Compact Sized LME49810/11 +ThermalTrak Amp

[panson_hk]

Hi everyone,

This is my new 49810/49811 + ThermalTrak Amp intended for low output power (Class-AB/B, <= 60 W). Output stage is driver + emitter follower. I have not put the heatsink for LME and driver (too noisy to do metal work in the midnight).

[klewis]

Panson,

I got both my Leach bais configuration and Cordell bais configuration for thermal trak working. The Cordell is tweaked a bit from his diagram - I decided to simulate both schemes to see how they react to temperature.
The values shown were adjusted to 2.93v at 27deg c. Then I changed the simulation temperature to 68deg c and ran it again. You can see that the Cordell version stays closer to the baseline Vbe which assumes that the Vbe transistor has a perfect thermal connection to the output transistors.

Ken

[panson_hk]

Originally Posted by klewis

Panson,

I got both my Leach bais configuration and Cordell bais configuration for thermal trak working. The Cordell is tweaked a bit from his diagram - I decided to simulate both schemes to see how they react to temperature.
The values shown were adjusted to 2.93v at 27deg c. Then I changed the simulation temperature to 68deg c and ran it again. You can see that the Cordell version stays closer to the baseline Vbe which assumes that the Vbe transistor has a perfect thermal connection to the output transistors.

Ken

Hi Ken,

Nice work! Have you tested the bias control in real circuit?

[klewis]

Originally Posted by panson_hk

Hi Ken,

Nice work! Have you tested the bias control in real circuit?

Panson,

Yes, using your modular approach, I now have three boards, the LME49811, with DC Servo, a board for the bias resistors, and the output board which contains the VBE transistor, predrivers, drivers and Thermal trak output devices. I put jumpers at the T Trak so that I can use either two, three or four of them in series. I had to make some mods to the driver board to go from the Leach bias scheme to the Cordell scheme. The photo shows the Cordell scheme. Note that I had to put pairs or resistors in series to get the correct resistance. Seems you never have the right stuff on hand regardless of biweekly DigiKey and Mouser orders...

My shop computer with the highend sound card died, so, no THD tests for a few weeks. But, played it with both bais schemes and it sounds great, also, sine and square waves look smooth and clean on the scope.

Ken

[panson_hk]

The amp is in operation. I am going to put them in a case.

[panson_hk]

One set of compact amp is setting in the lab for daily use. It is now playing Santana YouTube - PansonAudio's Channel

The operating voltage is +/- 35 V. The speaker is NHT SuperTwo (http://www.stereophile.com/floorloudspeakers/1099nht/).

[klewis]

Originally Posted by panson_hk

One set of compact amp is setting in the lab for daily use. It is now playing Santana YouTube - PansonAudio's Channel

The operating voltage is +/- 35 V. The speaker is NHT SuperTwo (Stereophile: NHT SuperTwo loudspeaker).

Panson,

Nice little video, made me smile. I also looked at the cross over distortion video. I presume that you are showing the bais trimming out the distortion. Is this measured just with the scope? Or how do you set up for this measurement?

Thanks

Ken

[panson_hk]

Originally Posted by klewis

Panson,

Nice little video, made me smile. I also looked at the cross over distortion video. I presume that you are showing the bais trimming out the distortion. Is this measured just with the scope? Or how do you set up for this measurement?

Thanks

Ken

Ken,

Yes, it is. The cross-over waveform is the THD residual from the AP output (after the notch filter).

Panson

[klewis]

Originally Posted by panson_hk

Ken,

Yes, it is. The cross-over waveform is the THD residual from the AP output (after the notch filter).

Panson

Well, I guess I won't be doing that anytime soon...

[PeterLeest]

Originally Posted by panson_hk

Ken,

Yes, it is. The cross-over waveform is the THD residual from the AP output (after the notch filter).

Panson

What is the bias current to get the optimum result ?