I have a customer who recently experienced an issue with his Yamaha Disklavier HQ100. The HQ100 is the Japanese version of the MX100II. He primarily uses this instrument over MIDI, and finds the Disklavier quite useful.
This discussion should apply to other models of Yamaha Disklavier. The Yamaha MX100A and MX100B use different solenoids, but the designs are similar.
The symptom was that, when started, the Disklavier would play a single key at maximum velocity. It would sustain it for several seconds before releasing. Additionally, the system emitted a noisy whine while powered on. This noise turned out to be a failed power supply, and a replacement resolved that particular symptom. However, despite this, the system would still play one single key after powering on and then fail to respond to any subsequent commands.
The key that was playing was a low D (specifically, D1). I disassembled the instrument and inspected the solenoid for that key. The solenoid assembly was exuding an enormous amount of heat. The burned coil can be seen below.
There was also considerable damage to the inside of the coil assembly. The solenoid was unable to move fully back down into the rest position, as the assembly was badly melted.
I replaced the solenoid; however, this did not solve the problem. The new solenoid would activate and rapidly heat up. At one point, F1 activated instead of D1, adding an interesting twist.
I disconnected CN201, meaning that the Drive C solenoid drive board was still receiving power from the Disklavier but wasn’t receiving a signal. This isolated the problem to the Drive C board. Replacing the solenoids made it clear that they weren’t the issue. Therefore, the problem had to be somewhere on the PCB below:
I traced the circuitry back from the solenoid. In the image above, the D1 key solenoid is plugged into CN207. I verified that diodes D213-D224 were all behaving correctly (between 400-600 mV forward, and 0 back). Eventually, I was able to trace the behavior to the transistor array. My current theory is that the transistor array failed, causing current to run to the solenoid continuously. The solenoid eventually melted the plastic sleeve. This would also explain why the F1 solenoid activated in place of the D1 during later tests—both key solenoids are attached to the same failing transistor array.
