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

Optimising Bass Microphone Servo control
Optimising Bass Microphone Servo control
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 6th December 2017, 10:25 PM   #11
weltersys is offline weltersys  United States
diyAudio Member
 
weltersys's Avatar
 
Join Date: Jan 2008
Location: Florida
Quote:
Originally Posted by wgh52 View Post

Maybe a possible trick here would be to filter the Mic-signal with a low pass equivalent to the 180 Hz cross-over low-pass, so the feedback circuit does not "hear" the mid-driver taking over at 180 Hz? Would that be "a huge overkill" or wrong or counterproductive or a potential improvement?
The mic works with the inverse-distance law, each doubling of distance looses 6 dB level. At "ground zero" the woofer level is likely far louder at the acoustic crossover than the relatively distant mid-driver. If you lower the mic low pass, it won't "hear" harmonic distortion as well, so can't correct it as effectively.
  Reply With Quote
Old 7th December 2017, 01:04 AM   #12
bentoronto is offline bentoronto  Canada
diyAudio Member
 
Join Date: Jan 2008
Location: Toronto and Delray Beach, FL
Optimising Bass Microphone Servo control
The 4n7 is clearly operating in the audio range. As you enlarge the operating band, you risk entering a zone of instability. Manufacturers of motional feedback systems live in fear of instability and blown drivers: otherwise they would have gone to a smaller filter cap too. Not saying what a DIYer should do with a closely monitored home system, just want to be careful.

This may sound heretical, but multi-speaker FR charts can be misleading due to phase/interference effects and in relation to the REW mic and musical content. In other words, pure swept sine wave results may not correspond to the audible loudness on complex sounds. So the middle FR with the dip may sound better than the 2n2 on the right.

B.
__________________
HiFi aspirations since 1957. Currently working on motional feedback again... the final frontier in audio (also, see recent post with data on a 17-foot labyrinth sub)
  Reply With Quote
Old 12th December 2017, 03:22 PM   #13
Sonce is offline Sonce  Macedonia
diyAudio Member
 
Join Date: Jan 2014
Location: Macedonia
Quote:
Originally Posted by bentoronto View Post
Using the figure-8 reception of that mic (once I paid 50-cents for that kind of capsule),
That kind of electret mic has omnidirectional (kugel) characteristic.
  Reply With Quote
Old 15th December 2017, 10:58 PM   #14
bolserst is offline bolserst  United States
diyAudio Member
 
bolserst's Avatar
 
Join Date: Jun 2005
Location: Fort Worth, Texas
Quote:
Originally Posted by wgh52 View Post
… not really sure if the above solution as such is "state of the art", accommodating for all the possible or necessary servo optimizations, i.e. can achieve what is possible. Therefore would like to gather input for servo bass improvement of my speakers, including how I would confirm by (acoustic and oscilloscope) measurements.
Looks like you have an interesting speaker in your hands

To determine what B&M achieved in the circuit and what optimization is possible, you would need to measure the open loop response. To do this, you need to break the feedback loop. The most convenient place to do this is the connection between the preamp and power amp as shown in the attached figure by the red X. Input a low level test signal into the power amp(10 – 50mVrms) from measurement software(ARTA, REW, etc) and measure the electrical frequency/phase response coming out of the preamp where you broke the loop. Dividing this by the input signal will gives the open loop response, which tells you how much feedback is being used and what the gain and phase margin is.

While you have the feedback loop broken, you might also measure the acoustic output of woofers while using the same direct input into the power amp. This will show you what the natural response of the woofer/cabinet is before feedback is applied.


Quote:
Originally Posted by wgh52 View Post
…There are a few details I don't understand about the circuit myself, though, e.g. what is the function of the 4n7 C across the microphone terminals in the actual circuit?
The 4n7 capacitors across the microphones form a first order low pass filter with R157, rolling off the feedback signal above about 1kHz.

You mention seeing a 6dB change in woofer level when changing the values of these capacitors. I don’t understand why that would be unless one of the 2 microphone connections was bad before or after the change. (ie factor of 2 = 6dB) Where are the two capacitors physically located? And how are the connections made between mic, capacitor, and preamp board?
Attached Images
File Type: png Open_Loop_Response.png (84.5 KB, 115 views)
  Reply With Quote
Old 16th December 2017, 12:53 AM   #15
bentoronto is offline bentoronto  Canada
diyAudio Member
 
Join Date: Jan 2008
Location: Toronto and Delray Beach, FL
Optimising Bass Microphone Servo control
I'd add - to the always super help from Bolserst - you might also chart the feedback signal itself. Like most of the signal of interest, it is relative to whatever is also motivating the system.

B.
__________________
HiFi aspirations since 1957. Currently working on motional feedback again... the final frontier in audio (also, see recent post with data on a 17-foot labyrinth sub)
  Reply With Quote
Old 3rd November 2018, 07:44 PM   #16
wgh52 is offline wgh52  Germany
diyAudio Member
 
Join Date: Feb 2011
Hello folks,

first of all my sincere apologies for leaving the subject. I was completely distracted by family related issues, i.e. am now a grandpa.

In the mean time major changes happened to the Omega project and the complete old speaker electronics are is being replaced with ChipAmps, new power supplies employing active rectification, a transformer starter circuit and transformer snubbing, remote on/off, LM317/337 supply, symmetrical/balanced inputs and ...

Which also brings me back to the acoustic feedback issue covered here.

So thanks for all your valuable posts! I'm going to thankfully take the advice posted by bolserst (and others) seriously as soon as I have the Chipamps working with the new power supply.

The main issue with the new Poweramp section is, that the Amp gain is not 35 dB as in the original B&M implementation, but only 26dB. This would mean that I need 10 dB more gain elsewhere in the loop, so would probably have to change the diff. Amplifier gain accordingly. right?

Now... Recently I remembered a schematic in the good old Tietze/Schenk electronics engineering study-book from decades ago and derived the idea to achieve optimal feedback dimensioning without a lot of math : A feedback system which adjusts the proportional (gain), integration and differentiation parts of the control loop adjustable individually. Well, the differentiation part is not needed here (omit the "R3" branch), but adjusting gain and integration individually, should enable finding an optimum feedback configuration relatively easily with a square wave input and measuring to acoustic signal ... right? The results would then be transferred into the simple one-OPAmp-circuit shown earlier in the thread.

Optimising Bass Microphone Servo control-pid-regler-tietze-schenk-800-png



Thinking a bit further f(I) is already known with 32Hz from the original design and may be OK to stay, so the adder gain and the microphone gain seem to be the real "tuning knobs" ...

So, it seems I'm on a possibly successful path. Or am I overkilling?
What's the advice, what's the thinking here folks?

Thanks and Regards,
Winfried
Attached Images
File Type: png PID Regler Tietze-Schenk 800.png (433.4 KB, 147 views)
  Reply With Quote
Old 8th November 2018, 12:28 PM   #17
TBTL is offline TBTL  Germany
diyAudio Member
 
Join Date: Oct 2013
You would need to measure the open loop response and tune that to some goal. It appears that the original open loop response is a band pass with first order slopes: a second order highpass (woofer in box, say at 60 Hz) combined with a first order low pass (electronics, say at 30 Hz). The woofer to microphone transfer function is likely to have resonances at high frequencies, so somewhere in the electronic part of the loop there must be a low pass filter.

For better control at some frequency range you need to increase the gain at the range at which the control is active. One way to accomplish that is by changing the first order low pass to a second order one and place it at or above the woofer resonance frequency. Now you have a band pass response with second order slopes. These have plus and minus 180 degree phase shifts. You cannot have 180 degree phase shift at the gain crossover frequencyies so you have to bend the phase a bit back towards 0 degrees using lag and lead compensators. 45 degrees shift (resulting in a net +-135 degrees) is sufficient.

An easier way is the PI controller you mentioned + a low pass. Let's analyse the open loop response. Start with the woofer to microphone transfer function, which is a second order high pass (~60 Hz) plus some resonances at high frequencies (~1 kHz). A second order high pass will have a 180 degree phase shift below the cut off frequency (~60 Hz). The 'P term' of the PI controller sets the open loop gain. The 'I term' works as a lead compensator to reduce the phase shift at the lower gain crossover frequency (~30 Hz). Then a low pass filter makes sure the open loop gain decreases below 0 dB before the resonances occur (~1 kHz).

Note that the open loop response includes high pass filters in the amplifiers. It's easiest if these are far away from the lower gain crossover frequency, otherwise they bend the phase even further away from 0 degrees at the lower gain crossover frequency.

Last edited by TBTL; 8th November 2018 at 12:50 PM.
  Reply With Quote
Old 8th November 2018, 05:56 PM   #18
bentoronto is offline bentoronto  Canada
diyAudio Member
 
Join Date: Jan 2008
Location: Toronto and Delray Beach, FL
Optimising Bass Microphone Servo control
A good way to see if MF is working is to examine the signal to the driver(s). You should see it is NOT flat and with some wiggles and there's a cut where the driver flaps (sealed or tuned box esp region below tuning) and a boost at low freq where a sealed box holds the drivers back. Also, and this will sound odd, the signal should show extra HD as a corrective to the driver. That's how you know MF is ALIVE.

Using REW to track electrical signals is more stable than loose acoustic measures - although no substitute for sound output. You can work with the simple basic comparison with is simply to a standard amp with no MF pre-amp driving the speaker, rather than worry about disabling the MF circuitry, just unplug it.

If working, it will be a very obvious A/B improvement percussive instruments.

Original circuit seems pretty good. I'd try to get performance measurements with that. You aren't moving forward until you start testing, then you can apply the comments of Bolserst and TBTL.

The electrets may be acting as accelerometers more than catching the breeze as mics, or just choice of name.

If being a grandpa is a valid excuse for procrastination, I can be a 14X procrastinator. Good excuse or not, congratulations.

B.
__________________
HiFi aspirations since 1957. Currently working on motional feedback again... the final frontier in audio (also, see recent post with data on a 17-foot labyrinth sub)
  Reply With Quote
Old 9th November 2018, 11:30 AM   #19
wgh52 is offline wgh52  Germany
diyAudio Member
 
Join Date: Feb 2011
Hi folks,

sincere thanks for your elabotate comments and advice!

One Q of understanding:
Quote:
... the signal should show extra HD as a corrective to the driver ...
What does the abbreviation "HD" mean?

I'm in the middle of building the new active electronics, like Power Amp and Supply boards featuring chip-amps, balanced inputs, active rectification and transformer snubbing. So, it will be a couple of weeks until I can show progress.

Thanks a bunch!
Winfried
  Reply With Quote
Old 9th November 2018, 11:46 AM   #20
bentoronto is offline bentoronto  Canada
diyAudio Member
 
Join Date: Jan 2008
Location: Toronto and Delray Beach, FL
Optimising Bass Microphone Servo control
Quote:
Originally Posted by wgh52 View Post
What does the abbreviation "HD" mean?
Harmonic distortion.

When you run a freq plot with REW, it does the freq plot and without further bidding, shows you the HD plot too.

REW - Room EQ Wizard Room Acoustics Software

REW - Room EQ Wizard Room Acoustics Software

B.
__________________
HiFi aspirations since 1957. Currently working on motional feedback again... the final frontier in audio (also, see recent post with data on a 17-foot labyrinth sub)
  Reply With Quote

Reply


Optimising Bass Microphone Servo controlHide 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

Similar Threads
Thread Thread Starter Forum Replies Last Post
Infinity Irs servo control rvito Multi-Way 0 11th May 2017 03:16 PM
Simple Servo Control KatieandDad Everything Else 0 2nd January 2014 02:19 PM
DIY Servo Sub- need help with control circuit flyingtele Subwoofers 1 28th January 2013 08:57 PM
Does anybody know how to use the MCU to control the CD servo? zhangzq71 Digital Source 0 19th December 2008 04:52 AM
Servo Bass mrfeedback Subwoofers 32 2nd January 2006 08:51 PM


New To Site? Need Help?

All times are GMT. The time now is 06:51 AM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 14.29%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio
Wiki