Piccolo Piccoboard gyro

Piccoboard overview
The piccoboard consists of a circuit board containing two ESCs (Electronic Speed Controllers) and a solid state piezoelectric gyro. Some Piccoboards are also packaged with a receiver in the same shrink wrap package, but the receiver is a totally separate component electrically.

The gyro is a separate component from a company called Tokin. It is connected to the Piccoboard by four soldered pins.

Both ESCs incorporate a feature that automatically sets the throttle zero point to whatever the circuit “sees” when it initializes. This autoset routine can compensate for a wide range of initial throttle settings, but it has its limits. And if the gyro output exceeds this limit or is unstable during initialization, the Piccoboard will fail to initialize.

Gyro failures (also see "update" below)
Starting perhaps in late 2001, we started seeing an abnormal number of gyro failures reported on the Ikarus BBS. Until recently, no one has come forward with any explanation for these failures. However, when I started repairing Piccoboards, I started investigating the failures and I think I have an answer. Since a schematic is not available, I have had to resort to reverse engineering, which is difficult on a multi-layered board like the Piccoboard. What I am about to present below is as close to correct as I can get without access to the designers of the board.

The original design Piccoboard (marked "Piccoboard Ikarus 9905" and lacking a wire jumper between the two caps near the servo leads) incorporated a 5 volt Tokin gyro marked 5AB1. This was apparently a custom version of the then-standard CG-L35 gyro, with a 5 volt Vcc and 2.4 volt signal and reference outputs. I have not been able to determine what was custom about it. See the Troubleshooting Page for how to measure these voltages.

Then at some point in time, the 5 volt gyro was replaced by a 3 volt gyro, marked 3AB0. Apparantly Tokin discontinued the 5V gyro in favor of the 3 volt version. The 3AB0 is a standard Tokin part.... their CG-L33, with 3 volt Vcc and 1.3 volt signal and reference outputs. Also, the output per degree of rotation is lower than the original 5AB1 parts. From what I can determine, at least some of these gyros may have been specially selected for some parameter. The standard Piccoboards that I have seen so far with 3AB0 (ie., 3V) gyros have all the same components as the original Piccoboards, so they should not work with the new gyros with lower output voltage.... but obviously they did until they failed. I suspect that selection parameter was the output and reference voltage, but I am not certain.

But the important part of this is that the voltage that these Piccoboards apply to the gyro Vcc pin is still 5 volts. The applications engineer I spoke to at Tokin told me that continuous operation above 4 volts is not recommended and can damage the gyro. Hmmmm.... do we have the reason for the failures here? Possibly.....

Just recently, I have come across some evidence that on some intermediate boards, the gyro Vcc was lowered to 3.5 volts, but this is so far unconfirmed. I am researching this to see A) was it done? B) how? C) did the reduction of Vcc have any other impact on the rest of the Piccoboard? I have recently found that the original Piccoboards (such as pictured on the Piccoboard Troubleshooting Page) used two 5V regulators; one for the BEC and one for the Piccoboard logic. Then at some point the logic regulator was removed and the logic Vcc was jumpered to the BEC regulator output. This is the wire jumper present on some boards. It is possible that there are some boards out there that have a 3.5V regulator for the logic instead of a 5V device, and I am trying to confirm this. If so, it addresses the 3V gyro issue but raises some concerns about the rest of the board. The PIC chips are specified for operation at 3V, but the LMC660 OP amp used in the gyro area is only specified to 4.75V. There is a lower voltage version, but I have not actually investigated a board that may have this configuration.

OK, so what to do? What I have done is develop a rework to the board that A) lowers the gyro Vcc to a safer level, and B) changes the gain structure so that the Piccoboard will initialize correctly with the lower signal and reference voltage on the 3 volt gyros. This modification has been tested successfully on two boards as of this writing and I am ready to apply it to boards.

Please note that the Piccoboard Plus has had all necessary circuit modifications by design, and the 3 volt gyros will work OK in them.

I am also starting to suspect that in addition to the possible voltage issue, there is a reliability issue with the bonding wires that connect the gyro ceramic element to the internal gyro circuitry. This may in fact turn out to be the major cause of gyro failures. Many defective gyros have been found to have broken wires and re-soldering them in many cases restores operation. Unfortunately, repairing a bonding wire is no easy task. They are about the size of a human hair and about as fragile. It can be done but unless you have a soldering iron with a micro tip, the iron will look like a railroad spike next to the solder joint.

Isn't reverse engineering fun?

UPDATE 10 December, 2002
Some of the newest Piccoboards seem to be showing up with a new gyro, the Tokin CG-L43. This gyro is supposed to be much more rugged.... but it is a surface mount component and therefore is not a drop in replacement for the CG-L33 (3AB0).

I have obtained a supply of these new gyros and they work great. They are fairly easily installed without board modification if the pins are bent under the gyro slightly.

Piccoboard repair page
Piccoboard troubleshooting guide
Piccolo main page
My home page