TGM Amplifier ?

[Bigun]

The psu is now assembled, chokes installed on the backside of the pcb along with some additional ceramic decoupling caps.

First power up in the chasis - no sparks provides +/-26V

[Lazy Cat]

I really like your latest PCB-s (TGM-5 channel too) ... they look sooo HF.

Congratulations Gareth

[Bigun]

Thanks LC. Well some more progress today. Finished the psu, installed front panel LED indicator. And completed some upgrades to the amplifier:

1) added output inductors, some nice ones recovered from a Samsung amp I think with two concentric coils, one inside the other (photo attached)
2) upgraded VAS by removing BD139 and replaced with 2SC3423
3) reduced phase lead comp from 22pF to 5pF silva mica - sims convinced me the old value was 'too heavy', left Cdom at 68pF


some more upgrades may be forthcoming but I need some darn bolts and washers before I can properly re-mount the amplifiers to their heatsinks.

[Bigun]

I've completed some further refinements.

I added emitter degeneration to the LTP, just 4R7 so that I can measure their voltage drop and check to see how well the current flow is balanced between the two LTP devices. I then adjusted the resistor in the collector of the LTP that drives the VAS, until the balance was good. The balance achieved was much better than the original build, with a mismatch of 1.8% on one channel, 3.6% on the other.

I adjusted the series feedback resistor to better match the value of the input resistor - thus ensuring low dc offset on the output. The dc offset is in the range 5mV to 20mV. I used the same values on both channels to keep gains the same. The bias and dc offset appears stable with temperature.

As this is going to be used mainly for near field listening I reduced the overall gain by increasing the value of the feedback shunt resistor from 2k7 to 5k1. This does increase the feedback factor slightly. Simulated OLG is around 48dB and CLG around 20dB.

The speakers used are my own design MARTELLO full range ported boxes.

To my ears these changes have brought about some significant improvements in the sound of the amplifier. It's cleaner without being 'too clean' - simply very nice to listen to. I imagine this is now much closer to the sound that the original AKSA produced. ALSA was the inspiration for this TGM1 design. It's not as precise as my TGM2 or TGM5 but very easy to listen to, relatively easy to build, easy to set up. I'd recommend this design for anyone who normally suffers from listening fatigue to SS amplifiers - you don't get any with this one.

It's also really very quiet, no hum. The trafo is the quietest EI trafo I've ever come across too. Some noises on turn-off.

The photos show the rest. Simple black cube, single volume knob on the front, power switch and LED power indicator. The old TGM1 modules are installed either side of the trafo, power supply mounted against the inside of the front panel. The rest of the space used for wiring. There is a simple switch on the back panel as an input selector. The box is closed with a perforated black metal cover.

Now that my kids have discovered this upgrade to the sound of our iMac it's unlikely I'm going to get my hands on it again for some time

[Bigun]

Listening impressions from the last month, listening through these:MARTELLO enclosure for FR88EX

Fantastic, I have a renewed respect for this relatively simple topology and numerous refinements guided by Hugh Dean.

No fatigue. Speakers sound a lot larger than they are. Not the last word in transparency, but so nice to listen to.

[Bigun]

I thought I should post a schematic as a record of what the final version looks like.