Something louder than a Ruby?

[davidallancole]

Aavid Thermalloy makes a heatsink for the 8 pin DIP chip - http://www.aavidthermalloy.com/cgi-bin/stdisp.pl?Pnum=580100w00000g  It will increase the power dissipation to 1.8W instead of the 1.25W.

I honestly think increasing the voltage on these chips with a 4-8 ohm load will not actually increase your power out.  BTL wouldn't increase the power either.  I hooked up a test chip this morning and these are the results I found using a 4 ohm load:

Vs    Vout
6V     1.7Vp
7V     1.8Vp
8V     2Vp
9V     2.1Vp
10V   2.1Vp
11V   2.2Vp
12V   2.2Vp
13V   2.3Vp

If you take these results and multiply the voltage by 2 to get peak to peak, you will see they follow closely to the chart called "Peak-to-Peak Output Voltage Swing vs Supply Voltage".  From these numbers I noticed that even tho I increased the voltage from 6V to 13V, the output only increased by 0.6V. 

I believe this chip current limits around ~0.5-0.6A.

[Mark Hammer]

If you open up this file - http://hammer.ampage.org/files/Miniamp.zip - you'll see the commercial heatsink I used on my LM380-based amp.

Quite frankly, if you have a bit of sheet aluminum and are handy with tin-snips, or even a nibbler, you can make your own.  Whatever you fashion, just make sure there is maximum contact between sink and chip so that the sink can efficiently suck that heat away.

Just as a sidenote, I would encourage those who whip up layouts for small battery-powered amps to leave room around the power-amp chips for heatsinks, so that folks who want to experiment with higher supply voltages and such can do so conveniently.  That adaptation would presumably include a means/space/holes to secure the heatsink to the board, and ground it.

[Top Top]

Aavid Thermalloy makes a heatsink for the 8 pin DIP chip - http://www.aavidthermalloy.com/cgi-bin/stdisp.pl?Pnum=580100w00000g  It will increase the power dissipation to 1.8W instead of the 1.25W.

I honestly think increasing the voltage on these chips with a 4-8 ohm load will not actually increase your power out.  BTL wouldn't increase the power either.  I hooked up a test chip this morning and these are the results I found using a 4 ohm load:

Vs    Vout
6V     1.7Vp
7V     1.8Vp
8V     2Vp
9V     2.1Vp
10V   2.1Vp
11V   2.2Vp
12V   2.2Vp
13V   2.3Vp

If you take these results and multiply the voltage by 2 to get peak to peak, you will see they follow closely to the chart called "Peak-to-Peak Output Voltage Swing vs Supply Voltage".  From these numbers I noticed that even tho I increased the voltage from 6V to 13V, the output only increased by 0.6V. 

I believe this chip current limits around ~0.5-0.6A.

This makes sense to me. I was using a 9v battery on my ruby, and then switched to a power supply that was reading 14v, and there was no particularly notable increase in volume.

[davidallancole]

Thanks for verfying this Top Top.  It looks like if you need more power than the LM386 provides, the LM380 like what Mark Hammer used in his miniamp is the way to go.

Has anybody looked at the TDA7267A or TDA7267T?  They use power dip chips like the LM380.  You can use a copper pad on a PCB for the heatsink with these and they only require 3 capacitors to function (plus supply bypass)

[doc_drop]

I recently rebuilt a Crate amp that I had blown the electronics on....don't ask how....

Anyway, I built one of these kits http://www.vellemanusa.com/us/enu/product/view/?id=521723

I added a ROG Tonmender circuit in front for the EQ portion. I also put a Sparkle Boost between the power amp and Tonemender as an extra gain/overdrive, but that was overkill. Anyway, at 3 Watts, the circuit is easily powerfull enough for the 8 Ohm speaker. With the Tonemender I can get a lot of sounds out of it. As long as I gain stage it correctly it is pretty clean as well. But it can really scream if I push it....

Just another option to think about.

[davidallancole]

doc_drop,

The manual for that kit says it uses the TDA7267A chip.  Do you remember the price for that kit?

[doc_drop]

Hmmm. I found it in one of the few Brick and Mortar electronics stores left here in the San Francisco Bay Area... I think it was $13 or something like that. It was very cheap, so I figured I would use it to experiment. It worked great, so I decided to use it to rebuild my Crate.

[Top Top]

Hmmm. I found it in one of the few Brick and Mortar electronics stores left here in the San Francisco Bay Area... I think it was $13 or something like that. It was very cheap, so I figured I would use it to experiment. It worked great, so I decided to use it to rebuild my Crate.

What is the name of the store, out of curiosity? I am in the Bay Area.

[Brymus]

Aavid Thermalloy makes a heatsink for the 8 pin DIP chip - http://www.aavidthermalloy.com/cgi-bin/stdisp.pl?Pnum=580100w00000g  It will increase the power dissipation to 1.8W instead of the 1.25W.

I honestly think increasing the voltage on these chips with a 4-8 ohm load will not actually increase your power out.  BTL wouldn't increase the power either.  I hooked up a test chip this morning and these are the results I found using a 4 ohm load:

Vs    Vout
6V     1.7Vp
7V     1.8Vp
8V     2Vp
9V     2.1Vp
10V   2.1Vp
11V   2.2Vp
12V   2.2Vp
13V   2.3Vp

If you take these results and multiply the voltage by 2 to get peak to peak, you will see they follow closely to the chart called "Peak-to-Peak Output Voltage Swing vs Supply Voltage".  From these numbers I noticed that even tho I increased the voltage from 6V to 13V, the output only increased by 0.6V. 

I believe this chip current limits around ~0.5-0.6A.

This makes sense to me. I was using a 9v battery on my ruby, and then switched to a power supply that was reading 14v, and there was no particularly notable increase in volume.

Try setting up 1 x386 running at 6V then use a LM317 with a pot so you can dial in the voltage from 6-18+V on 2 x 386 in PP BTL
And then tell me there is no audible difference between the two set ups.
Of course you may not have noticed the .2 watt difference between 9-14V
It takes a factor of ten to double the percieved volume.
Hence a 386-1 at 6 V makes .325 watts with 10% THD not very impressive.
take-2 x 386-4 at 16V makes 3.4 watts or 3400milliwatts at 3% THD  Thats a facter of almost 10 which gives a doubling of the loudness.
AT a lower THD rate so it sounds not only louder but cleaner too.
As far as voltage swing I show a max P-P swing of 2V with a 4V supply (386-1) and a max P-P swing of 10V with a 16V supply (386-4)
thats more than double.
Are you guys actually breadboarding the chips?
And taking your readings,cause IDK how our findings could be so different.

[doc_drop]

Top Top, I'm always happy to help out a homie.

It is Electronics Plus on 4th Street in San Rafael. Here is their web site: http://www.electronicplus.com/

I don't buy much there since it is very, very expensive compared to our buddies like Small Bear and Futurlec, but it is fun to browse. They have an amazing array of enclosures that I will never buy because of the price. Besides, before I had a clue about buying this stuff, I went there with a couple of parts lists. Not knowing enough to double check, they happily sold me a number of wrong componants and things like metal power jacks. You know the kind of parts we can't use in our stomps... I don't blame them too much since they don't specialize, so they don't know stomp box quirks. But they should know the difference between an A pot taper and a B taper.

[davidallancole]

So I breadboarded the single chip amp as well as the BTL setup.  (I did breadboard the single chip with the 4 ohm load were I got my previous results from Brymus I didn't just make those values up.)  It looks like an 8 ohm load does not follow as closely to the curve shown in the datasheet "Peak-to-Peak Output Voltage Swing vs Supply Voltage".  These voltages are all with an 8-ohm load and the LM386N-1.

1 Chip at 9V – 6.2Vp-p out (0.6 Watts rms)
2 Chips in BTL at 9V – 8.8Vp-p out (1.2 Watts rms)

1 Chip at 9V – 6.2Vp-p out (0.6 Watts rms)
1 Chip at 12V – 8 Vp-p out (1 Watts rms)
1 Chip at 15V – 8.8 Vp-p out (1.2 Watts rms)

It looks like you do get more power going with the BTL or higher supply voltage.  I can't test the perceived loudness playing my guitar through it right now since all my babies are sleeping, but I can't see the difference being that great. 

The original poster was looking for an amp that would be louder than his Ruby, so to do that he will need something closer to 5-6 Watts to hear much of a difference.  Just boosting the voltage on his Ruby or making it BTL will probably not boost the volume very much.  Something like the LM380, LM384 or that Velleman kit that doc_drop posted in my opinion would be the way to go.

[Brymus]

Hey David
I see the reason now,partly my fault for not being clearer.
I was talking about not just upping the voltage and going BTL but going from a single LM386-1 (NJM386B) to two LM386-4 (NJM386BD)
Creating the big difference in volume,as most people buy an LM386-1 at Radio Shack and use an inefficient spkr then they dismiss it as "not that good" or being too quiet to be useful.
Thats not what the original poster said but its been said in these forums before.
I was just trying to show how to get more volume from a Ruby, or a 386 based amp.
As per my attempts the last 4-5 months trying to extract the most from this versatile little chip amp.
I agree for a decidedly much louder amp a higher power chip would be better like a TDA2003 or the other ones you guys mentioned.
        NO DISRESPECT INTENDED, My apologies if it came across that way. I have found your posts useful on many occasions.

[davidallancole]

Not a problem.  These forums make it hard sometimes to communicate effectively.  Something gets lost in translation it seems.  My apologies as well for anything I may have said that was inappropriate.

I see what you are saying now about the 386.  I looked back at some of your other posts in this thread and re-read them.  I agree that there is a large difference in volume between 1 386 running at 6-9V pushing a 4" or 5" speaker compaired to 2 386's in a BTL pushing a 4x12 or something similar.

I have a few questions for you:

1.  Have you been using 16 ohms speakers with these chips?
2.  You said that you ran 2 chips in BTL with 18V supplies.  Did the chips get very hot running like this?
3.  Have you tried any heatsinks with these chips?

Thanks.

[petemoore]

  Speaker, at these watts, it's like moped speed, exactly how good are the tires and is the chain oiled influenced the top speed #.
  Higher effeciency here = that was a good place to start.
  Bass, takes lotsa current to drive a speaker to "SPL'' up some bass, sacrifices here may provide greater rewards to the upper range freq levels or clarity, it is worth having a bass rolloff if trying to push it.
  After a little of this, a little of that, I begin to think 'Pile It', create an amp pile, a pile of amps.
  Which then leads to the other amp, do another power supply / heat sinks etc.
  That pretty much takes me to the realm of the Firefly or the other "Loud-dinky-watt amps", I have made plenty of amps, but I just haven't made one small enough yet. At 5watts [even 2watts] that is a LOUD high effeciency 12'' speaker in a large room, I think I need more like <1watt

[davidallancole]

I feel the same petemoore.  For the volume I need in my basement, 50-100mW does it for me.  I have built a few amps with unity gain and it winds up that I have my pedal turned right down to get the volume in the right area for me.

[Brymus]

Not a problem.  These forums make it hard sometimes to communicate effectively.  Something gets lost in translation it seems.  My apologies as well for anything I may have said that was inappropriate.

I see what you are saying now about the 386.  I looked back at some of your other posts in this thread and re-read them.  I agree that there is a large difference in volume between 1 386 running at 6-9V pushing a 4" or 5" speaker compaired to 2 386's in a BTL pushing a 4x12 or something similar.

I have a few questions for you:

1.  Have you been using 16 ohms speakers with these chips?
2.  You said that you ran 2 chips in BTL with 18V supplies.  Did the chips get very hot running like this?
3.  Have you tried any heatsinks with these chips?

Thanks.

Hey no problem I am glad we can exchange ideas here 
1) YES,but I went nuts and cleaned the local thrifts outa PCspkrs and wallwarts a few months back so I have tried spkrs from 1-12"
up to 64ohms in many configurations.
I tried Jensen Jets Electric lightning,Jensen Mods 5",Eminence Lady Luck,Celestion clones,Crate custom,vintage alnico,vintage ceramic,home stereo spkrs,car spkrs ,
square TV spkrs,my Polk Audio refrence spkrs,4,5,6" home theater spkrs (some sounded pretty good)telephone spkrs,and some I am probably forgetting.
The guitar spkrs sound the best of course.
2)Yes when you upped the volume or gain to make em louder than at a lower V,I would put one finger on each chip and strum some notes and turn the volume up,
They actually get pretty hot before they start sounding like poo,but then you have to shut em down right away or you will release
the "magic smoke" like I did several times (melted part of a breadboard too )
3)Yes but it was a PIA and after comparing the cost of heatsinks and chips VS watts I decided a more powerful chip amp was a better approach.
Still I plan to come back to this and try adding slave chips to try and get 5-6 watts,I want to get as close as I can to the volume and tone of my EVJ,just for fun.
To some it may seem like a waste of time doing all of this but it was very educational to me,and I like the way the 386 sounds.

[davidallancole]

I always wanted a Epiphone Valve Junior, but I figured I would never be able to play it loud enough to get the benfits of the tube distortion.  One of the downsides of being married I guess.

I definately never fault a guy for trying something thats against the norm.  It may be cheaper/more efficient to buy a bigger power chip, but its not quite as fun as going crazy and building what you want.

[arawn]

I am not to sure about alum inum foil as a heatsink sounds really dicey to me.
I have seen exectly what happens wjhen you overvolt an n-1 386 and It is not pretty! Basically explodes!
bettter off to look at heatsinks at mouser and use a self adhesive type. Be sure to use a heatgun and heat the sink and the chip before apply. just don't pick up the heat sink with your bare hands.

[Mark Hammer]

I am not to sure about alum inum foil as a heatsink sounds really dicey to me.
I have seen exectly what happens wjhen you overvolt an n-1 386 and It is not pretty! Basically explodes!
bettter off to look at heatsinks at mouser and use a self adhesive type. Be sure to use a heatgun and heat the sink and the chip before apply. just don't pick up the heat sink with your bare hands.

If you're referring to my suggestion to use "sheet aluminum", I did not mean aluminum foil (which also comes in a sheet, so I can see the basis for the misunderstanding).  What I meant was 16 gauge aluminum or something similar; the sort of aluminum "sheet" you would see cover plates made from.  For that matter, you can make it out of tin cans.  The key things to aim for are to provide lots of surface area, make sure it can fit flush against the chip without shorting out against any pins, and that it can be affixed to the board somehow in an unobtrusive way.

[earthtonesaudio]

I ran a LM380 as a square wave oscillator, and injected audio to the input to make a super simple class D amp.  These chips are not built for this mode of operation, and it did get hot.  But it didn't die.  I never tested it thoroughly so I don't know if it's worth doing.