0 $0.00

RANDAKK'S SENSOR OUTPUT TESTING

The chart below shows a summary of my recent Turbo sensor testing. It confirms the research done circa 2013 by Leo Knijnenburg and Ko Maljaars of the Netherlands.

Details on their Pb Sensor, etc. research and recommendations is documented here on the CX500 Turbo Forum: Honda Sensor Upgrades. This is highly recommended reading!

 

These photos show the method I used to test sensors.

Testing Suzuki Sensor

Testing Honda Pb Sensor

 

Observations:

  1. The Honda specs were published in "psi absolute" terms. I have converted to "inches of mercury" (Hg) for convenience. 
  2. Honda assumes a rather liberal input voltage range of 4.75 - 5.25V. I tested off the bike with a tightly regulated and accurate power supply that delivered an actual 5.00V
  3. Neither of my original Honda sensors were actually “out of spec” even they were on the lower end of acceptable output voltage.
  4. In addition to the two “failing” Honda sensors on my bike, I tested several other Honda sensors purchased used. They all had comparable values, but my 2 original sensors had the overall lowest overall voltages.
  5. I tested a large lot of Suzuki #15620-35F00 sensors. All were purchased used. Some were “dusty” with deteriorated paint  and some looked brand new. All were represented as “from a running bike.”  All Suzuki sensors I tested yielded virtually identical output voltages. There was no “scatter” in the data whatsoever. This is evidence of the long-term stability and reliability of the Denso sensors. The values posted above are an average of all Suzuki sensors I tested.
  6. I put the 2 best looking Suzuki sensors into service on my bike. All Pb and P1 performance symptoms disappeared immediately. Perfect EFI calibration was restored!

Conclusions:

  1. Failure of these sensors is not usually catastrophic. It’s usually more insidious over time. These sensors are engineered to convert vacuum readings into predictable voltage output signals that the ECU references on a “map.” The math never changes. ECU errors are highly unlikely.
  2. These calculations were perfectly calibrated by the Honda engineers when the machines were new. If the ECU receives a signal that is “in range” based on what it was programmed to expect, it deems the signal to be “reasonable,” even if the signal is sub-optimal.  The Honda sensors’ ability to take accurate measurements, respond quickly and convert to a reliable voltage signal unfortunately degrades over time. But things have to be very far out of whack to trigger the Fuel System Warning light. The fact that your Fuel System Warning light stays off does not necessarily mean that everything is order with your sensors! My failing Honda sensors never triggered the Fuel System Warning light.
  3. In this type of early EFI system, sensors send voltage signals to the ECU which (simplistically) translates the voltage value into a predetermined optimal decision on the degree of richness or leanness that the engine would require at that instant. There is no “on-the-fly” feedback correction. For this type of system to work well, the sensors have to deliver reliable, accurate, instantaneous values.
  4. Practically speaking, the higher the voltage, the “richer” the request to the ECU. Looking at the table above, you will notice that the Suzuki sensor voltage output is consistently higher than the Honda sensor - typically about 0.3V throughout the range of vacuum readings. Seat of the pants riding confirms that the mixture is indeed a bit richer. Just enough to eliminate annoying, bad “over-lean” personality traits …especially below 3,000 rpms. But, not enough change to introduce “over-rich" problems.
  5. I suspect one of the other issues with the Honda sensors is not only the actual voltage output values, but also the responsiveness and accuracy of the signals as well. Degradation of a 45+ year old electronic/mechanical device’s capabilities is not hard to believe.
  6. Electrical engineers could no doubt explain the technical design, manufacturing and engineering superiority of the Suzuki sensors.  In the real world, field testing of the Suzuki sensors on CX500 Turbos since 2013 has proven that they are well-calibrated, accurate, and reliable.
  7. Why did I test sensors for output voltage under +20 psi (boost) pressure? Simple. The Pb sensor is actually a MAP sensor. MAP is an acronym for “Manifold Absolute Pressure.” Obviously, this means that the Pb (MAP) sensor measures pressures in the manifold.  Mainly, this means vacuum pressures. But, in a turbo application the manifold (plenum) routinely sees positive boost pressure. In the case of the CX500 Turbo, boost pressure can achieve a designed +19 psi! The Pb sensor experiences this boost pressure just like the pistons do. The ECU obviously needs to know actual instantaneous boost to regulate fueling to the cylinders. Honda could have used only one sensor to monitor vacuum and boost, but it chose to split these duties between two sensors. So, the ECU ignores boost values from the Pb. But, it’s still important to know how the sensor responds to boost pressure. More importantly, how quickly and reliably does to it respond to vacuum readings after momentary boost episodes are finished?  My testing revealed that all the Honda sensors were sluggish and somewhat unreliable when they were cycled from boost pressure back to “0.” No doubt the elasticity of the “diaphragm” membrane loses its flexibility over time. I believe this translates into “weirdness” in everyday riding …especially if you’re like me and ride in a spirited manner enjoying the boost on every ride. By the way, none of the Suzuki sensors exhibited this bad trait.

Advice:

  1. The Honda Pb and P1 sensors gradually fail over time. With age, they lose their ability to deliver quick, reliable and consistent output voltages to the ECU.
  2. The symptoms of a bad sensor are more pronounced with the Pb sensor, but a bad P1 sensor degrades performance as well.
  3. Replacing both the Honda Pb and P1 sensor with better Denso sensors from Suzuki is an excellent upgrade.
  4. If you buy used Suzuki sensors, test (at +5V!) and compare your readings to the verified “good” readings in the table above. Your values should match very closely.