Say you've got multi-tap output step-up transformer like so:


and you want to be able to change output voltage on-the-fly (without disconnecting output power to switch taps), and in increments smaller than 40V at a time.

Since there is only 120V potential difference between the lowest voltage output and the highest, could you take a 120V variac:



and slap it across the high side output taps like this:


?

Seems to me as it should work and give you variable 900V-1020V output so long as the variac case is not in any way connected to the input or output (remove internal case grounding wire).

I ask because (shamefully) I've never even laid hands on a variac and want to make sure there isn't some fundamental flaw in my logic that experienced variac operators would know about.

 

variacs are not insulated for those voltage levels. just put the variac into the primary.

 

The voltage levels the variac would sense would only differ by 120 volts, IF it is isolated from ground and insulated from contacting any other circuit parts or human hands..
Should your load short to ground all bets are off and the full 1000 volt tension will be placed across the variac.
Doable but not a good practice. If it for a one off test of some kind then no problem. Keep one hand in your pocket

 

I agree with alfacliff, use the vairac in the primary.

 

variacs are not insulated for those voltage levels. just put the variac into the primary.

I agree with alfacliff, use the vairac in the primary.

I posed the question using using convenient voltages and an ideal scenario. I wanted to start off with the most basic possible concept before muddying the waters.

The actual application is more complex; The multi-tap transformer is a 34kVA 3-phase transformer, 380/400/420/440/480/500V -> 3.0/3.1/3.2/3.3/3.4kV.
The 3.0-3.4kV powers a motor at the end of a 5km cable.
I need to be able to adjust the output voltage (without interrupting output) to compensate for variable voltage drop across the 5km cable, which will change with variable motor load.
Input to the multi-tap 3-3.4kV transformer is 3-phase 480V, 70A. output is 3,400V, 10A
The only variac I can find rated for 3-phase 480V 70A is this one for $14,758 USD.
Placing a giant $15k variac on the transformer primary is less than desirable. Not only money, but space constraints.

So I am considering ideas for adjusting the output voltage in other ways, among those ideas is this one; using a much smaller, 0-480V 3-phase variac on the secondary.
If I could use a 3-phase variac just to trim the output, I would only need this $2,600 variac which would fit inside the existing enclosure.

 

The bigger question is how much voltage drop are you experiencing at the motor from no load to full load?

After that what is its maximum and minimum supply voltage ratings?

 

The bigger question is how much voltage drop are you experiencing at the motor from no load to full load?

The voltage drop will depend on cable resistance. It will not always be the same cable; different cable each time it's used. Also, will depend on cable temp.
Typical cable resistance will be in the range of 5 to 30 ohms per conductor.

After that what is its maximum and minimum supply voltage ratings?

3,000V, no over/under spec. I want to keep it within 3000-3100V, never less.

 

what you actually need is a large variable frequency drive. most are adjustable to output voltage, if you dont want to use a variac or other adjustable transformer. check the local power company for availability of a "tap changer:" transformer used as a voltage regulator on power lines.

 

The voltage drop will depend on cable resistance. It will not always be the same cable; different cable each time it's used. Also, will depend on cable temp.
Typical cable resistance will be in the range of 5 to 30 ohms per conductor.

So you have multiple 5 KM long cables to use? Why not leave it with one dedicated one instead?

3,000V, no over/under spec. I want to keep it within 3000-3100V, never less.

Industry standard is for induction motors to work with any voltage within +- 10% of their spec voltage at full load and can go substantially below that at lighter load conditions. Going by that shooting for a 3300 volt no load voltage would give you at least 600 volts worth of drop to work with at full load and even more if the motor is not being ran at its maximum capacity.

To me what you are planning sounds like you may be over engineering something for no reason.

More info on what size of motor and what it is driving would be helpful.


Also going by common wire resistance charts a 5 - 30 ohm resistance at 5 Km would equate to an aluminum wire sizing between #3 and #11 gauge or #5 and #9 gauge copper of which I do not know of any utility that uses that small of wires for that length of three phase runs.

Is this some sort of special 'don't ask don't tell' off the official regulations and records application?

 

he said it is a three phase transformer, but didnt say about the motor. sizing the wires between supply and motor would be the way to go, sized for proper current load.

 

That's what I am curious about. Given the limited info he has supplied so far I have too many questions that at the present suggest their answers fall outside of conventional design standards.

He wants a very narrow working voltage at the motor regardless of load yet plans to have widely varying wire sizes connecting the motor to its primary power source some 5 KM away.

I can't figure how anyone could have multiple sizes of medium voltage powercords 5 KM long that could be easily rolled out and picked up as needed. Many years ago I worked at one of our local coal mines as a student electrician and we had large three phase power cords for running assorted equipment in remote locations and I can say for certain that 5 Km of power cord is not a trivial job to just roll out and wind up when done.

We had special cable spooling tractor units just for the task and even then doing a multi mile cable lay out or pick up was at least full day job and the size of cable for what load was going in service was always properly planned out in advance.

 

If you did use a Variac on the output you would need to mount it in an insulated (plastic) box with the adjustment shaft through a hole in the box so no one can come in contact with the case (which should be isolated from ground to avoid arc over).
The control knob should be a large insulated type with no exposed screws.
There should also be a properly rated fuse in series with the Variac output.

 

If you did use a Variac on the output you would need to mount it in an insulated (plastic) box with the adjustment shaft through a hole in the box so no one can come in contact with the case (which should be isolated from ground to avoid arc over).
The control knob should be a large insulated type with no exposed screws.
There should also be a properly rated fuse in series with the Variac output.

And then there should be a well-written set of schematics and text explanation of how the system works and why the safety features are in place - in the engineering office and a second set attached to the box. 10 years from now some poor guy who is not expecting anything like this will open the box, see a 120v variac and take all the safety precautions that experienced electricians normally take with 120v instead of a 1kv circuit.

 

And then there should be a well-written set of schematics and text explanation of how the system works and why the safety features are in place - in the engineering office and a second set attached to the box. 10 years from now some poor guy who is not expecting anything like this will open the box, see a 120v variac and take all the safety precautions that experienced electricians normally take with 120v instead of a 1kv circuit.

Good point.

The standard Danger - High Voltage sign should also be plastered all over the box.

 

So you have multiple 5 KM long cables to use? Why not leave it with one dedicated one instead?

To me what you are planning sounds like you may be over engineering something for no reason.

More info on what size of motor and what it is driving would be helpful.

Also going by common wire resistance charts a 5 - 30 ohm resistance at 5 Km would equate to an aluminum wire sizing between #3 and #11 gauge or #5 and #9 gauge copper of which I do not know of any utility that uses that small of wires for that length of three phase runs.

Is this some sort of special 'don't ask don't tell' off the official regulations and records application?

The application is a subsea saw built by my company; we perform a service, we do not sell the saws. The saw and the controls are provided by my company, the cable (subsea umbilical) is provided by the customer. So every time we go on a job with this saw, with a different customer, we will be using a different umbilical, with different numbers of different sized conductors with different ohm/km.

The umbilical owned by the customer is typically used for an ROV; we repurpose it for our saw. For the ROV, the ROV manufacturer and the umbilical manufacturer get together and design the umbilical around the ROV; all loads and voltage drops are known and factored into the design of the umbilical.

For us, we must work with what we have, which is a cable not designed for our saw, and with God-knows-what specs. We must be flexible.

Industry standard is for induction motors to work with any voltage within +- 10% of their spec voltage at full load and can go substantially below that at lighter load conditions. Going by that shooting for a 3300 volt no load voltage would give you at least 600 volts worth of drop to work with at full load and even more if the motor is not being ran at its maximum capacity.

Industry standard for induction motors doesn't have much to offer me. Industry standard applies to motors and the circuits feeding them, which typically dictates that the wiring installed in a building not drop more than 5% voltage to its load. This is fine for a motor, of which you say can handle +/-10% voltage error.

But let's say I deem +/-10% voltage "OK" for my motor. According to this, a 10% voltage drop will result in a 10% current increase. In a typical building installation, that's probably fine; in a typical building installation, motors are fed by short leads (short, relative to 5km long leads) and as long as you follow NEC guidelines on branch conductor AWG, lead resistance is usually negligible. But throw 30 ohms of series resistance per phase into the mix and re-run that [-10%_V],[+10%_I] scenario. A 10% drop in voltage results in a 10%increase in current which results in more voltage drop which results in more current increase and so on... what I describe is an avalanching failure resulting in motor stall and destruction.

 

Okay that makes a lot more sense now.

Thanks. Neat project.

Given that I see no reason why the variac would not work provided it has the proper HV isolation in place to keep above the common ground plane of the system.

 

Should your load short to ground all bets are off and the full 1000 volt tension will be placed across the variac.

Can you explain? I don't see it. I see this:

.. an infinite parallel load, that would be just as damaging to the transformer and everything else in the circuit. But I don't see 1kV across the variac.

 

Can you explain? I don't see it. I see this:
View attachment 94465
.. an infinite parallel load, that would be just as damaging to the transformer and everything else in the circuit. But I don't see 1kV across the variac.

If the output is shorted to ground then there will be ≈1000V from the top of the Variac to the wiper point on the Variac winding which could cause winding insulation breakdown.
A fast acting fuse (1000V rated) at the Variac output hopefully would prevent any damage from this.

 

This is what I see....The red is high voltage path

 

This is what I see....View attachment 94466The red is high voltage path

Good call. But, what is the voltage drop from the 1020V input to the wiper? Hopefully the drop is only 120 volts. However, when the motor is turned off, the inductive kick could be 1kV - right?