MFB and CFB using Piratelogics Electronic
- By esl 63
- Subwoofers
- 78 Replies
In this first post I will update all important stuff continuously for your convenience
New input interface for ACH-01 and stiffer
carbon fiber support (28 may) post 66 (better!!)
Add stiffnes post 69 (better)
Add "dust cap" stiffener... is it better?? Post 71
Background: After have been experimenting a lot with Servo subwoofers or MFB motional feedback in the 90-ties with success I went to dipole subwoofers and QUAD ESL 63 electrostatics (modified). Everything was fine until I realized that sometimes the ESL 63 playing on their own, had a magic in upper bass that can not be achieved with dynamic woofers... nope nix njet nej ikke.
8-10% distortion in a woofer does not blend well with QUAD or other electrostatics with <1% distortion in bass.
So only way out is to reduce distortion with a factor of TEN ! (Or go back to dynamic speakers..😱 naa.. )
And here is how it is done 😎
Circuit: First I am using ACH-01 accelerometer and using Piratelogic Electronics EVE 2020 circuit boards which is wery convenient, however this circuit is not designed directly to fit ACH-01 so some tweaking on some values has to be done... also using CFB is changing frequency responce and phase a lot and on top of that dipole woofer, the frequency correction and phase response will be TOTALLY different from using the EVE 2020 the "normal way"
Which is to drive voltage amplifiers with Starbass accelerometers and small woofers in small boxes. If you want to do so there is another thread by chriscam on this forum.
So be careful, the manual for using EVE 2020 will not guide you the right way for this combination... but nice and easy.
We will get it working! And if you want to make it more easy use Star bass accelerometer. They seems to work great.
The benefits with MFB CFB and Dipoles: Using motional feedback to a Dipole woofer system is double or maybe triple compared to a closed box!
First: Distortion comes with longer strokes, which you have a lot more with dipoles than any other principle.
Second: Since the dipole compensation 6dB/octave raise the low fundamental tones, those fundamental tones has corresponding harmonics which will be raised several dB! If you play a 100Hz tone, the 2nd harmonic of this tone is also lifted 6dB, and worse... the 9th harmonic 900Hz will be lifted >18dB!!
Third: The 9th harmonic 900Hz is much more disturbing and noticeable since your ears are much more sensitive in those frequencies.
Dipole woofers are hard to integrate with low distortion midbass of an electrostatic speaker!
Reed the theory papers below and in nr 3, you find the benefits of combining CFB and MFB.
Distortion reduction and measurements part 1: Post 30 shows 2-3 times lower distortion, but it will be improved!
I have reached a more than ten fold reduction in distortion or 22dB in a prototype, and here i will guide you trough the building, schematics, measurements setup and all problems that you may face during the journey, I know, I have done all mistakes there is, and will try to cover them all.
SERVO subwoofers or MFB theory:
1. http://rmsacoustics.nl/papers/whitepaperMFBtheory.pdf
2. http://rmsacoustics.nl/papers/whitepaperMFBdesign.pdf
3. https://www.rmsacoustics.nl/papers/Impulse Compensated Driver setup.pdf
Input stage: The black and white part of the schematic is the accelerometer itself and the buffer J310 68k and 820 ohm resistors, mounted on the woofer, these are SMD components glued to the cone and hard wired under a microscope... crazy idea from the beginning, and even more stupid when the J310 was fried :-(
A part of the schematic EVE 2020.0 is shown, and a table with some assumptions. With ACH-01 you should add an impedance buffer see below.. (Starbass has a buffer already built in)
Updated 220511
Accelerometer gain stage and loop mixer: Post 4 is misleading regarding J310 FET.. see picture below.
I have seen the application note schematic for ACH 01, but was mislead by another users design.
The ACH 01 shall have a negative bias as shown on the application note, to minimize distortion.
I will do a last try with ACH-01 Otherwise I buy ClingOn from Chris, and just get it working...
Big thanks to Robert-H Munnig Schmidt at RMS acoustics for explanation!! 🏆
Rob has achieved 30dB loop gain and 45deg phase margin... is that a world record? (some people is impressed by Messi and Ronaldo....)
Stability, Phase margin, and OLTF measurements: Post 34
Stability at low frequency is achieved but at around 3kHz it looks really messy. So I am reaching out for any help, here.
Below is the input side of the accelerometer on EVE 2020 board. From here you will see that everything has to be made different with Dipoles.
How it should look like!
I did a last try with ACH-01, with stiffer carbon fiber support and new electric interface in post 66 and it works much better!
New input interface for ACH-01 and stiffer
carbon fiber support (28 may) post 66 (better!!)
Add stiffnes post 69 (better)
Add "dust cap" stiffener... is it better?? Post 71
Background: After have been experimenting a lot with Servo subwoofers or MFB motional feedback in the 90-ties with success I went to dipole subwoofers and QUAD ESL 63 electrostatics (modified). Everything was fine until I realized that sometimes the ESL 63 playing on their own, had a magic in upper bass that can not be achieved with dynamic woofers... nope nix njet nej ikke.
8-10% distortion in a woofer does not blend well with QUAD or other electrostatics with <1% distortion in bass.
So only way out is to reduce distortion with a factor of TEN ! (Or go back to dynamic speakers..😱 naa.. )
And here is how it is done 😎
Circuit: First I am using ACH-01 accelerometer and using Piratelogic Electronics EVE 2020 circuit boards which is wery convenient, however this circuit is not designed directly to fit ACH-01 so some tweaking on some values has to be done... also using CFB is changing frequency responce and phase a lot and on top of that dipole woofer, the frequency correction and phase response will be TOTALLY different from using the EVE 2020 the "normal way"
Which is to drive voltage amplifiers with Starbass accelerometers and small woofers in small boxes. If you want to do so there is another thread by chriscam on this forum.
So be careful, the manual for using EVE 2020 will not guide you the right way for this combination... but nice and easy.
We will get it working! And if you want to make it more easy use Star bass accelerometer. They seems to work great.
The benefits with MFB CFB and Dipoles: Using motional feedback to a Dipole woofer system is double or maybe triple compared to a closed box!
First: Distortion comes with longer strokes, which you have a lot more with dipoles than any other principle.
Second: Since the dipole compensation 6dB/octave raise the low fundamental tones, those fundamental tones has corresponding harmonics which will be raised several dB! If you play a 100Hz tone, the 2nd harmonic of this tone is also lifted 6dB, and worse... the 9th harmonic 900Hz will be lifted >18dB!!
Third: The 9th harmonic 900Hz is much more disturbing and noticeable since your ears are much more sensitive in those frequencies.
Dipole woofers are hard to integrate with low distortion midbass of an electrostatic speaker!
Reed the theory papers below and in nr 3, you find the benefits of combining CFB and MFB.
Distortion reduction and measurements part 1: Post 30 shows 2-3 times lower distortion, but it will be improved!
I have reached a more than ten fold reduction in distortion or 22dB in a prototype, and here i will guide you trough the building, schematics, measurements setup and all problems that you may face during the journey, I know, I have done all mistakes there is, and will try to cover them all.
SERVO subwoofers or MFB theory:
1. http://rmsacoustics.nl/papers/whitepaperMFBtheory.pdf
2. http://rmsacoustics.nl/papers/whitepaperMFBdesign.pdf
3. https://www.rmsacoustics.nl/papers/Impulse Compensated Driver setup.pdf
Input stage: The black and white part of the schematic is the accelerometer itself and the buffer J310 68k and 820 ohm resistors, mounted on the woofer, these are SMD components glued to the cone and hard wired under a microscope... crazy idea from the beginning, and even more stupid when the J310 was fried :-(
A part of the schematic EVE 2020.0 is shown, and a table with some assumptions. With ACH-01 you should add an impedance buffer see below.. (Starbass has a buffer already built in)
Updated 220511
Accelerometer gain stage and loop mixer: Post 4 is misleading regarding J310 FET.. see picture below.
I have seen the application note schematic for ACH 01, but was mislead by another users design.
The ACH 01 shall have a negative bias as shown on the application note, to minimize distortion.
I will do a last try with ACH-01 Otherwise I buy ClingOn from Chris, and just get it working...
Big thanks to Robert-H Munnig Schmidt at RMS acoustics for explanation!! 🏆
Rob has achieved 30dB loop gain and 45deg phase margin... is that a world record? (some people is impressed by Messi and Ronaldo....)
Stability, Phase margin, and OLTF measurements: Post 34
Stability at low frequency is achieved but at around 3kHz it looks really messy. So I am reaching out for any help, here.
Below is the input side of the accelerometer on EVE 2020 board. From here you will see that everything has to be made different with Dipoles.
How it should look like!
I did a last try with ACH-01, with stiffer carbon fiber support and new electric interface in post 66 and it works much better!
Attachments
Floating supply amplifier (backward, transnova, qsc)
- By UltimateX86
- Solid State
- 4 Replies
Attachments
Old school simple project and some thoughts on opamps
- By m10chalk
- Analog Line Level
- 0 Replies
I see lots of very sophisticated projects here with people doing smart stuff with Arduino and Pi and all sorts. So I feel suitably chastened in posting this but may be of some interest to someone.
Having finished my first Chipamp I was left with a Cambridge Audio A1 case and some of the internals and decided to repurpose them into a preamp to feed the Chipamp. All very simple stuff. I sourced a modest 30VA 15-0-15 toroid. I stripped out the old regulator system (just Zeners) from both the main preamp board and the input board and installed a proper linear regulator based on L78/7915. I did some rerouting on the input board to cut out the the tape loop and give myself just another line level input. Then obviously I had to rewire everything together and in doing so tried to adhere to some better routing than the original design allowed (given that it had a comparatively massive power amp board in there).
So the design (as per original) is simple. An inverting buffer on input, feeding a shunt feedback Baxandall tone circuit (so again inverting), feeding a pot and then straight to output. So this last stage is a 'passive' preamp. I use short cables and the pot is 20k (ln) so no issues there, but I will be adding a final buffer stage just for versatility to allow longer cables if wanted (may enable both buffered and non-buffered outputs).
On to opamps. So much written I know but I will keep it brief. It came with TL072 in both stages. I am a great fan of TL072 from long back and in these low or unity gain circuits I didn't see a compelling reason to change, but I had socketed the boards and so experimented a little. Tried 4558, NE5532 and AD712 (as used in more upmarket Cambridge Audio A3i). I measured all of them for noise (pretty negligible in all cases but measurable differences) and did some listening tests.
OK my ears are not good and I am not therefore a great guide on sound quality - whatever that means. In all honesty I really couldn't reliably tell them apart. Its quite useful that the tone circuit is switchable so its possible to switch one op amp out of circuit and this did reveal a very slight degradation due to the 4558 - very slight, but female vocals just didn't sound as convincing (at least not to me). For the others I couldn't have reliably told whether the tone circuit was on or off with a blind test. So on this criteria alone the 4558 drops out of contention.
On to noise. Interesting. The differences are tiny and all versions are getting close to the noise floor of my ancient Levell muV meter but here it is. The noisiest (in this circuit) was the TL072, the quietest was the AD712. The total measured difference (full volume, tone on) was only around 10muV from best to worst but given the low levels I was measuring (20-30muV) I guess this is quite a large percentage. However .... it was impossible to tell them apart in practice unless my ear was right up against the speaker. In that case I would say the TL072 noise was a bit more objectionable / noticeable as in being more popcorn like.
Then I am going to add a criteria. DC offset. No surprise here the NE5532 was worst by a country mile. So what? Well the NE5532 implementation would be unusable without some switch on delay relay / switch off mute. I say that because if ever by accident you were to switch off that preamp first (or switch it on last) with NE5532 on board there would be very loud, very unpleasant and possibly damaging loudspeaker cone excursions. I tried it without tone enabled and volume turned down and it was bad enough. The TL072 by comparison was relatively benign (but you would still want to avoid a mistake of powering down preamp with power amp on). The AD712 was just a dreamboat. I couldn't measure any DC offset and it starts up and shuts down almost silently.
Perhaps no surprises - a circuit designed with JFET opamps works better with JFET opamps. But then the performance of the AD712 in comparison with TL072 did take me aback. Received wisdom is this is an ancient 80s opamp and offers no real benefit over the TL072. And I hadn't realised the benefit of DC precision is potentially simplifying a design. It was therefore measurable better and much more practical. In fact for the audiophile approach, DC offset is so low i could easily remove at least two DC blocking capacitors per channel if I wanted to. Picture of new internals just for the hell of it.
Having finished my first Chipamp I was left with a Cambridge Audio A1 case and some of the internals and decided to repurpose them into a preamp to feed the Chipamp. All very simple stuff. I sourced a modest 30VA 15-0-15 toroid. I stripped out the old regulator system (just Zeners) from both the main preamp board and the input board and installed a proper linear regulator based on L78/7915. I did some rerouting on the input board to cut out the the tape loop and give myself just another line level input. Then obviously I had to rewire everything together and in doing so tried to adhere to some better routing than the original design allowed (given that it had a comparatively massive power amp board in there).
So the design (as per original) is simple. An inverting buffer on input, feeding a shunt feedback Baxandall tone circuit (so again inverting), feeding a pot and then straight to output. So this last stage is a 'passive' preamp. I use short cables and the pot is 20k (ln) so no issues there, but I will be adding a final buffer stage just for versatility to allow longer cables if wanted (may enable both buffered and non-buffered outputs).
On to opamps. So much written I know but I will keep it brief. It came with TL072 in both stages. I am a great fan of TL072 from long back and in these low or unity gain circuits I didn't see a compelling reason to change, but I had socketed the boards and so experimented a little. Tried 4558, NE5532 and AD712 (as used in more upmarket Cambridge Audio A3i). I measured all of them for noise (pretty negligible in all cases but measurable differences) and did some listening tests.
OK my ears are not good and I am not therefore a great guide on sound quality - whatever that means. In all honesty I really couldn't reliably tell them apart. Its quite useful that the tone circuit is switchable so its possible to switch one op amp out of circuit and this did reveal a very slight degradation due to the 4558 - very slight, but female vocals just didn't sound as convincing (at least not to me). For the others I couldn't have reliably told whether the tone circuit was on or off with a blind test. So on this criteria alone the 4558 drops out of contention.
On to noise. Interesting. The differences are tiny and all versions are getting close to the noise floor of my ancient Levell muV meter but here it is. The noisiest (in this circuit) was the TL072, the quietest was the AD712. The total measured difference (full volume, tone on) was only around 10muV from best to worst but given the low levels I was measuring (20-30muV) I guess this is quite a large percentage. However .... it was impossible to tell them apart in practice unless my ear was right up against the speaker. In that case I would say the TL072 noise was a bit more objectionable / noticeable as in being more popcorn like.
Then I am going to add a criteria. DC offset. No surprise here the NE5532 was worst by a country mile. So what? Well the NE5532 implementation would be unusable without some switch on delay relay / switch off mute. I say that because if ever by accident you were to switch off that preamp first (or switch it on last) with NE5532 on board there would be very loud, very unpleasant and possibly damaging loudspeaker cone excursions. I tried it without tone enabled and volume turned down and it was bad enough. The TL072 by comparison was relatively benign (but you would still want to avoid a mistake of powering down preamp with power amp on). The AD712 was just a dreamboat. I couldn't measure any DC offset and it starts up and shuts down almost silently.
Perhaps no surprises - a circuit designed with JFET opamps works better with JFET opamps. But then the performance of the AD712 in comparison with TL072 did take me aback. Received wisdom is this is an ancient 80s opamp and offers no real benefit over the TL072. And I hadn't realised the benefit of DC precision is potentially simplifying a design. It was therefore measurable better and much more practical. In fact for the audiophile approach, DC offset is so low i could easily remove at least two DC blocking capacitors per channel if I wanted to. Picture of new internals just for the hell of it.
Replace pinch rollers in Becker BE1480
I have a Becker BE1480 tape deck in my car and lately it has eaten two cassettes. I'm thinking that I'd need to replace the pinch rollers in this 32 yo deck which would seem to be long overdue.
When I open the deck I see a tiny plastic disc holding the pinch roller in place. My question is: can I pry/pop off the disc in order to replace the pinch roller or should I not tamper with this piece?
When I open the deck I see a tiny plastic disc holding the pinch roller in place. My question is: can I pry/pop off the disc in order to replace the pinch roller or should I not tamper with this piece?
Filament Hum Bypass
- By edwardjurich
- Tubes / Valves
- 12 Replies
I'm sure others have come across this when selecting a power transformer. For a project I'm working on, I found a power transformer with the perfect voltage and current high voltage winding and a perfect filament current rating. Problem is the filament winding does not have a center tap. I've had the same situation before and a solution I have found that cancels background filament hum is to add bypass caps from each side of the AC filament line to chassis ground. I find ceramic caps work best for bypassing and 50V multilayer ceramic caps are small enough to fit anywhere. I ground the caps to the chassis and not directly to audio ground circuits. The 120V AC power inlet third prong ground also connects to the chassis and not directly to audio grounds. Although audio grounds do connect to the chassis, all audio grounds are wired together. Copper wire is a better conductor than a chassis giving some degree of isolation between audio ground and AC ground. I have never had a hum problem (so far 😊).
Aluminum electrolytic capacitors manufacturers in china
- By topcb
- Vendor's Bazaar
- 0 Replies
Asia'x capacitors,Yunxing Aluminum Electrolytic Capacitors,Yunrui capacitors agent.TOP Quality,ISO9001 Certified RoHs & REACH Approved,Cooperated with PHILIPS lighting for many years, no customer complaint.
https://www.super-capacitors.com/
https://www.super-capacitors.com/
Black Core Fr4 PCB,VT-42 Laminate
- Vendor's Bazaar
- 0 Replies
Hello Everyone,
We can make a beautiful PCB boards with FR4 black core board,also we can make different soldermask like transparent,purple or the other colors you like...
If are interested in,you can contact us!
Sophia
We can make a beautiful PCB boards with FR4 black core board,also we can make different soldermask like transparent,purple or the other colors you like...
If are interested in,you can contact us!
Sophia
Help finding a budget mono amplifier
Hello all.
I am posting after a long time . First i am a complete noob on this amplifier thing.
2nd sorry if i did/about to do any mistake out of my lack of knowledge. I intend to buy this particular tda 7293 board because it is cheap and convenient.
100W Digital Power Amplifier Board TDA7293 / 7294 High Power Dual 12-32V Single Channel Audio Amplifier Module for DIY Speaker
a.aliexpress.com
My subwoofer is a 75w RMS old Philips one.
Can i drive it optimally with it?
I intend to run this amp with a 35-0-35 VAC EI transformer (yep i am mr. Cheap 😛 ). One of the top most priorities is to keep it economical .
Other options:
1. Tda8954th btl mode
2. STK 4048
PLZ help
I am posting after a long time . First i am a complete noob on this amplifier thing.
2nd sorry if i did/about to do any mistake out of my lack of knowledge. I intend to buy this particular tda 7293 board because it is cheap and convenient.
100W Digital Power Amplifier Board TDA7293 / 7294 High Power Dual 12-32V Single Channel Audio Amplifier Module for DIY Speaker
3.62US $ 18% OFF|100W Digital Power Amplifier Board TDA7293 / 7294 High Power Dual 12 32V Single Channel Audio Amplifier Module for DIY Speaker|Operational Amplifier Chips| - AliExpress
Smarter Shopping, Better Living! Aliexpress.com
My subwoofer is a 75w RMS old Philips one.
Can i drive it optimally with it?
I intend to run this amp with a 35-0-35 VAC EI transformer (yep i am mr. Cheap 😛 ). One of the top most priorities is to keep it economical .
Other options:
1. Tda8954th btl mode
2. STK 4048
PLZ help
Tilting drivers.
- Multi-Way
- 25 Replies
Focal in their top range likes to tilt their mids and woofers.
Is there any pro to do this?
Maybe time alignment or lobbing?
Is there any pro to do this?
Maybe time alignment or lobbing?
The X-FoBooster is a transformer-like in-line booster
- By Elvee
- Analog Line Level
- 3 Replies
Here is another project inspired by Danny: it is based on the same principles as the BoosterPhone, low supply requirements, rail-to-rail output on a low impedance, in this case 600 ohm and decent audio quality without too much complications.
It is designed to match older sources, delivering 150~200mV to modern equipments requiring 700mV or more. It thus mimics a passive step-up transformer, but without the impedance penalty: a 1:4 step-up transformer increases the impedance by a factor of 16, which can be problematic. Instead, the output impedance is much lower than the input.
Although the X-FoBooster requires a supply, unlike a copper & iron one, it has no power switch and lasts forever on a single charge: the supply current is comprised between 55 and 85µA.
First, the simplest version optimized for 3.7V:
This is the graphical version, based on the real prototype. The sim uses the native transistor types of the LTspice.
Now the 5V version:
Here is the non-inverting version: a bit more complex, but it has a self-centering quiescent output bias, and a THD of 0.03% against 0.1% (at 0dBm/600ohm, actually measured)
The values shown as comments are the ones used in the physical circuit. The others are suitable for the sim.
All the transistors are also 2SA1015/2SC1815, except the lower output transistor Q6, which is a BC547C.
The output is referenced to the + supply, and it requires a DC link to work properly. If it used on a cap-coupled input, it is necessary to include a dummy load of 1 to 2K.
I didn't build the various versions as complete projects, because I have no use for them (for the moment), but I breadboarded them, and the measurements are based on the breadboarded circuits.
The 3.7V version can use the same battery management as the Boosterphone, and it will last even longer
It is designed to match older sources, delivering 150~200mV to modern equipments requiring 700mV or more. It thus mimics a passive step-up transformer, but without the impedance penalty: a 1:4 step-up transformer increases the impedance by a factor of 16, which can be problematic. Instead, the output impedance is much lower than the input.
Although the X-FoBooster requires a supply, unlike a copper & iron one, it has no power switch and lasts forever on a single charge: the supply current is comprised between 55 and 85µA.
First, the simplest version optimized for 3.7V:
This is the graphical version, based on the real prototype. The sim uses the native transistor types of the LTspice.
Now the 5V version:
Here is the non-inverting version: a bit more complex, but it has a self-centering quiescent output bias, and a THD of 0.03% against 0.1% (at 0dBm/600ohm, actually measured)
The values shown as comments are the ones used in the physical circuit. The others are suitable for the sim.
All the transistors are also 2SA1015/2SC1815, except the lower output transistor Q6, which is a BC547C.
The output is referenced to the + supply, and it requires a DC link to work properly. If it used on a cap-coupled input, it is necessary to include a dummy load of 1 to 2K.
I didn't build the various versions as complete projects, because I have no use for them (for the moment), but I breadboarded them, and the measurements are based on the breadboarded circuits.
The 3.7V version can use the same battery management as the Boosterphone, and it will last even longer
