Amp 2 pwr supply & upgrade ?s

[mr. ray]

In anticipation of receiving my AMP 2 kit which I ordered 8 days ago, I have already built a 1.5Kva, + / - 63v pwr supply in basic form. Since the T amp is way outside my realm of knowledge, I'm wondering if there are any quirky necessities the AMP 2 may have with respect to it's supply needs.
Also, read 41hz.com's references to needed upgrades for incr pwr output. Any specifics on what those are???? or is that revealed after the kit is received?
Thanks in advance for any replies!

rb

[Corax]

Hi Mr. Ray,

I hope your already build power supply is a regulated one, otherwise you might run into trouble (and finally damaging the Tripath chips) because the absolutely maximum rating for the power supply voltage is +/-70V. So take into account, that a non-stabilized power supply, in case you build such, will have a higher idle voltage. Don't forget to add a safety margin caused by line voltage fluctuations (You won't have always exactly 110V). Worst case considerations will in deed prevent costly failures.

Now to the power upgrade:
With the PCB I received from Jan it is almost impossible to upgrade since this PCB is the original Tripath design (have a look at the documentation for the Tripath TK2350 reference board). The most obvious difference is the utilization of only 4 output MOSFETs while Jan's "improved" design has space for 8 output MOSFETs (two at each channel for high- and low-side are working in parallel). The 8 MOSFET design is recommended for higher power and bridged outputs. Unfortunately in this case the Tripath driver chip (the larger one closer to the middle of the reference PCB) is a little "overstressed" (actually out of specs) and needs definately a small heatsink. Sooner or later I might be able to upload some photos in the forum which shows my mechanical and electrical work on these terrific amps.
If you really plan to build a high-power amp or using it in bridged configuration make sure you receive from Jan his "improved" or "advanced" design. However I heard some rumors that he took that specific design from the market for a while caused by a not specifically known issue (as far as I know/remember). On the other hand I heard also the rumor that he is working on it to get rid of that issue in an upcoming revision of his 8 MOSFET board. Let's see and wait what Jan can/will do about it.

Greetings from Corax

[mr. ray]

Corax,
Thanks kindly for your insights. I did receive the kit since posting and upon reading specs, the same thought occured to me. I had also since read your posts with another user (name I do not recall) about SMPS's. I have little experience with these, so I chose to avoid (especilly not know how one might behave with the particular T-Amp operating principles).

My kit IS the 4 MOSFET version and I DID notice some of the info sent showed the 8 version.

My 1.5 KVa supply is much overkill for this board version I can now see! but i've already sunk the money into the big toroidal xfmr. My only thought in reducing a bit would be to string about 6 - 20 Amp diodes off each side of the FWB rectifier befor reaching capacitor array. Depending on exact specs, that would pull my unloaded voltage down to a bit below + / - 60v.

Any thoughts on this?

You seem very knowledgable on this topic.

Again - Thanks!!

mr. ray

rb

[mr. ray]

or am I better off taking away some windings from the toroid secondary??

rb

[Corax]

Hi mr. ray,

to string a couple of rectifier diodes for a necessary voltage drop is just "uncool". It will work if you take care of a certain minimum current flowing through the diodes (in this case you need a closer look to the diagram where the forward voltage in dependency of the forward current is shown. Since the idle current of the TK2350 is very small the voltage drop across the diodes might not be sufficient in this case.

Well, I would suggest something more similar to these designs: http://www.welbornelabs.com/ps3.htm or http://sound.westhost.com/project102.htm

The regulator (in detail, the power transistor) does not dissipate that much power as the first thought might simulate in your brain . With less or moderate output power of the AMP2 the average current multiplied with the voltage drop might give you a power rating of 30...50W which will not be much higher when the AMP2 is working at full-power since the transformer will drop with the output voltage too if the load rises. On the other hand I must admit that such a big "overkill" toroid will not drop very much. Of course all the advantages of a class-D amplifier with it's low power loss will be "killed" by using such a circuit, just because the transformer output voltage is slightly to high.

If you have the chance to unwind some turns from the transformer, be my guest. Actually I wouldn't do this (right away), because you have to remove the insulation material before you can unwound the toroid.
My suggestion instead would be to add a winding with some turns on the transformer and connect this winding in series to the previously used secondary winding in that manner, that it will be subtracted from it (it's just a matter of polarity ). This way you don't need to disassemble your toroid and you can experience much better with the number of necessary turns. If you're then still in the right mood to change the original turns of your toroids secondary, just go ahead.

Another method of determining the desired output voltage drop (secondary winding turns derating), is the following:
- Measure and write down the output voltage of the secondary winding (i.e. 46V) and the desired output voltage (i.e. 42V).
- Attach a (precise) single additional winding on the toroid and measure the output voltage of that single winding (i.e. 0.8V/turn).
- Divide the necessary voltage drop of the original secondary winding by the measured voltage of the single winding (i.e. (46V - 42V) / 0.8V = 5).
- Remove this number of windings from the secondary (i.e. 5 turns) - that's it.

Greetings

Corax

[Corax]

Hi mr. ray,

to string a couple of rectifier diodes for a necessary voltage drop is just "uncool". It will work if you take care of a certain minimum current flowing through the diodes (in this case you need a closer look to the diagram where the forward voltage in dependency of the forward current is shown. Since the idle current of the TK2350 is very small the voltage drop across the diodes might not be sufficient in this case.

Well, I would suggest something more similar to these designs: http://www.welbornelabs.com/ps3.htm or http://sound.westhost.com/project102.htm

The regulator (in detail, the power transistor) does not dissipate that much power as the first thought might simulate in your brain . With less or moderate output power of the AMP2 the average current multiplied with the voltage drop might give you a power rating of 30...50W which will not be much higher when the AMP2 is working at full-power since the transformer will drop with the output voltage too if the load rises. On the other hand I must admit that such a big "overkill" toroid will not drop very much. Of course all the advantages of a class-D amplifier with it's low power loss will be "killed" by using such a circuit, just because the transformer output voltage is slightly to high.

If you have the chance to unwind some turns from the transformer, be my guest. Actually I wouldn't do this (right away), because you have to remove the insulation material before you can unwound the toroid.
My suggestion instead would be to add a winding with some turns on the transformer and connect this winding in series to the previously used secondary winding in that manner, that it will be subtracted from it (it's just a matter of polarity ). This way you don't need to disassemble your toroid and you can experience much better with the number of necessary turns. If you're then still in the right mood to change the original turns of your toroids secondary, just go ahead.

Another method of determining the desired output voltage drop (secondary winding turns derating), is the following:
- Measure and write down the output voltage of the secondary winding (i.e. 46V) and the desired output voltage (i.e. 42V).
- Attach a (precise) single additional winding on the toroid and measure the output voltage of that single winding (i.e. 0.8V/turn).
- Divide the necessary voltage drop of the original secondary winding by the measured voltage of the single winding (i.e. (46V - 42V) / 0.8V = 5).
- Remove this number of windings from the secondary (i.e. 5 turns) - that's it.

Greetings

Corax

[mr. ray]

Corax
Thank you again for your thoughtful and meaningful advice. I mentioned the diode string idea to friend and was rebuked in a way that would be not unlike as if I had voted for Mr. Bush. In other words - a very bad idea

I have already tried and succeeded with your suggestion of adding another secondary winding - trial and error led me to 12 turns (6 vac - translating to approx 8 rectified dc volts for each side) i.e. approx 0.5V per turn. Now running witout load at +/- 55vdc. using large 10 ohm resistors on both sides (approx 5.5 A @ +/- 55v (approx 550 - 580 watts out)) the voltage only drops to 53.3 v +/-.

Now I only need to find time to build the amp. My mechanical skills in this arena would qualify as 'good' however, never worked with smd's.

Again,
Thanks.
Ray

rb

[mr. ray]

Corax
Thank you again for your thoughtful and meaningful advice. I mentioned the diode string idea to friend and was rebuked in a way that would be not unlike as if I had voted for Mr. Bush. In other words - a very bad idea

I have already tried and succeeded with your suggestion of adding another secondary winding - trial and error led me to 12 turns (6 vac - translating to approx 8 rectified dc volts for each side) i.e. approx 0.5V per turn. Now running witout load at +/- 55vdc. using large 10 ohm resistors on both sides (approx 5.5 A @ +/- 55v (approx 550 - 580 watts out)) the voltage only drops to 53.3 v +/-.

Now I only need to find time to build the amp. My mechanical skills in this arena would qualify as 'good' however, never worked with smd's.

Again,
Thanks.
Ray

rb