What is a decent spindle relay ?

[Doddy]

Okay - bit of info about the CP0-10V board. John was kind enough to post this to me to eyeball.

The power-supply stage is a LM2576HV-12 device, designed to take a DC supply from 14-60V to provide a switched-mode regulated board supply of 12V for the analogue subsystem. A local linear 7805 provides machine-side logic supply, and an isolated DC-DC convertor provides a PC/Parallel interface supply. This last bit is kind of interesting and intended to isolate the ground from PC to machine. There's no substantial filtering.

John's board - it is the LM2576HV-12 device that's failed. The device is obsolete - and difficult to source. There's reference on Google to dodgy relabelled devices coming out of China. It is pin compatible with the non-HV variant, or the more modern LM2596, however the latter requires an inductor change on the board.

View from the bottom of the board,



It's another board that requires pin 18 and pin 25 of the parallel port to be shorted through the parallel port of the connected PC, or controller interface, to power up the PC-side interface. Notionally a safety interlock, it's necessary to be aware of this when integrating a non-PC based controller solution. I hate these - and would chose to hay-wire this to remove the "protection". However - it is what it is.

The board does provide a charge-pump interface, powered from the board rails. This influences the ENA outputs and relay outputs. Testing with a 12.5kHz source confirms this operation. There's a slide-switch to disable this function if necessary.

Why did it fail? Hard to guess. The 2576 is a robust design - with overcurrent, over temperature protection. I'm not convinced of any suggestion of user-error - it's more robust than the interfaces presented by the board (they'd burn out before the 2576 is damaged). Handling error ESD during board assembly?, manufacturing fault with the device itself are both possibilities. Voltage spikes presented to the input of the device by a common PSU driving the stepper drivers - that's a possibility, or possibly grey supply chain and illicit device is a distinct possibility. But that's all conjecture. What I can say is the device was as dead as the proverbial parrot.

So, solution agreed with John was to remove the LM2576HV-12 and short the input/output pads on the board. The board therefore now requires a regulated 12V supply to operate. It's a simple job to replace with low-voltage variants of the 2576, or the pin-compatible 2596, but John's use-case requires a >40V tolerant supply, which these devices do not support.



Not a good picture, but shows that the power LED is now on. I've tested all digital outputs to the axis step/dir and ENA pins and relay with/without charge pump. All satisfactory.