TGM8 - my best amplifier, incredible bass, clear highs, no fatigue (inspired by Rod Elliot P3a)

The bipolar are the outputs, operating in Class AB. The power FETS operate in Class C (normally off) and they slide in gently to help when more power is needed (either when you crank up the signal or if the load has impedance dips). This has the following benefits (in my opinion), a) retains great sound of single bipolar output pair at normal listening levels with 'normal' speaker loads, b) no parts to match, c) protects bipolar devices from beta-droop and high currents, d) limits Current draw from the drivers thus protecting them and hence a less demanding thermal control loop

I have yet to fully compare the TGM7 and TGM8. They are both superb sounding to my ears. TGM8 runs much cooler of course and has the convenience of built in speaker protection and power filter caps.
Soundwise I thought TGM7 would be unbeatable but much of the design was not mine. I set myself the one-time challenge of trying to beat it. The TGM8 might just do that, not only is the TGM8 bass strong and powerful but the mids and highs have more air, more nuance. I will use TGM7 for my HT and TGM8 for music duties. I'll be able to share more opinion once I've had more listening time.

Sounds like you heard the same things as me with my SKA GB150. The mids and highs are slightly dumbed down a bit compared to my other amps. Many folks probably like that. They call it "less fatiguing" of something to that affect. Looking forward to trying the TMG8 if you can work it out for me.

Blessings, Terry
 
Sorry Lordearl, difficult to help as I am on travel, in Hong Kong with phone only last week and now in UK with a borrowed iPad access only, will be back in Canada next week.

No worries Gareth - enjoy the travel! In the meantime I have downmixed to mono (using JRIVER) and am listening through the good channel.

... not only is the TGM8 bass strong and powerful but the mids and highs have more air, more nuance. I will use TGM7 for my HT and TGM8 for music duties. I'll be able to share more opinion once I've had more listening time.

My initial impression (mono only) is the same, the low frequencies are crisp and punchy - never heard anything quite like it. The treble easily rivals my 6L6 ultra linear push pull amp with Tango output transformers, plenty of air, no fatigue (also no apparent dumbing down of the sound!). Makes you realise why you loved audio in the first place!
 
I have downmixed to mono (using JRIVER) and am listening through the good channel.

I should warn you that after I started listening to mono more often I now find that I like it. For me, regular stereo sometimes sounds a little less natural, like an 'effect'. It is probably because I have never managed to set up a suitable listening room for stereo, so nowadays I listen to mono as well as stereo. I do use stereo consistently when sitting at a desktop computer because the listening position is enforced. With mono I am able to hear the music without having to enforce a strict listening geometry and find it works well when I want to fill the house with sound as I move around and for that I use a DIY 'boffle' speaker which has some features of an open baffle so the sound field integrates nicely into the open spaces. It is a fun hobby :)
 
I'm a bit confused about your figures. Could you please write them on your correct schematic, then I will examine. I've been on the wrong schematic, I believe.

Thanks Hugh, I've attached the schematic which the boards were fabricated from (Christian - could you confirm?). Q1 and Q2 are both 2N5551. I couldn't tell which pins/legs were the base/emitter so I measured each of the pins/legs with reference to ground.

For ease of reference, results were as follows:

Good channel -
Q1; -0.731v, -108.1mV, 36.64v.
Q2; -6.07v, -5.4v, 25.2mV.
LED; 1.356v (red->centre as green is n/a)

Bad channel -
Q1; 17.51v (slowly keeps increasing), 17.31 (also increasing), 38.42v.
Q2; 18.8v (increasing), 18.9v (increasing), 19.2v (increasing).
LED; 1.178v (red->centre)[/QUOTE]

Hope this helps?

Cheers,
Tom
 

Attachments

  • image.jpg
    image.jpg
    697.7 KB · Views: 754
Tom,

Check R2 is grounded, with other side on base of Q1, then set VR1 so there is around -6V wrt ground on emitter Q2, and -12V wrt ground on the zener.

Much of this problem is generated by Q2 VR1 setting, I believe. You should have around 6V across R6 (1K5) and this is around 4mA.

Verify that there is 0.63V across R4.

Cheers,

Hugh
 
Tom,

If you have a variac to halve the supply, then replace the fuse and have another go.....

If no, wire up a 12V 21W car brake light in place of the fuse and supply the reduced current.

It's a good method, if passing only 100mA the filament will cool and drop very little voltage across it. You will need a lamp in place of on EACH fuse for each rail.

Cheers,

Hugh
 
Ok well some more testing done but not much headway unfortunately.
The amp is drawing some serious current - I put a 20w globe in place of the fuse on each rail and they blew straight away. Because of this I can't take any meaningful measurements (without the fuses in place the readings are very different from the good channel compared to the bad channel, but on the good channel alone the readings differ when the fuses are in line versus when they are not).

I also replaced the LED then inserted fuses again but they blew straight away too. In particular F2 (the fuse on the negative rail) seems very aggressive, sparking furiously when I put it in and take it out!
 
Hello
I think you have to start to replace some of those semiconductors
Probably one or more got damaged, when you measured with DDM look s OK but under heavier load start acting weird.
No other pars can do that.
If you measured everything and all OK and mounted to right position only possibility left.
My humble opinion only. Please remove those mosfets if you have them, the amp work with out so less semis to be suspicions..
Greetings Gabor
 
I'm back from travel but it's difficult to dive in here with the problem solving - I am pleased to see the community spirit (Christian and Hugh) filling the gap :)

Lordearl - debugging and repair work is a challenge unto itself, if you fancy the fun of doing it then you are in good company, otherwise I would advise you take up Christian on his kind offer to repair them for you.