Gas discharge tubes and earth referencing

[captainscarlet]

Hi,

I have a 6 wire 12VDC stepper motor mounted to a case that might be subjected to a high voltage shock (10's of kV).

I've built a network of 6 gas discharge tubes that should short the wires to earth if a surge is detected on any of the lines.
The GDTs have a DC  sparkover of 75V.
The stepper motor driver and microcontroller is powered from a 12V isolated power supply. I chose this because I didn't want the board's ground referenced to earth in case a surge.

My problem is that when the device is powered on, some of the GDTs start to glow. Only 1 or 2 out of the 6 experience this.
I've measured about 15V across GDT

The motor still performs as expected, and if I short the earth and ground together the glowing stops.

Is this a referencing fault or something else?
Is there a way to overcome it?

GTD datasheet: http://www.farnell.com/datasheets/1640975.pdf


 

[rs20]

How did you measure 15V? What device, what settings?

Do you have an oscilloscope to hand?

Is it possible that inductive spikes from the motors are causing the GDTs to glow?

[captainscarlet]

I used a DMM to measure volts DC.
I  don't have an oscilloscope to hand at the moment, so I can't get a clear idea of what's going on.

You're suggestion of inductive spikes could be valid, as when the motors are running, the 6 GDT's flicker.
When the motor's are not running, only 1 or 2 GDT's light up, and not to equal intensity.

[KJDS]

Any chance of a diagram?

[rs20]

If you can remove a GDT and apply between 16 and 30V DC to it, and see no glowing, you can be sure that what you're dealing with is not pure 15V DC, but something more interesting with an average of 15V DC.

You can also try measuring AC volts, that's another way to see if "15V DC" is a correct description of what the GDT is experiencing.

If you have a diodes and a capacitor around, you could try building a peak detector and see what the peak voltage on the GDT is, which is really what matters.

[c4757p]

If it is indeed inductive spikes causing it, then it's probably harmless. Not a terrible way to dissipate them actually, if you're not going to use a driver that recirculates the current.

As for the ones glowing a bit when the motors are not running - are they "off"? Or have you just set the speed to zero in software? It's possible that your setup is still putting out very narrow drive pulses.

[captainscarlet]

I've attached a crude diagram.

Thanks for the suggestions.

I disconnected the motor and same one or two GDT's are still lit.

It's definitely coming from the controller and leaking someway to earth.

[KJDS]

I think others have nailed it, there are some narrow spikes coming from the controller even when it's "off"