Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

L here, L there, L Everywhere *-*
L here, L there, L Everywhere *-*
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
Thread Tools Search this Thread
Old 24th March 2011, 09:30 PM   #11
T101 is offline T101  Bulgaria
diyAudio Member
Join Date: Mar 2010
Location: Sofia-Bulgaria
With different values of the primary and secondary windings it gets even more interesting. You can adjust the slope, depth of the first level region and the depth of the second step down.
Different bypass and loading resistor values lead to many opportunities along with the possibility to avoid the bypass resistor.

Too much examples and we (at least I) still don't have the general rule or principle. That makes posting more examples useless until someone gives some kind of theory.

Only thing that i found out is that with many attempts the output of this thing can be quite tailored.

I'll do some real experiments on second and third order crossovers in the loading of the secondary winding. - This will be done if a common ground can be used in order to accommodate the elements in parallel.

Best regards to all!
  Reply With Quote
Old 26th March 2011, 06:15 AM   #12
dave slagle is offline dave slagle  United States
diyAudio Member
Join Date: Mar 2003
Location: NYC
Transformers are just complex bandpass filters. If you make the coupling factor in your sim 1 rather than .99 you will see the second corner disappear. In simplest terms for an ideal model, the K of .99 simply puts a small percent of the total inductance in series with the primary which represents leakage inductance.

your version of the transformer can also be modeled with two series inductors and give you the same result. Of course the non-ideal nature of transformers could be used to your advantage but that really opens up a hit or miss situation unless you are able to know and spec the exact transformer you need.

It seems if someone wanted to follow this approach it would be best attacked with simulations and known easily measurable parameters which the series non-coupled inductors provide. As an extra plus, you could tap the inductors to allow for ease of tuning.

  Reply With Quote
Old 13th April 2011, 03:22 PM   #13
T101 is offline T101  Bulgaria
diyAudio Member
Join Date: Mar 2010
Location: Sofia-Bulgaria
You are right Dave. I tried that and the smaller is the coupling the bigger is the second step no matter of the loading.

For instance under 0.9 the loading resistor only alters the slope of the first step, but almost does not alter the depth of the correction.

Anyway, thanks for the input. But I have to say that i don't feel that I am the right person to investigate this kind of circuits and to share an all situations valid conclusions.

Never the less I progress Today I got two sets of laminated E-transformer cores with 36x36 mm central cross sections. The cores are handled to the engineers who will make the transformers and will measure their characteristics, which i will share.

The finished correction circuits will be incorporated in the project in my signature (where I could use and will appreciate any help on the bass enclosure ) and I will share further data and listening impressions.

Here-below I apply a photo of the material for the cores -and choke coil.

Best Regards to all!

EDIT (forgot to ask) What coupling coefficient should be expected from this type of transformer and core? The segments are laminated (I read that in the Wikipedia, that the segments should be electrically separate in order to keep the losses and etc. low)
Attached Images
File Type: jpg transformercores.JPG (64.8 KB, 141 views)

Last edited by T101; 13th April 2011 at 03:39 PM.
  Reply With Quote
Old 13th April 2011, 04:54 PM   #14
thoriated is offline thoriated  United States
diyAudio Member
Join Date: Nov 2003
Location: flyover country
I used this type of approach with an air core transformer which I created by winding a 20awg secondary with about a 4:1 turns ratio on a large 12awg inductor form with about a 12mH primary winding. This allowed me to do some relatively narrowband attenuation EQ of the Iron Lawbreakers in the mutual coupling region of the air core transformer which was centered at about 150hz with only an additional resistor and film capacitor.

This worked out very well since the air core transformer doubled as a low pass series inductor above the mutual coupling range and allowed optimization of the frequency response at the same time greatly flattening the impedance throughout the bass to lower midband range with minimum parts count and low signal loss. In fact, this approach, in addition to special internal speaker damping I used that included adsorbents virtually eliminated the upper BR impedance peak (18 ohms +/- 20% from 40 hz ~ 250hz without any 'Zobel' networks) which I believe results in significantly clearer sounding bass. The 'downside' was that the whole development process was very empirical - there were few if any 'guidelines' to work from for using an air core transformer as part of a speaker xover.

Last edited by thoriated; 13th April 2011 at 05:12 PM.
  Reply With Quote
Old 15th April 2011, 05:05 PM   #15
T101 is offline T101  Bulgaria
diyAudio Member
Join Date: Mar 2010
Location: Sofia-Bulgaria
Transformers ready!

Primary winding 120 turns of 1.2 mm wire;
Secondary winding 120 turns of 0.8 mm wire;
Resistance of primary - 0.33 ohms;
Resistance of secondary - 0.79 ohms;
Inductance of both windings in final variant with 1.5 mm spacer - 10 mh;

Inductance with fully closed core (E+I) 31.7 mh;
Inductance with only E core - 4 mh;
inductance with full core (E+I), but with 1.5 mm separator between the top and bottom sets of E+I lamellas - 10 mh;

4.7 to 10 Kohms load on the secondary leads to maximal primary inductance of ~10 mh;
Short circuit of the secondary leads to almost none inductance of the primary;

Experiments with secondary winding with 30 turns have been carried out and dully documented as follows:

Secondary on short circuit leads to primary inductance of 0.235 mh;
Secondary loaded with 10 ohms leads to 8 mh of primary;
Secondary loaded with 27 ohms leads to 8.92 mh of primary;
Secondary loaded with 37 ohms leads to 9.33 mh of primary.

I am very happy with the results!


Because I am curious and my friend who made speakers with the same bass drivers as those from the project from my signature was curious too, we tried connecting the 10 mh as a simple BSC with 6.6 ohms bypass resistance and optionally with 1 mf capacitor. - We are very pleased with the results. We got ~29 hz at 0 db this way the bass became better articulated and detailed too.


Experiments with variable loading of the secondary winding in my project are coming (maybe on Easter...) as well as listening impressions...

I apply some photos.

p.s. By the way, i have never heard a BSC before... - good thing is this!
Attached Images
File Type: jpg SimpleBSC10mh6r6and1mf.jpg (77.5 KB, 106 views)
File Type: jpg thetrasformer2x10mh.JPG (60.4 KB, 105 views)
File Type: jpg VKN 12311 in 127 l bassreflex.JPG (53.5 KB, 105 views)
  Reply With Quote
Old 17th April 2011, 06:38 AM   #16
mondogenerator is offline mondogenerator  England
diyAudio Member
mondogenerator's Avatar
Join Date: Sep 2009
Location: City Of Villans

with a ratio of 4:1 it would be easy to replicate your circuit using a isolation trannis. 400-110. dont know about SQ though, I have one here with 3 ratios between 3.8 ,4 and 4,2, roundabouts...
Balancing the things I must do, with the things I'd like to do...This is a skill (or a luxury) amongst many that I do not possess.
  Reply With Quote
Old 10th May 2011, 10:44 AM   #17
T101 is offline T101  Bulgaria
diyAudio Member
Join Date: Mar 2010
Location: Sofia-Bulgaria
From series I went to parallel, because I found out that with the software used for simulations, serial circuits are not being simulated correctly. - the flat zone of the full range below crossover frequency should not be there.

Another interesting fact is that 90 degrees of phase difference at 1250 hz i quite a challenge for alignment. Both drivers in phase lead to ~ 20 to 32 cm (8 to over 12 in) required acoustical center offset. And full range out of phase leads to about 6 cm required offset. Curious fact is that on this frequency the delays are very easily detectable and noticeable.

Also I moved the constant inductance from before the variable BSC transformer to downstream from it.

It seems that with varying the loading of the secondary winding and thus shaping the electrical signal energy distribution, it is required to alter the value of the capacitor for the high pass for the full range. This is due to a probable actual crossover point shift.

Very funny such a simple circuit with so few parts to present such complicated challenges. Another challenge is the flat area in the electrical response of the variable BSC, with small value load all is normal and the signal falls with 6 db/oct after say 1000-1200 hz, but with 8 or 15 or greater load there is a flat plane where before was a slope.

Probably I will have to consult an engineer again for those issues as I am very optimistic about the results from a well engineered network with that device implemented in it. The listening tests were quite promising with only one flaw apart from the required acoustical centers offset - discrimination of complicated signals around crossover frequency - sound from falling thin metal objects is greatly jammed - this is ringing, tangling, jingling and etc. - great example is the Dire Straits song "Private Investigations" - there is such passage in it...

I apply a photo of the test crossover auditioned. Later it received a powerful potentiometer (two out of three leads of an adjustable L-pad used) and it made even more obvious the difference between settings.
Note that in the photo you'll see a zobel network not shown on the simulated circuits above, because the software is not able to simulate an R (driver) bypassed by a zobel. The zobel is 10 ohm and 8.2 uf and is for the bass driver.

Best regards to all!
Attached Images
File Type: jpg DIYBSC.JPG (66.8 KB, 71 views)

Last edited by T101; 10th May 2011 at 10:50 AM.
  Reply With Quote


L here, L there, L Everywhere *-*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

Forum Jump

New To Site? Need Help?

All times are GMT. The time now is 02:16 AM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 15.79%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio