Voltage Control

Hi Mate,

Sounds like your plant is drawing more current from the two weilding machines than it's rated incoming utility transformer (i.e if you have one). You either use equipment that are to the rated loading of your plant to avaoid voltage dips or probably increase the loading of your plant by means of installing a much bigger rated incoming transformer to meet increasing loads.

Regards,

Ron

PNG

Thanks Ron but this is not the case, The current drawn is momemtary hence it is not possible to install a transformer for a momentry load.

it is usually when heavy loads are started like large drives, the starting current increases several times the normal current. Because the supply and the cabling of the installation are designed for normal running current, the high starting current leads a voltage dip in both the supply network and the installation. The effect magnitude depends on how strong the network is.

One solution, if applicable, is to fit a soft starter so that starting current is limited to a lower value. Another solution is to discuss with the supply company for a lower impedance connection.

How are you going to fit a soft start to a welder?

there is a condition given, if applicable.

Momentary or not, the welder is drawing too much current for the service provided. This leads you to only a few sound choices.

• Follow Ron's advice (a GA from me ) and get sufficient power so that these welders will not deplete your power service. I suspect that if you do this you will find that the welders work better than before. You are also adding new machinery to this facility. You will require more power to support these new machines anyway.
• You could modify the welder so that there is no surge current. This can be done by adding a large capacitor to store energy in the welder but that involves knowing more about how the welder is used and designed than what we now know.
• Add an un-interruptible power supply for the voltage sensitive circuitry of the new CNC machines. However, most UPS are not really designed for providing frequent cycling.

They are all right. Get the problem to the building fixed. Yes, size it for the momentary power draw. Not only is it causing the problem, but is bad for everything electrical in the building. EVERYTHING.

A boost/buck transformer might be better than a UPS. Much cheaper for a given size, and CNC machines can be pretty big. A UPS would be better if you need very tight voltage control, like in a lab environment.

I agree with the position as mentioned by other posters.

You need to do a proper demand calculation for the facility and determine what needs to be done to "stiffen" the available power so to minimize the voltage drop to your distribution system when the load of the welders hits that system.

Now...

Not all areas of the CNC equipment will be prone to such drastic voltage fluctuations. A study of the the proposed CNC equipment in conjunction with the maker of the equipment may allow you to connect the sensitive portions of the gear (more then likely the logic controllers) to constant voltage power conditioners that will help you "ride through" your power transients.

Here is a link...

http://www.solahd.com/products/powerconditioning/index.htm

This should be viewed as an interim solution, until, the full blown load demand survey is conducted and a plan to improve the power supply at this facility is promulgated and implemented.

I agree with everyone who agreed with Ron.

Quote neeravg 2000 "Thanks Ron but this is not the case, The current drawn is momemtary hence it is not possible to install a transformer for a momentry load"

It is possible and that is what you have to do. In addition the utility circuits may not be stiff enough to handle your load. You are getting a large VD in the transformer and possible through the transmission line that feeds the transformer. You need an upgrade to these facilities.

Thanks all of you for your valuable comments.

Has any body heard about thyristor switched capacitors for controlling this type of fluctuations and will it be usefull in this case? If yes please let me know the technical details thereof and if possible, any vendor providing such services in India.

Thanks in advance.

As Ron says, you probably need to install a higher KVA incoming utility xformer.

If you can't afford it, then purchase SCR based welder controls, Weltronics, Technitron, Unitrol etc. these have the option of programming up-slope and down-slope currents (yes, like the Kiev Kid said a "soft start").

Also you might want to take a look on the welders transformer ratio and VKA, the higher the capacity and lower the secondary volts, the lower primary currents.

And you must also check on all insulations of the supporting arbor of the "live" electrode of the secondary, to prevent any current leaks to the chassis that can easily reach several hundreds of Amps. and affect the quality of the weld.

Yahlasit

The resistance welder is probably single phase, line to line connected.

The unused phase to neutral voltage will sag much less, - measure the lines.

The voltage sensitive CNC parts are likely to all be on one phase, appropriate phase selection should be a first step.

14kA seems unlikely corresponding to 5V or 7V at 75 or 100kVA respectively.

The two welders should be interlocked avoid simultaneous operation and be on the same phase to minimise sag on the 'cleaner' phase.

How momentary? How many volt-amps? AC or DC? Spot or seamless? Ask the CNC supplier. Lots of robots weld. Factories are full of transients. Chances are that a quality CNC has engineered past this problem.

I am using 14 such resistance welding m/cs along with CNC machines without any problem. My weld cycle time is 120 to 140 Millie seconds and secondary currents up to 15kA. Our process is battery cell inter-connect stab squeeze welding. to help you I need the following data:

1. What is your mains in-comer KVA rating; your mains KV & transformer rating.
2. What is the make of your welding m/c and weld controller make.
3. Is weld controller SCR based or inverter based.
4. How long is LT cable connection to your welder from mains PCC/ substation/ mains transformer.
5. What is the cable size to welder.
6. what is the application (product)

This data is required to calculate impact on CNC.

Regards.

Design system is not mentioned but when you talk of milliseconds so it seems a hi-speed Welding machine. Surely controlled by Electronic-controls; which normally use SCRs.

The current drawn is at the the mains supply voltage or welding voltage; which are normally much lower voltage.

I have the oppurtunity to work on hi-speed X-Ray machines.

It was the tuning of Line impedance of Power Source & the Input Impedence of machine.

It was partially the resposibillity of Electric Supply company to provide a specific line ZO and adjustments were made in the machines' input power side to match for Max Power Transfer.

Unless it not done problems in surges were there.

I suggest to consult the Welding machine supplier / manufacturer.

It may help. because such problems are in supplies in under developed countries.

Thanks all of you for your valuable comments.

Has any body heard about thyristor switched capacitors for controlling this type of fluctuations and will it be usefull in this case? If yes please let me know the technical details thereof and if possible, any vendor providing such services in India.

Thanks in advance.

Control the reactive current in this case will help stabilize the voltage. The best solution for such applications is STATCOM or STATCOM+ energy storage type devices. But a thyristor switched banks applicable also. We have experience in solving similar problems using a such devices. The thyristor switched banks can be performed as for symmetrical and for unsymmetrical modes of load. If'll be need you can contact me via our www.

I have designed and implemented Thyristor switched capacitor system for various applications. Also supplied STATCOM.

There are basic differences in TSC and STATCOM. like

STATCOM compensates harmonics and reactive power as well.

TSC supplied only reactive power. also can compensate for Half Cycle (10mSec).

I suggest go for TSC if you want to compensate otherwise go for STATCOM, whose response is very good and can compensate for instantaneous requirement.

If you need any more help log in to EPCOS or Crompton Greaves website. They have both system. I can help you to optimise the system.

Cheers

Paras Patodi

Thanks for the information, please give the web site address.

www.inteco-electro.ru (rus only) info@inteco-electro.ru

We do produce fast capacitor switches using thyristors for single or three phases loads. It is likely that it would solve your problem if sized properly.

Your location is a problem as we do not have a technical representative in your part of the world. We cover North America and Europe.

Contact me directly by leaving a mail in my profile on this site if you don't find a local source. (Click on my name)

Good luck.

Two common methods for dealing with this problem are:

1) Upgrade your mains to accommodate the KVA requirements of your equipment. This can be expensive because in worst case you will need to install a new sub-station.

2) Switch from AC welding to capacitor discharge welding. We build welders that deliver 150KA and use less than 15A from a 110 volt main. http://mysite.verizon.net/tom.es

It doesn't seem like good engineering to upgrade your plant mains to accommodate a manufacturing process that only requires short bursts of high current when your average requirement is very low. If you were running arc melting furnaces, it would be a different problem. I think it's a shame that the welding forum has restricted itself exclusively to arc welding because I think this discussion really belongs in that venue.

What about your neighbor dus it affect him, or is it only in your shop.

Over the years I have seen many fixes to this type of problem.

Please note that the New CNC equipment may be sensitive to more than the sags caused by the welders. The CNC equipment may not like noise or harmonic distortion on your lines.

I would recommend that you get with your CNC equipment vendor & see if the Computer portion of the equipment could be delivered & installed early,by doing so you will find out if there are any other power issues.(CNC power unit dropping off,computers halting or looping)

Check the wiring setup of your welders. These things have several wiring options. Chose the wiring option that disturbs your supply the least. This means use multiphase connections vs single phase connections & use the highest applicable voltage tap.

It would not hurt to trace the supply wiring out & check the electrical connections at all of the panel boards in the circuit. How long has it been sense the connections have been inspected with infrared & re-torqued.

The problem with short sags in voltage may not be entirely your problem. Talk to your electrical supplyer & see is they have a stronger feed line that thay could hook up to your transformer.

If you have more then one transformer feeding your building the put the the new CNC equipment on the cleanest transformer.

If you change out your present welders for new capacitive discharge type that may fix the problem but then you will have to change all of the process documentation for your work which would be a pain.

A UPS system could be installed at the CNC equipment. The batteries are a maintenance issue & when systems start failing the UPS may be overlooked as the cause.

You could install a tap switching boost/buck transformer at the CNC equipment. If the spikes from the switching do not disturb the CNC equipment. I have seen problems with these where the switches burn up because they could not make up there minds about which tap to use.

I think that a servo controlled variable transformer would not react fast enough to catch your sags. So that will not work.

The approach that I like is to use is a constant voltage transformer. These are a resonate transformer that is run in saturation. If the line voltage goes up a bit the output does not move. If the line voltage goes down, it must go down so far that the transformer comes out of saturation before you would get a output change. The problem with constant voltage transformers is that they do not like non-linear loads & must be properly selected. (non-linear loads meaning switching power supplies such as used by most computers.)

In the end I would put power monitoring equipment on the lines to see what kinds of problems are present. Only with this information will you know what needs to be fixed so you can form a plan to address problems.

good luck

Doug

thyristors

Doug,

It seems that their problem is that the welder's intermittent load has a very low PF and the KVA is very high for the transformer capacity.

Any tap changing technology will be too slow.

A resonant transformers for the CNC machine might work but would not help the rest of the plant.

The only solution that I have seen working for this type of problem is a tuned fast capacitor bank switch triggered by the welder current. In order to minimize the flickering and droops of voltage, the response time (on and off) must be in the sub-cycle range. This eliminates any mechanical contacts.

Since this is one of the products that I design, I am a little biased...but it works well in these difficult applications where other techniques have failed.

Thanks Doug for an excelent exposition.

Two further points. Many current-delivery computers have unity power factor type power supplies, (largely used to reduce neutral current in large arrays on three phase mains). This feature on the cnc machine would work better with the constant voltage transformer than the usual bridge feeding hundreds of microfarad capacitors in many supplies.

For short interuptions all power supplies have a hold-up time (not specified on the label of any PC supply I have seen). This is, in usual practice, some 30 to 70 ms. Increasing the energy storage/reservoir capacitor's value by five or ten times would mean the sags from welder could no longer be an issue.

Some computer power supplies have wide range (100 to 240V) without connection change, if the rectifier feeds the resivoir capacitor directly therewill be up to ten times more hold up time at 240V than at100V minus 10%. A lightly loaded supply also could double hold- up time. It would be best to bring up the concern and suggestions with the cnc supplier early to consider this cheap reliable and compact approach.