[KorV1nex]

Hi, I need help with my Mazak QT 10N lathe using the Mazatrol CAM T2 control or a similar one. The Mazak lathe isn't calculating the rotations correctly after entering the cutting speed. For example, when I input a 30 mm drill and a cutting speed of 15, it should calculate 160 rotations, but the display on the control panel shows 270, and the spindle rotates at that speed. This issue persists across all operations. For instance, when I set the lathe to VC 120, it actually operates at VC 180. The percentage difference is consistently about 60%. Does anyone know what might be causing this? Is it a parameter issue (if so, which parameters) or an electrical problem?

[MrMetric]

I don't know that control at all, but I am guessing this is a parameter issue. This is based on a few qualifiers. First, I am assuming that the physical rotation speed of the spindle accurately matches the displayed speed on the screen. If it doesn't, then that could be either a parameter issue (a multiplier might be wrong) or hardware in nature. If the speeds match, then that would tend to validate both of those and make the issue you are encountering likely be completely software in nature. There are two sources of fault there. The first would be a parameter, which is what I think is likely the case, and the second would be bit rot. The firmware for Mitsubishi (all ?) controls resides in EPROMs, which are memory chips that are "burned". But it isn't a physical change, like burning a hole in a piece of paper, that occurs when "burning" a chip. It is an electric charge and that charge can slowly leak. Most EPROMs have a meantime before failure of 20 years or so which basically makes them ticking time bombs; there are millions of these little memory cells, although only those that are "burned" are potentially suspect.

All of that said, I've actually seen very few EPROMs actually fail. Most of my machines were built in the 90s which would put them at about 30 years old. But, as I said, only a bit burned as zero can fail (default state is one), so it is possible some bits. *would*. have failed but it didn't matter because they were already one (don't quote me on the simplistic failure case... I think it might technically be possible to have a one fail too, but not as common).

From an architectural perspective, I'm pretty sure (but not positive) that much of the Mazak system is a Mitsubishi Meldas core with a Mazak graphical user interface. IF this is true, then I further believe that no checksum is performed on the firmware when the system starts. A checksum is essentially an addition operation that is performed across each memory location and then the result is compared against a the same calculation performed when the firmware was first burned. Those values should always match and if they don't, that is indicative of a failure someplace. It isn't uncommon for systems to perform checksum verifications simply as a means of ensuring there is no bit rot, but I don't think the Mitsubishi does this. That means that bit rot *could* occur and the control would never "know."

So.... Bit rot is a possibility. Now let's talk about parameters.... I *do* think that a checksum is performed on the parameter data (which is stored in EEPROM, or Electrically Erasable Programmable Memory). The Mitsubishi engineers likely did this because corruption in this area is more likely due to the memory being capable of being changed by the software in the machine (burning an EPROM is done during manufacturing of the control). So, parameters are more volatile and, therefore, verified. This is significant because it would tend to mean that whatever oddities you are seeing are probably a result of an actual *change* to the parameters (i.e. one that was intentional). So... what is the history of the machine? When was it last running correctly? Did anyone muck with it between the time it was running correctly and now? You should probably collect answers to those, and similar, questions.

If you read through the parameter list, you can probably find some type of multiplier or other customization that will affect the calculation. You may also want to contact Mazak and see if they can help. It is likely they could tell you the parameter very quickly.

Note: I am not a certified CNC repair person and I'm not familiar with a Mazak at all. I am pretty familiar with the guts of the Mitsubishi system, although these change depending upon the age of the control.