Power supply voltage exceeding 160V

[ikke1983]

Hi,

I have a Mikron WF51 CNC milling machine, the machine is in excellent mechanical condition but the original Siemens controller is not working. Instead of finding out what's wrong with the controller, I am freeing myself of future controller related headaches and converting the machine to run with mach3. Because the original servo drives are analog drives, they have to be changed by digital ones, I bought a VSD-XE drive from Granite devices. The sevo's on the machine have a voltage rating of 167V and a current rating of 11.5A. So I figured I needed quite a big power supply to run 3 such servo's, luckily I had a big transformer laying on the shelves that could possibly do the job. The transformer is a 3 phase 380V/120V with a power rating of 6500VA. 380V used to be the voltage for 3phase power in Belgium, but it has changed to 400V a couple of years ago. This means the AC output voltage of my transformer would become 126VAC and when rectified 178VDC, this is 18V more then the 160Volts maximum of the VSD-XE drive. Is this allowed? And when not, is there an other solution than buying/building a 3 phase controlled rectifier to reduce the voltage? Does a higher input voltage reduce the regenerative braking power of the drive.

Thanks
Jonathan

[Al_The_Man]

You could look at reducing the windings, take turns off, on that size of Txfr, the wire gauge will be fairly substantial, the secondary's are typically wound on last so it may be a fairly easy task?
The advantage with three phase rectified supply is you do not need anywhere as much for smoothing cap values.
Al.

[Xerxes]

Another solution to lower transformer voltage is to connect smaller and lower voltage transformer in series with the main transformer in reverse polarity.

For example if your transformer output rating is 10A and 126VAC and want to reduce it to 10A 110VAC, then get a 10A 16VAC transformer and connect it in series in reverse polarity. In case of 3-phase transformer you may need 3 pcs 1-phase transformers to do that. Getting polarity correct may require trial and error method (with electronics disconnected).

However, reducing windings it the cheaper way of course.

[ikke1983]

Sorry for the late reply, but my account didn't work, found the problem, needed to activate it.

Thanks Al, the secondary winding where indeed on top, and easy to unwind.
At the moment of unwinding the secondary windings, I had exactly 160V DC after rectifying. But as the grid voltage tends to be higher at some moments (lots of solar panels in our street that boost the voltage, all trying to deliver to the grid), I have 162-163 V sometimes.

Because I have big motors (lots of inertia) I get an over voltage fault every time the motor deccellerates, setting the accelleration/deccelleration value's extremely low in Mach3 solved the problem, so it was clearly a regenerative voltage problem.

Installing a brake resistor should solve the deceleration/over voltage problem.
But my question is, does the somewhat to high voltage (163V DC) make the drive more sensitive in triggering an over voltage fault? If so, I can unwind the secondary windings a bit more until I get 155V, or is this unnecessary and can the drive handle these few volt above 160V?
Another question is, when configuring the drive with GDtool, there's a field called "HV upper limit", what is is the maximum value I can user here? Since the components on the VSDE drive are 200V rated, can I use 200 000mV? The rectifier handles 660V and capacitors 350V, so there's no problem there. Or is it useless raising this value since I have to use a regenerative resistor anyway.

Jonathan

[Al_The_Man]

I cannot speak to the drive set up but every volt higher you have on the P.S. increases the tendency for over voltage to occur on the decel moves.
If you have large motors and want high decel rates, then a automatic parallel dump circuit may be needed.
Al.

[ikke1983]

YouTube - VSD-XE servo drive powering 1,8kW siemens servo motor
Here you can see how it runs now, very low acceleration and deceleration. But this wil be solved once I have the brake resistor installed (dump circuit). The servo motor has a 11.5 amps current rating, but the drive only delivers 10A. It can deliver 14A but then you lose the regenerative braking option, which is worth more to me than the 1.5A I'm losing now. The 10amps give some nice accelerations, I've tried that already, but then I have to reset the drive after every move because it decelerates at the same rate and triggers an over voltage fault.

Jonathan

[born2bewilder]

I've got a ~3kW servo I was planning to use with a VSD-E drive @ 160v, so this is very interesting. Please do update when you get a regenerative resistor installed, I'd like to see that in action...

[Xerxes]

Jonathan, nice video!

Sorry, i somehow just now notice your questions. My answers:

But my question is, does the somewhat to high voltage (163V DC) make the drive more sensitive in triggering an over voltage fault? If so, I can unwind the secondary windings a bit more until I get 155V, or is this unnecessary and can the drive handle these few volt above 160V?

The maximum settable overvoltage threshold is 170V, so the closer you operate that voltage, the more prone it is for overvoltage faults. I.e. at 169V it takes just 1V to cause overvoltage stop.

Another question is, when configuring the drive with GDtool, there's a field called "HV upper limit", what is is the maximum value I can user here? Since the components on the VSDE drive are 200V rated, can I use 200 000mV? The rectifier handles 660V and capacitors 350V, so there's no problem there. Or is it useless raising this value since I have to use a regenerative resistor anyway.

Drive firmware limits this to 170V. Operating at 200V would be bad idea for reliability (any additional overvoltage could kill the drive).

[born2bewilder]

While we're on the topic, I had a couple related Q's perhaps Tero (or others) could answer.

Firstly, concerning regenerative braking resistors, I was wondering how the ohm value affects the performance of the the braking, and/or voltage rise during braking. I was planning on buying a VSD-E to drive some oversized servos, and therefore may be pushing the drive to it's limits: 160v @ 10a continuous, 20a peak. The manual states that a braking resistor should not allow current to exceed 10A, which in this case means 16 Ohms total resistance.
If one is aiming to maximize deceleration performance, is it important to keep resistance as close to this minimum as possible?
What value resistor might be required to keep PSU voltage rise to 5-10 volts over nominal, and thereby avoid overvolt faults?
Would having a 32ohm braking resistor limit servo deceleration to half of possible acceleration rates? Or would it be even less, accounting for peak current values allowed?
Any light you could shine on the subject would be appreciated.

Secondly, the drive documentation allows that a single braking resistor can serve multiple axes on a system. Does this mean that for 3 VSD drives, the AO output lines for each could, or should, be tied together?
Also, the 10A limit, what is the reason for this? Is it due to PCB traces, an internal component (relay perhaps), or due to wiring? Would there be any value to connecting both the AO lead as well as the A channel motor lead to the braking resistor?

Thirdly, a question on implementation. I was eyeing some 100ohm, 100watt ceramic wirewound resistors. Assuming three servos on VSD-E's, running at 160v and 10A, would six of these in parallel be suitable for use? or perhaps 8 in a 2s4p arrangement.

Lastly, if anyone can comment on the suitability of lamps or heating elements as braking resistors. I've seen some comments on either side, but perhaps someone has some thoughts on what the ramifications of such a setup would be. I was thinking of 2x300w, 120v halogen lamps in series, or possibly an immersed 1000w water heating element (I have one laying around). The latter has a cold resistance of 15ohm.

-b2b

[ikke1983]

As regenerative resistors I ordered three (one for each axis) 47ohm resistors which can dissipate 300 Watts. I connected one today, and it solved the overvoltage problem perfectly. (see video)

YouTube - DC Servo Test
@b2b: The manual states that one regenerative resistor could be enough, but I think it also depends on the application. If you are going to be doing a lot of 3D milling operations, there might be the need of braking all three motors simultaneous. In that case it seems better to have three separate resistors.

The braking power needed depends also on machine construction. If you have ball screws with a pitch of 5mm, 10mm or even worse a rack and pinion system. The reflected inertia to the motor changes dramatically. In case of the 5mm pitch, a lot of inertia is taken by internal friction (the machine table reverse driving the ball screw). Where a 10mm pitch makes it a lot easier to convert the linear motion (inertia) back to rotational motion (inertia), which has to be stopped by the servo motor.

Jonathan

[born2bewilder]

Thanks Jonathan. Nice vid. I can see it's a world of difference.

My thought on sharing resistors is: if three separate resistors work for a machine, sharing the load between three in parallel can only be better still, in terms of dissipation capacity.

I had neglected to account for many of the mechanical losses of the system, however I was considering a direct drive belt system, which would be similar to rack & pinion in terms of inertial load on the motor.

[Xerxes]

10A thru braking resistor at 160V means that motor needs to generate over 1600W braking power to make that resistor insufficient. It happens only if you run motor full speed and start decelerating at absolutely maximum rate.

If your machine has total braking power more than 1.6kW, you may need to add resistors to multiple drives. I haven't seen this to be the case with anyone, yet