First attempt at transformer experiment failed
- Thread starter 11bgrunt
- Start date
- Status
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
160530-0742 EDT
kwired:
The induced voltage in a conductor is v = K*N*dPhi/dt. Where v is the instantaneous voltage induced in a conductor from a changing magnetic field dPhi/dt, K is some constant, and N is the number of turns. K can be considered to account for all sorts of coupling effects.
I have two coils that are used to make a transformer. Each is 1500 turns of #33 wire and the coil resistance is measured at about 87 ohms. The center hole is 0.6" x 0.65". Coil length is 0.6", and the outside is 1.2" x 1.25". These dimensions include potting and wire entry.
5 V 60 Hz was applied to one coil. Note 5*5/80 = 0.31 W, worst possible power dissipation in the coil. No problem for this physical size coil. The two coils were placed in direct contact to provide greatest coupling. Voltage measurement was no load on secondary.
Air core conditions --- secondary voltage 0.17 V, ratio = 0.034 .
1/2" steel bolt core --- secondary voltage 2.1 V, ratio = 0.42 . Still nothing close to 1 to 1. As the coils were moved apart on the bolt the secondary voltage diminished as is to be expected. Note the bolt does not provide a high permeability low magnetic impedance closed path. The best possible transformers for accurate coupling use tape wound toroidal cores (no air gap) where the windings are wound together.
The steel bolt I used was some low grade 1/2" diameter 4" long part. Ptrobably a 1020 steel.
To achieve a near 1 to 1 ratio there must be very good magnetic coupling.
.
kwired:
It is not true that theoretically the voltage ratio should be 1:1 for my definition of a transformer. My definition of a transformer is the magnetic coupling of two or more current paths. My definition does not require good magnetic coupling.
The induced voltage in a conductor is v = K*N*dPhi/dt. Where v is the instantaneous voltage induced in a conductor from a changing magnetic field dPhi/dt, K is some constant, and N is the number of turns. K can be considered to account for all sorts of coupling effects.
I have two coils that are used to make a transformer. Each is 1500 turns of #33 wire and the coil resistance is measured at about 87 ohms. The center hole is 0.6" x 0.65". Coil length is 0.6", and the outside is 1.2" x 1.25". These dimensions include potting and wire entry.
5 V 60 Hz was applied to one coil. Note 5*5/80 = 0.31 W, worst possible power dissipation in the coil. No problem for this physical size coil. The two coils were placed in direct contact to provide greatest coupling. Voltage measurement was no load on secondary.
Air core conditions --- secondary voltage 0.17 V, ratio = 0.034 .
1/2" steel bolt core --- secondary voltage 2.1 V, ratio = 0.42 . Still nothing close to 1 to 1. As the coils were moved apart on the bolt the secondary voltage diminished as is to be expected. Note the bolt does not provide a high permeability low magnetic impedance closed path. The best possible transformers for accurate coupling use tape wound toroidal cores (no air gap) where the windings are wound together.
The steel bolt I used was some low grade 1/2" diameter 4" long part. Ptrobably a 1020 steel.
To achieve a near 1 to 1 ratio there must be very good magnetic coupling.
.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
160601-2329 EDT
I ran some additional tests on my 1500 turns coil. This time to just get some rough information on the coil with and without the steel bolt. For these tests the secondary was not needed and was not involved. Note: an unloaded secondary reflects nothing back to a primary.
As before the DC resistance was measured at 87 ohms. A fairly large series resistance was used to measure current, 225 ohms. This is large compared to the coil resistance but does not matter because voltage was measured across the coil for voltage and across the 225 ohms for current.
With no bolt the phase shift was 10.8 deg, and with the bolt fully inserted to the bottom side of the carriage bolt head the phase shift was 39.6 deg. measurements at 60 Hz.
No bolt --- coil voltage 4.31 V and coil current 0.0484 A. Z = 89.05 ohms. AC resistance = 89.05 * 0.982 = 87.4 ohms. Good correlation with DC measurement. Xl = 89.05 * 0.187 = 16.7 ohms. L at 60 Hz is about 0.043 H. Inductance measured at 1 kHz on a GR 1750 bridge was 0.045 H. This is a good correlation as should be expected with an air core coil in this small frequency range.
With bolt fully insereted --- coil voltage 5.871 V and coil current 0.0454 A. Z = 129.3 ohms. AC resistance = 129.3 * 0.771 = 99.7 ohms. This is greater than the DC resistance because of the losses in the iron core (bolt). Xl = 129.3 * 0.637 = 82.4 ohms. L at 60 Hz is about 0.215 H from this voltage and current measurement. Inductance measured at 1 kHz on a GR 1750 bridge was 0.14 H. This is not a good correlation, and probably shouldn't be.
To reduce Eddy current (also called Floucault) losses in the core use many small diameter, possibly 0.05", iron wires to make the core, This has no effect on hysteresis losses.
.
I ran some additional tests on my 1500 turns coil. This time to just get some rough information on the coil with and without the steel bolt. For these tests the secondary was not needed and was not involved. Note: an unloaded secondary reflects nothing back to a primary.
As before the DC resistance was measured at 87 ohms. A fairly large series resistance was used to measure current, 225 ohms. This is large compared to the coil resistance but does not matter because voltage was measured across the coil for voltage and across the 225 ohms for current.
With no bolt the phase shift was 10.8 deg, and with the bolt fully inserted to the bottom side of the carriage bolt head the phase shift was 39.6 deg. measurements at 60 Hz.
No bolt --- coil voltage 4.31 V and coil current 0.0484 A. Z = 89.05 ohms. AC resistance = 89.05 * 0.982 = 87.4 ohms. Good correlation with DC measurement. Xl = 89.05 * 0.187 = 16.7 ohms. L at 60 Hz is about 0.043 H. Inductance measured at 1 kHz on a GR 1750 bridge was 0.045 H. This is a good correlation as should be expected with an air core coil in this small frequency range.
With bolt fully insereted --- coil voltage 5.871 V and coil current 0.0454 A. Z = 129.3 ohms. AC resistance = 129.3 * 0.771 = 99.7 ohms. This is greater than the DC resistance because of the losses in the iron core (bolt). Xl = 129.3 * 0.637 = 82.4 ohms. L at 60 Hz is about 0.215 H from this voltage and current measurement. Inductance measured at 1 kHz on a GR 1750 bridge was 0.14 H. This is not a good correlation, and probably shouldn't be.
To reduce Eddy current (also called Floucault) losses in the core use many small diameter, possibly 0.05", iron wires to make the core, This has no effect on hysteresis losses.
.
What about a fan regulator as a variac?
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
160602-21026 EDT
Sahib:
Close the magnetic path with a high permeability material with negligible air gap, then the open circuit output voltage will be very closely related to the turns ratio. If I can find my stash of core material, then I will run an experiment. I don't want to take a transformer apart for a core.
This type of test should never be run at a frequency near any of the resonant frequencies of the transformer.
.
Sahib:
Close the magnetic path with a high permeability material with negligible air gap, then the open circuit output voltage will be very closely related to the turns ratio. If I can find my stash of core material, then I will run an experiment. I don't want to take a transformer apart for a core.
This type of test should never be run at a frequency near any of the resonant frequencies of the transformer.
.
- Location
- Placerville, CA, USA
- Occupation
- Retired PV System Designer
A fan regulator (speed control?) is not a linear, sine wave output, device.
PS: A Variac (TM) is a variable autotransformer. Not any other type of variable AC voltage output device.
I must say that I have found the responses to be quite interesting.
I still maintain that is is a core problem resulting is high Eddy currents.
I would go with s shell form type transformer where the core is oblong with straight sides. Simply use a long strip of plain sheet metal, not galvanised.. varnish the surfaces and wrap the strip into and oblong shape. With straight sides. The varnish insulates the layers electrically reducing core loses as a result of eddycurrents.
After being we could cut the core in half across each leg in the middle. Then file the cut surfaces smooth such that they joint back together flush.
Wind a primary coil of a given number of ampturns and secondary coil of a ration you have decided of the first coil, it could be 1:1, 1:2, etc.
The coils should be just page enough to just slip over each led of your core. Then strap the two parts of the core back together tightly.
The key in a laminated core.
I still maintain that is is a core problem resulting is high Eddy currents.
I would go with s shell form type transformer where the core is oblong with straight sides. Simply use a long strip of plain sheet metal, not galvanised.. varnish the surfaces and wrap the strip into and oblong shape. With straight sides. The varnish insulates the layers electrically reducing core loses as a result of eddycurrents.
After being we could cut the core in half across each leg in the middle. Then file the cut surfaces smooth such that they joint back together flush.
Wind a primary coil of a given number of ampturns and secondary coil of a ration you have decided of the first coil, it could be 1:1, 1:2, etc.
The coils should be just page enough to just slip over each led of your core. Then strap the two parts of the core back together tightly.
The key in a laminated core.
junkhound
Senior Member
- Location
- Renton, WA
- Occupation
- EE, power electronics specialty
...file the cut surfaces smooth such that they joint back together flush...
If you are assuming the OP is a master tool and die maker to be able to file like that*, might as well tell OP how to mix Zn, Mn, and Fe powders and how to bake it to make ferrite, then he can run at 10s of kHz. <G>
* or has a platen belt sander, or ....
0r, leave the air gap there and make a flyback transformer and use a couple of IGBTs and PWM drive....., etc..
or, while we are going way out, tell OP to go to the farm supply store and get a big 50 pound coil of 19 gauge vinyl coated fence tension wire. Then it only takes threading a few dozen turns thru the spool to make a pretty decent 60 Hz transformer.
If you are assuming the OP is a master tool and die maker to be able to file like that*, might as well tell OP how to mix Zn, Mn, and Fe powders and how to bake it to make ferrite, then he can run at 10s of kHz. <G>
* or has a platen belt sander, or ....
0r, leave the air gap there and make a flyback transformer and use a couple of IGBTs and PWM drive....., etc..
or, while we are going way out, tell OP to go to the farm supply store and get a big 50 pound coil of 19 gauge vinyl coated fence tension wire. Then it only takes threading a few dozen turns thru the spool to make a pretty decent 60 Hz transformer.
Last edited:
Yes, it may be a stretch to assume that those in the err electrical trade has any knowledge in regard to the basics of electricity and the basic skills to be able to use basic tools. This is one reason why the OP transformer failed for the most part because of the core. He has the right idea regarding the windings but without an acceptable core design transformer coupling is slim to none. Even a basic relay, contactor, etc with an AC coil will have a laminated iron core. You would be more successful energizing an AC coill with DC that trying to energize a DC coil build with a solid core with AC.
- Location
- Massachusetts
Wow, everybody sucks except for you.
kwired
Electron manager
- Location
- NE Nebraska
Sometimes people blow instead of suck, just depends if they are inhaling or exhaling at that moment
Wonderful. Nice response.
My response was to that of junkhound who may have been sarcastic in his opening sentence but it did made sense though.
My previous assignment was with a custom dry type transformer. I have to apologize for my being amused with the replys to the OP's question when it is obviuou that his design lacks an acceptable core.
I have a couple custom 1500kva core and coil that was shipped to Australia for installation in a nitrogen infused totally enclosed no ventilated enclosure, 4 padmounted transformers ieach in 6'x6' enclosures at the Guam Intl AP, a transformer to power the chiller for the George Washington University auditorium, a custom designed encapsulated liquid cooled transformer with a 1/4" welded steel enclosuremounted on a underground rail car custom encapsulated transformer on an oil rig off the coast on East Africa, a transformer that powers the X- Band radar ft or Israel's mistake defense, a number a current limiting sir core reactors consisting of 3 individual 10'x10'x10' enclosures.
I also observed my winder disassembling a transformer to do an autopsy for a cause of failure.
These are just a few of the projects as I can recall where I had to work with my customer and my engineers to design and build a transformer for their project.
It may be that what I find to be obvious is not so obvious to others when it come to transformer basics.
There is an error that spell correct did its own thing on. It was suppose to be " Israel's missile defense ."
- Location
- Placerville, CA, USA
- Occupation
- Retired PV System Designer
And "sir core reactors"??
Thanks, I read right through that one even though I tried to order it word for word. The aforementioned was within a 3 year period. It was a most awesome opportunity and experience. With a phenomenal engineering staff and an excellent manufacturing facility it permitted me to turn opportunities into orders.
- Location
- Massachusetts
It is how it came across to me.
Last edited:
- Status