The MONGREL (supersym II)

[keantoken]

I also thought the SS9012/9013 were interesting. The 9014's might be a really good low noise VAS.

So where can I find these Baxandall papers?

- keantoken

 

[ostripper]

I also thought the SS9012/9013 were interesting. The 9014's might be a really good low noise VAS.

So where can I find these Baxandall papers?

- keantoken

Baxandall papers -
The Baxandall Papers: Transitional Miller compensation

SS9014/15 = low Vceo ... maybe a "iddy bitty" amp OR in a cascoded VAS used as the active device.

BTW , the "goldmund killer" (CX1.3) gets stuffed tommorrow AND lookie what I gotz .... (BELOW)
OS

 

[keantoken]

Nice. I haven't been feeling very smart the last few weeks and took a handful of vitamins yesterday. Better, but I still hate ordering parts online, especially when I have to do it from multiple vendors. But I think I'll be ready to order tomorrow...

- keantoken

 

[dadod]

Hi Os,
Could you take a look at this simulation?
It is quite simple amp and simulation shows 7ppm at 3kHz and 17ppm at 10kHz.
Third harmonic is a bit prominent but at -136dB I think it is not so important.
This is not bootstraped VAS because TT diodes need more stable current tnen bootstrap can provide.
dado

 

[AndrewT]

what is 9ppm?

Hi,
9parts per million is 9/10^6 = 0.000009 = 0.0009%

I see 0.000009% as the simulated/predicted distortion on Ost 900 pic.

 

[ostripper]

Hi Os,
Could you take a look at this simulation?
It is quite simple amp and simulation shows 7ppm at 3kHz and 17ppm at 10kHz.
Third harmonic is a bit prominent but at -136dB I think it is not so important.
This is not bootstraped VAS because TT diodes need more stable current tnen bootstrap can provide.
dado

Dadod , assume we have NONE of your models.My default models are gone, so include ALL models. I could not get to work without subbing for models I had. It works good...otherwise.

Think about it , at 120db or lower ... what is left to improve on ?? Input noise levels , thermal and Hfe matching , passive component choices , EVEN psycho acoustic considerations. As far as CCS vs. bootstrap , consider that TMC will correct the slight biasing inconsistencies that the t-traks are supposed to alleviate. YOU DON'T NEED THEM. The bootstrap , especially using a high quality Panasonic EE grade , will have a more "forward" bass response and still retain crystal clear detail at the high end...REALLY!

Use your $1.50 usd "run of the mill" 2sc5200/1943 , and still have class A Xover free operation. Also , feel free to add the cascode and experiment with FET or low noise BJT inputs.

OS

 

[ostripper]

Hi,
9parts per million is 9/10^6 = 0.000009 = 0.0009%

I see 0.000009% as the simulated/predicted distortion on Ost 900 pic.

So what I have is .9 ppm ..WOW! I was basing from .1% not a full 1%
OS

 

[Melon Head]

What does this circuit look like into 3 Ohm load?

 

[ostripper]

What does this circuit look like into 3 Ohm load?

That would depend on your choice of OPS. Single pair , you might go "pooof" ... with my new 8 device NJW21193/4 PB250 , a whole different ballgame.

With my 8 device NJW0281/0302 board , testing with 2 pair 8R 15" 3-ways in parallel was standard (3R or lower). Never saw any VAS loading/heating , sound was the same at 20w or 200w.

The PB120 or your own 4 device output stage would be recommended for 6-8r loads medium duty. For 40C continuous operation (like I did with mine) into 3R
8 devices (Both the PB250 "N" and "H"). Very large heatsinks , too. (.1C/W) or better.


OS

 

[Melon Head]

I am only interested in max output of up to 40V peak to peak into 3R.
I don't go any louder than that. Actually 28V peak to peak is probably what I listen to most of the time (that is peak levels not average levels).

Generally speaking I always use at least two pairs, but it ends up being 3 pairs most of the time.

 

[ostripper]

I am only interested in max output of up to 40V peak to peak into 3R.
I don't go any louder than that. Actually 28V peak to peak is probably what I listen to most of the time (that is peak levels not average levels).

Generally speaking I always use at least two pairs, but it ends up being 3 pairs most of the time.

With 3 pairs you might want to go the IRFP240/9240 route. My BX/CX ,etc. will run these as well , I just have not made the appropriate layout.

OS

 

[ostripper]

ALL READY NOW... my "frankensteins" will LIVE.. ah haha

Finished the "goldmund killers" , much harder / more parts than the simple blameless's. I know they will not blow up , ampslab's bi240's (very similar) are proven. I know they are more stable and faster than the infantile goldmund designs . This is the 3' rd pair of "symasym" type voltage stages I've attempted , If these don't impress , I guess I'm a blameless man. I still have the "leech" (LX) to perfect/design/build ... so there is more to come. PHOTO's of perfect builds below !!!

OS

 

[dadod]

Hi Os,
I did inlcude all models in the zip file, did'nt I?
By the way here is simulated distortion at 10kHz
Fourier components of V(vout)
DC component:-0.236072

Harmonic Frequency Fourier Normalized Phase Normalized
Number [Hz] Component Component [degree] Phase [deg]
1 1.000e+04 1.378e+01 1.000e+00 -0.91° 0.00°
2 2.000e+04 5.863e-07 4.255e-08 -50.57° -49.66°
3 3.000e+04 2.223e-06 1.613e-07 139.13° 140.05°
4 4.000e+04 1.913e-07 1.388e-08 34.14° 35.06°
5 5.000e+04 1.428e-07 1.036e-08 -43.06° -42.14°
6 6.000e+04 1.716e-07 1.246e-08 7.23° 8.15°
7 7.000e+04 3.496e-07 2.537e-08 -14.09° -13.17°
8 8.000e+04 2.032e-07 1.475e-08 -18.79° -17.87°
9 9.000e+04 2.242e-07 1.627e-08 -7.17° -6.26°
Total Harmonic Distortion: 0.000017%

At 3kHz it was 0.000003%
dado

 

[dadod]

by the way, we should not call an amp with bootstraped VAS blameless, as D.S. used CCS in all his blameless creations.
dado

 

[dadod]

Here are some photos of DX-TMC amp.

 

[Macleod]

Hello,

I like your thread even if I didn't have time to read it all...

Do you really intent to build all the various versions?

 

[Bigun]

OS is the new DX...😀

I'm not ready to face strong emotions 😀

 

[AndrewT]

I thought Ost named it Bx and said to avoid confusion he would never use Dx in his Mongrel.

 

[smartx21]

Yes, he did.
But OS and DX are on honeymoon!

 

[BMW850]

Hallo OS,

Maybe it’s better I go for the The MONGREL supersym II (OS and DX) than for the Goldmund.
When all is finished can I buy a completed module with all the parts on it?

Rudy