100 amp low voltage power source

[Tmoose]

I'm looking for a low dollar method to provide 100 plus amps, either AC or DC, probably for an extended period, to test some bolted connections' resistances. With the connection likely having a small fraction of an ohm resistance I think I don't need more than a volt or 2.
I've been thinking of tranformers or one of those 6 volt car battery with the external series connector links cut and run in parallel.

I bet many of you have much more clever ways to accomplish this.

Thanks,
Dan T

 

[waross]

You could try a high ratio current transformer with a few turns of heavy cable through the window. Energize the 5 amp winding with a variac. The combined resistance of the cable and the joint under test must be low enough that the 1^R of the circuit does not exceed the rated burden of the CT.
respectfully

 

[ScottyUK]

A big CT is a definite possibility - ideally make sure it a Class X type with a high kneepoint voltage so you can connect it to the mains supply through a Variac. The thing to look for is basically cross section area of the core and the weightof the CT. More cross section = better. More weight generally = better.

I recently picked up an Agilent 6681A 8V 580A DC power supply for £750 if the DC route is more appealing. It is 'transportable' rather than 'portable' at over 50kg, if that is a factor.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[rcw retired EE]

AC test current will give you impedance readings, not resistance unless you do some fancy phase angle measurements. Use DC test current.

There are low resistiance tests sets available for rent that are designed to measure resistance of high-ampacity, low-resistance bolted connections. (Biddle's -Digital Low Resistance Ohmeter - DLRO is one unit.) They read directly in micorohms or milliohms by pushing 10A or 100A DC current through the connection, measuring the millivolt drop and calculating the connection resistance. The portable tst sets are momentary units and cannot be used to heat up a connection with a continuous current.

 

[sreid]

Arc Welder?

 

[waross]

I like the welder. That and a mili-volt reading with a DVM.
I would not expect a bolted connection to exhibit much induction but if it did I may be more interested in the impedance than just the resistance. Also, the impedance will always be equal or greater than the resistance so if it passes an AC test (based on voltage drop across the joint under test) it should also pass the DC test. Granted, in many instances such as motor or transformer windings the difference between a DC test and an AC test is considerable.
respectfully

 

[Tmoose]

waross,

Were you refering to the .707 differenced in hipot testing with AC vs DC? That makes sense to me, since it is the voltage that stresses the insulation, and an AC signal is briefly 1.4X the nominal voltage.

In my test the insulation is of no concern. We just need to be sure the joint/contact resistance is low and stays low.

 

[waross]

Hi Tmoose;
I was referring to the difference between impedance and resistance. In a motor or transformer winding there is a significant difference between resistance and impedance.
An AC test will measure impedance, a DC test will measure resistance.
Impedance (Z) = (R^2 + X^2)^.5
Resistance = R
Reactance = X
A bolted joint is not expected to display reactance so the impedance should equal the resistance.
respectfully

 

[LionelHutz]

I think was Waross is saying in a round about way is that if these bolted connections are being built to carry 60hz AC current then you should test with 60hz AC current. If they are being built to carry DC current then test with DC current.

Buy a 200VA, 120VAC to 1VAC transformer and feed it with a variac. Might be cheaper than the CT.

 

[ScottyUK]

A resistive joint - negligible reactance - will measure the same using an AC or DC source. At power frequencies a bolted joint in a busbar will have no appreciable reactance. An AC source will be easier to build cheaply than a low ripple DC source.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[Zogzog]

Buy a microohmeter! Or IR scan the connections under full load coniditions.

 

[Tmoose]

I think we suspect a low resistance measured by tiny test voltages might not force the joint to reveal its true feeling about high current.

That's why we are pursuing the big amp power source.

The technical leader just upped the ante, to ~1000 amps, for a many hours. (!!)

Considering I (a short timer) had not heard anyone mention aggressively testing our joints before I started whining about it, this is in some ways a surprising development.

 

[ScottyUK]

Depending what size busbar you are measuring a micro-ohm meter might not be good enough. We struggle to get reliable readings on generator busbars unless there is something desperately wrong with the joint. A nano-ohm meter would be nice, but commercial products are rare and have poorer resolution than a heavy current source and a laboratory class multimeter. Watch for thermal EMFs at low volt-drops if you use a DC source.

I've successfully used the big DC supply to measure the resistance of the 2500A conductor on a 275kV bushing at the request of the manufacturer and I needed every amp I could throw at it. They were demanding 10n[Ω] precision to give us a go / no-go on a brazed joint which had failed on one of the other bushings from the batch - easy in their test facility with a 7.5 digit meter, a stable temperature, and a huge DC source, but bloody hard with the bushing on the transformer!


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[oldfieldguy]

NETA World, the magazine of the interNational Electrical Testing Association, recently ran an article talking about using AC millivolt drop testing in lieu of the traditional method of using DC via a digital low range ohmmeter.

The gist of the article is that there can be sizable inaccuracies hidden in the AC millivolt drop test due to the unmeasured power factor components. A relatively small increase in the total millivolt drop may be entirely due to a significant change in the resistive component of the measurement.

In other words, mV drop readings of 130, 150 and 130 may have within them resistive components of 15, 35 and 15 micro-ohms resistance, and this might be overlooked instead of investigated and corrected.

old field guy

 

[waross]

Hi Scotty
Thanks for the heads up on the thermal voltages. Am I correct in surmising that the thermal voltages will reflect the difference between the temperature of the joint and the temperature of the meter?
Respectfully

 

[rcw retired EE]

Go back to why are we testing? If the concern is how well has the joint been assembled in a design that has already been proven, then a resistance measurement of a joint may be all that is needed. Results can be compared to tests on similar joints or compared bewteen phases.

If you are trying to decide if a new connection design will carry 1000 amps AC in your machine, running 1,000A AC through the joint and monitoring temperatures with thermocouples or infrared would give better data.

The micro-ohmeter check we used was always to verify the condition of a joint or a contact in a system whose design had been proven. We were just checking the quality of the installation or the condiiton of some aged equipment. (Examples- checking intercell connections on battery banks, contact resistance on disconnect switches, bolted joint connections on smaller bus bar.)

 

[ScottyUK]

olfieldguy,

It really comes down to understanding the item you are testing and picking the appropriate test method. It is certainly possible to get erroneous results though poor choice of test method.

rcwilson,

Can I add to your excellent question 'Why are we testing' some others: 'What exactly are we testing' and 'How accurate do you need the results'. The latter inevitably leads to a third question, 'How much are you prepared to pay for that accuracy'?

waross,

The thermal EMFs occur due to the thermocouple effect of dissimilar metals in contact where the test probes meet the object being measured, where the probes meet the test leads, where the leads meet the instrument, etc. A few microvolts of thermal EMF causes havoc when it is large percentage of the volt-drop you are looking for! Good commercial micro-ohm meters use an automatic technique where the voltage probe connection is reversed during the test so the thermal EMF cancels when the two readings are averaged. Home-built test rigs for special or awkward jobs don't have the luxury of this feature so it has to be done manually.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[roydm]

Simple, find an old transformer about 1 KVA, remove the secondary and replace with a few turns of welding cable in parallel. If you have a variac available this will give you a nice smooth power supply and should put out about 1,000 Amps for a short period. I have used this to test switchgear. Figure out the Volts / turn by counting the secondary turns or with about 10 turns of 18 gauge wire
Roy

 

[ScottyUK]

After six months does he still care? And I think this solution was suggested in the first reply, with some good discussion of some of the shortcomings of using an AC test source followinbg in later posts. The problem is 'Simple' if you don't care about accuracy.

Did you read any of this thread?


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[capuchi]

Advanced Test Equipment Corp. offers on rental low voltage but high current test set for circuit breakers. Maybe they can help.