SOZ Choke input filter?

[Super_BQ]

Hello FETHead,

For a former I used an old ceramic sewer pipe with a couple of custom boards pieces on each side. Yes I actually HAND wound them which took me quite some time.

The enamel coated copper wire is the same stuff they used to build my transformer. The transformer shop here in Christchurch (NZ) is only a 2 min drive and they only charge by the weight (for ANY size guage wire). Including GST I think it's about $14/kg NZ$ (around $2.75 US / LB or $6 US/Kg). They will even let me borrow the whole roll - take as much off it and return it - then pay the difference.

BQ

[Variac]

Nelson Pass recommended the Erse 14 Ga Super Q coils
at Zalytron.com

They do have a metal core but
I think he said they didn't saturate when he was running over 6 amps DC through them. Since you need 2 for each channel (one per rail) that means 12amps I think.
Could someone confirm that they split the load?


That is a pretty good sized SOZ (about 15 watts/ch?)

A great advantage of them is the DCR of a 2 mh is .10
They also only cost 7.52 each.

Super BBQ:

How many feet did it take to make your monster 10 AWG coils?
I don't think my store will let me take the roll home

Thanks,
Mark

[Super_BQ]

Mark:

Using the following inductor calculator

http://www.colomar.com/cgi-bin/inductor_proc

I never bothered measuring. What I did was when I thought I had enough wire wrapped around, I would bring out my LCR meter and measure it's inductance. Then I would continue winding until I coil looked pretty (balanced) and until the right value was reached.

I'm pretty confident with the amount of heatsinks I used - I could do at least 40 watts x 2. But then with 25w x 2 right now - it's pretty loud on a pair of 90db 1w/1m efficient speakers.

BQ

[Nelson Pass]

Keep in mind that this thread was started with
common mode inductors, which only filter the
noise common to two lines. These can be quite
small but have high current ratings, but will not
filter non-common mode current, and most of what
you are going to want to filter is not common mode.

This type of coil is good for common mode RF suppression,
not audio range supply filtering.

[Variac]

Super BQ:

I don't know much about electronics, but it appears to me that the huge coils you made don't really solve some of the problems
that the SOZ tends to create.

I went to the web site you mentioned and they state that
to get a 3.18 Mh inductor with AWG 10, and you wind a diameter coil the size they recommend (smaller), you get a DCR of .22 ohms

Your huge coils have a DCR of .60 I guess because of the tremendous length of wire you used

I guess they cool better because they have more surface area
proportionally, but you are wasting a LOT of power with that much resistance. And creating a lot more heat that needs to be
dissipated.

A kindly person on this site gave me these Ohm's Law formulas:
If they are wrong please tell me:


You have a 25w/ch SOZ with 2 coils.
Lets use one channel data, and use the specs for one coil.
(although the two coils supply 2 channels)

One channel:

From the Pass charts re 25 watt SOZ:

600watts/38volts supply= 16 amps

.6 ohms DCR *16= 9.6 volts
So you lose 9.6 volts across the inductor?
9.6 volts*16 amps= 154 watts needed to dissipate

I think this is reallly conservative, because you will need to crank up the voltage to compensate for the losses.


If you use the recommended coil diameter:
.22 ohms*16= 3.52 volts
This way you only lose 3.52 volts
3.52* 16= 56 watts needed to dissipate.


If you get 3 Erse 14 ga. mag core 9 MH inductors in parallel:
.243 ohm DCR/3= .08 ohms DCR


.08 ohms*16= 1.28 volts
This way you only lose 1.28 volts
1.28* 16= 20.5 watts needed to dissipate.(7 watts apiece)

3 of the Erse inductors are $36

If you run the SOZ under 20watts, you could probably get by
using 2 1.5 mh in series for a cost of $15 'cause the smaller
values are cheaper.

Also Mr. Pass says that 2MH is adaquate
Of coures the theme for SOZ is : More is better!!! Buahahahaah!

Someone please correct me if these thoughts are wrong, because I am about to order the inductors


Thanks, Mark

[Super_BQ]

Hi Mark,

I know this thread is getting off topic...

"I don't know much about electronics, but it appears to me that the huge coils you made don't really solve some of the problems
that the SOZ tends to create. "

The large inductor certainly solved the problem I had - and that is to prevent overheating by reducing the amount of heat dissipation. Using a regular digital multimeter, the original MCM inductors had a HIGHER resistance value then the hand made ones. I don't claim that the digital MM is accurate (ie RMS rated) - but I did use the same meter for both coils and do know that the 2.0mH MCM ones dissipates more watts.

"I went to the web site you mentioned and they state that
to get a 3.18 Mh inductor with AWG 10, and you wind a diameter coil the size they recommend (smaller), you get a DCR of .22 ohms"

The diameter of the coil (inside diameter) is as close as to what the chart recommends (over 6"). Please keep in mind that the .22 ohm value is extremely difficult to achieve if you don't have perfect lay widings and right distance for each wind - so naturally the wire is going to be a LOT longer. A proper winding machine would do the job but I don't think it's feesable to make or buy one if you're only going to make a pair.

"If you get 3 Erse 14 ga. mag core 9 MH inductors in parallel:
.243 ohm DCR/3= .08 ohms DCR "

1st thing that comes to my mind on magnetic core inductors is thier saturation point (loss of inductance as the DC Bias current increases). Most places don't advertise these specs so I would rather stick with unsaturateable air core inductors. There is a whole area of inductors that will do circles around my mind (ie. core loss (Eddy currents), Leakage Effects, Magnetostriction) and one can find this info off the net.

With a 38V 1000VA transformer, you can easily get around 25 amps. Now the question is can 3 of the 9 mH steel core inductors handle that kind of current without being saturated? Even more important, how effective are these in filtering noise/ripple (the audiable noise) that air cores are known to filter out?

BQ

[Variac]

Howdy BQ

Well I agree that the 10 AWG ones you made are better then your 18 AWG originals, and have less DCR . My point was maybe making them so big is contraproductive, because the DCR goes way up. The site with the calculator does say that the inside radious should be 3", So I agree that you made them on the right core, but they come up with a DCR of .22

This discrepancy between their DCR and yours, is the difference
between me wanting to make a coil and not wanting to make a coil. I guess the perfect wound ones have some advantages. As I stated about the cored inductors: NP says they are good for over 6 amps ea. , so we do know that they work. That was part of my point.

I also pointed out that the 3 were for 25 watts/ch power output because you said it was silly to run the amp higher than that.
It's possible that you can't get max power out of your amp anyway due to the coil resistance in your power supply.

So we now have 3 options: machine wound, hand wound,
Erse mag cores. Pick the one that works for you!!

I am also very concerned about the saturation, because I am using the inductors right after the rectifier, to lower the voltage. If they saturate, I figure that the voltage will jump up to 1.2 times the transformer output. You can be sure that I will be gingerly turning up my Variacs to find that saturation point!!!

I hope to have my SOZ come out as well as yours when I'm done.
At the moment it is all attached to a 1x6 plank and is no candidate for the Pass Gallery!!
Mark

[Variac]

Here is a photo of inductors made by Northcreek.
They are a very respected maker of crossover componants
http://www.northcreekmusic.com/COILS.html
The photo on the right is an 8 AWG coil (!)
They don't seem to feel it is necessary to make a 6" core size.

I almost feel like the coil calculator people:
http://www.colomar.com/cgi-bin/inductor_proc
have made some sort of mistake to spec such a big core.
Those monster coils aren't the proportions I am used to seeing
for inductors. I'm wondering the if anyone knows what the
tradeoffs would be for a 3" DIAMETER core vs. a 3" RADIUS core.
It would sure as hell have a lot less DCR!

NELSON!!! Did I take your name in vain when I claimed you
said the Erse 14 AWG inductor would take 6 amps or more?
I did a search and can't find that statement anywhere.

[Super_BQ]

Actually there is a good reason for having a large donut. I can't remember the exact reason but it did come out of this Air Core inductor handbook I read last year (from the local library). The book mentions that a good air core inductor should proportional to the size of the donut hole. It is also important that the width & length of the windings itself (not the air core) is to be in a square proportional. By using a smaller donut hole would mean having more windings and the w x l wouldn't be optimized (possibly more likely to be saturated?).

I do know that in Solen coils, their donut holes for higher mH and larger guage wire values increases. Maybe NorthCreek is encomizing in size for shipping? For a 3mH 10awg coil I think their price is insane ($74.10 usd EACH!). Maybe they use some super high grade copper wire to generate such a low DCR value? But for power supply filtering, I would much rather use a higher voltage spec. transformer and who cares about the amount of watts gobbled up by a high DCR value inductor.

If it were for passive xovers in speakers - then that's a horse of a different colour.

[FEThead]

Um, Guys,

I hate to bring you back to reality, but in the original post gnomus asked:



My question is how many amps can I expect from a 30 Volt DC supplied SOZ?

Has anyone tried a Choke input SOZ? Or can share the general difference one
can expect betwween choke input VS PI?

Back to the present train of thot, in-duck-dance vs. hole size: I can not find the formula for donut inductors, but for solenoid coils, the inductance is roughly (diameter^2 * number of turns^2)/(18*diameter + 40*the length) (from a radio handbook) So diameter squared is in the numerator and diameter is in the denominator. And the length of the wire is the circumfrance which is diameter * pi. So it looks like the bigger the hole the better.

I have not seen all of these iron core chokes you guys have been discussing, but if a choke is wound around a bar of iron or iron plates it is unlikely to saturate because most of the magnetic path is thru the air. Too much current in these will usually just cause a foul oder in the room.

So, for gnomus, how many amps can you expect from a 30 volt SOZ?