Inside the Aura 2300 *PIC*

[envisionelec]

Provided Link: http://home.woh.rr.com/envisionelec/Aura2300/


Ok, this amp requires a LOT of current. I have a decent 13.8V supply and the requirements exceed by a wide margin. I hardly get 50W out before the instantaneous draw is too great.

I would attribute this to the toroid windings. I think that they could be a bit more efficient, but the designers certainly made up for that by fusing them with dual 40A fuses. The underside of the boards explain why the entire top layer is a split ground plane - it's absolutely packed with thoughtful layout techniques and adequate unmasking (where the solder forms a 'river' between lands). This increases current capacity by a large margin, but is a compromise because wide open solder fills tend to flow around when they get hot causing parts to fall out. Then again, if you don't abuse the amp and make sure the power connections are tight (a great heat source when loose), you'll have no problems.

On to the pics - you might want to print the list for reference:

1. Bottom of board under power supply and terminals.

2. Bottom of board under preamp/filter/gain sections. Lots of thought went into the layout to minimize (but not eliminate) above board jumpers. The long vertical solder strip is a fat ground trace for the preamp section. Note that it does not follow into the output section - a true star connection!

3. Lower left is speaker-level input. Lower right is +/- 15V zener/transistor pass regulator for preamp. Note large 2W 270 ohm resistors. These get hot and stink a little. It smells like burning plastic, but don't be alarmed. Upper left is a JRC13700 OTA that controls bass level when the "knob" is plugged in. Upper right is a TL074 opamp handling buffer duties after the 13700. The blue 1% metal film resistors are used appropriately rather than everywhere or nowhere.

4. Middle of pic is 5532 opamp handling gain duties. The pot to the left is the gain setting. As I stated earlier, the input section is quasi-balanced using the 5532 to "decode" the audio signal and ignore noise traveling down a ground loop. Inherent to this topology is a great gain structure that requires the pot to set actual opamp gain rather than shunting the audio signal to ground (which would defeat the purpose of a balanced input). The benefit to this is that as you turn the gain up, the overall noise of the system is masked by the increased sensitivity to the input signal. If you're a cardiophile like myself, you keep the gains down as low as possible to maintain the best S/N ratio possible.

5. 6 gang potentiometer for 18dB/octave crossover filter. TL074 to the right handles filter gain. Above that is a pair of BA4560 for bass boost and are only active when LPF is engaged. A single BA4558 sums the boost with the filter amp. A 4558 isn't all that bad when used with 1% resistors as it is in this case.

6. A triplet pair of TIP35C TIP36C transistors is used here. Bias compensation sensor located under foam piece between the pair pictured.

7. An orthogonal view of the entire output section. It is a bit hard to tell here, but the topology is conventional dual LTP/dual current source with triple darlington outputs. I have no problems with this, especially for driving low impedance loads.

8. This is a great look at the 2nd tier of the triple darlington kingdom. The transistors here are 2SC3117 and 2SA1249 Sanyo epitaxial 1.5A 160V parts. These parts will allow EACH of the TIP35/36C to deliver 37.5A of current before their drive voltage begins to droop. So that's around 110A total per channel. As you can see, the output section WILL ONLY be confined by the rail voltage and not suffer from final drive current depletion as is found in SO MANY car amps.

9. 3300uF x 2 63V capacitors per rail. Plenty at the switching frequency (agh, I didn't check THAT, LOL). I guessimate it's around 40kHz.

10. This part gets a laugh from me. A perfectly designed SMD board jutting out of the mainboard. This contains the TL494 and a quad comparator for protection. So, the high tech is relegated to a 1.5 square inch board. Heh.

11. IRFZ44N. Used in millions of car amps since 1994. Rugged, dependable and cheap. I would install IRFZ48V just cuz I like 'em. There are four per power supply (8 total). I personally would use 6 per rail for the power levels expected, but they'll do OK under normal audio conditions.

12. Toroids. Two of them. Between the toroids is a complementary totem pole MosFet driver. These increase the efficiency of the MosFets by decreasing switching time and increasing gate drive current. High Dv/Dt (for a 12V device, anyway, makes this a requirement.

13. US Quarter placed on top of a toroid for size comparison. BTW, size of toroids makes less difference if the core material is of correct permeability. I think these are fine.

14. Buss Bars on the right carry current from the bank of MosFets to a toroid. I initially found the arrangement odd, but now see that the idea was to keep the toroids' primary center tap close to the power input terminals while dividing the MosFet's current into four across multiple buss bars. Neat!

15. Finally, a view of the emitter feedback resistors of 0.22 ohms.

[erniecornick]

[robk]

...what's your final verdict on this amp as is, in non-modified form. I'm currently running a Rockford Punch 225 into my MTX thunderform with 2 MTX 10's. Would the Aura 2300 be a good replacement as far as power and is it better quality? I'm not sure of the year of the Rockford but it looks like the one below.

[dlsbobby]

> Ok, this amp requires a LOT of current. I
> have a decent 13.8V supply and the
> requirements exceed by a wide margin. I
> hardly get 50W out before the instantaneous
> draw is too great.

> I would attribute this to the toroid
> windings. I think that they could be a bit
> more efficient, but the designers certainly
> made up for that by fusing them with dual
> 40A fuses. The underside of the boards
> explain why the entire top layer is a split
> ground plane - it's absolutely packed with
> thoughtful layout techniques and adequate
> unmasking (where the solder forms a 'river'
> between lands). This increases current
> capacity by a large margin, but is a
> compromise because wide open solder fills
> tend to flow around when they get hot
> causing parts to fall out. Then again, if
> you don't abuse the amp and make sure the
> power connections are tight (a great heat
> source when loose), you'll have no problems.

> On to the pics - you might want to print the
> list for reference:

> 1. Bottom of board under power supply and
> terminals.

> 2. Bottom of board under preamp/filter/gain
> sections. Lots of thought went into the
> layout to minimize (but not eliminate) above
> board jumpers. The long vertical solder
> strip is a fat ground trace for the preamp
> section. Note that it does not follow into
> the output section - a true star connection!

> 3. Lower left is speaker-level input. Lower
> right is +/- 15V zener/transistor pass
> regulator for preamp. Note large 2W 270 ohm
> resistors. These get hot and stink a little.
> It smells like burning plastic, but don't be
> alarmed. Upper left is a JRC13700 OTA that
> controls bass level when the
> "knob" is plugged in. Upper right
> is a TL074 opamp handling buffer duties
> after the 13700. The blue 1% metal film
> resistors are used appropriately rather than
> everywhere or nowhere.

> 4. Middle of pic is 5532 opamp handling gain
> duties. The pot to the left is the gain
> setting. As I stated earlier, the input
> section is quasi-balanced using the 5532 to
> "decode" the audio signal and
> ignore noise traveling down a ground loop.
> Inherent to this topology is a great gain
> structure that requires the pot to set
> actual opamp gain rather than shunting the
> audio signal to ground (which would defeat
> the purpose of a balanced input). The
> benefit to this is that as you turn the gain
> up, the overall noise of the system is
> masked by the increased sensitivity to the
> input signal. If you're a cardiophile like
> myself, you keep the gains down as low as
> possible to maintain the best S/N ratio
> possible.

> 5. 6 gang potentiometer for 18dB/octave
> crossover filter. TL074 to the right handles
> filter gain. Above that is a pair of BA4560
> for bass boost and are only active when LPF
> is engaged. A single BA4558 sums the boost
> with the filter amp. A 4558 isn't all that
> bad when used with 1% resistors as it is in
> this case.

> 6. A triplet pair of TIP35C TIP36C
> transistors is used here. Bias compensation
> sensor located under foam piece between the
> pair pictured.

> 7. An orthogonal view of the entire output
> section. It is a bit hard to tell here, but
> the topology is conventional dual LTP/dual
> current source with triple darlington
> outputs. I have no problems with this,
> especially for driving low impedance loads.

> 8. This is a great look at the 2nd tier of
> the triple darlington kingdom. The
> transistors here are 2SC3117 and 2SA1249
> Sanyo epitaxial 1.5A 160V parts. These parts
> will allow EACH of the TIP35/36C to deliver
> 37.5A of current before their drive voltage
> begins to droop. So that's around 110A total
> per channel. As you can see, the output
> section WILL ONLY be confined by the rail
> voltage and not suffer from final drive
> current depletion as is found in SO MANY car
> amps.

> 9. 3300uF x 2 63V capacitors per rail.
> Plenty at the switching frequency (agh, I
> didn't check THAT, LOL). I guessimate it's
> around 40kHz.

> 10. This part gets a laugh from me. A
> perfectly designed SMD board jutting out of
> the mainboard. This contains the TL494 and a
> quad comparator for protection. So, the high
> tech is relegated to a 1.5 square inch
> board. Heh.

> 11. IRFZ44N. Used in millions of car amps
> since 1994. Rugged, dependable and cheap. I
> would install IRFZ48V just cuz I like 'em.
> There are four per power supply (8 total). I
> personally would use 6 per rail for the
> power levels expected, but they'll do OK
> under normal audio conditions.

> 12. Toroids. Two of them. Between the
> toroids is a complementary totem pole MosFet
> driver. These increase the efficiency of the
> MosFets by decreasing switching time and
> increasing gate drive current. High Dv/Dt
> (for a 12V device, anyway, makes this a
> requirement.

> 13. US Quarter placed on top of a toroid for
> size comparison. BTW, size of toroids makes
> less difference if the core material is of
> correct permeability. I think these are
> fine.

> 14. Buss Bars on the right carry current
> from the bank of MosFets to a toroid. I
> initially found the arrangement odd, but now
> see that the idea was to keep the toroids'
> primary center tap close to the power input
> terminals while dividing the MosFet's
> current into four across multiple buss bars.
> Neat!

> 15. Finally, a view of the emitter feedback
> resistors of 0.22 ohms.

I don't know jack about car audio, so don't scream at me. But isn't that kind of high? Or, is that typical of car audio? Or, I could be missing something. Robert.

[wg_ski]

> I don't know jack about car audio, so don't
> scream at me. But isn't that kind of high?
> Or, is that typical of car audio? Or, I
> could be missing something. Robert.

Typical of power amplifiers. What the maximum power is depends on how much distortion (clipping) it's driven to. 1% is visible clipping on an oscilloscope. These amps probably have .01% or so when not clipping - it's not all that hard to get to these days.