AMP15-PS-XP, a beefup of AMP15

Jan

9 Comments

AMP15-PS is very high performance in terms of audio quality. It is also a very practical design, an all integrated amp with power supply and auxiliary power. But it has a limitation in a moderate maximum output current.

Now, a new version has been developed: AMP15-PS-XP with significantly higher output current capacity, making it suitabe also for 4 ohm loads

Maximum power with 4 ohm load in the 2x300W class
The new AMP15-PS-XP has four 30Amp transistors instead of the dual 9Amp transistors used on AMP15-PS
The footprint is the same as before, 160x100 mm
AMP15-PS-XP is suitable for 4 ohm loads, with up to +/-60V supply (2x42VAC)
Essentially the same layout, with some modifications done to fit the higher power transistors
On the AMP15-PS-XP, there are toroids, wound with Litz wire, instead of ready-made inductors
As before, there is an on-board, integrated power supply with heavy duty 27.000uF capacitors and onboard supplies for 5V and VN10.

In terms of THD+N, the AMP15-PS-XP shares the excellent performance of the AMP15-PS, THD+N measures near 0.005%THD+N at 10W, 1kHz into eight ohm load!

Like with the AMP15-PS, the noise floor is typically 120dB below the main signal. Dead quiet.

The AMP15-PS-XP significantly outperforms the original Tripath TA3020 designs, in some respects, with almost an order of magnitude!

Measurements on the original Tripath TA3020 reference design, shown here as a comparison. THD+N is typically in the order of 0.02%

Like other Tripath based amps, the AMP15-PS-XP has very high efficiency, low losses, but this one is perhaps exceptional: while testing, the amplifier was running at 50Wrms for long periods without any heat sink at all, just bare transistors with nothing attached, and they still they remained at a temperature of around 40-45C!

AMP15-PS-XP will start selling in the beginning of october. The price will be the same as for the current AMP15-PS.

---

9 Comments

Scratchy - 03-Oct-10, 02:29
- Reply
Nice!
This makes my AMP2's obsolete (as it includes the 5V on-board as well and is now 4ohm capable).
I was never a fan of the Truepath due to additional houeskeeping voltages required.
This one has it all on-board, just what I'm looking for!
Is this one bridgeable as well?

FFF - 03-Oct-10, 22:38
- Reply
I would like to see a AMP-2 performance test as well, with the new MosFets. Still seems (to me) a winner (even if old).
(with some graphs, like amp15-ps, amp15-ps-XP, amp9-basic, if at all possible).

Jan - 04-Oct-10, 15:02
- Reply
I am gradually doing measurements on all models as I work along. Hopefully, I can present them all side-by-side some time in the future. It is a bit time consuming.

At THD levels well below 0.1%, it does become important that the measurements are done exactly the same way to be comparable. It is easy to change sample rates, block lengths, averaging, load, grounding etc in a way that does show as a measurable difference in the diagrams, but hard to evaluate unless you are very experienced and know all he exact details.

I am looking for original Tripath demo boards with the original components to see if my measurements are comparable to the ones they did for the datasheet. All I can see now is that the sum of all little changes DO make a difference.

One thing has surprised me a bit with the AMP15-PS-XP is that THD is behaving so well in the tweeter range also with 120nS BBM setting.

Bridging; yes, you can (always) do it, but there are limits; Keep the Ipeak=Z/U (load impedance / supply voltage) reasonable. For this amp, keeping the Ipeak at around 20A max, would be a reasonable starting point.
- With bridged outputs, the maximum voltage is the rail-rail voltage
- With a single ended outputs, assume the max output voltage is one rail-ground
Example +/-60V, single ended (stereo) 4 ohms => Ipeak 60/4 A, approximately 15A for each channel,
Example +/-60V, bridged (stereo) 8 ohms => Ipeak 120/8 A, approximately 15A

Maximum power in these two cases:
Single ended +/- 60V: (60/1.41)^2/4 = 450W x2
Bridged +/- 60V: (120/1.41)^2/8 = 900W x 1

The power ratings here are too high, as that would be the outputs very hard driven, with high distortion. In these two cases, 300W and 600W with reasonable THD are perhaps better ratings.

For bridged 4 ohm loads and +/-60V or 2 ohm loads and use like PA or musical instruments playing long hours at very high average power, I'd still suggest AMP7

kifysara - 04-Oct-10, 15:43
- Reply
It'd be interesting to see how the 2050 high power boards compare to the 3020 ones, measurement-wise!

Scratchy - 05-Oct-10, 05:04
- Reply
Thanks Jan.
I certainly like the simplicity of the AMP15.

I'm building bridged AMP2's with +-48V rails with the stock current limiting resistors.
Should be good for 400W into 8ohms if driven hard, although I'm building it for SQ.
If this is successful, I'll have to build an AMP15-PS-XP and do a head-to-head sound comparison.

androuski - 08-Oct-10, 17:14
- Reply
Originally Posted by Scratchy

Thanks Jan.
I certainly like the simplicity of the AMP15.

I'm building bridged AMP2's with +-48V rails with the stock current limiting resistors.
Should be good for 400W into 8ohms if driven hard, although I'm building it for SQ.
If this is successful, I'll have to build an AMP15-PS-XP and do a head-to-head sound comparison.

Hello, nice boards
What is the lowest voltage required to make the board work properly ? I've got a 2x18V 225VA transformer, would it be sufficent ?

Jan - 08-Oct-10, 23:26
- Reply
2X18VAC / 2X25VDC is quite enough to run this amp. The minimum is probably around 2x10VAC.
If using it below 2x35VDC you need to change the undervoltage sensing resistors. Changing the gain would also be recommended. In alll 8 resistors need to be changed. When orrdering, let me know and I will throw in the right resistors for your voltage.

androuski - 09-Oct-10, 09:28
- Reply
Originally Posted by JanAF

2X18VAC / 2X25VDC is quite enough to run this amp. The minimum is probably around 2x10VAC.
If using it below 2x35VDC you need to change the undervoltage sensing resistors. Changing the gain would also be recommended. In alll 8 resistors need to be changed. When orrdering, let me know and I will throw in the right resistors for your voltage.

Thank you,
I guess that the gain should be increased ? Is this really necessary with speakers that have high sensivity such as 94-95db@2,83v ?

Jan - 10-Oct-10, 00:25
- Reply
Optimal gain setting is done in two stages, and is the same for all models that allow gain adjustment;

1) Input stage; independent of power supply voltage, you set the input sensitivity so that with maximum volume, the input signal times input gain is equals around 4V ptp. This means that the inputpt stage is fully used, up to near its 5V limit.

2) Set the power stage gain so that with fulll input volume (4Vptp) the power stage output ptp is near the rail voltage. This means the powerstage is optimally used.
- So the input stage optimum depends on the signal level from your source
- The power stage gain depends on your supply voltage. With a lower supply voltage, it is therefore better to lower the power stage gain. If the gain is too high, the power stage will clip (reach maximum power) also when the input stage is at low voltage, so it it not optimal from a noise point of view.
I made a posting about this a long time ago. I have to search for it. Fund it;

http://www.41hz.com/forums/content.p...Selecting-Gain

AMP15-PS-XP, a beefup of AMP15

vBulletin Message