So the max you would run this chip in in 2...i remember on your notes something about the 4886 or 7 whatever the stereo version is being capable of additional power albeit small. Also i believe you made a statement about pin 7 *i would have to go over your notes* being the critical pin and the traces being as short as possible for it. Opamp wise i have a ton of LM6171 now, a few of the cheaper TL082s, a few tl555CPs, tle2072a1, and some 555s. I believe you can buck boost voltage swing with a 555 chip, is doing that going to trigger SPIKE in 3886? Are you not a fan of solid state designs? i hardly heard people favor class D to an AB other than perhaps for subs here and there. What are your thoughts on the LME49811 as a high powered hybrid? Do you feel the SMPS are able to source a enough current for a powerful solid state design?
Gotcha - so it is sort of like watts to db. I thought there was another variable as well making that not an exact science more more of an approximation. Please correct me if i am wrong.
If you also consider that many amplifiers double power into half the load, when you go from one driver to two of those drivers in parallel you get +3dB from doubling of the driver area (what I have shown above) AND +3dB from doubling of the power for a total of +6dB. For the same power output by the amp you only get +3dB when you double drivers.
See for more info:
Frequently Asked Questions
Correct but a 4 ohm load is more taxing on the amp and distortion figures on the amp side are typically worse correct?
Yes, a 4ohms speaker draws more current than an 8ohms speaker.
The amplifier will perform better on the higher impedance because it is drawing less current.
The amplifier will perform better on the higher impedance because it is drawing less current.
Is there a point of diminishing return? for the 500W 4 ohm LFET amp i have, running the speakers (8ohm nominal) at 4 ohms still allows ample head room at max values - if i am not mistaken (off memory) 256w put them at 125 db spl -running in 8 ohms would basically double the wattage and just coming up shy of the 125 db at max - running into a 16 ohm load basically makes it a 125w mono block and should further improve the sonics and at max basically puts me at a little under 121db. Is the trend of load vs THD/SNR,etc linear or is a changing value were eventually it is as good as it will be.
Also, the harder things work, the more power is lost by heat, how am i able to identify the ideal max wattage for an amp per driver to minimize this effect?
better performance... well we should clarify this a bit.
0.000000000001% THD+N vs 0.00000000000099 % THD+n , clearly there is an improvement, a better performance.
question is, is that performance increase by any means detectable by humans ?
it differs from person to person. there is no way to tell if an improvement in performance will make any difference at all for a particular person or not.
// i must allso add tht some amps actually have worse performance with light loads. so it is quite impossible to answer your question, as it involves the topology of the amplifier, the person in question, the sound level you want, the speakers you use, and naturally the environment you gona set them up. so the answer might as well be 42.
0.000000000001% THD+N vs 0.00000000000099 % THD+n , clearly there is an improvement, a better performance.
question is, is that performance increase by any means detectable by humans ?
it differs from person to person. there is no way to tell if an improvement in performance will make any difference at all for a particular person or not.
// i must allso add tht some amps actually have worse performance with light loads. so it is quite impossible to answer your question, as it involves the topology of the amplifier, the person in question, the sound level you want, the speakers you use, and naturally the environment you gona set them up. so the answer might as well be 42.
Last edited:
That's interesting. Is that because the amp is providing too much current for the speaker? I know bjt front ends are great for different input and output impendences between a pre and amp. And that higher current makes 4 ohms sound better. Is this more up to the toroid and its capabilities or can it be compensated for in the current gain stage?
You may want to spend a little time with a quality amp and a distortion analyzer. I've spent extensive time with a high-end amp (Modulus-86) and the second-best distortion analyzer in the world (Audio Precision APx525) I have generally found the THD differences ranging from nearly none to 5-6 dB.
My Modulus-86 provides 0.000061 % THD at 40 W, 8 Ω (just before clipping) and 0.000067 % THD at 65 W, 4 Ω (just before clipping). This is a measurable difference and is the result of the output stage of the LM3886 having to deliver twice the output current.
In the Parallel-86, I measured 0.00020 % THD at 55 W, 8 Ω and 0.00033 % THD at 110 W, 4 Ω. That's a 4.3 dB difference!
That is a very relevant question. In case of the Modulus-86, I will argue that the THD is not audible in either the 8 Ω case or the 4 Ω case. However, as Andrew pointed out, there are amplifiers where the change from 8 Ω to 4 Ω will cause an audible shift in THD. With a poor layout it is easy to get that type of degradation due to the doubling of the load current, thus doubling the error voltage dropped across the ground connection(s).
Tom
Last edited:
Tom, who sells a good distortion analyzer and is there a software version? I have a lot of high end recording equipment and getting an almost picture perfect image of the sound would be pretty easy. I have a few good amps. Best one is a nice solid state AB discrete amp. While I appreciate a lot of things especially your knowledge of the lm3886 distortions vs sonic bliss vary from one to the next. I happen to think a lot of things should be a mirror reflection where your amp will be ideal. I also do enjoy the distortions of a great tube as well. So THD and THD+N don't seem to really matter as much as the sonic qualities of the amp itself while SNR power supply noise and other things of the ilk would pretty much never make a solid design. Just a wierd subjective game all the way around. Basic to my prior question about high power/gain and clean amps and being a fan of LFets got me thinking of a few things. Lfets are easy to drive but do need a few things to sound perfect and from a design issue it is hard to implement. The best ones without complex circuits offer or come with elevated driver rails. Can I use some bjt chip running at 15v for the front end while 60vs are pushing the driver circuit for more juice and a minimalist approach w tight tolerances easily repeatable results and cleaner sound then an all discrete amp? Basically it would be a headphone amp on steroids at end of the day.
The best in the world is the Audio Precision APx555. The base model starts at about $32k as I recall. The second best in the world is what I have, the APx525.
The Quaint Asylum QA400 is an interesting box. Undoubtedly, it's not as flexible as the AP, but it's nearly 40 dB less expensive.
A software version? What do you mean? Software can't measure anything without hardware. If you're looking for a soundcard-based audio analyzer, I suggest looking at True RTA.
Tom
The Quaint Asylum QA400 is an interesting box. Undoubtedly, it's not as flexible as the AP, but it's nearly 40 dB less expensive.
A software version? What do you mean? Software can't measure anything without hardware. If you're looking for a soundcard-based audio analyzer, I suggest looking at True RTA.
Tom
just read my post again, and again till its well understood.
3 dBL less noise , when the noise was allready under the detection level of the listening person is meaningless. as it simply makes no difference at all what so ever.
the threshold level is different from person to person.
it must be accounted.
but once you are under the level that can be detected, it makes no difference if you are under it by 1000 dBL, or just 1.
the result is the verry same.
3 dBL less noise , when the noise was allready under the detection level of the listening person is meaningless. as it simply makes no difference at all what so ever.
the threshold level is different from person to person.
it must be accounted.
but once you are under the level that can be detected, it makes no difference if you are under it by 1000 dBL, or just 1.
the result is the verry same.
I agree that if the THD is well below audible, it doesn't matter if it's 40 dB below audible or 43 dB below audible. I think your extreme exaggeration on the "0.00000000000000001%" THD diluted your message, however. I also disagree on the magnitude difference between 8 Ω and 4 Ω that you claimed as I've measured much larger changes. Nearly 6 dB in some cases. A 6 dB change is a 6 dB change.
Tom
Tom
Last edited:
sarcasm was intended yess 🙂
i do agree with you that generally a higher impedance load does indeed reduce distorsion, the 6 dBL change is 6 dBL, no argument about that eighter.
but will it make a difference for every listener ?
most likely a lot of them will simply not hear the difference at all.
it surely exsists, but its below threshold.
i do agree with you that generally a higher impedance load does indeed reduce distorsion, the 6 dBL change is 6 dBL, no argument about that eighter.
but will it make a difference for every listener ?
most likely a lot of them will simply not hear the difference at all.
it surely exsists, but its below threshold.
soild state amps and load impedance.
Look at it from the other direction.
Start with a high load impedance. The amp works well.
Halve the load impedance. The amp generally works well, sometimes one hears some audible oddities. Are these only when the music/voice becomes loud?
Halve the load impedance again. The audible oddities are much more frequent. They seem to be occuring during some of the not so loud passages.
Halve the load impedance again. The amp if well designed shuts down. If not well designed it damages itself !
Look at it from the other direction.
Start with a high load impedance. The amp works well.
Halve the load impedance. The amp generally works well, sometimes one hears some audible oddities. Are these only when the music/voice becomes loud?
Halve the load impedance again. The audible oddities are much more frequent. They seem to be occuring during some of the not so loud passages.
Halve the load impedance again. The amp if well designed shuts down. If not well designed it damages itself !
if even the half of the load impedance is within the specs what the amp was designed for , i would still say the difference will be marginal.
surely if you have a lod impedance that the amplifier can not drive properly then i will distort like hell or explode or shut down, that is no rocket sciense.
surely if you have a lod impedance that the amplifier can not drive properly then i will distort like hell or explode or shut down, that is no rocket sciense.
- Status
- Not open for further replies.