Tuning a SYM Citycom 300i

Sluggish Throttle Response.

This is partly due to the airbox and can be easily improved with no fuelling changes and very little cost.P1020006 The quickest change is simply to remove the 32mm diameter intake pipe. The airbox is in two halves so remove all the philips head screws and at the front of the outer half, there is a rubber tube. It reduces intake noise but also limits air flow so twist it out. They made it very easy to do.


The back half of the airbox is bolted onto the swingarm/transmission. Behind the air filter, it has an outlet pipe of 42mm diameter, connecting the airbox to the throttle body, as shown in this photo :

airbox connecting pipe

Because this connects directly to the throttle body, it is effectively the velocity stack. The design of the velocity stack affects throttle response, torque spread and maximum power. This stack is very long – improving torque at lower rpm, but lowering high rpm power. No wonder she doesn’t do as well as an XR250. The stack needs a bell mouth to optimise throttle response and this pipe has a tiny bell mouth with a radius of less than 2mm. For a 42mm pipe, the bell mouth radius should be at least 10.5mm. So I made a bell mouth – not a great shape but still very effective.

First I added a “C” shape of 8mm tube. The tube was cut and held into shape by using fishing line threaded inside and tied in a loop. The white stuff is epoxy putty. It dries very hard and strong and is very sticky. That can make it hard to shape but it’s easier with wet fingers. After that, I added some 15mm tube cut into semi-circular halves. They are held in place with super glue until the epoxy is ready.


The epoxy putty is smoothed with wet fingers and shaped to create a smooth gently-curved mouth. You can see the white tip of the air temperature sensor to the left of the intake. Since the velocity stack has a diameter of 42mm, I opened up the airbox intakes to get a slightly larger intake area so that the intake doesn’t choke flow for the velocity stack, but not so large that the airbox lowers its internal pressure. I cut away the plastic that was used to hold the removable intake tube. I added putty to smooth the area and more to smooth the gas flow over the square-edges


airbox tabP1010427

And I removed the intake tab and filled the cavity next to it, once again to create adequate intake area  and keep the flow as smooth as possible. BTW smooth flow not only helps to improve torque, it lowers noise too. Remove the tab, and fill the cavity.

The airbox is de-restricted and gas flow is as good as the design allows, save for changing the air filter. And yes, the throttle response to a roll-on is much more snappy. It is definitely well worth the effort to do, and it costs very little. It sounds good too – a deep purr thanks to the smooth intake.


P1010480I got an exhaust for a CBX 250 (?) and an interface from Taobao. The rivets look cheap but the rest seems fine.  The interface inner diameter was 1mm too tight so that got filed down to fit. The outer diameter was 2mm too small so I added some thin steel pipe, with a split along its length to allow clamp tightening. The exhaust then bolted on without interfering with the original exhaust mounts, or sticking out from the side of the machine. It fitted very well in fact. I got very, very lucky with that. :)

The exhaust sound is very deep but attention-getting loud without the decibel killer fitted. She ran a little leaner, and based on experience, by about 4%. However, the spark plug showed she ran a little rich as standard so I didn’t think she needed a fuel change for the exhaust. The original exhaust with its bolts weighed around 8kgs and the replacement weighs 2.5kgs, so that’s significant, especially since the 5.5kgs reduction is unsprung mass at at the back of an already back-heavy machine.

Air Injection Solenoid Valve (AISV)

aisvThere was some popping on the overrun – below about 3,500 rpm, air is sucked into the exhaust to help emissions. It comes from inside a frame tube, goes through a tube into an air filter canister, and then another tube into the AISV. The AISV is open by default, so the air flows through another tube into the cylinder head cover where there is a reed valve. The output from the valve flows out of the head and through a metal tube to the exhaust port. Above about 4,000 rpm, the ECU (orange/black wire) grounds the supply voltage (red/yellow wire) so the AISV closes to stop air flow. I removed all of it and plugged the frame tube, and the head entry to the reed valve. The popping disappeared. I made some some flat plates to bolt where the pipe had been.

aisv plates 2aisv system

In total, the parts removed weighed 550 grams, mostly unsprung weight. This is the complete system removed save for the reed valve, with the AISV (bottom left corner) upside down, oops 😉

If the AISV is disconnected, the yellow warning light stays on and that’s annoying. So I used the power for the front sidelights since I have the dip beam on all the time. The ECU wants a minimum load of around 250mA or 4W and the AISV uses 500mA or 6W so that’s the limit. The front LED sidelights use two 12V T10 LED bulbs, each using 2 watts (4W total).


Air Filter and Fuelling Modifications.

There is a  DNA filter available http://www.motorcycle-exhausts.co.uk/DNA_AIR_FILTER_SYM_CITYCOM_300_201014–product–16243.html#.U4NpoiimU1I and it can flow three times the volume, but it is very expensive to get it to Hong Kong.

DNA FCD air filter flow: 202.00 CFM (Cubic feet per minute) @1,5”H2O corrected @ 25degrees Celsius
SYM stock paper filter: 62.10 CFM (Cubic feet per minute) @1,5”H2O corrected @ 25degrees Celsiusk&N filter

What volume of airflow does this engine need ? Based on this


…. Airflow (m3/min) = (Engine Size (Liters) x RPM) x VE / 2000 …. so the engine is 263cc or .263 litres, let’s say rpm is 6,000 , and VE is 0.90 for a 4 stroke, so that’s a flow of .263 x 6000 x 0.9 / 2000 = 0.71 cubic metres a minute. 0.71 x 35.3 = 25 cubic feet per minute. Pulsation factor is 2.1 so that means 50 CFM is at 6,000rpm, and a maximum of 75 CFM at the red line of 9,000. The calculation suggests the stock filter is restricting performance a little, however, after replacing it with a homemade K&N air filter, I found out that the stock air filter is limiting performance significantly.

To make my own K&N, I looked around but it’s hard to find one that will fit in the space, and that is available locally, and is re-usable,  and will flow at least 50% more. The best option was a K&N for a Honda Silverwing 600 part HA6002. The SYM paper costs HK$220 (US$28) and is replaced every 6,000 kms, whereas the K&N only needs cleaning and re-oiling every 20,000 kms and costs around 3x the SYM so it’ll pay for itself in time. The elements are a similar size so it was an easy job to cut them so the K&N fitted the SYM frame, and was fixed and sealed with epoxy putty. I found that the air filter needs cleaning every 12,000 or so because so much crap gets into the airbox. So I added a pre-filter using gauze from a fish tank cleaner and this worked at stopping the larger particles and can be quickly replaced. Currently, I’m experimenting with using two pieces of flexible pipe (from a vacuum cleaner) to get the airbox to draw air away from the area near the rear wheel. This seems to work well at reducing the dirt drawn into the airbox.




As a temporary way to enrich the air/fuel mix, I added resistance in series with the TA sensor – air temperature sensor in the airbox. I used a Bourns 10K ohm 20-turn potentiometer. P1010857P1010864The ECU has an internal resistance of around 2.2K ohms. I took her to the SYM dealer and plugged in the diagnostic computer. The temperature that day was 25 Celsius and the computer said 1 deg – a drop of 24, or 8.4% extra air density for 4.9K resistance. Every reading was normal except for the MAP (manifold absolute pressure) which read 32, the lower limit, and Baro (barometer reading) was 97kpa with 98kpa as minimum. So I made an op amp circuit using an OPA735 op amp in a non-inverting negative feedback circuit to set the MAP output with a gain of 1.075, or 7.5%. This mod worked very well up to about 9% extra voltage, and above that she ran fine but chugged when started from cold, and fuel economy got worse. To fit it, I cut the output wire red/black and soldered the copper. I removed the insulation (melted then trimmed) from the 5V supply yellow/black and the ground green/red. The 5V supply is used to power the op amp (red), which as a very low quiescent current so doesn’t add any significant extra power demands, and this is grounded through the black wire. The blue wire takes the output of the sensor and passes it directly to the positive input of the op amp. The op amp is difet so it has a very high input impedance and doesn’t place any strain on the sensor. The gain is adjustable by the blue 10K ohm negative feedback potentiometer. The only other components are power supply capacitors – 4.7uF 35V and 0.1uF X7R ceramic. This helps stabilise the supply so the op amp performs reliably.

A few hundred kms later, I took out the spark plug to see how the engine was doing. Here’s the plug, with a new CR9EIX to compare:

cr8eix 27_03_15New CR9EIX and used CR8EIXta sensor values





The next step was to replace the added resistance on the TA sensor with a thermistor to get a consistent amount of fuel added in summer and winter. A fixed resistor will add more fuel as the temp rises and the TA sensor’s resistance drops – and it’s non-linear – as the graph shows. So to avoid adjusting the resistance every time the temp changes, I found a suitable thermistor on Ebay to replace the Bourns 10K pot resistance (Vishay, P/N NTCLE100E3152JB0). I had frozen the TA sensor in ice with a temp probe and measured the values as it warmed up, then worked out what I needed to get a consistent gain. This xl file has my hard-to-follow calculations: temp sensor.

I’ll keep updating this page as I make progress…but so far so very good. I’ve got 7.3% extra on the TA sensor and 8.9% extra on the MAP and she has much more power and instant throttle response. I’ll try to get an optimum balance of fuel economy and power. I have to say I’m very happy with the extra performance and now I just wish the chassis was better.

This is the spark plug at 16,000kms. Looking very healthy and hard to do better.


I recently re-geared her with some heavier DR Pulley roller weights ( 25mm diameter, 22mm width, 20gram: 25x 22 x 20) and although this reduced acceleration, it lowered the engine speed too so the scoot now does an indicated 110kmh at 6,000rpm – the gearing at highway speeds is now 17kmh/1,000rpm compared to a little over 16kmh/1,000rpm before. This started dropping soon after because the belt was coming to the end of its life. The belt was 23mm wide at 18,500kms, (24mm new) and by 21,870 the width was the same but revs had started rising while cruising on the highway. So I fitted a new belt, plus I cleaned the K&N air filter and cleaned out the crankcase breather tubing because a little oil was in the airbox from over-filling the engine oil. After the new belt was fitted, she went back to 110/6,000, and gave 120/6,700 and a top speed of 148/8,200 on the flat into a headwind. I reckon she’d do 150 indicated, which is 135kmh real speed according to GPS.

I was thinking about modifying the stock exhaust muffler to try to create a simple version. I measured the inlet and outlet tubes and I suspect the muffler is the same design as a Zuma, so I should be able to create this:

Before modifying the stock, I tried out a longer exhaust to see what effect it would have. It was very cheap – designed for a CB400 and there’s a huge number of them on Taobao. Not only does it fit well, it performs better, with better throttle response and more low down torque. Without the dB killer, it is very very loud and unfortunately the dB killer resonates so it’s not as loud but drones at an unpleasant tone. To kill the resonance, I fitted a short slash cut stainless tip into the 1″centre hole – and it worked! She’s louder than stock but not obnoxiously and the tone is deep and soft. Nice.



The maximum MAP gain is hard to know but here is a guesstimate. The highest air pressure recorded on Earth is around 1085 hPa and Hong Kong’s highest recorded is 1035, with the highest monthly averages of 1020 in January, so 4.8% gain (1085/1035) is fine, and then add a margin of error for the sensor component, which is likely +/-5%. So +10% is fine. The other main factors affecting air density are altitude (I’m at sea level – the highest pressure so that altitude is still managed by the MAP sensor), humidity and temperature. A humidity change from 50% to 90% reduces air density by 0.6% – that’s barely enough to notice and about the same as a 1 deg C change. Temperature changes are still managed by the temperature sensor, albeit in a lower range due to the extra resistance.

atmospheric pressuresensors