lineup said:
Hey man, if there's some of the signal that's un-desired and all other means have failed, it might be a good time to admit this, and improve the efficiency by slagging that screaming sucker's NFB. Like this:
From -input, a resistor from 22k to 27k
From this resistor, a capacitor from 10nF to 15nF
From that capacitor, a resistor of about 150k (or a dial)
From that resistor, the speaker +
And that is a tone control--gage to use it at the point where it "just barely" takes audiable effect and you'll have wiped out the screech without seriously harming the music. I'm saying "just barely take the edge off," because if you went farther, it wouldn't be hifi.
This is on the side of the NFB that can say "don't do it" to the amplifier.
Removing un-desired signals improves efficiency.
There are also artificial means to clear up the smashed bits of the signal, without dropping the gain on them (like we just did).
Parallel amp and/or blackgate input filter capacitor (or brainy little matching device) can hash/divide up the signal going in so that it won't get full of sine waves going out. Theoretically, any multipath arrangement can do this, and its up to you whether its a distortion or whether its a more-desired signal.
On the capacitance you mention, one way to avoid trouble is to avoid promoting it. Larger contacts are less condusive to capacitance and that most certainly includes the spacing inside your resistors. Some people believe to pass as much signal as possible. But, in this case, I'd like less trash signal and more audio signal. The popular band-aid fix is a large Caddock or the equally good 1/2w 1/2cent carbon film. Both are carbon, with larger spaced contacts and both are tube-amp centric. If these are mixed with metal film, you have a voltage divider that promotes the errata of the larger carbon resistor (hopefully accentuating its ability to remove or reverse a noise). Its this technique a distortion or is it quantity versus quality? Once again, its not for me to decide what sort of signal is the most efficient. That's up to you.
But, the LM3886 itself is just fine. If you can't hear its lovely bass and baritone, then cause a situation where you can. Cheers!
Did I just say that LM3886 + terrible cheating "could" make better fidelity than any TDA7294? Yes. And, I apologise if that offends anyone. The comment was directed towards Lineup because he and I both have similar audio preferences (of what is and isn't a desired signal).
EDIT: I think that "favorite" is about desire of relating to the efficiency of a "most desired" signal.
EDIT: I think that "favorite" is about desire of relating to the efficiency of a "most desired" signal.
danielwritesbac said:
Right, even if conditions look the same at first sight, they often are not if you look into details. So I think the whole discussion here will not help to find "the best" chip, because we have allot of different impressions (a question of taste) and allot of different conditions... which powersuply, which external components, which layout, which music and which loudspeakers have been used? So my comment was just my impression and everybody can make his own research. Maybe someone will see different results or will prefer somethig different. I did the test before I started to build my amplifier and finaly i ended up with the LM3875 for the Mid and High chanels and pralell LM3886's for the bass on a 3way system.
Seems different combination of component will have different sounds?
But for my own opinion, I believe that "simple is the best", if too much component inside the circuit, will pickup more noise, is it?
I agreen with "Tranceformer", as we are using different power supply, different component, so it is hard to compare, can we have a way to test different sound of each chip?
But for my own opinion, I believe that "simple is the best", if too much component inside the circuit, will pickup more noise, is it?
I agreen with "Tranceformer", as we are using different power supply, different component, so it is hard to compare, can we have a way to test different sound of each chip?
dynabok said:
I think your thread is really valid just like it is. For instance, Lineup and I and many others, have a lot of fun causing LM1875 and TDA7294 to mimic the sounds of tube amplifiers. EDIT: Those chips have facilited efficiency via helping create "a more-desired signal" for their given application.
And, without those chips, I'd be buying tubes because I'm just not smart enough to trick a transistor.
EDIT: So howabout the chipamp and its application listed together, like Tranceformer's inspired application of both LM3875 and LM3886, showing how they are used?
My votes are for the LM1875 and the OPA549
The OPA549 is very transparent but must be run at lower gain so you need a good pre. It can deliver a pile of current with the right power supply.
The thing I like most about the 1875...apart from the sound is only 5 pins (as are some old STK3875 chips manufactured for Sony).
In my experience the 3875 is quite punchy. I was not fond of my parallel LM4780....but it makes a great sub amp.
They are all fun to try, I'm sure the differences are mostly in the individual application, as any of them can be made to sound good. I don't truly believe there is a "best".
G.
The OPA549 is very transparent but must be run at lower gain so you need a good pre. It can deliver a pile of current with the right power supply.
The thing I like most about the 1875...apart from the sound is only 5 pins (as are some old STK3875 chips manufactured for Sony).
In my experience the 3875 is quite punchy. I was not fond of my parallel LM4780....but it makes a great sub amp.
They are all fun to try, I'm sure the differences are mostly in the individual application, as any of them can be made to sound good. I don't truly believe there is a "best".
G.
dynabok said:
danielwritesbac said:
Like dynabook and others,
I want to keep it as simple as possible, while still make it good audio.
My rule is
Yeah, some would use even Schumacher Formula One Ferrari car to go shopping.
It will possibly work. But is this the best solution?
Is this even one desired, wanted way to do it?
I can use one gigantic truck weight 10 tons (10000 kg)
to move away some old junk from my garden.
Junk, that I can put into the back of my Toyota Corolla.
And transport it where I want it.
-----------------------
danielwritesback
thinks Tranceformer have a good idea:
So how about the chipamp and its application listed together,
showing how they are used?
I think this is the best idea
Just tellling: This or that Transistor is best one.
Really does not give any good information, at all.
We should put the device in its CONTEXT.
In what situation, in what audio chain, environment,
would this chip LM1875 perform very good?
OK Guys, I can tell a bit more about my Chipamp.
In the beginning I used a big 600W transformer, fast diodes, big capacitors and so on. I designed a PCB and started to run some tests. Beside listening to music I tested the noise with a open input. I connected a headphone and spent allot of time by optimising the cabeling and shielding the transformer. But I was not 100% happy with the results because I still did hear some 50Hz hum and noise on the headphone. The next stage was a regulated power supply which was an improvement but not the final design. After several experiments I ended up with switch mode power supplys and a new PCB design.
I agree to what dynabok said : simple is the best
So I changed my first design to a inverting design which saved a cap in the signal path and I removed the pots and used resistors for a fixed gain. I am using a digital x-over inside the amp to get the signal for my 3-way speakers. In the whole path from the digital input (spdif) to the loudspeaker there is only one capacitor.
As a conclusion I would say the same that people here said before: THE DESIGN ARROUND THE CHIP CAN BE MORE IMPORTANT THAN THE CHIP ITSELF.
In the beginning I used a big 600W transformer, fast diodes, big capacitors and so on. I designed a PCB and started to run some tests. Beside listening to music I tested the noise with a open input. I connected a headphone and spent allot of time by optimising the cabeling and shielding the transformer. But I was not 100% happy with the results because I still did hear some 50Hz hum and noise on the headphone. The next stage was a regulated power supply which was an improvement but not the final design. After several experiments I ended up with switch mode power supplys and a new PCB design.
I agree to what dynabok said : simple is the best
So I changed my first design to a inverting design which saved a cap in the signal path and I removed the pots and used resistors for a fixed gain. I am using a digital x-over inside the amp to get the signal for my 3-way speakers. In the whole path from the digital input (spdif) to the loudspeaker there is only one capacitor.
As a conclusion I would say the same that people here said before: THE DESIGN ARROUND THE CHIP CAN BE MORE IMPORTANT THAN THE CHIP ITSELF.
Attachments
You are using switching power supply, is it will noise? I also think about to use switching power before, but I am afraid it will pickup a lot of noise? how can you do it?
By the way, I come up with a idea, is it possible that we can make them as module, for example: make power supply as a module, make regulator as a module and make chip amp as module(1875 chip amp as a module, 3875 chip amp as a module, 3886 chip amp as module).
so that we can change the chipamp and hear different sound from different chip and different circuit design? and also we can change the different regulator to see how it can affect the sounds, how about this idea?
By the way, I come up with a idea, is it possible that we can make them as module, for example: make power supply as a module, make regulator as a module and make chip amp as module(1875 chip amp as a module, 3875 chip amp as a module, 3886 chip amp as module).
so that we can change the chipamp and hear different sound from different chip and different circuit design? and also we can change the different regulator to see how it can affect the sounds, how about this idea?
What kind of noise would you expect? As you can see on the picture the powersupplys are very close to the aplifier PCB's. But from my experiance the results are better than everything I tried before. Even when everything is quiet (by night) and the volume is set to normal values I hear absolutely no noise out of the speakers. ( I hope this has nothing to do with my hearing )
This kind of powersupplys work with frequencys around 100KHZ which is out of the audio range. So I think it is better than 50 or 60 Hz of an ordenary linear PWS configuration.
Or imagine you would try to filter 100KHz and 50Hz with the same capacitor. The filter effect will be much higher on 100KHz.
)This kind of powersupplys work with frequencys around 100KHZ which is out of the audio range. So I think it is better than 50 or 60 Hz of an ordenary linear PWS configuration.
Or imagine you would try to filter 100KHz and 50Hz with the same capacitor. The filter effect will be much higher on 100KHz.
If you had used the switching power supply OR batteries for the test, you might could learn the differences between the chipamps; but, the test would not reveal any information about which is the best because the chipamps would not have been optimized for their given application.
Yet these differences could be useful towards revealing application specific information.
Yet these differences could be useful towards revealing application specific information.
1
I started with TA2020(amp6,from 41hz).It had much more separation in the low frequency,compared to entry level HT amplifier.Still too bright for my taste.
2
Next was an TDA2052 amp,made from scrap parts for workbench use.Tested it one day with proper size speakers and was hooked.The sound was smooth as velvet.Unfortunately the bass wasn't as good as with the D-class.
This amp was in use while the next project was completed.
3
Now I'm listening, surprise surprise,Lm1875 gainclone with regulated PSU.This one I like a lot.The low frequency is taken care of with amp6.Best of both worlds...
All is subjective ofcourse.
I started with TA2020(amp6,from 41hz).It had much more separation in the low frequency,compared to entry level HT amplifier.Still too bright for my taste.
2
Next was an TDA2052 amp,made from scrap parts for workbench use.Tested it one day with proper size speakers and was hooked.The sound was smooth as velvet.Unfortunately the bass wasn't as good as with the D-class.
This amp was in use while the next project was completed.
3
Now I'm listening, surprise surprise,Lm1875 gainclone with regulated PSU.This one I like a lot.The low frequency is taken care of with amp6.Best of both worlds...
All is subjective ofcourse.
Yep :
different chipamp = different sound? = .. i don't know by now
but different speaker = different sound? = hell yes!
why don't try a different pre - amp?.
the 5 pin series are my favourites, for the easy instalation.
i remember when i did hundreds mods before finalize my first lm1875 amp.
last time i'm finished my inv lm2030 + pedja's buffer in the same board, the same psu for the buffer,without regulator, n sounds just great. Amen.
well, sometimes i finaly find their perfect utility,
1875 for big n good efficient drivers,
surprise surprise 3886 good for small severe 4' 2 way
TDA1514a is heard so much like 3875,average.
TDA7394 dual bridge is best for car's.
bridged TDA7294 good for sub-woofer,
tda2005 stereo otl sounds better than its btl.
lm386 + nokia charger + light 4' paper just enough for pc's.
maybe someone will help tells what chipamp sounds close to class A?
different chipamp = different sound? = .. i don't know by now
but different speaker = different sound? = hell yes!
why don't try a different pre - amp?.
the 5 pin series are my favourites, for the easy instalation.
i remember when i did hundreds mods before finalize my first lm1875 amp.
last time i'm finished my inv lm2030 + pedja's buffer in the same board, the same psu for the buffer,without regulator, n sounds just great. Amen.
well, sometimes i finaly find their perfect utility,
1875 for big n good efficient drivers,
surprise surprise 3886 good for small severe 4' 2 way
TDA1514a is heard so much like 3875,average.
TDA7394 dual bridge is best for car's.
bridged TDA7294 good for sub-woofer,
tda2005 stereo otl sounds better than its btl.
lm386 + nokia charger + light 4' paper just enough for pc's.
maybe someone will help tells what chipamp sounds close to class A?
TDA
After a class D amp with 2 darlingtons (BDX66,67)/channel, for me, the best is the amp what I've build after the producer's schematics with TDA2030A as driver for the pair BD911,912.
For my sub, (150W, 8 ohms) I use a bridge with 2 TDA7294 and it works VERY good. (power supply: +/- about 30V at nearly ful load, from a transformer 2X25V ac)
After a class D amp with 2 darlingtons (BDX66,67)/channel, for me, the best is the amp what I've build after the producer's schematics with TDA2030A as driver for the pair BD911,912.
For my sub, (150W, 8 ohms) I use a bridge with 2 TDA7294 and it works VERY good. (power supply: +/- about 30V at nearly ful load, from a transformer 2X25V ac)
favourite chip amps
at the moment i'm using tda7294 as the main amp in my system, i'm using 45volt rails and as long as you stay within the limits of the protection circuits it is very detailed amp not even using any eq! pretty serious dynamics. i have a little lm1875 and i find the sound warmer than the tda, more relaxed anyone else have the same experience?
at the moment i'm using tda7294 as the main amp in my system, i'm using 45volt rails and as long as you stay within the limits of the protection circuits it is very detailed amp not even using any eq! pretty serious dynamics. i have a little lm1875 and i find the sound warmer than the tda, more relaxed anyone else have the same experience?
LM3875 question
Some weeks ago i "found" a Technics SA AX720.
Seems to work fine, i connected every speaker i had to it and it sounded pretty decent.
the 720 however requires an amplified subwoofer.
I wonder if i can hook up a home made 3875 amp (would be my first 3875) onto the designated connector
i don't need a woofer that will shake the neighbor's dinner, jsut enough to reflect the low notes.
Thanks
fyi i did build several 1875's and they turned all out to be decent amps, they are all except for one, spread all over Seattle
Some weeks ago i "found" a Technics SA AX720.
Seems to work fine, i connected every speaker i had to it and it sounded pretty decent.
the 720 however requires an amplified subwoofer.
I wonder if i can hook up a home made 3875 amp (would be my first 3875) onto the designated connector
i don't need a woofer that will shake the neighbor's dinner, jsut enough to reflect the low notes.
Thanks
fyi i did build several 1875's and they turned all out to be decent amps, they are all except for one, spread all over Seattle
Re: LM3875 question
Yes, of course, but you might enjoy LM3886's remarkably clear bass and its abilities with 4 ohm (subwoofer) loads. After all, your movie entertainment style unit will make demands and either those are met or there's clipping. I think that LM3886 will be able to meet the demands of the movie-centric subwoofer output channel effects.
Sometimes you see it doing just that:
http://www.mcmelectronics.com/product/50-6277
uvodee said:
Yes, of course, but you might enjoy LM3886's remarkably clear bass and its abilities with 4 ohm (subwoofer) loads. After all, your movie entertainment style unit will make demands and either those are met or there's clipping. I think that LM3886 will be able to meet the demands of the movie-centric subwoofer output channel effects.
Sometimes you see it doing just that:
http://www.mcmelectronics.com/product/50-6277
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.