This is Electrovalue, in the UK, listing ETD49 cores at 1.73 BP, and bobbins for it at 75 pence.
http://www.electrovalue.co.uk/pdf/FER.PDF
In a bind, I guess it might do.
Dexter Magnetics , will cost you less, and they have all cores and bobbins at decent prices, but minimum order is $100.
Ultimately, any core shape will do; the only glamour about ETD cores is their special geometry which is optimized for the most power with the least amount of copper.
Since you may not know the grade of ferrite in your toroids, you can go trial-and-error: You test the prototype under reduced supply voltage and well fused. As you increase the voltage, monitor the Xformer temperature. If the core gets too hot to touch, there may be too much core loss. (wrong grade, wrong frequency). If the copper or the transistors get too hot. the core may be saturating (freq too low for ferrite grade) If nothing heats up too much, and power output is low, frequecy could be too high for the ferrite grade.
Since this experimentation is often accompanied by smoke, I try to avoid unknown material grades for SMPS power transformers. I figure 6-8 hours of experimentation, accompanied by cursing, smoke, injury, and some semiconductor losses, are a higher cost than 6 bucks paid for a known good material.
IMHO, the only unknown ferrite safe for an SMPS power transformer is a ferrite that comes from an SMPS power transformer.
http://www.electrovalue.co.uk/pdf/FER.PDF
In a bind, I guess it might do.
Dexter Magnetics , will cost you less, and they have all cores and bobbins at decent prices, but minimum order is $100.
Ultimately, any core shape will do; the only glamour about ETD cores is their special geometry which is optimized for the most power with the least amount of copper.
Since you may not know the grade of ferrite in your toroids, you can go trial-and-error: You test the prototype under reduced supply voltage and well fused. As you increase the voltage, monitor the Xformer temperature. If the core gets too hot to touch, there may be too much core loss. (wrong grade, wrong frequency). If the copper or the transistors get too hot. the core may be saturating (freq too low for ferrite grade) If nothing heats up too much, and power output is low, frequecy could be too high for the ferrite grade.
Since this experimentation is often accompanied by smoke, I try to avoid unknown material grades for SMPS power transformers. I figure 6-8 hours of experimentation, accompanied by cursing, smoke, injury, and some semiconductor losses, are a higher cost than 6 bucks paid for a known good material.
IMHO, the only unknown ferrite safe for an SMPS power transformer is a ferrite that comes from an SMPS power transformer.
Ferrite and Powdered-Iron Toroids
I have several Ferrite Toroids, and there is not much of a way to tell the different FERRITE materials apart from each other, but you CAN tell the difference between a Ferrite and Powdered-Iron (PI) core. Most manufacturers, (Amidon, Fair-Rite, Ferroxcube and others) use a standard color code for their PI Toroids.
For example, Amidon Coproration's Jan 2000 Catalog shows the color code for their P-I Toroids:
Material Color
0 Tan
1 Blue
2 Red
3 Grey
6 Yellow
7 White
10 Black
12 Green/White
15 Red/White
17 Blue Yellow
26 Yellow/White.
The BIG toroid found in most AT and ATX SMPSs is most often the Yellow/White #26 P-I material. 26's magnetic properties are in the catalog.
For the Ferrites, I am not aware of any color code that exists for the different types of materials. However, Amidon's catalog has all the charts and graphs of all the ferrite materials' magnetic properties.
The material most often used iin the main power Xfmr in half-bridge AT/ATX boxes is the #77 mix, with a u=2000. I can't speak to the flyback types.
For a very cut-n-dried process on how to figure the number if turns for a specific application, check out George Chryssis' book "High Frequency Switching Power Supplies", (c) 1989 edition. I can't remember the chapter number, but the chapter on transformers gives a very good procedure for both forward transformers and flyback transformer-chokes. I have used this design procedure many times winding toroids, with great success.
For winding P-I Toroids, I use the A(L) value charts to figure the number of turns for a desired inductance.
Check out www.amidoncorp.com and ask for their latest catalog or spec sheet.
Hope this helps,
Steve
I have several Ferrite Toroids, and there is not much of a way to tell the different FERRITE materials apart from each other, but you CAN tell the difference between a Ferrite and Powdered-Iron (PI) core. Most manufacturers, (Amidon, Fair-Rite, Ferroxcube and others) use a standard color code for their PI Toroids.
For example, Amidon Coproration's Jan 2000 Catalog shows the color code for their P-I Toroids:
Material Color
0 Tan
1 Blue
2 Red
3 Grey
6 Yellow
7 White
10 Black
12 Green/White
15 Red/White
17 Blue Yellow
26 Yellow/White.
The BIG toroid found in most AT and ATX SMPSs is most often the Yellow/White #26 P-I material. 26's magnetic properties are in the catalog.
For the Ferrites, I am not aware of any color code that exists for the different types of materials. However, Amidon's catalog has all the charts and graphs of all the ferrite materials' magnetic properties.
The material most often used iin the main power Xfmr in half-bridge AT/ATX boxes is the #77 mix, with a u=2000. I can't speak to the flyback types.
For a very cut-n-dried process on how to figure the number if turns for a specific application, check out George Chryssis' book "High Frequency Switching Power Supplies", (c) 1989 edition. I can't remember the chapter number, but the chapter on transformers gives a very good procedure for both forward transformers and flyback transformer-chokes. I have used this design procedure many times winding toroids, with great success.
For winding P-I Toroids, I use the A(L) value charts to figure the number of turns for a desired inductance.
Check out www.amidoncorp.com and ask for their latest catalog or spec sheet.
Hope this helps,
Steve
In order to check for core material properties, build a simple current-limited flyback converter that allows to regulate duty cycle and frequency with a potentiometer. Wrap several turns on the core as a primary winding and use a very low value non-inductive resistor as a current sense element. Power the whole thing from an adjustable low voltage current limited power supply.
The primary current versus time waveform will tell a lot of useful information. Iron powder will show very low permeabilities so primary current will skyrocket if not enough primary turns are used. Power ferrites will show moderate permeabilities and progressive saturation. Signal ferrites will show high permeabilities and very sharp saturation. Saturable reactors will look a lot like signal ferrites but they will need an adjustable DC current source (in opposite direction to switch current) in order to get biased and to adjust saturation time.
For iron powder cores (and gapped ferrites), inductance may be calculated by dividing the applied voltage by the observed current slope in amps/second . Current versus time waveform will also show saturation current level. Inductance and saturation current may be extrapolated for other turn counts with the following formulas:
L2 = L1 * N2^2 / N1^2
Is2 = Is1 * N2 / N1
(Note that this gets less and less accurate for lower permeability materials and longer air gaps due to leakage inductance).
For ferrite cores (ungapped), the time the core takes to saturate for a given amount of turns and voltage may be easily extrapolated for different conditions, allowing to choose optimum turn counts for unknown materials. Time will be proportional to turns and inversely proportional to voltage. Note that saturation time will be doubled when bipolar excitation is employed (push-pull converters.
I use this empirical method even when I actually have material specifications. I feel it's better to actually check transformer and inductor properties before using them instead of just relying on equations and material parameters.
The primary current versus time waveform will tell a lot of useful information. Iron powder will show very low permeabilities so primary current will skyrocket if not enough primary turns are used. Power ferrites will show moderate permeabilities and progressive saturation. Signal ferrites will show high permeabilities and very sharp saturation. Saturable reactors will look a lot like signal ferrites but they will need an adjustable DC current source (in opposite direction to switch current) in order to get biased and to adjust saturation time.
For iron powder cores (and gapped ferrites), inductance may be calculated by dividing the applied voltage by the observed current slope in amps/second . Current versus time waveform will also show saturation current level. Inductance and saturation current may be extrapolated for other turn counts with the following formulas:
L2 = L1 * N2^2 / N1^2
Is2 = Is1 * N2 / N1
(Note that this gets less and less accurate for lower permeability materials and longer air gaps due to leakage inductance).
For ferrite cores (ungapped), the time the core takes to saturate for a given amount of turns and voltage may be easily extrapolated for different conditions, allowing to choose optimum turn counts for unknown materials. Time will be proportional to turns and inversely proportional to voltage. Note that saturation time will be doubled when bipolar excitation is employed (push-pull converters.
I use this empirical method even when I actually have material specifications. I feel it's better to actually check transformer and inductor properties before using them instead of just relying on equations and material parameters.
I have a question for you SMPS experts here. I did not want to start a new thread, and I figured that it is a beginner's question, so here goes:
On the homepage of amp manufacturer LAB.Gruppen they say the following about the SMPS in their (now discontinued) amplifier model LAB1300C (a 2x600W class AB amp):
'' In contrast to many of our competitors, who use so-called push-pull converters and other designs, which require current limiting, we have designed the LAB 1300 to obtain the same caracteristics as a conventional power supply. It works by of injecting a magnetic pulse from the AC-line during a pulse time segment, which is separated from the output charge current pulse. Because of this no current limiting is needed for the secondary side of the transformer, and high peak power and good bass response are guaranteed. ''
What kind of circuit do you think they are talking about here? In contrast to QSC, there are no schematics available on the homepage of this company, but the corresponding user manual gives a clue that sense windings are used to achieve the above operation. Are schematics for a similar smps available somewhere?
On the homepage of amp manufacturer LAB.Gruppen they say the following about the SMPS in their (now discontinued) amplifier model LAB1300C (a 2x600W class AB amp):
'' In contrast to many of our competitors, who use so-called push-pull converters and other designs, which require current limiting, we have designed the LAB 1300 to obtain the same caracteristics as a conventional power supply. It works by of injecting a magnetic pulse from the AC-line during a pulse time segment, which is separated from the output charge current pulse. Because of this no current limiting is needed for the secondary side of the transformer, and high peak power and good bass response are guaranteed. ''
What kind of circuit do you think they are talking about here? In contrast to QSC, there are no schematics available on the homepage of this company, but the corresponding user manual gives a clue that sense windings are used to achieve the above operation. Are schematics for a similar smps available somewhere?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.