Hello DIY Audio Community!
I am currently building an active crossover for my Maggie SMGs. The Crossover can be seen here
http://sound.westhost.com/project09.htm
The article says that an output buffer is highly recomended, however It says to use an expensive op amp. I was wondering if it were a better Idea to use something like a Darlington Pair.
If a Darlington pair is not practical, would someone please point me in the direction of a suitable op-amp for the output buffer, As I am completly new to Op-amps
Thank you all so much
-Moose
I am currently building an active crossover for my Maggie SMGs. The Crossover can be seen here
http://sound.westhost.com/project09.htm
The article says that an output buffer is highly recomended, however It says to use an expensive op amp. I was wondering if it were a better Idea to use something like a Darlington Pair.
If a Darlington pair is not practical, would someone please point me in the direction of a suitable op-amp for the output buffer, As I am completly new to Op-amps
Thank you all so much
-Moose
hi alexmoose,
I'm in the process of building the exact same project. The board looks well designed -- and unless I got something wrong there is a space on the board for the output buffer.
For the opamps, I decided to go for the LM4562 for the whole board. These opamps are fairly recent, unity gain stable, and there is plenty of info (mostly very positive) on them in these forums. Digikey (order code LM4562NA-ND) has them for $5.50 each, which makes $38.50 for the 7 needed for the whole board.
other parts I will order from digikey for the P09 are:
3296W-200LF-ND - Bourns 20k trimpot
3013PH-ND - 100nf bypass caps
+ the various frequency caps
I'll get the resistors from Mouser -- Vishay/Dale RN60 1/4w
would be nice to hear what you decide on.
I'm in the process of building the exact same project. The board looks well designed -- and unless I got something wrong there is a space on the board for the output buffer.
For the opamps, I decided to go for the LM4562 for the whole board. These opamps are fairly recent, unity gain stable, and there is plenty of info (mostly very positive) on them in these forums. Digikey (order code LM4562NA-ND) has them for $5.50 each, which makes $38.50 for the 7 needed for the whole board.
other parts I will order from digikey for the P09 are:
3296W-200LF-ND - Bourns 20k trimpot
3013PH-ND - 100nf bypass caps
+ the various frequency caps
I'll get the resistors from Mouser -- Vishay/Dale RN60 1/4w
would be nice to hear what you decide on.
Okay, I was going to use Dirt Cheap Op-amps for the crossover, but expensive ones for for buffer stages (As the website seggested). I'll look into the one you seggested
I will probably get the resistors, and crossover capcitors from Tubedepot. my crossover only needs to be 2-way, which means 1 quad op-amp per channel.
Keep me informed on how it goes!
-Moose
I will probably get the resistors, and crossover capcitors from Tubedepot. my crossover only needs to be 2-way, which means 1 quad op-amp per channel.
Keep me informed on how it goes!
-Moose
Tim__x said:
Thank you for reminding me about one of the fundemental rules of DIY, and of life itself.
You say the NE5532 is perfectly suitable. Sounds great!
I do not mean to disbelieve you, but I am completly new to op-amps (i'm building a Solid State system after building a tube system). what Makes an Op-amp suitable for this? how does one determain what op-amps should be used for this property (perhaps something in the datasheet?) I do not know, and would forever be in your dept if you were to enlighten me.
Thank you
Moose
alexmoose said:
Me too.
alexmoose said:
from what I can gather, the ideal opamp for this design should be unity gain stable.
Rod says the "TL072 is the most economical option" -- though not necessarily the best.
There are also some threads on opamps in the
ESP forum , including recommendations in various places for the OPA2134, NE5532 and LM1458.
NYopu cannot use NE5532/34 for this purpose.
Input impedance is about 300kohm wich might sound like a lot, but in these filters this could affect the attack frequency and precision. You MUST use FET-input opamps for the filtering itself.
I will recommend OPA2134 as a cheap and good opamp for this purpose. Input impedance is almost infinite.
For output buffer you can use NE5532/34, darlington transistors or what you feel is right.
Input impedance is about 300kohm wich might sound like a lot, but in these filters this could affect the attack frequency and precision. You MUST use FET-input opamps for the filtering itself.
I will recommend OPA2134 as a cheap and good opamp for this purpose. Input impedance is almost infinite.
For output buffer you can use NE5532/34, darlington transistors or what you feel is right.
Nrik said:
Thx for the info on impedance -- checked the datasheet of the LM4562 I was planning to use and get two figures:
Differential Input Impedance: 30Kohms
Common Mode Input Impedance (–10V<Vcm<10V): 1000Mohms
presumably the common mode figure is the one to go by?
I checked on the NE5532 datasheet for the figure you mentioned, and there is only one figure -- the 300Kohm you mention.
I would say no. 30kohm is even worse.
If the frequency-determining resistors in your filter was say 3kohm you would shift the attack frequency 10%. If your resistors are of higher values (wich is not that hard to imagine) it would be far worse. And your attack frequencies will have dips because of the parallel-resistor in the input of the opamp is causing dividing networks in the lowpass, and a lower Q in the high pass filter section.
I realise that many people are fans of the sound made by NE5534, LM4562 and other bipolar input opamps, used as an amplifier, buffer, I/V converter etc.
But for filter circuits like these you really need the perfect opamp, with infinite input impedance and zero ohms output impedance, and that is just for the cicuit to operate properly.
We are not discussing sound quality here, just basic function.
It is not for fun that Rod Elliot recommends the TL072 (the preferred FET input opamp back in the days).
But hey: To find out yourself if I am right or wrong, use sockets, and try out the different opamps. Then you could also swap for even better types in the future.
OPA2134 datasheet
http://focus.ti.com/lit/ds/symlink/opa2134.pdf
Input impedance (both common mode and diff mode) 10^13 ohm or 10.000Mohms
If the frequency-determining resistors in your filter was say 3kohm you would shift the attack frequency 10%. If your resistors are of higher values (wich is not that hard to imagine) it would be far worse. And your attack frequencies will have dips because of the parallel-resistor in the input of the opamp is causing dividing networks in the lowpass, and a lower Q in the high pass filter section.
I realise that many people are fans of the sound made by NE5534, LM4562 and other bipolar input opamps, used as an amplifier, buffer, I/V converter etc.
But for filter circuits like these you really need the perfect opamp, with infinite input impedance and zero ohms output impedance, and that is just for the cicuit to operate properly.
We are not discussing sound quality here, just basic function.
It is not for fun that Rod Elliot recommends the TL072 (the preferred FET input opamp back in the days).
But hey: To find out yourself if I am right or wrong, use sockets, and try out the different opamps. Then you could also swap for even better types in the future.
OPA2134 datasheet
http://focus.ti.com/lit/ds/symlink/opa2134.pdf
Input impedance (both common mode and diff mode) 10^13 ohm or 10.000Mohms
You're welcome.
The OPA2134 sounds in my ears as almost nothing. It does not add or subtract much. So you can even use it as output buffer/amp.
If you want to go nuts and spend some more money, the OPA2604 sounds even more neutral.
But use sockets, and you are ready to find your own preferred opamps.
If you experiment, then use any opamp that:
- Have high input impedance
- Are unity gain stable
- Uses voltage feedback (not current feedback like some)
That's it. Good luck.
The OPA2134 sounds in my ears as almost nothing. It does not add or subtract much. So you can even use it as output buffer/amp.
If you want to go nuts and spend some more money, the OPA2604 sounds even more neutral.
But use sockets, and you are ready to find your own preferred opamps.
If you experiment, then use any opamp that:
- Have high input impedance
- Are unity gain stable
- Uses voltage feedback (not current feedback like some)
That's it. Good luck.
These are such great responses, thank you very much!
I have already doubled my knowladge of Op-amps.
I am used to tubes, where every little tiny part has a profound impact on the sound. are there any Op-amps that measure poorly, but everone loves the sound of?
by the way
10^7 would be 10 Megaohms
10^13 would be 10 Teraohms
I have already doubled my knowladge of Op-amps.
I am used to tubes, where every little tiny part has a profound impact on the sound. are there any Op-amps that measure poorly, but everone loves the sound of?
by the way
10^7 would be 10 Megaohms
10^13 would be 10 Teraohms
alexmoose said:
....would you really want that?
Well you can't really say that any opamps have really bad specs. Only the really old ones, type 741/1458/324 are a bit slow, but no-one really likes the sound of those.
...Except for use in distortion-pedals for guitars. I have substituted some decent audio opamps for these poor sounding ones on request from my guitarplaying coworker, and he loves it. There are webpages outthere for guitarfreaks explaning those freaks who wants to 'upgrade' their pedals, to find scrapped ghettoblasters to find these opamps.
And thanx for the whats-in-a-number lesson.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.