M86 powered by Li Ion bat

Status
Not open for further replies.
Jump to Latest
P

Philfr

Member
Joined 2012
#1
Hi all,
After built an LM1875 powered by Li Ion batteries :
http://www.diyaudio.com/forums/chip-amps/273559-lm1875-p2p-solar-powered.html
I want study new project with "ultimate" sounding 🙂
LM1875 works well, better than some of my LM3886, for two reasons : low power and Li Ion. I test some other bat like NiMh, acid, GEL.., and for me the best are Li Ion cause they are not "noisy". Kaneda use this technology for example.
So I'm thinking about M86, maybe with LM3876 cause I have. Tom said that is no difference between 3886 and 3876.

Actually with LM1875 I have about 10 hours of listening, when bat is full they give 15 volts, after 10 hours only 11 volts.
How much hours I can expect with M86 with two bat 14,4 v 1500 mA ?
The only remaining problem is recharge, time for that is near 90 minutes. But I have four of them and I can quickly change discharging batteries.
Thank's for advices or some comments..
Phil.
 
#4
Philfr said:
So I'm thinking about M86, maybe with LM3876 cause I have. Tom said that is no difference between 3886 and 3876.
Click to expand...

Allow me to clarify a bit. There is one difference: LM3886 can supply 7 A (11 A TYP). The LM3876 supplies 4 A (6 A TYP).

If you're driving 8 Ω loads on up to ±32 V rails, there's very little practical difference between the two. I measured identical performance on the MOD86 with one channel running LM3886 and the other LM3876. They sounded the same to my ears as well.

For 4 Ω loads, the game changes. With a 4 Ω load, the LM3876 can run at up to ±16 V rails before it would run out of output current near the peaks of the output swing. The LM3886 can run up to ±28 V.

I may be a tad pessimistic here as I'm using the worst case numbers, but I prefer to make promises I can keep and statements I can back up with numbers.

Some people use the lower output current drive of the LM3876 as a safety feature in amps driving tweeters. Others select the LM3876 as it has lower worst case noise (8 uV RMS vs 10 uV RMS for the LM3886). In practical terms, the LM3886 and LM3876 are completely interchangeable if you're driving 8 Ω loads on up to ±32 V rails. For low power applications, in particular, there'd be no discernible difference between the two.

Tom
 
Last edited:
#5
abraxalito said:
14.4V won't put the opamp supply regulators (LM317L/337L) into regulation - they're expecting at least 16.5V.
Click to expand...

You could change the resistors setting the output voltage of the LM317/337s. I list the relevant values in the design documentation.

Another option, if one really wanted to open up a can of worms, would be to use a buck/boost converter to generate the supply rails. This way the amp would keep running at the same performance level even as the batteries discharge. The buck/boost would be set up to shut down once the batteries have been discharged (that's a common feature on buck/boost ICs).

Just a thought.

Tom
 
Last edited:
#6
Hi Tom,
Thank's..
So in my case I can use LM3876. Speakers are 95/97 db, and I'm happy with 5/10 watt 🙂
See below impedance curve of speakers.
Buck/boost, I'm afraid it change sounding, more noise or something like this (distorsion), I don't knaw ?

What would be the lower voltage limit, and the consumption of M86 with 15 volts ?
Phil.
 

Attachments

  • impedance_dad_fe126en_new.jpg
    impedance_dad_fe126en_new.jpg
    148.2 KB · Views: 244
#7
In theory, you can operate the MOD86 down to ±10 V. I say, "in theory", as you'll get rather low output power. You can probably count on about 7-8 Vpeak swing under those conditions, so about 3-4 W into 8 Ω. ±15 V would get you to 10 W into 8 Ω.

The LM3876 should be fine for your speaker on ±15 V rails. No issues there.

Tom
 
Status
Not open for further replies.
Top Bottom