Power supply/preparation unit

[atifeh]

Hello,

From what I have learned and gathered here and in other discussion groups, I am arranging bits and pieces of a puzzle, hoping to map a featured power supply unit. Although getting a power unit from PMDX, CandCNC or CNC Resouce will surely be justifiable in the sense of money, time and reliability, but there are some restrictions that I can not get them.

My intention is to incorporate those features that have been discussed in different threads and delineate a generic configuration of a power supply unit for use with stepper motors. After establishing the integrity of this design, different power requirements could be attained by changing the components specs within this design. For this reason I have not allocated figures to the components and the priority is given to its overall functionality. My current project is a gantry router which will be using HobbyCNC board with 1.8 N.m stepper motors. With regard to the size of my project many features that I have in mind might seem redundant, or at least there might be easier solutions or alternatives to do that, but I want to be able to use the same configuration for more powerful boards and motors.

So far these features are:
An aux AC outlet to connect to the spindle/router motor.
Using multiple filter capacitors to reduce capacitor heating.
Active bleeder resistors which would only heat up on shutdown.
An Over Voltage Protection cicuit to dischange of the DC capacitor based upon the following concept: http://zeus.eed.usv.ro/misc/mirrors/...t.htm/0198.htm
Non-electronic shutdown for limits and e-stop and providing a means to notify the software to shutdown.
Indicators for main power and capacitor dump.
Diode/fuse combination to provide a path for motor emf, in case the fuse blows.

Are there any other features missing?

Since my knowledge of electrical and electronics is very limited, any comment, correction, suggestion or direction will be greatly appreciated.

Description of the circuit:
Main 0/1 switch will activate the panel fan, the AC outlet for the router and light the MAIN ON indicator. (Question1: how about adding a switch to activate/deactivate the router outlet as well?).

A momentary N.O. pushbutton activates a Double Pole Double Throw AC contactor (relay1). The relay coil should remain active by jumping between terminals A and 4. Instead, a Solid State relay is placed between these two terminals which is connected to the hard limit switches and the E_stop mushroom button (all N.C.); once opened, should break the circuit.

A resistor will limit the current in this circuit.
(Question2: does this resistor need a filter capacitor? If yes, how should it be attached?)

The SSR is fed by a 12 volt DC power supply which gets its voltage from terminals 6 and 9 of relay1. The 12 volt DC power supplier is intended to drive future relays if required; otherwise I suppose 5 volts might be sufficient.
(Question3: does the snubber circuit do any good for the SSR protection?)
(Question4: how can I tell the software that an E-STOP has happened?) I think using the 1 and 3 terminals of relay1 (already vacant) and using proper filtering might be feasible fo this purpose.

Terminals 6 and 9 of relay1 also provide the feed AC voltage to the transformer.
The transformer's secondary voltage passes through an NTC to control the in-rush current.
(Question5: I have seen NTC's been used at the DC side of the power supply(for example in Ajax unit), but I recall Mariss suggested to use it on the AC side. How does it make difference?)

Connected to the capacitors are two high wattage bleeder resistors which are attached to terminals 1 and 3 of another Double Pole Double Throw AC contactor (relay2), being parallel to transformer's primery winding. It will be activated when relay2 goes off, cutting power from the system, hence dumping the capacitors. An indicator lamp will illuminate until the capacitors are dumped.
(BIG Question6: Do the contactors have enough response time regarding their pull-in and drop-out times during their interactions or during energy dump? Am I right using these contactors?)
(Question7: how would you compare this dumping circuit with Mariss’s energy dump circuit? What are the pros and cons?)

Heatsinks for the SSR and the bridges and shield of microswitch cables are connected to the chassis in a star configuration (not shown).
(Question8: are there any other points that should be grounded?)

The AC outlet is isolated from the rest of the circuit by a fuse, in case a fault happens in the external circuit.
(Question9: where else should be protected by fuses? To protect what?)
Diode/fuse combination is used on each track to driver.

Attachments are:
A wmf file (zipped) which is capable of resizing without losing its clarity and a dwg file (in Autocad 2000 MDT version) that contains the components as "block" which you might find useful in your designing.

Thanks
Nader

P.S. main is 230V, 50hz.

[NC Cams]

Have you ever heard of the K.I.S.S. method???

I have a factory produced Bridgeport mill with a factory CNC package that is servo based that has a very simple p/s - transformer, bridge rectifier, capacitor, bleeder resistor connected all the time - PERIOD.

I have a factory produced stepper controller for a grinder that uses the EXACT same type of P/S.

Both work reliably and has done so for nearly 10 years of daily operation (mill) and even longer for the stepper system.

It is admirable to try to come up with an "active" p/s that will do everything but make bread. However, it really isn't necessary.

If you do persist in your admirable endeavors, I'd seriously revisit the idea of using relays - high power switching of voltage/current with relays is not without its challenges. One severe issue to deal with is that of contact arcing. Perhaps the use of an SCR or mosfet may offer better reliability and be much easier to switch with electronic circuitry.

On the other hand, a simple power supply like that shown in the following links may prove to be easier and more reliable to construct:

Linear P/S design/construction for steppers:
http://www.campbelldesigns.com/files...ply-part-1.pdf

Servo amp P/S design:
http://www.elecdesign.com/Articles/A...7635/7635.html

Servo motor, servo amp, powersupply sizing for CNC
http://www.rutex.com/pdf/Mystique2.pdf

My personal opinion is that you're generating complexity that is in dire search of a need...

[Al_The_Man]

Just skimming over it briefly, there are a couple of concerns I would have,
1/ The e-stop SSR look a bit of a convoluted way of retaing the relay and by rights an E-stop circuit should not be done with electro-mechanical devices, if the SSR shorts
But if you go that route, I would eliminate the 12v supply if all you are using is the SSR, then you could use the PC 12v from one of the hard drive connectors then this would also not allow the system to start unless the PC was fired up.
I would use a contact on relay 1 in series with the SSR output to retain Relay1.
Also, it would appear the supply across Relay 2 coil will be the same potential, I don't see any neutral to the coil? The only place I see the Neutral connected is to B of relay 1?
Also the bleeder resistor connects back through a N.C. contact to the primary of the main transformer?
Also the set of contacts on relay1 actually shunt the primary winding of the 12v tfmr when the relay is operated.
What you should do is draw a schematic first and post it here.
Al.

[atifeh]

NC CAMS,
Thank you for your comments and also the links. I really need from time to time that someone reminds me to keep it simple, although when I read different aspects and views, it leaves me no choice to go deeper in the subject. And you know well what happens next... what if? ...how about? ...

The reason I turned towards AC contactors was this source while I was researching on snubber circuits.
http://www.allaboutcircuits.com/vol_4/chpt_4/2.html
Here is an excerpt which describes the arcing problem:
The reason for this is the self-extinguishing tendency of an alternating-current arc across an air gap. Because 60 Hz power line current actually stops and reverses direction 120 times per second, there are many opportunities for the ionized air of an arc to lose enough temperature to stop conducting current, to the point where the arc will not re-start on the next voltage peak. DC, on the other hand, is a continuous, uninterrupted flow of electrons which tends to maintain an arc across an air gap much better. Therefore, switch contacts of any kind incur more wear when switching a given value of direct current than for the same value of alternating current. The problem of switching DC is exaggerated when the load has a significant amount of inductance, as there will be very high voltages generated across the switch's contacts when the circuit is opened (the inductor doing its best to maintain circuit current at the same magnitude as when the switch was closed).
With both AC and DC, contact arcing can be minimized with the addition of a "snubber" circuit (a capacitor and resistor wired in series) in parallel with the contact.

This description made me think of contactors implemented only on the AC side, while avoiding the use of relays and DC current switching.

Thanks,
Nader

[atifeh]

Al,
Thank you for your comments and suggestions. Hinted by your notes I revised the circuit totally and made it much simpler.

I would use a contact on relay 1 in series with the SSR output to retain Relay1.

I tried different arrangements but without any success. I replaced the jumper back, but the relay1 will only be deactivated when the coil is shorted; that is, the SSR should be used in series with a line leading to A or B coil terminals. On the other hand, the AC side of the SSR appears as a N.O. switch before its DC activation, therefore could not be used in a line to feed the coils. It seems that putting it in place of the jumper is the only way. What problem might arise with this arrangement? I would appreciate your help.

Also the bleeder resistor connects back through a N.C. contact to the primary of the main transformer?

For the bleeder I followed this circuit:
http://www.edn.com/article/CA90760.h...dustryid=44217
The bleeder is connected to the primary of the main transformer when the relay coil is/goes off i.e. the relay is deactivated. I think the relay schematic should be N.O. type, otherwise I am missing your point. Please clarify them for me.

Also, it would appear the supply accross Relay 2 coil will be the same potential, I don't see any neutral to the coil? The only place I see the Neutral connected is to B of relay 1?

You were right and I found along similar flaws. Please check the revised version.

Description of the circuit (ver 03a):
Main 0/1 switch will activate the panel fan, the AC outlet for the router which is protected by a fuse, a 12 volt DC supply and the MAIN ON indicator.

A momentary N.O. pushbutton activates a Double Pole Double Throw AC contactor. The relay coil should remain active by jumping between terminals A and 4. Instead, a Solid State relay is placed between these two terminals which is connected to the hard limit switches and the E_stop mushroom button (all N.C.); once opened, should break the circuit. I wonder if the AC side of the SSR will only see a closed switch or else?

The SSR is fed by a 12 volt DC power supply which gets its powered before the relay goes on. The snubber circuit has been removed and a NTC is added. Also A resistor will limit the current in this circuit. If I find there is no other use for this 12 volt DC power supplier I will change it to a 5 volt.

Terminals 7 and 9 of the contactor are connected to the primary of the transformer.
The transformer's secondary voltage passes through an NTC to control the in-rush current.

Connected to the capacitors are two high wattage bleeder resistors which are attached to terminals 1 and 3 of the contactor, parallel to transformer's primery winding. It will be activated when relay goes off, cutting power from the system, hence dumping the capacitors. An indicator lamp will illuminate until the capacitors are dumped.

Diode/fuse combination is used on each track to driver.

My questions are now limited to the followings:
How can I tell the software that an E-STOP has happened?
I have seen NTC's been used at the DC side of the power supply(for example in Ajax unit), but I recall Mariss suggested to use it on the AC side. How does it make difference?
How would you compare this dumping circuit with the energy dump circuit that Mariss proposed? What are the pros and cons?

Thanks
Nader

[Al_The_Man]

I still see a DC charge-dump resistor going back to the 110vac? If you follow the circuit through the N.C. contact, it will be connected to the 110vac via the indicator light.
For clarity it is much better to draw a system up in schematic form rather than wire layout form as you have it, this is usually done after the schematic is proved.
It is much clearer to follow this way.
If I get chance later today I will try and draw up a schematic based on what you seem to want, I suspect there may be other preferencial ways of acheiving what you are after.
Snubbers should also be used across AC coils as they can be a source of voltage spikes. Generally there is no problem using DC relays if a reverse EMF diode is used across the coil.
Al.

[atifeh]

Thank you Al. I very much appreciate that. I do not know much about electrical and electronic schematics, this is the best I know. I will review the circuit again.
Nader

[atifeh]

Thank you Al,
I got what you meant about the dump resistor, and I think I have corrected it. Please take a look. Also added the snubber.
Nader

[Al_The_Man]

The one comment I would have is I do not consider it very good practice to dump the capacitors into the primary of the transformer, consider this, if they were to dump a large charge that happened to weld the contacts, when the relay pulled in, it would put 230vac across the capacitors.
Al.

[atifeh]

Al,
About the last comment that you made: I see it that the primary of the transformer is attached to the common terminals while the bleeders is attached to the unactivated state terminals. In case the contacts happen to weld, I presume the coil will not be able to pull it in, or maybe I’m not seeing your point again.

Anyway, I revised the drawing. Actually the contactor is 3PDT and I used another set to wire the bleeders through. Also I changed the NTC location and tagged the components’ spec.

I would very much appreciate to have any comments for you or others about this version, since I will be starting to assemble it by tomorrow.
Thanks
Nader

[Al_The_Man]

Originally Posted by atifeh Al, About the last comment that you made: I see it that the primary of the transformer is attached to the common terminals while the bleeders is attached to the unactivated state terminals. In case the contacts happen to weld, I presume the coil will not be able to pull it in, or maybe I’m not seeing your point again.

A contactor would probabally not pull in if the N.C. contacts were welded, but an ordinary relay would, due to construction.
I think you will need to put the indicator across the Caps, as in series it is going to limit the current and defeat the 10w resistors.
Otherwise it appears to be OK.
One other thing, if your servo drives have enable input, I would disable them in case of an e-stop. by a small relay in parallel with the contactor or some contactors can be fitted with aux contacts.
Al.

[atifeh]

Thank you Al, your comments are very helpful. So far I made the change. Also, I am using stepper motors. Is your comment about disabling the enable input also valid for stepper motors?

I have some other questions which I appreciate your response:
1) In order to notify the software that an E-stop has occurred, I thought of two alternatives:
a- Use an extension wire from the last limitswitch, reducing the current to suit parallel port and wire it to the PC.
b- Use a relay which the coil is driven from the same ~5 VDC supply and attach the contacts straight to the PC parallel port (the drawing shows the relay alternative).
Which one is suitable?

2) Also I bought two 8200mf 71 volt caps. I think they are very small in size, compared to what I have seen in other configurations (about 20mm diameter x 60mm length). It does not have a screw terminal and has soldering legs instead. I thought of soldering a 3 sq.mm (AWG 10) solid copper wire to make the connections, along with a means for using a screw terminal. Will this caps do the job, or should I go for the big screw-on type?

Thanks
Nader