The company I work for makes high current circuit breaker test sets, using a transformer which can produce pulses of 60,000 amperes with an open circuit output of about 12 VAC. Today, our production assembler brought it to our attention that one of the C-core laminations had very visible gaps. It appears that the tension on the steel strap that holds the assembly together is insufficient to provide proper pressure, or perhaps the laminations retained too much springiness. There is a jagged opening along one edge of the core, showing where it had been cut in a staggered pattern.
The manufacturer looked at pictures we took and replied that sometimes they opened like that during dipping and baking, but the exciting current readings were OK.
I would not think the exciting current would be adversely affected, and in fact would probably be less, but I think there will be other problems.
My reading about gaps in transformer cores shows that even small air gaps greatly increases inductance and reduces efficiency. It is important for these transformers to be able to provide very high peak currents and to exhibit low internal impedance. I am quite concerned about this, and I am afraid that this transformer will probably perform poorly, and also likely degrade in time due to thermal and mechanical stresses. It will also probably be electrically and acoustically noisy.
I have built smaller high current transformers using toroidal primary cores and copper bus bar secondaries, and they have proven to work quite well. The C-core design was originally cheaper, but I think a toroidal design would be quite competitive and also have advantages of smaller size, less weight, and greater efficiency, as well as low noise.
I have posted a photo of the transformer:
formatting link
I'd like to send the whole batch of transformers back and make our own from toroids. What do you think?
The lams are not properly glued together and the gap is not flush but staggered.
It is a poor man's excuse for one.
** Who's idea of " OK " is that ?
** Imag increases with the presence of an air gap.
** No - it reduces inductance and increases I mag.
** It may work more or less OK, despite the flaw.
It is a cert to be noisy though.
** The thing is, THAT is not a conventional " cut core" C-core.
Looks like strips have been simply cut to fit in bunches of about 10, then tied in place with a lot of hope that dipping and baking would fix all.
This the guy that designed it - shown here trying his hand at welding ?
I would have thought that introducing air gaps decreased the permeability, decreased the shunt inductance, and therefore increased the exciting current. Depending on the winding geometry the leakage inductance may also rise.
[snip]
I have only ever used flat-faced C-cores.
It looks as though those stepped cores have been so loosely assembled that varnish has got down between the lamination sections, has been overheated, bubbled, and pushed the segments apart.
Check whether there should have been some temporary banding to hold the steps together during varnishing?
"Phil Allison" wrote in message news: snipped-for-privacy@mid.individual.net...
I'll inform my client of your observation. I concur completely. The earlier versions of this tranny looked pretty good. Somebody's getting sloppy, and we pay over $6000 each for these beauties!
What I read, copied at the end of this reply, mentions core flux fringing due to air gap and increasing inductance. But later they explain how an air gap reduces inductance. Thanks for setting me straight on this.
Now *that* is an uggly tranny. I didn't know you could make one out of clay?
Thanks, Phil. Much helpful advice. Maybe I can get them to go toroidal.
Toroidal Transformer Advantages. Higher flux density Is possible In a toroidal transformer because the windings are wound symmetrically Fig. 4. If an air gap is present, more magnetizing force (H) is needed to produce a given core flux (B). I
Fig. 5. Influx fringing. a small amount of core flux escapes the core at the air gap. decreasing the useful flux path and increasing core inductance, over the gapless core. That symmetry results In smaller size and weight of the iron core. Also, because the windings completely enclose the core flux, stray magnetic fields that could Interfere with other circuitry within the en-closure are greatly reduced. Much less shielding Is required for use with sensitive or high-gain electronics. The noise (hum) In a conventional transformer Is due to core magnetostriction, which is a very small deformation of the core iron under the influence of the magnetic field induced by the AC primary current. Be-cause the windings completely envelope the core in a toroidal transformer, audible hum Is reduced to 10% to 15% of that of a conventional transformer. Because It has lower losses, the toroid transformer Is more efficient and runs at a lower operating temperature. It also has better load legulatlon than a conventional transformer of the same power rating.
The Effects of Air Gap. Much of the lower loss can be attributed to the reduced air gap of a toroidal transformer core. While intentional air gaps are designed into Inductors to pre-vent DC saturation, air gaps produce a number of undesirable effects In power transformers. Par the typical E-l-or C-core power-transformer laminations, the air gap is 0.002 Inches. In a toroidal transformer the effective air gap is extremely small (typically less than 0.00001 Inch) and can be Ignored for design purposes. The lack of a discrete air gap minimizes losses, leakage. and-flux-wave -distortion, and decreases the mmf needed to produce a given level of flux In the transformer. The Ac inductance Is determined by the number of turns, the Impressed voltage, and the core cross-sectional area. The magnetic flux path has two components, the core magnetic length and the air-gap length. Those two components are not equal, be-cause air and Iron have vastly different permeability's. Permeability Is the ratio of the change In magnetic induction (B) to the change In magnetizing force (H), and is equal to BIH. The permeability of air is constant at 1, while the permeability of silicon Iron depends on the degree of saturation in the core. At 80% saturation, silicon iron has a permeability of about 4000. Because air offers 4000 times more reluctance to flux changes, a very small air gap has a great effect on the magnetizing volt-amperes needed to produce a given output power. An air gap In-creases the effective length of the magnetic path, reduces the inductance and, as shown in Fig. 4, causes more slope In the B-H curve. That re-quires more magnetizing force, and thus more primary current, to generate a given core flux. Once saturation Is reached, no further Increase in flux can occur even If the magnetizing current Is increased by raising the primary voltage. Another disadvantage of an air gap Is flux fringing. Not all of the core flux remains within the core cross-section adjacent to the gap. A small per-centage curves outward near the edges of the core as shown in Fig. 5. causing core flux fringing. That fringing decreases the useful flux path area and increases the core Inductance
I think they should learn how to control their manufacturing processes better, and build you a proper batch, and not send out sub-par goods, regardless of notable effects or not.
If that is happening in baking, I would question the process or the varnish or whatever they are dipping it into.
That or they have settled on a problematic core design, and need to arrive at one that exhibits more consistent fabrication results.
One would think that a transformer company would already strive toward this goal.
$6000! The fact is that the mag inductance might still be OK like this, but it wont take much of additional opening for it to vanish and you're screwed.
If they insist on these trasnformers to be OK, then make them write and sign that they claim it to be OK, now and for the future, and in case the transformers later fail, then then fully pay for *all* the incuring costs (xfrmer, xfrmer replacing costs, customer damage, loss of production, loss of image,...) That should make them rethink their statement.
I would probably send them back. That looks like a problem that may get worse over time. The gaps have to be causing the inductance to go down excitation current up - even with a staggered lamination that has to be a problem. May be noisy too.
That's just unacceptable quality control - or lack of control. A company that lets that slide through may have other problems.
Winding toroids takes some pretty sophisticated machines so you'd have some capital investment there - but a toroid wouldn't have that problem and probably be smaller and less expensive in materials.
I take it by pulsing - you mean you can get 60,000 amps for a short period of time? and 12 volts is the open circuit voltage under no load? What power will it deliver continuously?
--
----== Posted via Newsfeeds.Com - Unlimited-Unrestricted-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+
Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----
"Sometimes" they open like that? Oh man. Send them back. With those kinds of pulses one needs to be a bit concerned about other equipment in the vicinity as well. It might not take the stray magnetic field lightly.
I wouldn't wait until the sparks are flying.
I love toroids. They don't "talk" and their cores usually remain in one piece. I am not exactly sure what your pulses have to look like but what if you bought large off-the-shelf toroid transformers that don't have their centers potted? Then run a huge piece of copper through the center and make that your secondary. Since the primary and all that is already on there it should be a rather quick assembly. Even if you have to ECO-release the whole set-up there should be thousands of Dollars in savings, enough for a nice company barbeque.
image shows clearly how the strapping should be wedged to prevent this extra gap from occuring. extra pressure should have been predicted at manufacture to prevent just this problem. expecting glue or adhesive, even modern engineered polymers, to sustain strength during time and the enormous excursions of lamnation travel inside such a hgh flux unit is simply unacceptable.
fermi labs in batavia illinois can offer u some graphic documents regarding the effect of failed magnets that sustain megawatt pulse impacts. they have done ring atom smashing for a while. also many others CERN, and the one at stanford/livermore in california.
get a written document from the supplier, guranteeing the performance. at least u will have some recourse when it comes time to sue the bums for your subsequent failed breaker test results. ( and believe me, some jerk down the line will sue you if any breaker is found to fail)
I have a design for a toroidal transformer using 8 sections at 2 kVA each, with tubular bus as conductors. I have designed it so that it could be liquid cooled, and I think it could be made quite compact, although the plumbing, pump, and heat exchanger may be bulky.
I recently worked on a huge DC test set on site at GE in Georgia. It was rated at something like 30 VDC at 30,000 amperes, and had 24 huge hockey puk SCRs in water cooled heat sinks with bus bars, manifolds, and hoses surrounding a monstrous three phase transformer. Originally it used SCRs and phase firing to adjust the output, but it was unstable and highly distorted. We replaced the defective SCRs and some 4000 ampere fuses, and used a motorized Powerstat to adjust the 0-560 VAC primary. That was a job!
I did that once when I had to test a 1kAmp current shunt. No way I could get equipment in a short time to produce that current, so I just grapped a toroid transformer and looped a single turn of 250mm2 cable trough. It worked well, but my analog watch almost died since the field was so powerfull.
I learned the lesson that sometimes a commercial part can be easily converted to an other use. If you are carefull with the construction (tension of connections and so on), it may even be possible to use a number of cheap toriod transformers with the secondary connected in parallel to achieve the high current
I use a small (400 VA?) toroid to produce high current for calibration into a 1000A 100mV shunt on my product, the Ortmaster, which measures recloser currents as low as 10 amps, and pulses up to about 10,000 amps. I can get about 500-800 amperes, which is enough to verify calibration at the lower ranges. See
formatting link
for more information.
I have about a dozen variable autotransformers (Powerstats) that had been damaged and were going to be discarded. Mostly the damage was to the brush assembly and parts of the exposed windings. They are rated about 10 amps
240 VAC. I removed all except the core and windings, and wrapped them with Mylar tape. They produce about 1 VAC per turn, so a single or multiple parallel bus bar through the hole would provide 2400 amperes continuous and at least 20,000 amperes for a short pulse. A toroid such as this would probably cost about $200, so a tranny using 8 of them would cost $1600 (plus the cost of bus and mounting hardware). Two such transformers in parallel would easily produce very nearly the same output as the $6000+ transformer we buy. When I get the time and energy, I may put one together and demonstrate it to the powers that be. I already have the cores and much of the copper, so I could make a few thousand $$ pretty quick.
If anyone is interested in building these trannies for delivery to Maryland, please contact me. We really don't want to be in the transformer building business, but I am very disappointed in the quality and the escalating cost of these units, and they really should not be too difficult to build.
Sometimes it can even be done with potted cores. Once I had only a huge potted version left so I carefully drilled a hole into the epoxy. But most of all it's important to take off wedding bands, metal watches and so on. A friend from Norway almost lost his ring finger when the wedding band accidentally shorted the loop.
$200 a pop is quite expensive. If you don't have to build lots of these you could look for large surplus EI transformers with outer secondary. Maybe one big enough that you don't need additional cores. Then use a bolt cutter to get rid of the secondary and replace it with bus bars. At today's copper prices the old secondary scraps might even fetch a pretty penny at a recycler for that company barbeque. But be prepared to show your ID :-)
"Paul E. Schoen" skrev i en meddelelse news:46932c9d$0$25605$ snipped-for-privacy@news.coretel.net...
I think you are right!
The flux will tend to enter the laminations "vertically" along the gap and cause increased eddy currents in the core laminate. If they used decent core material, it will be "directional" so the losses are very much lower when the field runs along the laminate and much higher losses for perpendicular fields.
The magnetic forces would tend to push the gap apart too so the transformer will run hot *and* be noisy.
It is a strange interleave BTW - What I have seen at ABB, in times past, is that the interleaves overlap (the same way as when one builds a brick wall). What you have there is sort of designed to fly apart eventually when the varnish cracks (which it will).
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.