Still trying to understand transformers

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Well, I don't know the specifics of what you are working with, as regards, the core dimension, type of steel, winding geometry and number of turns. Also, what type of VA rating are you trying to achieve? If you increase the primary voltage with a fixed number of primary turns eventually you will, at some point, drive the core into saturation, and magnetizing current will get real big, real fast. The knee may not be real pronounced. The lower the primary turns, the greater the core flux needs to be at idle and the magnetizing current will be larger than would you could get away with with a primary of more turns, and this drives up the copper losses. Sometimes cheap, buzz box welding transformers use this build, as it allows a greater secondary current, but for a shorter time, a short duty cycle results, but is is tolerated in a home welding product. If you see the magnetizing current climb and the core gets hot, you are approaching saturation, or are lightly saturated already. The solutions would include a greater core, or more primary turns, and you must be mindful of the primary wire gauge, to keep IR losses in check. IF you need a larger primary wire gauge, well then you need a larger window area, and larger core dimensions to keep reluctance low. Kinda like flying a helicopter, if you change one thing, everything else changes............
Tom
 
this is why it makes absolutely no sense to run the core at anything less than just below the "knee" in the flux curve/

just below the knee is about right for good regulation.....
but designers can choose to make their traffos at any point just below the knee....
after all any point below the knee is playable...

How do we determine from the transformer PriI vs PriV curve, what optimum voltage we should run at

my primary line voltage varies from 215 to 240 depending on the time of day...
therefore i design my traffos to run at 230 volts and compute primary turns based on 230 volts...
of course having more primary turns means less no load current and vice versa...
but there comes a situation wherein in the actual wind-up, and in the last layer,
i find that i will only use half of the layer...so what do i do? do i continue and fill up the layer?
or stop at half the layer? or just delete the half and go for lesser turns?

i choose to fill up the layer most of the time and adjust the secondary turns count,
my reason being, the added turns lowers idle primary current a bit....

the power transformer i built here is a 3 inch stack of 1.5 inch center leg laminations,
after 45 minutes of playing it the power traffo warms up a little bit, very comfortable to touch....

IMG_4898_zps5dea9f91.jpg
 
Andrew,
Not enough info for a numerical answer. That seems to be what you are after, I gave a generalized answer, not a specific one to you design question. All transformer design data has been around for years and off the shelf software is available if you are designing a specific transformer for a specific product. Run enough turns to get just below the flux knee as someone suggested.
Tom
 
Andrew,
You can proceed with empirical data if you know enough about the device under question, however it might require lots of testing and reiteration to come up with a satisfactory answer. I can tell you, most power transformers are operated close to the knee, EXACTLY where is not exactly a corporate secret, so I am not able to answer your question.
Tom
 
I test many of my transfoermers for no load primary current off the Variac.

The Curve does not have a pronounced knee. It just curves over nearly becoming horizontal @ 260Vac.
The increase in primary current from 244Vac to 254Vac can approach 2times.
that makes a transformer run very hot.

ok, that doubling of primary current is the sharp knee that i spoke of.
but if you measure watts with an analog watt meter, its even worse.

The knee's sharpness also depends on the impedance of your variac--lower is better. air gaps in the transformer core also wash the knee out
if your variac weighs less than the transformer you are testing, don't expect accurate info because its the third and higher harmonics that matter.

especially if you're winding a lightweight test coil to find out where you want the iron loss to be at, you have to account for the larger primary to have lower resistance losses.

also, if you subtract from your voltage graph, the resistive dc loss in the primary coil then your knee will sharpen up quite a bit.
better yet, measure the secondary voltage and primary current.. not primary voltage and primary current. the knee produced will be quite a bit sharper because you'll also be excluding primary leakage inductance.
 
Last edited:
ok, that doubling of primary current is the sharp knee that i spoke of.
but if you measure watts with an analog watt meter, its even worse.

The knee's sharpness also depends on the impedance of your variac--lower is better. air gaps in the transformer core also wash the knee out
if your variac weighs less than the transformer you are testing, don't expect accurate info because its the third and higher harmonics that matter.

especially if you're winding a lightweight test coil to find out where you want the iron loss to be at, you have to account for the larger primary to have lower resistance losses.

also, if you subtract from your voltage graph, the resistive dc loss in the primary coil then your knee will sharpen up quite a bit.
better yet, measure the secondary voltage and primary current.. not primary voltage and primary current. the knee produced will be quite a bit sharper because you'll also be excluding primary leakage inductance.
I will compare the priV vs PriI curve to the SecV vs PriI, since I have not measured nor plotted that.
See if it sharpens up the knee in the plot.

Up to now all the tests I have carried out show the expected S curve shown in the reference books.
I have NEVER seen a pronounced "knee" in the curve and thus no definable optimum Primary Voltage.
 
imho, heat is the single biggest consideration with transformers...

how hot can you tolerate your traffos in your amps....?
if you know the answer to this question, you will know what to do....

Yes, I was surprised at the post I saw in another Thread that transformer regulation gets worse as core flux is reduced.

yes this is true, lower flux means more turns which leads to higher losses....
really a compromise, which one can you live with? lower operating temperature or better regulation?
in the end it is all a matter of choice....and diy'ers are free to make choices....

So the predominate limiter in output power is the core size because it limits flux?

yes, as per RDH4, VA= [Ax5.58]^2, where A is cross section area of core in square inches, VA is volt amperes capacity
core material such as M6 can be run at higher flux densities, the M50 at much lower densities...
 
I will compare the priV vs PriI curve to the SecV vs PriI, since I have not measured nor plotted that.
See if it sharpens up the knee in the plot.

Up to now all the tests I have carried out show the expected S curve shown in the reference books.
I have NEVER seen a pronounced "knee" in the curve and thus no definable optimum Primary Voltage.

S curve?
this isn't an S curve, its a hockey stick showing the optimal voltage is 103 volts.
--saturation started at 90 volts.
http://johansense.com/bulk/motloss.png
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.