The Frugalamp by OS

homemodder · 2009-03-14 6:38 pm

Os youd be surprised how good car audio can be, when I say competition I dont mean the loudness SPL competitions. We dont favour any frequencies or use equalizers or active crossovers, systems are from source to amps to speakers, just like home systems, the subs have pasive crossovers. One has to look hard for finding someone using class D amps , maybe one here and there will use this to power a sub. We have only one hickup, I have a truck battery in the boot, a large car battery under the engine cap and a second costum made alternator which produces 120 amps at idle and 200 amps at cruise speeds.. The sound quality in my car is within 97 percent of the quality of my home based system. We put the loudness SPL freaks to shame when it comes to quality although some of our systems dont lack too much when it comes to power either. Most of the guys just use 400 to 600 watt total music power.

I ll have a look at the new stuff from fairchild.

Steve Dunlap · 2009-03-14 7:00 pm

Michael Chua said:
Hi OS

Mike will do.

I prefer metal cans for bass. Don't know why they sound different. May well be due to their construction. Basically, the MJ15024/5 sound more muscular with bass. This is one of the reasons for using them in my kits. The problem with these older TO3s are the mids and highs. It takes quite a bit of work to get them to sound smooth like the newer 30MHz power transistors.

For subwoofer use, I believe the TO3 outputs in your circuit will perform better. Your amp is quite "fast". I expect the bass to have a higher resolution - tighter and more dynamic with the MJ15024/5 or MJ21193/4.

Nothing scientific about this. A purely subjective opinion.

Mike

I have also found this to be the case with my amps. I don't know why, but there seems to be some agreement here among builders that have tried multiple types of outputs. Just coincidence?

Michael Chua · 2009-03-15 1:15 am

Hi Steve

No, I don't think its coincidence. I believe some can actually make out the differences simply by listening.

For myself, it is very obvious that bipolars sound different from mosfets. IRFP mosfets sound different from the 2SK laterals too. Don't know why. I suppose its due to the way they are manufactured.

NS sent me some opamp samples for testing last year. I was stunned by the LME49720. I said to myself "At last, an opamp that is as good as a discrete". Then they sent me another lot. This time the same opamp but in metal can (TO-99). I couldn't believe the difference. The metal can LME49720HA is miles better than the plastic DIP packing. I don't know why. They don't know why too and they are the manufacturers. Doesn't matter. Just keep on manufacturing them is good enough.

Audio is fun isn't it.

Cheers
Mike

ostripper · 2009-03-15 2:01 am

By M. chua - I believe some can actually make out the differences simply by listening.

I have both in the same room .. I must admit, the 15024/5's
are the "hands down" winner.. sub wise. I have not been able
to audition the plastic MJL 21193/4 , but in 15 days , I will..

I do suspect the die used is the main factor rather than the
package. The die of a plastic 21193 is the exact
same as the to-3 varient, the only difference might be the
package (the to-3 has more surface area?)

Most of the common TO-264 (plastic packages) are at reduced wattage (150-200w) - The 2sc5200/2sa1943 have a much smaller die (I have broke them open).

OS

Michael Chua · 2009-03-16 12:34 am

Hi OS

Glad you can make out the MJ15024/5. They not only have the oomph but you can actually get details in the bass.

I think it has more to do with just the die size. The Sankens MT200 should have pretty big dies inside too. Yet the MJ15024/5 is still better in bass.

My guess is the plastic perhaps slow the die from heating up when current is passing through and that somehow affects the bass performance.

Mike

ostripper · 2009-03-16 2:15 am

bY M. Chua -Glad you can make out the MJ15024/5. They not only have the oomph but you can actually get details in the bass.

It was a lot harder, I just hooked the wires up from my new
VGS to make the assessment. Laying out a TO-3 board
is more difficult from both the spacing/alignment considerations
and the total "footprint" of the layout. I would have to go to a
8" board to place 6 devices, but luckily I will build just 1.

My choice will be the MJ-21193/4 (TO-3), max SOA/wattage. :devilr:

I admire your "projects" section at ampslab , so ..for all the
beginners , a starter board will be posted. You can build
a C200 or a blameless/ bootstrap varient.. choice of
a current mirror or none , 1 or 2 pair of outputs...

Perfect for all tastes (attached).
OS

ostripper · 2009-03-18 3:38 am

NEW 4 in 1 UNIVERSAL "NEWBIE" BOARD

It was just a little harder to design a "universal" PCB, but
This one should do nicely for all the "schoolbook" amps.

With the "A" schematic , one can build a 60 or 100w
"blameless" , with performance matching a "brother of quasi"
or any other of the same topology. Also, one could also
variate the design.. omitting the current mirror (ampslab C200)
or running off 40 volt rails.

With the "B" schematic , a constructor can use the "bootstrapped'
method of sourcing the VAS and also have the flexibility
to use or not use the current mirrors. (similar to
DX amp , aska 55 , and many.. many others.)

Common to both designs are the choice of nested or
global HF feedback , fully scalable voltage requirements,
choice of 2/4 OP devices, and a fully symmetrical layout.

All heatsinking is done with 1/8" thick 1-1/2" "L" channel .

Now all 3 layouts ...

1. FA1_AB_4IN1_FULL.pdf
2. FA2T_FULL.pdf
3. KrillOPS_PCB.pdf

ARE NOW IN PDF , complete with silk (mirror) , main PCB , layout
and schematics/ HERE ..

http://ts1/pdf1/Electronics/Projects/Audio_amp/Frugalamp/

TO-3 and MOSFET versions and the BOM's (FA1 is done already)
are due to follow.
OS

blueskynis · 2009-03-18 3:51 am

Hi ostripper,

Should C4 be connected to negative supply line (-70V) to improve PSRR? By connecting it to ground you are only making it worse.

ostripper · 2009-03-18 4:02 am

I just looked at both c4 /schema and c4 PCB and it is attached
to "sig grnd." on the other side ...r10 , NFB and the base of
Q2.. that seems right. C4 is the DC component of NFB .. so
to (clean) ground it must go...

Look at self's book or any similar topology..

these are working models , so I hope there are no errors.
If you meant C5 that is "aka doug self" ..proper. (CCS decouple)
OS

blueskynis · 2009-03-18 4:12 am

No, I am talking about this capacitor (shown on picture).

ostripper · 2009-03-18 4:20 am

Thank you much , schema is wrong

.. luckily PCB ,silk and ,
layout are right...

this is good , for creative criticism.. caught that bug ger.
redownload in 2 min will be fixed.
OS

IT is FIxed

blueskynis · 2009-03-18 4:23 am

It happens to everyone...everybody makes mistakes, it's in human nature...

By the way, nice amps you are building

Cheers

ostripper · 2009-03-18 4:28 am

By the way, nice amps you are building

I thought you were referring to FA1 (C4).. that was the
misunderstanding.

I try to make them easy ,with no questions or errors .

BTW ,FA1 has been running for almost a year. and
2 a month.

Many more amps to come.. all free.

OS

AndrewT · 2009-03-18 9:36 am

ts1 could not be found.

Is this my browser settings or the address?

PB2 · 2009-03-18 2:23 pm

Considering all the claims here of differences in bass performance for different topologies, and output devices, it seems to me you all ought to have a get together. I'm sure some will claim that their ears are better than some of us more on the theoretical side, however it seems at most local meets, I'm the one who hears the problems with various systems, LOL!

Let me just point out that if you're going to make a claim that different toplogies or metal output devices have different bass quality you had better be sure that every LF time constant is the same in both designs, and that the power supplies are exactly the same. Otherwise, all bets are off IMO. I can hear the responses now ..... Don't know why I bother, LOL!

Any of you willing to do a blind A/B or A/B/X test?

I don't expect this to happen just throwing it out there.

Pete B.

AndrewT · 2009-03-18 2:40 pm

PB2 said:
.............Let me just point out that if you're going to make a claim that different toplogies or metal output devices have different bass quality you had better be sure that every LF time constant is the same in both designs, and that the power supplies are exactly the same. ...........

Yes, a valid comparison should be made.
In addition a 200degC 250W device has a much higher SOA than a 150degC 200W device.
Reducing the temperature swings of the junctions could have an effect on the sound quality of large current transients. Bass Power, or lack of, could be an audible effect that only shows in a comparison to the rest of the music spectrum, but is immeasurable as a frequency response aberration.

ostripper · 2009-03-18 3:51 pm

By andrew -Is this my browser settings or the address

Nope , It was my address error.. correct one is below
luckily, my ISP has let me have 71.203.210.93 for a long,
long time.
http://71.203.210.93/pdf1/Electronics/Projects/Audio_amp/Frugalamp/

By andrew -In addition a 200degC 250W device has a much higher SOA than a 150degC 200W device.

that is a valid point which I attemped to explore , since only 2
pair of mj15024/5's SOA match 3 pair of the njw's. Also I ran
wires to my to-3's, as the to-3p's were direct PCB.

I did the comparison on a FA1 without current mirrors
(no bootstrap) ,but I only swapped devices. So , it could
be a total current/SOA "effect" . Same 500VA trafo , same
12" peerless sub. I did listen to the to-3's on full range
speakers and could hear no difference, but on the sub ,
more objects in the room "moved". Until I have "matching" FA2T's
with both varients in the listening room , I won't be totally
convinced. :whazzat:

OS

ostripper · 2009-04-08 3:36 pm

Bigger power supply - newamp

Boy , there are many more considerations in "moving up",
as far as amplifiers go. :bigeyes:

The quality parts were piling up in my living room so I had to
build. Since I don't want to mess with small amps (and don't
have to) , I decided to do a little primitive layout work.