HELP on variable lab power supply

[fastline]

I want to use our lab power supply to power up a CNC machine we have that has a bad power supply. When I tested the outputs from the "off" state, to "on" state, the dialed in voltage seemed to spike, them settle down to where I wanted it. I am concerned that voltage spike could damage something. It is also possible that my meter was picking up on something else and it would be nothing to be concerned about. I have one ch set to 24V and it spiked to 30V when turned on.

 

[vne147]

It's really hard to say without knowing how the original CNC power supply worked. Can you give some more info on that? Voltage in, voltage out, max current, etc.

 

[Sceadwian]

If it's a lab supply, dial it up and plug it in, if something goes wrong it's a poor lab supply. What kind of CNC machine is this? If you don't know exactly what you're doing you need to be very careful, what you think is wrong may not actually be what's wrong, and tweaking around with the machine without knowing what you're doing could break what IS working making repairing it all the more expensive.

 

[fastline]

I appreciate the concern but we are pretty well versed as to the design parameters of the machine and done plenty of PCB level repairs before. We know the power supply is bad, we know what it powers up, and we know why it went bad but the only thing that concerns me is there is a signal voltage that is returned as a closed loop back to the CPU board for a systems check. If there is not voltage returned to that board, she won't go. That CPU board is VERY pricey so we do not want to be sending voltage spikes to it. There is also an unknown about of time one the system is energized for the CPU board to do a system check and clear any errors. Ie, there is a window of time for all systems to boot up before erroring out.

We are still trying to repair the PS. The frustrating part is not the actual repair, it is determining the what the components are so they can be tested properly. Obviously there are several components that fit in a TO-92 package for instance. We obviously could not accurately test or replace components until we know what they are or it is just a best guess.

I am really looking for a semi conductor failure but that is just a guess because nothing made smoke, nothing got hot, no current over loads tripped, it just lost 2 of the 4 outputs. I am really hoping we can find a fix for the OE PS and be done...

 

[Sceadwian]

fastline did you test the lab supply under load?

 

[fastline]

No, I was going to add a high value resistor across a channel and test that way. The OEM indicated that it should not spike voltage if there is a small load on it. I would think just energizing the wiring would be enough load the suppress any transients but have to make sure.

Another problem we have is our lab supply is not setup to be a common rail supply so not even sure if we can do that. Our OE PS has a common pin and then separate voltage taps. We need to configure in that fashion. We also have -5V to create and our lab supply only has a 3A 5V output so still not sure how we might make a + and - reference with a common for that output. This could get rather complicated fast.

 

[Sceadwian]

What are the full and total specifications for the power supply you're trying to replace?

 

[fastline]

Due to the lack of data on the PS, the best we can estimate +5V - 5A, 5A - 1A, +8V - 3A, 24V - 5A. Input is 190-220VAC. As far as noise filtering that is still a bit uknown. The estimated current values are based on trace sizing and some components but the actual trace thickness is not known. The PS is fused at 5A for the AC inputs so guessing the PS is 80-1000 watts total. This however does not mean the machine really needs all that power. It is common to use a PS for many machines and there are several output plugs that are not used.


Not sure if you were looking for other info but performance specs are just not known as of yet but we know this just is not a precision power supply based on the components involved and what the PS does in the system.

 

[Sceadwian]

Wait, due to the lack of data on the PS? So you don't even know what the CNC lines actually need as far as volts amps or voltage/current regulation? You need to contact the makers of the CNC unit to find out what it's requirements are, there's nothing anyone here can offer you as far as assistance goes if you don't even know what you're requirements are... We're not psychic =\

Can you at the least provide the make and model number of the CNC machine and if possible the power supply itself. Without that there's not even a shred of factual evidence to go by and you're dead in the water.

 

[fastline]

Loads on a power supply will vary greatly depending on what all is being used. The OEM of the machine is out of business. The machine runs 5 AC closed loop spindle drives, 5 AC closed loop servo axis systems, among a vast array of sensors, relays, etc. We have all the diagrams for the machine and certainly not new to CNC repair as we have several and learned how to fix them before ever running them. The estimates on the PS will prove rather close. We know what each channel is running and have a good feel for what is required. Disclosing the machine make/model will not get vary far because each machine is always a custom build to the customer needs.

Basically a 5 axis, Y axis, sub spindle, live tool lathe with 3" bar cap.

 

[Sceadwian]

Without knowing the specifics a replacement supply(s) (from what you've listed you're gonna need a few 'off the shelf' parts to fill in for the proper one) its gonna cost you an arm and leg.
If you can't give us more information you know more about it than we do... Custom built or not they tend to use standard power supplies, you haven't said what you've searched for on Google to try to find the power supply reqs, perhaps there are users here better on Google than you. You've at this point spent more time deliberately withholding the information you DO have to no aid of anyone, what harm is it going to do for you to share with us the information you do have that you don't think is any good?

 

[fastline]

The power supply is an SLPW ED139 and Machine is Hitachi. I am not trying to withhold information but rather know machines and know this is not as easy as searching google or ebay for parts. There is not a wiring diagram for this unit anywhere or a parts list. There is also not a spec list because you were supposed to call the dealer and grab your ankles when it went bad. We are down to a few proprietary components from Nipron and any PCB trace damage at this point.

 

[Sceadwian]

There is also not a spec list because you were supposed to call the dealer and grab your ankles when it went bad.

Measuring it for a typical wideband cycle during operation would have covered your ass as well.

I have this link bookmarked for further study, bought the best I can do is look it up later, you're likely gonna find an answer or solution before then.