[jcdammeyer]

Hi everyone,

Been a while what with work-work projects but I finally have had an opportunity to get back to doing some work on my never ending CNC project.

Henrik Olsson published a nice application paper on a power dump circuit for handling over voltage conditions generated by a decelerating Servo Motor.

I decided to take that one step further and I've made up a couple of circuit boards that do two things.

1. First board has a relay that when not energized does nothing to a resistor placed in series with the primary voltage going to the Servo Power transformer. This is the soft start board.

2. The second board uses the output of the Servo Power Transformer to signal a small PIC microcontroller that AC power has arrived. After 2 seconds, the PIC turns on the relay on the first board shorting out the resistor and allowing full primary current into the transformer.

The second board also monitors the DC Servo Rail across the filter capacitors. If the voltage goes above a threshold it turns on a FET which places a load dump resistor across the capacitors. After the transient, the PIC waits a bit longer and the turns off the FET.

There's also another relay across this dump resistor so that if the power to the PIC controller vanishes, the relay opens and the resistor is places across the capacitor. This way the power bleeds off the cap fairly quickly without drawing power when it's not needed.

A picture is attached. I'll post the schematic in a day or so after I've done some testing with an actual power supply and UHU Servo drive with motor. I have no idea how well it will all work. I'll post some connection drawings and test results when I have them.

John Dammeyer

[tenmetalman]

John, put me on your list for a minimum of a couple of PCB's at least. it looks like your fertile mind is at it again !
Paul

[jcdammeyer]

Hi,

I've attached the schematics of the boards as modified with a few extra parts added after the original boards didn't work quite right. That's what I get for rushing.

I'll write up a circuit description next. The biggest problem with the prototypes is that only 5V is used to turn on the Drain Control FET. It really should have been 12V. I'll have to add an opto isolator or an NPN drive transistor to isolate the PIC processor from 12V. I need to breadboard that yet.

The rework needed on the original prototype board is quite extensive but unless I can piggy back an upgraded PCB order with another one the proto boards will be adequate for my personal needs.

John

[rboeser]

Hello;

How has this worked for you? Looks like it could come in handy on over voltage / back emf from the servos.

I see the old schematic with 5v driving the m-fet. Do you have a new schematic with the 12v driving the m-fet?

Why not put all components on a single board?

Thank you, Ron

[jcdammeyer]

Hello;

How has this worked for you? Looks like it could come in handy on over voltage / back emf from the servos.

I see the old schematic with 5v driving the m-fet. Do you have a new schematic with the 12v driving the m-fet?

Why not put all components on a single board?

Thank you, Ron

Hi Ron,

I don't remember why I made it two boards.

I've finally finished the casting for the Size34 Stepper motor for my mill's knee. The Gecko has arrived. We'll see if it can run the 500 oz-in motor to move the knee upwards. Otherwise I have to wait for the 1200oz-in motor to arrive.

In either case, I'm planning on getting back to it all this summer. And, the schematics posted before are older and match the PC boards I built. The PC board now has lots of jumpers and holes drilled to make it match the new schematic.

John

[rboeser]

Hello John;

With your approval, I will do my best to make the time to redraw both schematics, combining them into one schematic in Eagle and draw a new PC board and material list.

I will then post the files and release it to the public domain with credit to you for the design.

It's been a few years since I had my last PCB made, it is a fun project for me.

Before I order any pc boards, I would appreciate you and others review my work to double check and correct any known errors.

I will also need a count for the number of boards to order, the more the better. (cheaper)

Thank you, Ron Boeser

[jcdammeyer]

Looking at the two drawings again I remember why it's two boards.

It's not good (or safe) design practice to mix the AC power side with a digital logic side. Even your power supply box or cabinet should be designed in such a way that there is a barrier between the lethal (not that 90VDC at 15A isn't lethal) AC side and the control side.

I don't recommend building it all onto one board. The system was designed so that the 16V supply transformer was not run through that relay. Things wouldn't work well if it was. So the DrainControl circuit is always alive while AC power at the circuit breaker or power switch is on. That provides the logic 5V and the 12V relay supply.

Once the DrainControl is active it turns on the relay to remove the AC input power resistor from the circuit.

The 5V regulator isn't needed on the Softstart board. And the J1 connector on the DrainControl is input for the 16VAC from the small transformer for logic (or some other power supply) and the AC from the output of the big motor power transformer. The U2 circuit is to sense when power is applied to the big transformer.

Often power is applied to that big transformer based on the state of the ESTOP signals. So if you are not in ESTOP, a relay supplies AC power to your Motor Transformer. The DrainControl senses that and a short time later shorts the inrush resistor.

If an ESTOP happens, AC is removed, the inrush relay is de-energized and the Drain Relay is turned on to make the power supply capacitors safe.

But in either case, the AC mains side should be kept away from the DC side.
John

[rboeser]

Hello John;

Thank you for the explanation of the two boards.

I agree that AC should always be separated from any DC, especially logic circuits.

However in some circumstances it appears the two do appear on the same board. IE: the newer servo drives that accept 80-240vac and yet contain the components to drive a servo.

I hope everyone treats ALL electricity with respect be it AC or DC.

With the proper ground planes and isolation, it is safe to do.

In reality, one would mount the "AC" board in an electronics cabinet, then 4 inches to the right of it will be the "DC" board mounted in the same cabinet.

I could design a nice filtered DC power supply with a board mounted x-former starting at 24vdc, then 12vdc and finally 5vdc. You could use the 24vdc to drive the other relays in the cabinet and not have to take the real estate or expense of a separate power supply.

Ron Boeser

[jcdammeyer]

Hi Ron,

Any consumer equipment is built with the AC and DC in the same box. But the regulations are pretty clear about keeping things separate. For that matter a switching power supply as AC on the input and DC at the other end of the PC board. So one could do all this on one board.

But, I still like the idea of keeping the AC mains end of things on one end of my metal box with some shrouds around it all so fingers etc. have no access.

Trouble is my 90 to 110VDC is also lethal so playing with those big servo's has a certain element of danger. Except, the metal frame of the mill isn't part of the return circuit. I won't be connecting the DC ground side to the frame of the mill. So to get a lethal shock one would have to have both hands on opposite sides of the power circuit. That's harder to do compared to touching something live with one hand while leaning against machine ground with the other.

John

[Al_The_Man]

I agree that AC should always be separated from any DC, especially logic circuits.
However in some circumstances it appears the two do appear on the same board. IE: the newer servo drives that accept 80-240vac and yet contain the components to drive a servo.

It is not always the case and it can be done, if done with thought.
For e.g. your PC MB P.S. (logic) is not isolated from your AC spindle.
Not unless you use a laptop PC.
PC. P.S. referenced to earth ground. AC spindle referenced to earth ground = no isolation.
In the case of most Servo drives that are mains fed the logic however is isolated from the power section.
Al.