I think the comment about Seas being inappropriate was based on efficiency and power handling. Tri-amping and a waveguide will take care of that to a alrge extent.
The 27TDFC ought to be good for down to 1500 in the MCM waveguide that Zaph used. Note that you don't have to go the route Zaph did to mount the tweeter. AJinFLA filled the back with bondo (auto body filler) and screwed the tweeter on from the rear. Mine are full of bondo, I haven't made the router jig yet.
If you want to go lower, you'll need a bigger waveguide, since this one starts rolling off around 1500 according to Zaph's graphs. I don't think I'd take the TDFC all the way to 1000 Hz, even with a cauer filter (recommended for 1500 Hz).
You could always make a waveguide to control your ribbon's dispersion. But, based on Zaph's tests I would prefer the 27TDFC route.
Don't forget you'll need to EQ that area where the upper end shelves off as the waveguide unloads. A first order shelving high pass will work for that.
I think that you misunderstood Calvin on baffle step. Open baffle speakers will have a 6 db/octave roll off startging at a frequency determined by the baffle width. The difference is that open baffle speakers keept rolling off where a closed box stops rolling off after 6 db.
So using a wide/U/H baffle will help delay the roll off until you cross to a sealed woofer. See Linkwitz for further reading. Yes, damping a U shaped baffle will help control response and the effects of floor bounce and room gain will reduce the apparent dipole loss. Look for the NAO for more info on u shaped baffles.
In case you hand not discovered this yet, the power supply is one of the most expensive parts of an amp. Especially if you need to ship transformers. Good luck on your repairs.
The 27TDFC ought to be good for down to 1500 in the MCM waveguide that Zaph used. Note that you don't have to go the route Zaph did to mount the tweeter. AJinFLA filled the back with bondo (auto body filler) and screwed the tweeter on from the rear. Mine are full of bondo, I haven't made the router jig yet.
If you want to go lower, you'll need a bigger waveguide, since this one starts rolling off around 1500 according to Zaph's graphs. I don't think I'd take the TDFC all the way to 1000 Hz, even with a cauer filter (recommended for 1500 Hz).
You could always make a waveguide to control your ribbon's dispersion. But, based on Zaph's tests I would prefer the 27TDFC route.
Don't forget you'll need to EQ that area where the upper end shelves off as the waveguide unloads. A first order shelving high pass will work for that.
I think that you misunderstood Calvin on baffle step. Open baffle speakers will have a 6 db/octave roll off startging at a frequency determined by the baffle width. The difference is that open baffle speakers keept rolling off where a closed box stops rolling off after 6 db.
So using a wide/U/H baffle will help delay the roll off until you cross to a sealed woofer. See Linkwitz for further reading. Yes, damping a U shaped baffle will help control response and the effects of floor bounce and room gain will reduce the apparent dipole loss. Look for the NAO for more info on u shaped baffles.
In case you hand not discovered this yet, the power supply is one of the most expensive parts of an amp. Especially if you need to ship transformers. Good luck on your repairs.
Yes for sure !
But the big filtering caps and the diode bridge are already integrated to the boards, so I "only" need toroidal transformers which are not so expensive (36€ for 330VA). Always cheaper than buying Rotel amps Anyway... Coming back to my speakers, I had a great idea this night : instead of a waveguide, why not use directly a compression driver with a real horn crossed at 800Hz ?
Compression drivers are said to give the best details possible. What do you guys think about this ?!
Well, the boards are pretty big, and the PSU part is separated from the amps part. Just that they are on the same physical board.
I have heard these amps and they beat anything I have already heard. They have both power, control and refinement.
What do you think of the DE250 driver from BC speakers ?
http://www.bcspeakers.com/download/prodotti/PDF/DE250.pdf
it looks pretty linear for a HF driver.
I've never heard it, but from a FR perspective I am sure you could do worse.
The dual hump impedance characteristic may be a result of the horn, I am not sure. However, even allowing for that there are a few ragged spots that seem to indicate resonance issues.
You might try it and if it sounds "like a horn" add some low density foam to the throat, ala Geddes to control higher order modes. This will require some EQ to compensate for the absorption.
High order modes are where the sound reflects from one side of the horn to the other in addition to the desired unified wave front. this is what Geddes indicates is the cause of horn sound.
The dual hump impedance characteristic may be a result of the horn, I am not sure. However, even allowing for that there are a few ragged spots that seem to indicate resonance issues.
You might try it and if it sounds "like a horn" add some low density foam to the throat, ala Geddes to control higher order modes. This will require some EQ to compensate for the absorption.
High order modes are where the sound reflects from one side of the horn to the other in addition to the desired unified wave front. this is what Geddes indicates is the cause of horn sound.
The 5532 and 5534 is not the best choise of op-amp for audio.... In theory, since many newer op-amps perform much better in measurements.
But audio is all but theory and measurements. Audio is about great musical performance, and somehow the 5532 and 5534 perform better than many of these new devices, despite worse technical spec.
Remember to use the "A" version that is specielly selected for audio-use. Also be aware, that the Philips devices sounds better than all the others. Don't know why, but maybe they use a different process in production.
Also did you know, that the 5532 series is one of the only op-amps ever made primary intended for audio??
But before ever using op-amps in audio, consider the discrete alternative. Even though there is great difference in sound of different op-amps, they all sound like op-amps.
try using real transistors, and consider not using negative feedback. This will perform worse in matter of distortion and noise, but your ears will love it
But audio is all but theory and measurements. Audio is about great musical performance, and somehow the 5532 and 5534 perform better than many of these new devices, despite worse technical spec.
Remember to use the "A" version that is specielly selected for audio-use. Also be aware, that the Philips devices sounds better than all the others. Don't know why, but maybe they use a different process in production.
Also did you know, that the 5532 series is one of the only op-amps ever made primary intended for audio??
But before ever using op-amps in audio, consider the discrete alternative. Even though there is great difference in sound of different op-amps, they all sound like op-amps.
try using real transistors, and consider not using negative feedback. This will perform worse in matter of distortion and noise, but your ears will love it
Notice that the AD797 will hardly be optimum here. I would suggest some good JFET opamp but beware of distortion introduced if the input impedances not are matched. This is important if the feedback not is lowohmish.youyoung21147 said:I plan to use some AD797 but they are really really expensive so I won't use them everywhere.
-106dB for NE5532 and -114dB THD for NE5534 (-134dB for AD797 !)
I'd be glad if someone could help because I'm really lost !
See page 5.108
http://www.analog.com/library/analogDialogue/archives/39-05/Web_Ch5_final_PtB_F.pdf
Good stuff here also:
http://www.analog.com/library/analogDialogue/archives/39-05/Web_Ch6_final_I.pdf
I just want to say that NE5532 is one of OP-s that play music with all it´s fullness, emotion and natural timbral balance.
Some have more explicit detals or briliance baut NE is all about balance and music from top to bottom.
My favorite.
I see it all the time in some of best sounding broadcast equipment.
I have Studer D731 beatuyfull sounding CD player and it is full of NE5532.
Some have more explicit detals or briliance baut NE is all about balance and music from top to bottom.
My favorite.
I see it all the time in some of best sounding broadcast equipment.
I have Studer D731 beatuyfull sounding CD player and it is full of NE5532.
I did not hear any difrences between various makers.
I have Philips , Label (S) and Texas instruments NE´s.
Only diffrence is that Tex. ins. have slightly curved corners housing.
My notes of sound quality of these and every other OP model goes on quality of voltage regulator in power supply, specific value of cap after reg. and shortest posible path between reg and OP.
Offcourse all OPs are very gratefull to have current buffers after output to help them driveing interconnects, earphones........
Biasing in A class is not needed anymore if buffer is in action, and you know, not all OP´s react same to class A push.
I have Philips , Label (S) and Texas instruments NE´s.
Only diffrence is that Tex. ins. have slightly curved corners housing.
My notes of sound quality of these and every other OP model goes on quality of voltage regulator in power supply, specific value of cap after reg. and shortest posible path between reg and OP.
Offcourse all OPs are very gratefull to have current buffers after output to help them driveing interconnects, earphones........
Biasing in A class is not needed anymore if buffer is in action, and you know, not all OP´s react same to class A push.
OK, I have very very little time these days now I'm in an engineering school. (I hope I'll soon learn about opamps lol, because maths really suck )
peranders : I will use low-omish feedback like suggested in the AD797 datasheet (surely 680R or so). What do you mean with "matched impedances" ?
Gasho : I hope I'll not be disapointed by NE5532/NE5534, your experience seems very positive !
I will supply this stuff with a Zener + emitter//follower reg, to get low noise & impedance.
Cheers !
peranders : I will use low-omish feedback like suggested in the AD797 datasheet (surely 680R or so). What do you mean with "matched impedances" ?
Gasho : I hope I'll not be disapointed by NE5532/NE5534, your experience seems very positive !
I will supply this stuff with a Zener + emitter//follower reg, to get low noise & impedance.
Cheers !
If you do regs like zenner based emmiter follower, place LC filter between zenner diode and transistor base. Choke can be small by dimension and high serial impedance is not a problem. Higher inductivity better filtration off zenner hum. Cap in paralel(from base to ground) can be 10-22uF and little higher working voltage than zenner voltage. After transistor you must put a resistor thah shunt´s the output of reg. It can be 2W resistor and his value depends on out voltage. That resistor is pushnig a reg in Class A mode. A current that resistor draws from reg must be little higher than the max needed curent for active cirucit.
Cap at output can be 100-150uf low ESR type like Rubycon ZA/ZL series.
Cap at output can be 100-150uf low ESR type like Rubycon ZA/ZL series.
Good luck finding a TO-99 packaged 5532 - and no, the LM833D isn't an adequate substitute.
You might try substituting a National LM4562, this is a new audio dual op amp that came out in late 2006 and is available in TO-99 packages! It won't be cheap in TO-99 but you should be able to get them from Digi-Key. National says Digi-Key has about 168 in stock at this writing.
http://www.national.com/pf/LM/LM4562.html
You might try substituting a National LM4562, this is a new audio dual op amp that came out in late 2006 and is available in TO-99 packages! It won't be cheap in TO-99 but you should be able to get them from Digi-Key. National says Digi-Key has about 168 in stock at this writing.
http://www.national.com/pf/LM/LM4562.html
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.