Tip of the Day 168: Of Pinched Piezo Cables Signal Drift

At a recent customer visit we were testing the tool on the bench. We were pushing on pins while watching values on the Cycle Values screen and noticed one changing when we were not touching it. The value was continuously going down.

The cause? A pinched piezo wire.

Moral of the story: Be very careful of piezo cables. Do not pinch, crush or kink them

                                                   (or fold, staple or mutilate).

Any damage to a piezo cable can cause drift in the signal. It also can cause other weird pressure data behavior like plateaus on the graph or other behaviors not natural to plastic.

Some Technical Discussion

If you recall from tip #165 any dirt or water that can provide a path between the signal wire and the shield can cause drift. This is due to the natural behavior of the piezo amplifier circuit injecting electrons across the dirt.

In the case of a pinched wire the electrons can get in from outside. Any hole punctured in the shield provides a path. A very small hole or pinch may leave a tiny hole with a large resistance. Furthermore the insulation resistance of a crushed cable will decrease because the insulating material is squashed thinner.

But the resistance with the hole is not as large as a sealed cable. Thus any pinched point, especially while it remains pinched, can create another path from ground into the inside of the cable. Hence the drift.

Note: Our specification for drift is less than 0.2% in 1 minute. Usually the Lynx amplifiers and cables are much better than that.

Tip of the Day 165: Visual Fault Detection for Piezo Sensors

In tip 164 we mentioned how the electronics do not automatically detect sensor problems for piezo sensors like they do for strain gage. Howerver the data provide useful indications if a piezo sensor is having problems.

Zero pressure

If the sensor reads zero pressure throughout the cycle but the parts are full and packed then the cable between the piezo adapter and the sensor itself may be disconnected.

1. 1. Add “Plastic Pressure / sensor location” to the Cycle Values window or open Raw Data Viewer and watch it there.

1. 1. Open the mold and press on the sensor face or ejector pin with a brass pin. The readings should go up and down. If they do not the cable may be disconnected.

Drifting Pressure

If dirt gets into the connections it provides a path for the electrons to move around the sensor and influence the reading. You can see this in a steady rise in the pressure value while the mold is open like this:

The pressure can also be falling. But the normal scaling on the cycle graph does not show this. Add the “Plastic Pressure / sensor location” to the Cycle Values window to see if it is drifting down with the mold open. Or scale the low end of the curve on the Cycle Graph to something below zero (say -1000 psi) to see if it is drifting down.

If either a dramatic up or down drift is occurring then there could be dirt or water in a connection somewhere. This could be in the Fischer connector to the piezo adapter or the threaded connection between the sensor and its cable.

The solution is to disconnect the connections, clean them, let them dry and re-connect them. The document for cleaning these connections is located on the RJG web site here.

Tip of the Day 164: Piezo Sensors Do Not Detect Disconnects

It’s all in the title.

Some of our customers have thought that because our Lynx Piezo sensors have the integrated case electronics that they would behave the same as the strain gage sensors in detecting disconnections or broken wires. This is not true. When a piezo sensor becomes disconnected or its wire cut it simply reads zero pressure. Here is why.

A piezo sensor is made with a crystal sandwiched between two wires. It puts out charge (electrons) in one direction or the other during change in pressure. When the pressure is not changing it just sits there, inert and doing nothing.

The crystal itself is a superb electrical insulator. The schematic symbol looks like this:

The crystal, when not being pressurized or depressurized, looks exactly like air. Disconnecting the wire like this…

…appears to the Lynx electronics just the same as the sensor when connected. Thus if you disconnect a piezo sensor or do not connect the cable from the piezo adapter to the sensor then the electronics sees no change and reports zero pressure.

The strain gage sensors, on the other hand, will drive the electronics to the limit when a wire is cut, shorted or broken. This is how the strain gage electronics can detect that kind of damage and report to the eDART that the sensor reading is not good.

Tip of the Day 159: Be Kind to Your Sensors: Load ‘em True

In recent weeks we have had some reports of sensor installations not up to standards. Specifically these were cases where the sensor was not supported or loaded on its center line. Like the three-legged stool, sensors will not work if you don’t have symmetrical support behind them.

Example 1: Pin Off Center

Using a force (ejector pin) sensor, the pocket holding the sensor was not centered under the pin. This example shows this problem in exaggerated form to make the point.

This causes the load to “twist” the front of the sensor creating inaccurate readings and eventual sensor damage. Also having the sensor tilted (non-parallel pocket floor) will cause the load to be off of the center axis.

Here is same example shown with a centered pin:

Example 2: Sleeve Cut Asymmetrically

While most customers use the threaded retainer nut sometimes a sleeve is used. In this example  the support sleeve for the sensor was sliced off on one side to make room for some mold component as shown here:

The force on one side caused the sensor to read innacurately.

Here is the same example with the sleeve with the proper small wire slot left at the back.

Conclusion:

Always install sensors so that the load is on the center axis of the sensor. Also ensure that the support behind the sensor is symetrical around the center. These requirements are called out on the sensor installation documents.

Tip of the Day 125: Always Ground Temperature Modules

We have discovered from time to time that customers find noise on their cavity temperature sensor signals. As noted in the install document it is critical that, when using a grounded thermocouple (as our TS-PF03-K sensor) the electronics that measure that sensor must be grounded to the same steel that the sensor is. In other words, the Lynx Quad Temperature module LQT should be bolted to the mold. If it cannot be bolted to the mold (as in hot molds) it must be grounded to the mold somehow. Some molds have insulating plates between clamp plate and platen. We cannot always depend on case-hardened clamps being a good ground.

When the LQT is not grounded you can get the kind of noise shown below. The eDART uses a 2° rise to detect the flow front arrival (for Process Time / Temp @… and for controls). Thus the noise with the LQT un-grounded completely confuses the eDART‘s ability to interpret the signal.

Additional Points re: Shielding

• Thermocouple wires must also be shielded. The mold makes a shield if they remain buried all the way up into the LQT. If not then they need slip-on shield.


• A “step” on the LQT (one digital value to the next) is 0.1 °F of temperature change. The electrical signal from the thermocouple for 0.1 °F  is only about 2 millionths of a volt (0.000002 volts, 2 μV). This is so tiny that it is like the voltage in your radio antenna. Any exposed wires act like antennas and pick up signals from lights and machinery; NOT the signals in which we are interested.


• Thermocouple extension wire used when making cables (as for attaching to a moving slide) must also be shielded.


• Use the black Lynx cables and metal JLX junctions that carry the shield all the way back to the eDART.

Tip of the Day 51: Lynx Devices v.s. Channels

Tip of the Day 50: The Ins and Outs of Lynx Devices

Tip of the Day 47: The Practical View

Tip of the Day 46: eDART Lynx Port Limits

Tip of the Day 18: How Not To Bang Your Head Against The Wall