Winding my own 1H and 100mh inductor

[Spencerleehorton]

Hi Guys,

im going to build an API-553 eq and wanted to wind my own 1H and 100mH inductors.
I will also wind my own 2503 transformer but i already have that one down.
could anyone help me with calculations etc, i have several different sized bobbins and lams but would like to keep these fairly small if possible like the originals.

regards

Spence.

 

[abbey road d enfer]

Hi Guys,

im going to build an API-553 eq and wanted to wind my own 1H and 100mH inductors.
I will also wind my own 2503 transformer but i already have that one down.
could anyone help me with calculations etc, i have several different sized bobbins and lams but would like to keep these fairly small if possible like the originals.

regards

Spence.

Are you sure you want to use iron core?
It is more common using ferrite cores for such values.
With ferrite, the calculation is quite easy:
L=Al. N², where Al is a parameter specified by the manufacturer and N the number of turns
Since Al is expressed in nH, the result is also in nH.
The best compromise is usually Al630 because the higher Al types have a very small gap, resulting in wider tolerance and lesser saturation characteristics.
1260 turns will result in 1H. You need to verify that those 1260 turns will fit in teh bobbin. All that is pretty well explained in the EPCOS-TDK and Murata websites.

For iron cores, it's a little more difficult because the Al is not known, and you must use a formula that is likely to cause errors because of the many zeros after the comma, so you'd better use a calculator such as this one
http://thoughtmountain.com/inductor_calculator.html
Just watch the zeros!
Beware this calc is for non-gapped cores. If you run DC in the inductors, you need a gapped core; then the calculaion is evn more error prone...

 

[PRR]

If you know a sample core-size and wire size, just start winding. Wind 1/10th the window area, and measure for 1/100th of the desired H. If that is not looking good, find a thinner wire and try again.

 

[Spencerleehorton]

bobbin

http://www.ebay.co.uk/itm/Plastic-Coil-Former-Bobbin-8-Pin-PQ2620-Ferrite-Cores-for-Inductor-/191685060703?hash=item2ca151e85f:g:kSAAAOSw4shX-7-Q

core

http://www.ebay.co.uk/itm/6-pairs-NEOSID-RM8-L2-AL630-ferrite-core-pair-23mm-x-8mm-deep-pot-/391624941723?hash=item5b2eaa609b:gSoAAOSwB09YLH2D

 

[abbey road d enfer]

bobbin

http://www.ebay.co.uk/itm/Plastic-Coil-Former-Bobbin-8-Pin-PQ2620-Ferrite-Cores-for-Inductor-/191685060703?hash=item2ca151e85f:g:kSAAAOSw4shX-7-Q

core

http://www.ebay.co.uk/itm/6-pairs-NEOSID-RM8-L2-AL630-ferrite-core-pair-23mm-x-8mm-deep-pot-/391624941723?hash=item5b2eaa609b:gSoAAOSwB09YLH2D

It is announced as Al630, so you need 1260 turns for 1H. Cross-section is about 28mm², which indicates a maximum square-section of wire of 0.022 mm² resulting in a diameter of 0.14mm.
Mean Length Turn is 42mm. Closest gauge is 0.125mm (AWG36), resulting in a DC resistance of about 75 ohms.
You would need only 400 turns for the 100mH, for a DCR of about 25 ohms.
Beware that adjusting the final value needs to be done in-circuit, checking the actual circuit's response. Most inexpensive LC-meters give very inaccurate values. If you need super-accuracy, you should wind a few percent more turns than the calculation and adjust for teh final value. Also beware that properly setting the two halves of the core is important; there could be as much as 5% variation between an not-too-well seated core and one that's properly set. You have to rub them against each other.

 

[Spencerleehorton]

How close will this get me?

http://www.ebay.co.uk/itm/10x-RM8-Pot-Core-Ferrite-Bobbin-Philips-Epcos-Coils-/361200418349?hash=item541938fe2d%3Ag%3AgIQAAOSwWKtUyQKT&clk_rvr_id=1143336834368&rmvSB=true&ul_ref=http%253A%252F%252Frover.ebay.com%252Frover%252F1%252F710-53481-19255-0%252F1%253Ftype%253D4%2526campId%253D5336876652%2526toolId%253D10001%2526customId%253Dix01oxw3d1000n0n00003%2526mpre%253Dhttp%25253A%25252F%25252Fwww.ebay.co.uk%25252Fitm%25252F10x-RM8-Pot-Core-Ferrite-Bobbin-Philips-Epcos-Coils-%25252F361200418349%25253Fhash%25253Ditem541938fe2d%25253Ag%25253AgIQAAOSwWKtUyQKT%2526srcrot%253D710-53481-19255-0%2526rvr_id%253D1143336834368&ul_noapp=true

I have some 0.120mm.

 

[Krcwell]

I made a few 553's using rm8 cores; AI400 for the 100mH, AI630 for the 1H. I'm pretty sure I used 38 awg wire for the 1H, can't remember off the top of my head if 36awg will fit on bobbin. User bruce0 posted a really handy google docs spreadsheet to assist with calculations a couple years back, it will calculate everything for you, and gives estimates on max # of turns per awg for a variety of core sizes. Fantastic spreadsheet. Bruce0, if you read this, thank you very much!

https://groupdiy.com/index.php?topic=55373.msg707217#msg707217q

 

[abbey road d enfer]

How close will this get me?

http://www.ebay.co.uk/itm/10x-RM8-Pot-Core-Ferrite-Bobbin-Philips-Epcos-Coils-/361200418349?hash=item541938fe2d%3Ag%3AgIQAAOSwWKtUyQKT&clk_rvr_id=1143336834368&rmvSB=true&ul_ref=http%253A%252F%252Frover.ebay.com%252Frover%252F1%252F710-53481-19255-0%252F1%253Ftype%253D4%2526campId%253D5336876652%2526toolId%253D10001%2526customId%253Dix01oxw3d1000n0n00003%2526mpre%253Dhttp%25253A%25252F%25252Fwww.ebay.co.uk%25252Fitm%25252F10x-RM8-Pot-Core-Ferrite-Bobbin-Philips-Epcos-Coils-%25252F361200418349%25253Fhash%25253Ditem541938fe2d%25253Ag%25253AgIQAAOSwWKtUyQKT%2526srcrot%253D710-53481-19255-0%2526rvr_id%253D1143336834368&ul_noapp=true

I have some 0.120mm.

These are ungapped Al2900 using N26 material, that I have not seen used in audio applications. These cores would require about 4.6 times less turns. I can only vouch for N48 material.

 

[protosardo]

https://groupdiy.com/index.php?topic=47526.20
http://powermagnetics.co.uk/calculator

 

[protosardo]

http://www.mouser.it/EPCOS-TDK/Passive-Components/EMI-Filters-EMI-Suppression/EMI-RFI-Suppressors-Ferrites/Ferrite-Cores-Accessories/RM-Core/_/N-bw7tb?P=1yzvhh7Z1yhy7dwZ1ytdx60Z1yxgwjo

 

[Spencerleehorton]

this would do for the 100mh right?

http://www.ebay.co.uk/itm/390565303198?_trksid=p2060353.m1438.l2649&ssPageName=STRK%3AMEBIDX%3AIT

 

[Spencerleehorton]

so if i use those cores and bobbins off ebay its 1260 turns  / 4.6 = 274 turns for the 1H wound with 0.120mm wire?

then i can just buy that other 100mH version, just need to check it will be ok

 

[abbey road d enfer]

this would do for the 100mh right?

http://www.ebay.co.uk/itm/390565303198?_trksid=p2060353.m1438.l2649&ssPageName=STRK%3AMEBIDX%3AIT

These are line chokes, for which the optimization compromises are very different than for audio filters. The ferrite material may not be the best. Also, the tolerance is quite large -30/+50%. It may more or less work, but it will never be as good as an RM core of N48 material.

 

[Spencerleehorton]

ah ok, i will stick with RM8 N48 then!! thats kind of what i wanted to know, nice one.

I have a contact at a transformer company locally so i will get some off them!! i teach guitar to one of the persons grandsons and he is very helpful.

 

[Spencerleehorton]

Looks like I might have some RM10 coming my way, will this be a problem?

 

[abbey road d enfer]

Looks like I might have some RM10 coming my way, will this be a problem?

No problem, the larger the better. But you don't say what AL...

 

[Spencerleehorton]

once i get them i will let you know details,  im still in the process of making the API 553 pcbs, so hopefully over the next few days i should have everything together and start building.