Goodbye Permanent Magnet Speakers

[Hari Iyer]

The current lockdown period allowed me to utilize the time to measure the loudspeaker more with less ambient noise outside my home. I wanted to experiment with time-alignment between mid-woofer and tweeter and spent most of the time reading technical articles about them. Also tried electrical time delay using RC network to delay the high frequency by frew msec. Also did a physical off-set of mounting the tweeter from behind and again checking the step response for integration and after experimenting for 10+ days with measurement and listening, finally will be settling with the sloping baffle method.

The tweeter offset gives serious diffraction issues which can only be tamed (not avoided) by the absorbing surfaces. The flat baffle made the tweeter sound much brighter than they should, but the sloping bafffle (around 18mm) gave me the right balance between the mids and highs. Here are some FR and step measurements,









If you notice step-1 and step-3 just by tilting the baffle, the tweeter energy is reduced significantly without much affecting the frequency response. This can be a useful tool for someone who finds their setup a little more airy or bright just by raking the front baffle by few mm. How much to rake can be tuned by listening too and would not require a measurement imo.

 

[kapvin]

The current lockdown period allowed me to utilize the time to measure the loudspeaker more with less ambient noise outside my home. I wanted to experiment with time-alignment between mid-woofer and tweeter and spent most of the time reading technical articles about them. Also tried electrical time delay using RC network to delay the high frequency by frew msec. Also did a physical off-set of mounting the tweeter from behind and again checking the step response for integration and after experimenting for 10+ days with measurement and listening, finally will be settling with the sloping baffle method.

The tweeter offset gives serious diffraction issues which can only be tamed (not avoided) by the absorbing surfaces. The flat baffle made the tweeter sound much brighter than they should, but the sloping bafffle (around 18mm) gave me the right balance between the mids and highs. Here are some FR and step measurements,

View attachment 45015
View attachment 45016

View attachment 45017
View attachment 45018

View attachment 45019View attachment 45020

If you notice step-1 and step-3 just by tilting the baffle, the tweeter energy is reduced significantly without much affecting the frequency response. This can be a useful tool for someone who finds their setup a little more airy or bright just by raking the front baffle by few mm. How much to rake can be tuned by listening too and would not require a measurement imo.

Hi Hari,

just some thoughts:

for the more technically inclined, a 20dB per division vertical chart is not very meaningful. you should expand scale to atleast 5DB per division or preferably 3dB per division.

have you set up REW correctly? there seems to be a bit of pre ringing in your step response graph. i am trying to understand how that is possible

what sort of gating have you used? I am curious, because, intuitively a OB cannot be measured the same way as a closed box system .

finally, the electromagnetic powersupply - how have you filtered it? again, thinking inductively, wont even the slightest bit of main AC current hum colour the response?

 

[Hari Iyer]

@kapvin - i am glad that atleast someone responded!

I am pasting the final implemented FR with a 5db scale for your view,

This pre-ringing before 0sec is normal and you can also find in all of stereophile measurements too and is a mystery even for JA.

I am using default gating and not gating the response to adjust the LF response and displaying as it is. You can see hump in the step response (two of them) which is due to the waves hitting the boundaries of the walls or furntiure which i am okay with as i am aware of that.

The FC powersupply is using a R (10 watt + 15 watt) incandesant lamp with a "C" filter followed by a 1H choke and a "C" filter. when i put my ears on the FC cone, its as good as its off. This method was suggested by the seller of the speaker. The incandesant lamp acts a current limiter and filter for the supply. i thought of adding one more "L" but did not add as anyway there is no hum and will be a overkill.

My DC voltage before current limiting is 260V and after adding the current limiter and choke(75 ohms + 1 H) it drops to around 75volts which is the recommended supply @ 70mA current which approximates around 6 watts constant current.

 

[kapvin]

@kapvin - i am glad that atleast someone responded!

I am pasting the final implemented FR with a 5db scale for your view,

This pre-ringing before 0sec is normal and you can also find in all of stereophile measurements too and is a mystery even for JA.

I am using default gating and not gating the response to adjust the LF response and displaying as it is. You can see hump in the step response (two of them) which is due to the waves hitting the boundaries of the walls or furntiure which i am okay with as i am aware of that.

The FC powersupply is using a R (10 watt + 15 watt) incandesant lamp with a "C" filter followed by a 1H choke and a "C" filter. when i put my ears on the FC cone, its as good as its off. This method was suggested by the seller of the speaker. The incandesant lamp acts a current limiter and filter for the supply. i thought of adding one more "L" but did not add as anyway there is no hum and will be a overkill.

My DC voltage before current limiting is 260V and after adding the current limiter and choke(75 ohms + 1 H) it drops to around 75volts which is the recommended supply @ 70mA current which approximates around 6 watts constant current.

View attachment 45022

what matters is that you are happy.

for the Mains hum, you might ( just to eliminate any latent paranoia) want to close the mike the woofer with the magnet activated / deactivated and measure.

 

[Hari Iyer]

what matters is that you are happy.

for the Mains hum, you might ( just to eliminate any latent paranoia) want to close the mike the woofer with the magnet activated / deactivated and measure.

Field coil speakers should not be played with magnets deactivated, this will burn the voice coil. Also by mistake on e I tried playing withoutactivating the magnets and there was almost no sound from the driver.

 

[kapvin]

Field coil speakers should not be played with magnets deactivated, this will burn the voice coil. Also by mistake on e I tried playing withoutactivating the magnets and there was almost no sound from the driver.

I meant with no signal other than the field coil active and inactive. The idea is to see if there is any sound that the coil is adding. Best wishes

 

[Hari Iyer]

I meant with no signal other than the field coil active and inactive. The idea is to see if there is any sound that the coil is adding. Best wishes

That will be the noise floor of my room without the amplifier on I have done that last week to check the ghesalt of my room. With the latest version of REW you can also display the noise floor along with distortion measurements

 

[Hari Iyer]

I have build many speakers past 15 years, but this is the first time after i started measuring my speakers 8 years ago, i have measured a square wave response @ 1KHz so perfect for any of my speakers till date. Thought of sharing.

 

[Vinodtherattil]

The virtue of a Field Coil

audio, tube, horn loudspeaker, high efficiency, loudspeaker, turntables, vinyl, field coil,

electravolt.blogspot.com

Congratulations Hari on your new purchase, fieldcoil drivers. These are made in East Germany. I have the same brand tweeters. Excellent ones.

In my listening experience fieldcoils have very low distortion than alnico magnets.

Just like tubes, TTs, vinyl have made a comeback, in the west the fieldcoils also gaining popularity.
GIP of Japan made some excellent replica of Western electric fieldcoils. Companies like ems and few others are also making very good fieldcoils. But this is an expensive n complex hobby.

I have a small collection of vintage fieldcoils ( Jensen, Western electric, RCA, Altec etc etc) also a modern one, EMS ( Electro magnet speakers, France).

Looks like 2020 may bring more discussions and explorations of fieldcoils in the forum.

My resonant cabinets for ems fieldcoil drivers are almost near completion. Will start a thread on them soon.
Cheers!!

https://imgur.com/Pl1eroh

Totally agree...

After owning line magnetic 755i speakers, I'm a huge fan of FCs....

Line Magnetic Audio LM 755i Field Coil Speaker Review

Back on January 15th and 16th Matt Rotunda of Pitch Perfect Audio hosted the first ever Greenhaus. It was Pitch Perfect Audio’s homage to the Monkeyhaus events that John DeVore of DeVore Fidelity held in New York. You may have read some of Art Dudley’s stories about these events. Matt wanted to...

www.dagogo.com

Please let me know if any one in Bangalore have FCs....

 

[kapvin]

I have build many speakers past 15 years, but this is the first time after i started measuring my speakers 8 years ago, i have measured a square wave response @ 1KHz so perfect for any of my speakers till date. Thought of sharing.

View attachment 45150

Hari, a square wave is displayed on graph which has amplitude on y axis and time on x axis. This graph is of frequency on x axis. So I am not sure what it shows. if the signal is 1khz then all this graph is doing showing a HUGE amount of side Band distortion—-For an octave in each direction.

Ideally to Show how your loudspeaker is playing a square wave, you need to feed the mic signal to software which can render the waveform, like an oscilloscope or similar plugin Which does the same.

not at all to say that your fc sPPPoE Akers are bad or good, just that this is not how you should measure Them.

On a related note, have you measured the T/S parameters of the speaker, at Different levels of drive of the field coil. That would a useful and informative experiment.

I have been fascinated by field coil theory But the only field coil I have actually heard is the focal Grand utopia ( the woofer has a Field coil. A field coil using using modern understanding could well provide the holy grail of a linear loudspeaker at reasonable cost (Underhung—long and strong magnetic field, short coil). are there any moden examples?

 

[Hari Iyer]

Hari, a square wave is displayed on graph which has amplitude on y axis and time on x axis. This graph is of frequency on x axis. So I am not sure what it shows. if the signal is 1khz then all this graph is doing showing a HUGE amount of side Band distortion—-For an octave in each direction.

Ideally to Show how your loudspeaker is playing a square wave, you need to feed the mic signal to software which can render the waveform, like an oscilloscope or similar plugin Which does the same.

not at all to say that your fc sPPPoE Akers are bad or good, just that this is not how you should measure Them.

On a related note, have you measured the T/S parameters of the speaker, at Different levels of drive of the field coil. That would a useful and informative experiment.

I have been fascinated by field coil theory But the only field coil I have actually heard is the focal Grand utopia ( the woofer has a Field coil. A field coil using using modern understanding could well provide the holy grail of a linear loudspeaker at reasonable cost (Underhung—long and strong magnetic field, short coil). are there any moden examples?

This wave is generated using a pulse pink noise and measured using a real time analyzer. I have not fed any 1 kHz square wave signal to measure them. Using the pink PN you can't setup any frequency and the 1 KHz is always defaulted. I don't have any other method of acoustic measurement of the pink PN other than the RTA.

 

[Hari Iyer]

I meant with no signal other than the field coil active and inactive. The idea is to see if there is any sound that the coil is adding. Best wishes

I did critical listening putting my ear on the cone with FC activated and deactivated. I can listen just a very very faint hum which I can easily ignore. This hum is even with the volume in zero position so it comes from the DC power supply. My FC has a DC supply of around 75 v and I faintly remember measuring around 4volts AC couple of months ago. Need to figure out what filter can bring down this AC getting into DC. I already have a R-C-L-C filter currently. Suggestions please.

 

[Hari Iyer]

Past 5 months spent some time discovering room dynamics and integration of my room with my OB. Did some measurements on RT60, EDT, Clarity and Definition, but finally trusted my ears more than the mic.
I have noticed that reflections from the front wall in an OB can cause some major phase errors that can interfere with your direct sound and smear the mids and highs resulting in listening fatigue. After reading many articles from Flyod Toole and other patented material finally designed my control dispersion OB box last week. This material is semi circular and will be on the rare side of the OB and will diffuse and disperse the rare wave from the mid range and prevent them from hitting the rare wall. They will actually disperse to the side walls and there by creating late reflections and enhance acoustics like that of an auditorium. Also have added a baffle on top of the current baffle for time alignment
This should be fully ready by end of this month if all goes well.

 

[Hari Iyer]

The modifications have undergone some changes, as i found the isophon tweeters not matching up to the field coil w.r.t SPL and dynamics. Moreover they begin to roll-off around 10KHz and suddenly dip as high as 20dB at 20KHz. Due to this the higher harmonics gets impacted and the upper dynamics and detailing gets impacted. I thought of many things to counter this, but nothing could be worked as its because of the frequency response of the isophon which was causing this issue.

I later was thinking if the isophon could be replaced with my original 8" Russian FR and if they can be good substitute. I had the archived measured response of the Russian FR and i immediately simulated them along with the field coil 8" to check if they compliment each other. Below are the results.



Red - Russian Driver, Black - Field coil.
You can see, the Russian driver has an extended response till 20KHz and the field coil roll-off is around 3KHz. Both the drivers may not be measured at the same level as they were measured in a different time zone of more than 6 months apart. Neverthless they give me a sense of hope that there is no serious cancellation happening if they are placed with the correct phasing. Both drivers complement each other quite well as per the simulation. Below is their combined response.



I have used a simple BSC and zobel to keep the impedance constant across the response to get a flatter response as below,



Below is the combined system impedance,



If you notice the Qms of both the drivers are reduced by both the drivers complementing each other for the same. If you notice the resonance of the russian driver around 38Hz is reduced to almost 3.5ohms because of the FC driver impedance and similarly the resonance of the FC driver is reduced to almost 4ohms by the impedance of the Russian driver. The nominal impedance will be around 4 ohms with minimium impedance around 2.45 ohms for the speaker system. This low impedance imo would not cause any serious amplifier load issues, as they are more of a resistive or inductive load.

 

[Hari Iyer]

I received the modified speakers early this week and assembled them. As there were to be connected directly without anything inbetween drivers and amplifiers, i could hook-up within 30 minutes. But then there were serious phasing/ polarity issues with the drivers, because both drivers being vintage there were no proper polarity marking on them. I need some 30 minutes more to figure out the correct polarity and could finally connect them in-phase. The controlled dispersion section consist of a parabolic reflector, which prevents the rare wave from the drivers to directly hit the front wall. The reflector will diffuse and disperse the rare waves and infact get diffused to the side walls creating late reflections. This gives auditorium like sound stage with precise imaging and clarity. This is how they now finally look,






I performed some measurements today and the frequency response look like this,




To my ears, they sound perfect and i don't think i shall add any compensation network because they are already quite balanced in the frequency domain. I can say that, this has been the best sounding speakers in my room till date, in my setup for the past 20+ years of my speaker building hobby for myself, my friends, relatives and of-course FMs.

 

[Hari Iyer]

some issues with the vertical scaling when i copy the images, it gets resized to 20dB vertical scale always and FM @kapvin has objection to this. So am trying to repaste with 5dB vertical scale,

 

[Hari Iyer]

For the DIY community...

The frequency response measured today after one month of break-in...

The impedance measured last month before break-in...

 

[Hari Iyer]

I am planning the below upgrade to my speakers,
- add a 12" Ahuja woofer on the current OB. I will need to add wings to both side of the baffle to prevent cancelling at low frequency.
- sell the current H-frame OB subwoofer as I need to create some space in my living room as per request from family members.
- the Ahuja will be crossed around 280 Hz to take care of driver positions and edge diffraction.
- the current FR drivers compliment quite well with the H-frame sub, but I need to experience the lows and mid-bass of the tube amplifier without getting them in a Chinese plate amplifier.

 

[Hari Iyer]

Last week, i had purchased the Ahuja 12" woofer and assembled them on the existing OB baffle. I removed the H-frame subwoofer and mated the Ahuja with my 2004 Denon AVR using a low pass filter card purchased from Amazon. I am using a temproary wings on them as i have yet to complete my OB design as the woofers were undergoing break-in. This how they look now,


After using them for around 4 days, my wife asked me if these wings on both sides are going to be permanent and i said it would be integrated with the baffle. Needless to say these were poor WAF and they haijack the already cramped space in my living room. It appeared as if they were better off with the sub-woofer than these wings. I tried all possible tricks to convince them about the wings but she was not getting covinced. My daughter who was listening through the argument told me to put a folding wing and use them only when required and fold them in when not in use. I thought for 10 sec and thought it was a brilliant idea and i looked at my better half and she nodded her head in agreement.

I am sure the Bose 301 would have some female team member inputs in their design. The Bose 301 sub-woofers might be designed by his male team members and the satellite by his female team members. If you have completely male team designing your speakers then you have something like @prem JBL 4343 -

 

[jenson]

Really happy to see a post without measurements..