing the very fast oscillations that are actually occurring in the mixture. By contrast, most aftermarket tail-pipe air/fuel ratio meters aren’t sensitive enough to “see” this behaviour. Closed loop operation does not occur in the following driving conditions: (1) during throttle lift-off; (2) when the engine is in warm-up mode; and (3) at wide throttle openings. At these times, the ECU ignores the output of the oxygen sensor, instead picking the injector pulse widths solely on the basis of the data maps programmed into it. When the throttle is opened wide, the air/fuel ratio becomes richer, holding at that level. For example, the green LED second from the end may light and stay on. If you accelerate even harder, then the very end green LED may light. On the other hand, back right off and it’s likely that all the LEDs will go out. That’s because the injectors have been switched off on the over-run and the air/fuel ratio is so lean that it’s off the scale. Watching the behaviour of a LED mixture meter really is a fascinating window into how an ECU is operating! The mixture meter is also a vital tool when undertaking engine modifications. For example, if a particular LED lights at full throttle before and after making engine modifications (eg, to increase power), then you can be fairly confident that the mixtures haven’t radically changed (under the same conditions, that is). Conversely, if the lit LED shifts two along after the modifications have been done, you can be fairly sure that the mixtures are different! A word of warning though – the Smart Mixture Display shouldn’t be relied on when making major engine modifications and/or working on expensive cars, In summary, fitting the Smart Mixture Display to your car has three major benefits – you can roughly track your mixtures in real time, you can see the operating modes of the ECU and you can be warned if there is an unexpected catastrophic high-load lean out. Sounds good to us! How it works OK, let’s take a look at the circuit details – see Fig.1. IC1 is an LM3914 dot/bar display driver. In dot mode, it drives the LEDs so that as the voltage at its pin 5 input increases, it progressively turns on higher LEDs. For example, at the lowest input voltage, LED1 is lit. At midrange voltages, LED4 or LED5 might be lit and at the highest input voltage, LED10 will be lit. Trimpots VR1 and VR2 set the voltage range for the LED display. Normally, VR2 is set so that its wiper is at ground and VR1 is set so that its wiper is at 1V. Thus, the LED display covers a 0-1V range which is the normal output variation of an automotive oxygen sensor. The LED brightness is set by the total resistance from pin 7 to ground and we vary this to dim the LEDs in darkness. In bright light, the Light Dependent Resistor (LDR1) is a low resistance and this provides current to the base of transistor Q1 which switches it on to set the LED brightness at maximum. Conversely, in darkness, LDR1 is a high resistance and so transistor Q1 is off. This sets the LED brightness to minimum. Trimpot VR3 adjusts the dimming threshold. If it’s set fully clockwise (ie, to minimum resistance), the LEDs will be dimmed at a relatively high ambient light level. As VR3’s wiper is rotated anticlockwise, the dimming begins at progressively lower ambient light levels until eventually, the LEDs are at maximum brightness in normal daylight. Op amps IC2a and IC2b are used as comparators to monitor the load and oxygen sensor signals respectively. As shown in Fig.1, IC2b monitors the oxygen sensor signal at its noninverting input (pin 5), while VR4 and its associated 10kΩ series resistor set the voltage at the inverting input (pin 6). If the oxygen sensor signal level is below the voltage on the inverting input, then IC2b’s output (pin 7) goes low and lights LED11. Comparator IC2a operates in reverse fashion. It monitors the load signal at its inverting input (pin 2), while VR5’s wiper sets the threshold for the noninverting input (pin 3). If the load voltage is above the level set by VR5, pin 1 of IC2a goes low and LED12 lights. When the outputs of IC2a and IC2b are both low, transistor Q2 is switched on due to the base current through 5.6V zener diode ZD4 and the 2.2kΩ resistor to ground. Q2 then drives the piezo buzzer. www.siliconchip.com.au Fig.2: this diagram shows where each of the components is placed on the main PC board. Use this diagram, the photos of the completed board and the parts list to help you assemble it correctly. Now consider what happens if one of IC2’s outputs goes high – ie, if the oxygen sensor signal goes above VR4’s wiper or if the load input signal goes below the VR5’s wiper. In that case, ZD4’s anode is pulled high via either diode D2 or D3 (depending on which op amp output is high). This causes transistor Q2 to turn off and so the alarm stops sounding. This means that the outputs of IC2a & IC2b must both be low for Q2 to switch on and sound the alarm. Note the 1MΩ input resistors in series with the oxygen sensor and load inputs. These prevent loading of the circuits they are connected to and ensure that the car’s ECU operation is not affected in any way by the addition of the Smart Mixture Display. The associated 10nF capacitors to ground are included to filter voltage transients on the inputs. Power for the circuit is derived from the vehicle’s +12V ignition supply. Diode D1 prevents damage if the battery supply connections are reversed, while the 10Ω resistor and 470µF capacitor provide decoupling and filtering. As a further precaution, 16V zener diode ZD1 is included to prevent voltage spikes from damaging the ICs. Construction The Smart Mixture Meter is straightforward to build, with all the parts installed on a PC board coded 05104041. Fig.2 shows the assembly details. Begin the assembly by installing the wire links and resistors first. Table 1 shows the resistor colour codes but it’s advisable to check each one with a digital multimeter as well, as some of the colours can be difficult to decipher. www.siliconchip.com.au The assembled PC board should look like this! Make sure that you observe the orientation of the 12 LEDs, two ICs, seven diodes and the electrolytic capacitor. Our prototype has rectangular LEDs for the mixture display but round ones are generally easier to mount in a panel. They can also be mounted remotely from the PC board to make it easier to package the meter in your car. Note that the LDR must be able to see ambient light, otherwise it won’t work! The diodes, capacitors and trimpots can go in next, along with the two ICs. Follow these with the two terminal blocks and the piezo buzzer. Make sure that you install the polarised components the correct way around. These parts include the diodes, ICs, transistors, piezo buzzer and the 470µF electrolytic capacitor. Follow the overlay diagram and the photo closely to avoid making mistakes. Finally, install the LDR and the LEDs. The LDR can go in either way, but the 10 bargraph LEDs must all be installed with their anodes (the longer of the two leads) to the left. LEDs 11 & 12 are installed with their anodes towards the top – see Fig.2. Note that you can use high intensity LEDs if you want but because these are more directional, they may in fact not be any easier to see than normal LEDs. You may also used round or rectangular LEDs – the choice is yours. We used rectangular LEDs in our prototype for the 10-LED mixture display and these were installed with their leads bent through 90°, so that they were in line with the edge of the PC board – see photo. Alternatively, April 2004  63 Air/Fuel Ratio Measurement & Oxygen Sensors TOPIC OF measuring the voltage TtheHEoutput of an oxygen sensor to quantify air/fuel ratio is surrounded by misin- formation. This is especially the case when people are attempting to perform critical tuning of modified engines while working within a budget that calls for the use of a low cost sensor. Most exhaust gas oxygen sensors have an output voltage of approximately 0–1V, depending on the mixture strength (or air-fuel ratio). In most cars, the oxygen sensor is used in a closed loop process to maintain an air/fuel ratio of about 14.7:1 (“stoichiometric”) during idle, light load and cruise conditions. In this way, emissions are reduced and the catalytic converter works most effectively. However, this project attempts to quan- Fig.3: the output voltage from an oxygen sensor changes rapidly as the air/fuel ratio passes through 14.7:1. The degree to which the response curve flattens on either side of this ratio determines how useful the sensor is at measuring mixture strengths away from 14.7:1. [Ford] Fig.4: the operating temperature dramatically affects the output of an oxygen sensor. Sensors mounted close to the engine are particularly affected by temperature variations. [Bosch] 64  Silicon Chip tify air/fuel ratios on the basis of the sensor output, which can be well away from the stoichiometric point. Commercially available air/fuel ratio meters utilising oxygen sensors - now widely used in automotive workshops – do the same thing. However, they use what are known as “wide-band” sensors, as opposed to the “narrow-band” sensors used in nearly all cars. So what are the performance differences when it comes to wide-band sensors and can narrow-band sensors still be used to provide useful information? The most common type of oxygen sensor is the zirconium dioxide design. In this sensor, part of the ceramic body is located such that exhaust gases impinge on it. The other part is located so that it has access to the atmosphere. The surface of the ceramic body is provided with electrodes made of a thin, gas-permeable layer of platinum. Above about 350°C, the ceramic material begins to conduct oxygen ions. If the proportions of oxygen at the two ends of the sensor differ, a voltage proportional to the difference in the oxygen concentrations is generated. The residual exhaust gas oxygen component is largely dependent on the engine’s instantaneous air/fuel ratio – thus the output voltage of the sensor can be correlated with the air/fuel ratio. Fig.3 shows the typical output characteristic of a zirconia oxygen sensor. As can be seen, the output voltage varies rapidly either side of the 14.7:1 stoichiometric point. This is the characteristic curve output of a narrow-band oxygen sensor, as used in most cars. What is generally not realised is that a so-called wide-band sensor also has a very similar output, with just a little more linearity in its response at both ends of the air/fuel ratio scale! In addition to the air/fuel ratio, the output voltage of a sensor is heavily dependent on its temperature. At very low temperatures – below about 350°C – the ceramic material is insufficiently conductive to allow the sensor to function correctly. As a result, the output signal of a “cold” sensor will be either non-existent or incorrectly low in voltage (note: the minimum operating temperature varies a little from sensor to sensor). To overcome this problem, a resistive heating element is often placed inside the sensor to quickly bring it up to minimum operating temperature. Once this occurs, the heater is the usually switched off, with the flow of exhaust gases then responsible for heating the sensor. The temperature of the sensor has a major bearing on the output voltage, even in the normal working range of 500-900°C. Fig.4 shows the change in output voltage characteristics of a sensor when it is at 550°C, 750°C and 900°C. (Note that here the air/fuel ratio is expressed as Lambda numbers – Lambda 0.75 is an air/fuel ratio of 11:1). As can be seen, temperature variations can cause the output signal to vary by as much as one third of the full scale! It is also important to note that as the temperature of the sensor increases, its reading for the same air/fuel ratio decreases. Specifically, one tested sensor had an output of 860mV at 900°C, which corresponds to an air/fuel ratio of 11:1 (which is very rich). The same output voltage at 650°C would indicate an air/fuel ratio of 14:1 (ie, much leaner). The temperature of the sensor also has a major effect on its response time. The response time for a voltage change due to a change in mixture can be seconds when the sensor is below 350°C, or as short as 50ms when the sensor is at 600°C. These temperature-dependent variations occur in all zirconia-based oxygen sensors – wide-band and narrow-band. So where does this leave us when we want to source a cheap sensor for use in measuring air/fuel ratios during tuning? First, an oxygen sensor which still has a variation in output well away from stoichiometric is required. Once that sensor is found, its temperature should be kept as stable as possible, while being maintained above 350°C during the testing. As part of a general research project into the characteristics of common oxygen sensors, mechanic Graham Pring (a modification enthusiast) and the author (Julian Edgar) conducted an extensive series of tests on professional air/fuel ratio meters and sensors, both (supposedly) wide-band and narrow-band. We found that there were major variations between the readings of professional air/fuel ratio meters and that the use of a slightly used sensor could make a dramatic difference to the reading. In short, when using zirconia oxygen sensors away from stoichiometric ratios, the professional meters were often not accurate to even one full ratio, let alone the one-tenth of a ratio shown on the digital displays. The best low-cost probe that we found was the heated NTK-manufactured Ford E7TF 9F472 DA sensor, which gave excelwww.siliconchip.com.au Parts List 1 PC board, code 05104041, 121 x 59mm 1 plastic case, 130 x 68 x 42mm 2 PC mount 2-way screw terminals with 5mm pin spacing 1 12V piezo alarm siren with 7.6mm pin spacing 1 Light Dependent Resistor ((Jaycar RD3480 or equiv.) (LDR1) 1 100mm length of 0.8mm tinned copper wire Fig.5: this diagram shows the relationship between the air/fuel ratio and the voltage output at different exhaust gas temperatures for the heated Ford E7TF 9F472 DA oxygen sensor (the best low-cost sensor we have found). This sensor is sufficiently wide-band that it can be used in conjunction with a digital multimeter to give a more accurate indication of mixture strength than is achievable with the 10-LED meter. lent results, even when compared with a new Bosch wide-band sensor. The E7TF 9F472 DA is the standard sensor from the Australian Ford Falcon EA, EB and ED models. To gain the best results from this sensor, it should be mounted at the tailpipe with its 12V heater active. Any testing should be consistent in approach so that the actual temperature of the sensor (due to both the internal heater and the exhaust gas) remains similar during each procedure. For example, the same warm-up and engine loading sequence should be undertaken for each test. By using the Ford sensor in this way, results are sufficiently accurate and a fast-response multimeter can be used to monitor the sensor output. However, realistically, an air/fuel ratio accuracy of only about 1-1.5 can be expected. With this warning kept in mind, Fig.5 gives an indication of the response curves of the Ford sensor, measured at three different exhaust gas temperature ranges: 250–300°C, 300-450°C and 450–650°C. However, tapping into the car’s standard oxygen sensor and using the 10-LED Smart Mixture Display as described in the main text will still give data that is very useful. In fact, the lack of a digital readout is actually an advantage, as it stops people putting too much faith in numbers which in all likelihood are not accurate to even a full ratio. The temperature of the exhaust reduces as it gets further from the engine. As this computer simulation shows, by the time it reaches the tailpipe it is typically only at about 200°C whereas close to the exhaust valves, the gas temperatures can be over 800°C! [Network Analysis] www.siliconchip.com.au Semiconductors 1 LM3914 display driver (IC1) 1 LM358 dual op amp (IC2) 2 BC327 PNP transistors (Q1, Q2) 3 16V 1W zener diodes (ZD1ZD3) 1 5.6V 400mW zener diode (ZD4) 1 1N4004 1A diode (D1) 2 1N914 diodes (D2,D3) 4 5mm red LEDs (LED9-12) 2 5mm yellow LEDs (LED1,2) 6 5mm green LEDs (LED3-8) Capacitors 1 470µF 16V PC electrolytic 2 10nF (.01µF) MKT polyester Trimpots 1 200kΩ horizontal trimpot (VR3) 2 100kΩ horizontal trimpots (VR4,VR5) 2 5kΩ horizontal trimpot (VR1,VR2) Resistors (0.25W, 1%) 2 1MΩ 3 2.2kΩ 1 220kΩ 2 680Ω 4 10kΩ 1 10Ω you can mount the LEDs vertically so that they later protrude through a slot (or a row of holes in the case of round LEDs) in the lid of the case. Another alternative is to use round LEDs which are mounted remotely from the board, to mimic the response curve of the oxygen sensor – see photo. Installing it in a case It’s up to you what type of case you mount the PC board assembly in. As it stands, the board is designed to clip into a standard plastic case measuring 130 x 68 x 43mm. Note that if your car is very noisy, you may want to mount the piezo buzzer external to the box – or even fit a louder one. The buzzer April 2004  65 up) and that the signal coming from the airflow meter rises when the throttle is blipped. Note that the 0V connection for the Smart Mixture Meter should be made at the ECU. Setting up The step-by-step setting up procedure is as follows: (1). Make sure that the “High” trimpot (VR1) is set fully clockwise and that the “Low” trimpot (VR2) is fully anticlockwise. (2). Start the car, let the oxygen sensor warm up and confirm that the LED display shows one illuminated LED. It will probably move around, perhaps quite quickly. (3). Go for a drive and briefly use full throttle. The end yellow LED should light up. Back off sharply – the end red LED should light and then the display should blank for a moment before resuming normal operation (ie, the over-run injector shut-off is visible). (4). Check that the illuminated LED travels back and forth when the engine is at idle (ie, the engine is in closed loop mode). Cars like this Ford XR6 Turbo are especially vulnerable to engine damage if the mixtures go lean under load. The Smart Mixture Meter sounds an alarm the instant there is a high-load lean-out, allowing the driver to back off. can draw up to 60mA without causing any problems to the circuit. Fitting You will need to make four wiring connections to your car. It’s easiest to do that at the ECU, so you will need to have a wiring diagram showing the ECU pin-outs. The four connections are: (1). +12V ignition switched; (2). chassis (0V); (3) oxygen sensor signal; and (4) airflow meter signal. Use the car’s wiring diagram to find these connections and then use your multimeter to check that they’re correct. For example, when you find the +12V supply, make sure that it switches off when you turn off the ignition. In addition, you have to confirm that there is a fluctuating signal in the 0-1V range on the oxygen sensor lead (the car will need to be fully warmed Adjusting the display to suit your oxygen sensor (1). If the end yellow LED never lights, even at full throttle, adjust VR1 so that it lights when the mixtures are fully rich. (2). In closed loop, the moving LED should move back and forth around the centre LED. If the oscillations are all down one end after adjusting VR1, adjust the “Low” pot (VR2) again to centre the display. Adjusting the Lean Alarm (1). Adjust the Load Threshold pot (VR5) until LED12 comes on at reasonably heavy loads. For example, in a turbo car, the pot should be set so that LED12 first lights when there’s a little boost showing on the gauge. Fig.6: this is the full-size etching pattern for the PC board. Table 1: Resistor Colour Codes o o o o o o o No. 2   1   4   3   1   1 66  Silicon Chip Value 1MΩ 220kΩ 10kΩ 2.2kΩ 680Ω 10Ω 4-Band Code (1%) brown black green brown red red yellow brown brown black orange brown red red red brown blue grey brown brown brown black black brown 5-Band Code (1%) brown black black yellow brown red red black orange brown brown black black red brown red red black brown brown blue grey black black brown brown black black gold brown www.siliconchip.com.au Silicon Chip Binders REAL VALUE AT $14.95 PLUS P & P In this installation, round LEDs have been used for the mixture display, mounted remotely from the PC board. Note how the owner has chosen to arrange the LEDs to mimic the response curve of the sensor. This is a great approach if there is sufficient room available. [Michael Knowling] (2). Adjust the Oxygen Level Threshold pot (VR4) until LED11 comes on for what would be regarded as a lean condition at the above load; eg, so that LED11 lights when the unit is showing the last green LED (LED8) before the red (LED9). (3). When LEDs 11 & 8 come on together, the alarm sounds. If this occurs when there’s no obvious problem, adjust VR4 until the alarm just no longer sounds when running high loads. Adjusting the dimmer (1). Turn the dimmer sensitivity pot (VR3) until the display dimming matches your preferences – clockwise will give a brighter display at night (so you need to cover the LDR to simulate SC night when you’re setting it!). Lambda vs Air/Fuel Ratio The ratio of the mass of air to the mass of fuel is the most common method of describing the mixture strength. So an air/fuel ratio of 13:1 means that there is a mass of 13kg of air mixed with 1kg of fuel. However, sometimes mixture strength is quoted as a Lambda (or excess air) value (λ). This is defined as the air/fuel ratio divided by the stoichiometric ratio (ie, on typical road fuels, 14.7:1). So an air/fuel ratio of 12:1 (rich) is 0.82 Lambda (12/14.7 = 0.82). Uhh, Ohhhh – Check Your Car First! In some cars, this meter simply won’t work and there can be several reasons for this. First, it needs an oxygen sensor that outputs a voltage between 0-1V, with higher voltages corresponding to richer mixtures. The vast majority of cars produced over the last 15 years use this type of sensor but there are exceptions, so be sure to use your multimeter to check the oxygen sensor output signal before buying a kit. Second, the car must use an airflow meter with an output voltage www.siliconchip.com.au varying between about 1-5V, with the higher voltages corresponding to higher engine loads. However, some airflow meters use a frequency output signal and this circuit won’t work with that type of design. Also, in non-turbo cars using a MAP sensor, the sensor voltage will go high whenever the throttle is snapped open. This may cause false alarms, as the air/fuel ratio won’t immediately go rich. By contrast, this design should be fine in turbo cars using a MAP sensor. Again, check the output of the load sensor with a multimeter first. These binders will protect your copies of S ILICON CHIP. They feature heavy-board covers & are made from a dis­ tinctive 2-tone green vinyl. They hold 12 issues & will look great on your bookshelf. H 80mm internal width H SILICON CHIP logo printed in gold-coloured lettering on spine & cover H Buy five and get them postage free! Price: $A14.95 plus $A10.00 p&p per order. Available only in Aust. Silicon Chip Publications PO Box 139 Collaroy Beach 2097 Or call (02) 9939 3295; or fax (02) 9939 2648 & quote your credit card number. Use this handy form Enclosed is my cheque/money order for $________ or please debit my  Visa    Mastercard Card No: _________________________________ Card Expiry Date ____/____ Signature ________________________ Name ____________________________ Address__________________________ __________________ P/code_______ April 2004  67