I've built LM4780 in a same chassis as LM3875, both using premium components and gave it to a friend. He preferred LM3875.
Initially, I had hard time choosing which sounds better, but eventually decided on 3875 as well.
The comments: "less control in bass/low midrange and less definiton in highs". 4780 seems to a bit loose was stated by two other people.
This with regards to parallel application of LM4780. However, there is much more drive and power with bigger chip, so use according to your preferences.
Initially, I had hard time choosing which sounds better, but eventually decided on 3875 as well.
The comments: "less control in bass/low midrange and less definiton in highs". 4780 seems to a bit loose was stated by two other people.
This with regards to parallel application of LM4780. However, there is much more drive and power with bigger chip, so use according to your preferences.
Nuuk said:What lightbulb thread Al?
http://www.diyaudio.com/forums/showthread.php?s=&threadid=45290
LM3875 vs LM3886 vs LM4780 vs OPA549
If properly implemented they will probably sound all very similar within normal levels (particularly on the 549), except that the 3886 and 4780 seem to handle difficult loads a bit better than the 3875.
Pick the one that is more practical to you. Just be careful with the supply you are going to use and the speakers you are going to drive.
Carlos
If properly implemented they will probably sound all very similar within normal levels (particularly on the 549), except that the 3886 and 4780 seem to handle difficult loads a bit better than the 3875.
Pick the one that is more practical to you. Just be careful with the supply you are going to use and the speakers you are going to drive.
Carlos
Re: Re: Re: LM3875 vs LM3886 vs LM4780
Drop me a line and I'll post a couple over, I have plenty
Carlos
While I totally agree on your comments on the LM series, I found the difference to the OPAs was greater at low levels. Hmm. I also found that the 3886 prefers a lot more capacitance at the PSU rails, IMHO 4700uf was best.
I really must get back to building the electronics side of my current active speaker project, as part of that is a big regulated supply I can use to try out different GC topologies.
ShieldJaguar said:
lol cheers pinkmouse, looks like i'll have to build one of these as well
Drop me a line and I'll post a couple over, I have plenty
Carlos
While I totally agree on your comments on the LM series, I found the difference to the OPAs was greater at low levels. Hmm. I also found that the 3886 prefers a lot more capacitance at the PSU rails, IMHO 4700uf was best.
I really must get back to building the electronics side of my current active speaker project, as part of that is a big regulated supply I can use to try out different GC topologies.
Re: Re: Re: Re: LM3875 vs LM3886 vs LM4780
On the tests we did with a friend, the 3886 behaved differently with small and high capacitance on the PSU. Up to 1500uF, response on treble and mid was cleaner, though bass was a bit "diffuse". More than 2200uF improved the bass a lot, but was not so good for the upper side.
A regulator with small capacitance at the output, such as that based on the LM338, should solve matters for all frequencies. Even so I am not so sure if using lots of capacitance, like 20000uF or so, would not be a better option on an unregulated PSU powering the bass side of an active system. Except if you over dimension everything, a regulator might compress the bass response.
A regulator will probably even response between all chips above. I wonder how much you can get, in watts, from the 549, as TI does not specify a figure for that. Strange.
Carlos
pinkmouse said:
On the tests we did with a friend, the 3886 behaved differently with small and high capacitance on the PSU. Up to 1500uF, response on treble and mid was cleaner, though bass was a bit "diffuse". More than 2200uF improved the bass a lot, but was not so good for the upper side.
A regulator with small capacitance at the output, such as that based on the LM338, should solve matters for all frequencies. Even so I am not so sure if using lots of capacitance, like 20000uF or so, would not be a better option on an unregulated PSU powering the bass side of an active system. Except if you over dimension everything, a regulator might compress the bass response.
A regulator will probably even response between all chips above. I wonder how much you can get, in watts, from the 549, as TI does not specify a figure for that. Strange.
Carlos
Carlos
To be honest I never really pushed my 549 amp, but it it felt like a comfortable 50W or so, and is currently working in a small commercial studio driving some NF monitors.
TI don't give any SOA curves as you say, so I can only imagine that if you have enough heatsinking it will run the full load!
To be honest I never really pushed my 549 amp, but it it felt like a comfortable 50W or so, and is currently working in a small commercial studio driving some NF monitors.
TI don't give any SOA curves as you say, so I can only imagine that if you have enough heatsinking it will run the full load!
chip sound
I build lm3886 with 1000uF PSU caps, and with my DIY 4ohm speakers sound is disaster. No bass and good treble. PSU is 250w transformer with 2x22v output. When I add some capacitors 8x4700uf +2x470uf on chip power supply pins with zobel on output sound is better bat still way behind my modded Pioneer 3xx amp. On same PSU TDA7294 amp sound much better. Maybe TDA chip is better solution for low impedance speakers?
I build lm3886 with 1000uF PSU caps, and with my DIY 4ohm speakers sound is disaster. No bass and good treble. PSU is 250w transformer with 2x22v output. When I add some capacitors 8x4700uf +2x470uf on chip power supply pins with zobel on output sound is better bat still way behind my modded Pioneer 3xx amp. On same PSU TDA7294 amp sound much better. Maybe TDA chip is better solution for low impedance speakers?
Re: Re: LM3875 vs LM3886 vs LM4780
Definately go with the 549. 541 seemed to sound good (and really shook my sub) but the chips kept starting on fire, so I'm staying away from them now.
Is the 549 a class AB or B? I've heard different things.
And, stex, the 549 doesn't mind low impedence. I've hooked up a 2ohm load to it in bridge mode (1ohm load per amp) and it didn't fail, got pretty warm, distorted and made a "popping sound" easily, but other than that sounded pretty good. I wouldn't call it stalble at 1ohm, but it was able to handle it. It was pretty stable with a 2ohm per chip load, though.
An 8, 4 or possibly even 2 ohm load with a low enough voltage works fine with a 549. Parallel it and you would get full power/stability into a 2ohm.
pinkmouse said:
OPA549
Definately go with the 549. 541 seemed to sound good (and really shook my sub) but the chips kept starting on fire, so I'm staying away from them now.
Is the 549 a class AB or B? I've heard different things.
And, stex, the 549 doesn't mind low impedence. I've hooked up a 2ohm load to it in bridge mode (1ohm load per amp) and it didn't fail, got pretty warm, distorted and made a "popping sound" easily, but other than that sounded pretty good. I wouldn't call it stalble at 1ohm, but it was able to handle it. It was pretty stable with a 2ohm per chip load, though.
An 8, 4 or possibly even 2 ohm load with a low enough voltage works fine with a 549. Parallel it and you would get full power/stability into a 2ohm.
Re: chip sound
Is it possible that your Pioneer has sloppy bass? And the chip amp gave less flab and overhang? Or was the LM3886 sounding limp and uncontrolled in the bass?
This worries me! I've just got 4 LM3886T chips, to experiment with (my first DIY amp or two), and I have 4ohm speakers!stex said:
Is it possible that your Pioneer has sloppy bass? And the chip amp gave less flab and overhang? Or was the LM3886 sounding limp and uncontrolled in the bass?
Re: chip sound
Yes, LM3886 doesn't sound as it could with 1,000uf caps, it gets severely starved and the bass is not good.
Go much higher in capacitance and it sounds "cheap" in the midband and treble.
Fortunately, there are solutions for this, and that's why I say that a chip alone doesn't make an amp.
The LM3886 sounds very good, but implementation is the key.
stex said:
Yes, LM3886 doesn't sound as it could with 1,000uf caps, it gets severely starved and the bass is not good.
Go much higher in capacitance and it sounds "cheap" in the midband and treble.
Fortunately, there are solutions for this, and that's why I say that a chip alone doesn't make an amp.
The LM3886 sounds very good, but implementation is the key.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.