Shielded or Toroidal Inductors for Class D output filter - Page 2 - diyAudio
Go Back   Home > Forums > Amplifiers > Class D

Class D Switching Power Amplifiers and Power D/A conversion

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 2nd February 2009, 07:32 AM   #11
Eva is offline Eva  Spain
diyAudio Member
 
Eva's Avatar
 
Join Date: Oct 2003
Location: Near the sea
Send a message via MSN to Eva
Not all powder cores exhibit high losses, only the higher permeability low-cost ones do. MPP (Molybdenum Permalloy Powder) achieves reasonably high permeabilities and moderate losses (26u, 60u, 125u are best). Micrometals RF materials like "2" exhibit very low losses at the expense of low permeabilities. Several other Micrometals materials are a compromise between low permeability and low losses, but "2" is the lowest cost option.

Gapped ferrites have two main drawbacks: If the gap is covered by the windings, winding losses in front of the gap become very high due to the stray magnetic field from the gap inducing Eddy currents in the wire/foil. This is partially solved by using Litz wire (expensive and hard to source). If the gap is not covered by the windings, an undesirable strong stray magnetic field is produced because there is nothing acting as a shield in front of the gap.

I don't recommend gapped ferrites, at least not for compact high power designs. On the other hand, they are great for high current "pi" filter inductors where flux is not high and leakage is not a problem.
__________________
I use to feel like the small child in The Emperor's New Clothes tale
  Reply With Quote
Old 2nd February 2009, 09:27 AM   #12
diyAudio Member
 
Ouroboros's Avatar
 
Join Date: Dec 2003
Location: Nottingham UK
Anyone looking for an off-the-shelf shielded inductor could look at the new SER2918H range from Coilcraft. This is surface-mountable and uses copper strip as the winding material.

Coilcraft (in the UK at least) will supply small quantities to hobbyists. Pricing is shown for most parts on their website.
  Reply With Quote
Old 5th February 2009, 07:49 AM   #13
diyAudio Member
 
soundcheck's Avatar
 
Join Date: Mar 2005
Location: D
Quote:
Originally posted by Ouroboros
Anyone looking for an off-the-shelf shielded inductor could look at the new SER2918H range from Coilcraft. This is surface-mountable and uses copper strip as the winding material.

Coilcraft (in the UK at least) will supply small quantities to hobbyists. Pricing is shown for most parts on their website.

Interesting. Extreme low DCR. Do you have any experience with these?


I also stepped over this:

Ferroxcube Class D OF


Hmmmh: By looking at the filter calculations in above PDF I realized that I need not only to change the C in the filter if working with higher loads - as it is recommended in most of the e.g. Tripath datasheets.
For e.g. going from 4R to 8R I'd also need to double the size of the inductor according to above.

Is this correct?? How about 16R? And what is actually the result - soundwise - if working with the wrong filter size?

Cheers
  Reply With Quote
Old 5th February 2009, 08:27 AM   #14
diyAudio Member
 
soundcheck's Avatar
 
Join Date: Mar 2005
Location: D
One more:

Also interesting to see in above pdf, that on low power (<15W) applications as delivered by e.g. by T2020/T2021B the differences (obviously) and losses are quite small.

Which brings back the question if the air-core inductor for these low power amps wouldn't be the best choice.

And another obvious conclusion would be, that (small) digi-amps work best on high efficiency speakers due to lowest distortions/losses.


Cheers
  Reply With Quote
Old 5th February 2009, 09:06 AM   #15
diyAudio Member
 
Ouroboros's Avatar
 
Join Date: Dec 2003
Location: Nottingham UK
I have four samples of the SER2918 in a 22uH size. I was looking at using these in a bridged 400W class-D (UCD style) amplifier to drive a 100V step-up transformer, but as they are very expensive, I'm going to use custom-wound inductors on RM12 cores instead.

The SER2918 range use cores of 3C92 ferrite, which is pretty good for use at class-D oscillation frequencies.
  Reply With Quote
Old 5th February 2009, 09:32 AM   #16
diyAudio Member
 
soundcheck's Avatar
 
Join Date: Mar 2005
Location: D
Quote:
Originally posted by Ouroboros
I have four samples of the SER2918 in a 22uH size. I was looking at using these in a bridged 400W class-D (UCD style) amplifier to drive a 100V step-up transformer, but as they are very expensive, I'm going to use custom-wound inductors on RM12 cores instead.

The SER2918 range use cores of 3C92 ferrite, which is pretty good for use at class-D oscillation frequencies.

What do you call very expensive?


I'd appreciate if you could comment on the dimensioning thing I brought up?
Would be interesting to know what inductance I'd need for a 8R/16R speaker, since Tripath datasheets just recommend to change the C value of the filter.

THX
  Reply With Quote
Old 5th February 2009, 09:50 AM   #17
diyAudio Member
 
Ouroboros's Avatar
 
Join Date: Dec 2003
Location: Nottingham UK
The on-line purchasing facility on their website shows the 1-off price at $5.55. The first price break is at 250-off, when the price drops to $4.63 each.

It is true that to get an optimum match between the filter and the load, the filter component values need to be changed to suit. There are a number of free passive filter design programs on the web which will design a second-order Butterworth filter for you if you specify the source impedance (low), the load impedance, and the desired -3dB frequency. If you have access to a SPICE program (such as the simple 'TINA' available free on the TI website) or the fully featured LT spice available free on the Linear Tech website, then you can see the effect on the frequency and phase response when alering the load impedance.
( The free version of TINA is severely reduced in types of components, functional blocks and analysis features compared with the bought version).

I only use UCD-style amplifiers with post-filter feedback, which greatly reduces the effect of the filter termination load on the frequency response.
  Reply With Quote
Old 5th February 2009, 10:02 AM   #18
diyAudio Member
 
soundcheck's Avatar
 
Join Date: Mar 2005
Location: D
THX.

One more quick question: What is the theoretical impact if the filter is not correctly designed. E.g. running a 4R config on a 16R speaker.
  Reply With Quote
Old 5th February 2009, 10:27 AM   #19
diyAudio Member
 
Ouroboros's Avatar
 
Join Date: Dec 2003
Location: Nottingham UK
A 27uH with a 470nF into 8 Ohms looks like this.
Attached Images
File Type: jpg 8r.jpg (25.6 KB, 635 views)
  Reply With Quote
Old 5th February 2009, 10:28 AM   #20
diyAudio Member
 
Ouroboros's Avatar
 
Join Date: Dec 2003
Location: Nottingham UK
A 4Ohm load gives this.
Attached Images
File Type: jpg 4r.jpg (23.9 KB, 606 views)
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Shielded inductors - where? Glenn2 Parts 6 29th October 2007 08:07 PM
Class T Output Inductors : Litz or Solid Core? Peteruk Class D 2 8th April 2007 01:03 AM
Dual differential mode inductors in a bridged output lowpass filter BWRX Class D 3 17th November 2005 03:39 PM
pi filter in class a amp: okay to parallel inductors? jwb Solid State 6 20th October 2002 07:56 PM


New To Site? Need Help?

All times are GMT. The time now is 07:36 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2