Internal signs of a solid built amp?

General car audio discussion area.
User avatar
Sawzall
New Member
Posts: 62
Joined: Sat Jul 14, 2007 9:21 pm
Location: Washington, DC

Postby Sawzall » Wed Aug 01, 2007 4:15 pm

First off, my comments are not intended to dismiss Class D/T as bad or worse - hell, I was stupid enough to even buy stock in Tripath once upon a time (and lost my ***, but thats another story) because I believe in the products. I just don't think they are ready for high end audio, full range.

Sampling was a bad choice of words - but without regard there is some transform happening in order to get to the end, which is switching on and off a transistor (usually a MOSFET). Which is where the confusion starts for most people - switching on and off does not equal 1 and zero (In real digital world, you almost never use absense of voltage to represent zero, that happens by accident. Typically you use something like under 2 volts equals zero and over 4 equals one, for example) - And in the most basic approach, there is not a true "sampling" occuring, but as you noted above when you mentioned TI's DSP, there is in a great deal in the actual applications of analog switching amps.

Switching on and off is gonna be noisy, see any push-pull design, valve or solid state. So we minimize it with A/B designs.

But my real problems are clearly shown in TI's app notes page 12 of their TAS5414 chip. Look at the THD+N at one watt. It shows that the higher frequencys noise is more than .5% already - this for a device rated designed for a car reciever I am thinking. It does not give info on what the components of that distortion are, but I bet little of it is 2nd order so unless it has tweeters that roll off hard at 8K, this radio is gonna suck.

(I have only heard a few class D for home at this point, mostly based either on the ICE modules or TriPath chipsets - I will grant that it has been a couple of years since I did most of my listening as it was part of my due dilligence for investing in TriPath. Since then I have only really listened hard to the ARC unit and I am not sure whos math they are basing their amp on.

And boy does topology count here. The DIY guys have not had a great deal of luck so far rolling their own - just because somebody makes a chipset and an app note does not mean the design is done. Which is why I guess the ICE guys sell only modules and not even to the home crowd then. But they have hot rodded the cheap tripath unit to death.)


BTW, them Hales are some sweet speakers.
User avatar
eezip
New Member
Posts: 1009
Joined: Mon Dec 17, 2001 12:43 am
Location: Orange County, CA
Contact:

Postby eezip » Wed Aug 01, 2007 10:44 pm

First, you can't be faulted for buying stock on Tripath. They had the amplifier equivalent of Apple's media machine whirring away when they started :)

I think you are confusing implementation with realization. You are absolutely correct that not just any Tom, Dick, or Harry will be able to make a class D amp work well. It's a skill that relatively few individuals have demonstrated. And even organizations like Mueta have promised the world for years now and can't seem to deliver any product. Some class D amps are bad, others are good. The PL380 hits 0.004% midband at about 1kW/channel with both channels going and has a noise floor about 108dB without weighting. The Hypex UcD modules have exceptionally flat frequency response vs. load impedance and also post great THD numbers with exceptional noise floor due to the extreme PSRR of self oscillators. However, Philips' UcD demo board pales in comparison. Which just goes to show you that it's all in the implementation. There are certainly bad performing and bad sounding class D amps out there - but the technology isn't doomed. If you need to power a cell phone piezo running off a battery, power draw is critical which class D excels at. And the low power dissipation allows great integration which you see in the TI amps. And if ultimate performance is key, I haven't seen anything to compete with discrete solutions. Personally, all Tripath amps I've heard sound awful. And ICEpower amps don't blow my skirt off either although they don't make me cringe. I can understand if class D doesn't exactly do it for you, as personal preference is a big factor, but until you've heard the entire gamut of class D amps there's no way you can discount class D as a whole. Don't knock 'em 'til you try 'em :)

BTW, I've measured the TI TAS5261EVM down to 0.009% THD and, being an open loop design, it can maintain this distortion nearly out to clipping unlike most class D schemes. Plus, the guys who designed TI's technology [originally Toccata] are some seriously smart guys who give nothing up to the best engineers I've ever met. There was an article in TAS a few months ago you might want to read along with the articles Bruno Putzeys has published in periodicals over the past years. I think some may be available at http://www.hypex.nl.
HT: Sony KV-27FS13 | Oppo DV-981HD | Denon AVR-2807 | 7x Cambridge MC100 | Acoustic Research ARPR1010 || Stereo: Dynaudio Audience 42 | Hales Concept 3 prototypes | wouldn't you like to know... :)

Home PC: Yamaha RP-U100 | B&W LM-1 | Mpyre 65x || Work PC: Yamaha RP-U100 | QSC AD-S32T | Sony MDR-V600
User avatar
Sawzall
New Member
Posts: 62
Joined: Sat Jul 14, 2007 9:21 pm
Location: Washington, DC

Postby Sawzall » Thu Aug 02, 2007 10:21 am

Well, the investment thesis mistake on Tripath was the high percentage ownership by the CEO and head math guy. And the absolute fraud committed by the CFO / Asian agent who lied about asian market sales didn't help alot. But he had the potential for a great product that could own the market but has fail miserably as a business. Once again proving that good managment trumps a great product every day.

At the end of the day, I think we can agree that we are disagreeing at the margins - but are in mostly in agreement that D/T is fine for 95% or more of all amplification jobs. And I do agree that its got a hell of a lot better in a fairly short time - much improvement has occured in much less time consumed in the development of early transistor amps for example. Early Darlington bipolar configs came out in what the late 50's but it took the Japanese guys till the 80's till they had them refined enough to stick in recievers by the millions for the masses. In the end I think an individual's acceptence of D/T amps will be a personal one like valve vs. transistor, or even Mosfet vs Bipolar power transistors. 99% of the world will go "huh?" and not care.
User avatar
Sawzall
New Member
Posts: 62
Joined: Sat Jul 14, 2007 9:21 pm
Location: Washington, DC

Postby Sawzall » Thu Aug 02, 2007 10:36 am

And more on-topic. Number one sign of a good amp vs bad amp. Look for the copper. Meaning the transformers and inductors in the powersupply (and outputs if so equipped). They, along with the heatsinks, are usually the most expensive parts in the entire amp and are the bean counters first target when saving money of the BOM. Having alot of it does not mean its a good amp all by itself, but not having enough makes it a bad amp no matter what else is going on inside.

In a related area - storage caps, in terms of quanity and quality, can be tricky. (I mean inside the amp - don't get me started on "stiffening caps" - what they stiffen is typically NOT the rail voltage). Too much is too much, but too little is alot worse. How much really depends on the needs of the amp - but high quality caps right next to each output transistor is a good sign.
User avatar
Davagio
New Member
Posts: 11
Joined: Fri Jan 11, 2008 11:01 pm
Location: Tulsa,Oklahoma

question about cheap amps

Postby Davagio » Thu Jan 17, 2008 3:12 am

Ok i will be honest here guys i do not now much if anything about internals of amps.So for years now i have only looked at the outside of amps and judged them on there fuse ratings (3 30Amp 2 40Amp Etc.)For instance if you take a cheaper amp like the Blade Series From Boss the 3000D if im not mistaken it has 4 40amp fuses on it,now i know that amp is not doing anything near 3000watts peak im guessing more like 800-900 Peak if that,but could you give me a simple formula (Im sure this had been covered,Somewhere in this post) so i could actually know how to find an approx number of real watts with any amp

Im simple guys what can i say so if you actually do reply please keep it semi retarted,just for my sake.After over 7 years i feel i should be able to know how to do this i always just hooked stuff up and physically tried them.But it would be nice to know the formula,(If the manufacturer doesnt bull**** on the tech specs,i know the do on the outside of the box.)Thanks guys for any help:cwm22:
User avatar
JimJ[VT]
Posts: 2872
Joined: Tue Aug 12, 2003 11:10 am
Location: Hillsborough, NC/B'burg, VA
Contact:

Postby JimJ[VT] » Thu Jan 17, 2008 10:07 am

but could you give me a simple formula (Im sure this had been covered,Somewhere in this post) so i could actually know how to find an approx number of real watts with any amp


Ohm's Law, but it requires that you actually have the amp in possession and can test it into a load of known & constant impedance. You can't tell just by looking at it...in the tube world things are a little different, if I see a 2A3 single ended triode amp I generally know it's going to be around 3.5 watts, a 45 around 1.5, 300B around 8, etc, etc...but even then that's dependent on the biasing and up to the builder as to what they did...
User avatar
eezip
New Member
Posts: 1009
Joined: Mon Dec 17, 2001 12:43 am
Location: Orange County, CA
Contact:

Postby eezip » Sun Feb 17, 2008 12:32 pm

Davagio wrote:Ok i will be honest here guys i do not now much if anything about internals of amps.So for years now i have only looked at the outside of amps and judged them on there fuse ratings (3 30Amp 2 40Amp Etc.)For instance if you take a cheaper amp like the Blade Series From Boss the 3000D if im not mistaken it has 4 40amp fuses on it,now i know that amp is not doing anything near 3000watts peak im guessing more like 800-900 Peak if that,but could you give me a simple formula (Im sure this had been covered,Somewhere in this post) so i could actually know how to find an approx number of real watts with any amp


Not really. There are different power supply and amp topologies that can drastically change the efficiency curve of a product. If the amp is linear, it will require somewhere in the neighborhood of 2x the power input to make the same output power. And I can adjust biasing on a linear amp and deadtime on a class D amp to change idle loss significantly. I wouldn't personally draw many conclusions about the amps output power from the fuse rating without knowing anything more.
HT: Sony KV-27FS13 | Oppo DV-981HD | Denon AVR-2807 | 7x Cambridge MC100 | Acoustic Research ARPR1010 || Stereo: Dynaudio Audience 42 | Hales Concept 3 prototypes | wouldn't you like to know... :)



Home PC: Yamaha RP-U100 | B&W LM-1 | Mpyre 65x || Work PC: Yamaha RP-U100 | QSC AD-S32T | Sony MDR-V600
User avatar
flakko
New Member
Posts: 171
Joined: Fri Jan 28, 2005 8:24 pm
Location: So Cal

Postby flakko » Sat Feb 23, 2008 5:00 pm

mr eezip, could you possibly take a gander at this here amp? its a stock JBL 4ch 140w one from my 92 lincoln (thats pretty much all i know about it) Let me know what you think about it. i tried to be as detailed as possible in the pictures as well as get all the components brand and serial numbers:

Image

Image

Image

Image

Image

Image

Image

Image

Image

Image

Image

Image

Image


now here is something interesting. there are two of these little modules soldered from under the board:

Image


and they lead up to these places on the top-side (you can reference from the first pictures):

Image

Image



thanks so much for your time :)
User avatar
eezip
New Member
Posts: 1009
Joined: Mon Dec 17, 2001 12:43 am
Location: Orange County, CA
Contact:

Postby eezip » Sun Feb 24, 2008 4:09 am

It's incredibly difficult to draw any reliable conclusions from pictures. And having something in my hands isn't all that much better. The only real way to know is to test the amp - and even audio measurements don't necessarily tell you if an amp will sound good or not. In general, though, if it measures good it sounds good but the opposite isn't true [witness tube amps].

At any rate, here is what I can see;
-it looks like your garden-variety cheap amp with little regard to SQ.
-The NE5532 is a great dual op amp for audio when you don't need anything particularly special. It's at the bottom of the 'acceptable for audio' class of op amps but that doesn't mean it's any slouch - some very well-respected and -performing gear uses 5532s.
-by the number of BUZ71s, the TL594 PWM controller, the common anode diode and the other part next to it [likely a common cathode diode], the pair of voltage regulators on heatsinks, the pair of electrolytic caps by the diodes I suspect are secondary rectifiers, and being that it runs off of 14V, the power supply is probably a simple push-pull converter with bipolar rails.
-the transformer looks like a ETD-core. That's impressive as they cost significantly more than toroids or E-cores but have usefully lower leakage inductance - an important consideration for push-pull converters.
-the modules under the board must be the audio amps. That's probably not good - I've never seen a module that performed better than a discrete amp, but they're often cheaper.
-power flow isn't all that great. The output of the power supply wraps back over itself to get to both amps and then to the output. The input section looks like it's probably not too bad, though.
-the technology I can see is nothing to brag about but probably perfectly fine for a 16-year-old factory system.

The amp doesn't look like it breaks any new ground, even when it was new, and I could be wrong just looking at pictures. With 2008 as my reference, the amp likely is pretty weak; that sounds mean but it's the honest truth. 16 years ago it might not have stacked up to poorly but I can guarantee it was far from the best-sound and -performing amp you could get at the time. Not to be a dick, but why are you interested in an amp you must have known was nothing special??
HT: Sony KV-27FS13 | Oppo DV-981HD | Denon AVR-2807 | 7x Cambridge MC100 | Acoustic Research ARPR1010 || Stereo: Dynaudio Audience 42 | Hales Concept 3 prototypes | wouldn't you like to know... :)



Home PC: Yamaha RP-U100 | B&W LM-1 | Mpyre 65x || Work PC: Yamaha RP-U100 | QSC AD-S32T | Sony MDR-V600
User avatar
flakko
New Member
Posts: 171
Joined: Fri Jan 28, 2005 8:24 pm
Location: So Cal

Postby flakko » Sun Feb 24, 2008 11:43 pm

eezip wrote:It's incredibly difficult to draw any reliable conclusions from pictures. And having something in my hands isn't all that much better. The only real way to know is to test the amp - and even audio measurements don't necessarily tell you if an amp will sound good or not. In general, though, if it measures good it sounds good but the opposite isn't true [witness tube amps].

At any rate, here is what I can see;
-it looks like your garden-variety cheap amp with little regard to SQ.
-The NE5532 is a great dual op amp for audio when you don't need anything particularly special. It's at the bottom of the 'acceptable for audio' class of op amps but that doesn't mean it's any slouch - some very well-respected and -performing gear uses 5532s.
-by the number of BUZ71s, the TL594 PWM controller, the common anode diode and the other part next to it [likely a common cathode diode], the pair of voltage regulators on heatsinks, the pair of electrolytic caps by the diodes I suspect are secondary rectifiers, and being that it runs off of 14V, the power supply is probably a simple push-pull converter with bipolar rails.
-the transformer looks like a ETD-core. That's impressive as they cost significantly more than toroids or E-cores but have usefully lower leakage inductance - an important consideration for push-pull converters.
-the modules under the board must be the audio amps. That's probably not good - I've never seen a module that performed better than a discrete amp, but they're often cheaper.
-power flow isn't all that great. The output of the power supply wraps back over itself to get to both amps and then to the output. The input section looks like it's probably not too bad, though.
-the technology I can see is nothing to brag about but probably perfectly fine for a 16-year-old factory system.

The amp doesn't look like it breaks any new ground, even when it was new, and I could be wrong just looking at pictures. With 2008 as my reference, the amp likely is pretty weak; that sounds mean but it's the honest truth. 16 years ago it might not have stacked up to poorly but I can guarantee it was far from the best-sound and -performing amp you could get at the time. Not to be a dick, but why are you interested in an amp you must have known was nothing special??


lol thanks man i appreciate your in-depth explanation of the circuitry. the reason i was asking is because i heard a buddies system with the exact same components (speakers and amps) and the only thing that was different from mine was that he had a cd changer in the back (note all stock equipment).

to be compeletly honest, it sounded amazing. the sound was warm, detailed, and didnt fatigue at all. i loved it. the only thing i didnt like was some of the midbass was coming from the back, but thats just speaker placement. I just wanted to know if there was anything special hidden in this amp.

again, thanks for your time man :)

Return to “Car Audio”

Who is online

Users browsing this forum: No registered users and 1 guest