• Disclaimer: This Vendor's Forum is a paid-for commercial area. Unlike the rest of diyAudio, the Vendor has complete control of what may or may not be posted in this forum. If you wish to discuss technical matters outside the bounds of what is permitted by the Vendor, please use the non-commercial areas of diyAudio to do so.

TSSA V1.6 docs and buyers list

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
@Mikelm

There are C4,C5 470n bypasses on C2,C3 AC/DC split gain feedback electrolytics. Did you experiment on those too? Starting without any direct bypasses and then compare bypass solutions maybe its an acceptable practice also. I remember when I had been suggested to bypass my LCR 15000uF PSU capacitors in a JLH 80W amp with MKPs. The tone got pinched and I ditched them very soon. Snubbing is best but needs original PCB space. Your constantan wire does the trick neater though.

Hi Salas,

I will check the FB cap in spice later but I suspect they will not suffer from this problem because they only have very low impedance on the ground side whereas the rail caps have low impedance environment on both terminals which allows very high Q at resonance - but I will report for sure on this later. However I suspect that just upping the FB bypass cap from .47uF to 4.7uF might give other sonic benefits. The epcos open frame stacked film range are great because they are so small.

P.S. What speakers you got and how is bass with TSSA at ~20dB feedback?

My current speakers are quite modest but actually I really like them. They are Modified B&W 602 S3 ( redesigned X-overs & cabinets )

Google Image Result for http://i157.photobucket.com/albums/t69/lyntonfitt/Bild200021.jpg

They give a clear insight to the music. The bass is fast & tight but may not be the best speakers to assess extreme low bass ( there isn't any ! ) 200mm cones . . . 90db / W / M.

Nelson Pass reports that subwoofers do need low impedance to drive them well otherwise he often prefers the sound of higher impedance o/p.

This amp has about 0.3R o/p impedance and with my speakers the bass there is, is fast, tight & very well defined :)
 
diyAudio Chief Moderator
Joined 2002
Paid Member
My diy speakers have tight S shaped double pleated rubber surrounds and they enjoyed my tube amp with 0.6R so it will be OK I guess since I will surely need your gain setting due to my vinyl source and buffer pre. Your info is enlightening, thanks for paving the road.
 
diyAudio Chief Moderator
Joined 2002
Paid Member
Those caps remind me of the Siemens branded silver ones in the 90s. I still have 2-3 in a bin. Transparent, silver innards, you could see the stacking at the sides where the leads are running half height externally. They were good HF decoupling caps.
 
Yes, but I think that it's having the VAS as the O/P stage which is the most radical element in TSSA design.

So far I have only compared my SE 3 stage CFB with this Symmetrical 2 stage so it is not 100% clear to me what change causes what effect. So I would very interested to hear impressions for those who have built and listened to:

SSA - 3 stage with follower O/P stage
vs
TSSA - 2 stage with with VAS O/P stage

The other logical option would be a SE two stage with VAS as O/P but that would probably have a current ceiling of 2 x O/P stage bias so would have to be another heat monster :D

I must take another look at Nelson's 2 stage JLH - that could be another contester in this arena.
 
Thanks Sonny,

I do feel I learnt a some useful things from this project particularly how important different topologies can be to achieve different "sounds".

It turns out that even though I would not include electrolytics in the FB path in my ideal amp their impact on the sound is way less than some other factors.

Western scientific method only works well if all involved have open minds and are happy to say bye bye to old theories as soon as they evolve.

Personally I am a little perplexed way anyone ever used bypass caps on power lines without snubbers. I repeated my spice research again to show to Salas and it confirmed again that my ears tell me whenever I have tried omitting them.

I Will present it below.
 
Hi salas,

here is my test circuit for bypass cap resonance
 

Attachments

  • test cct.jpg
    test cct.jpg
    33.2 KB · Views: 254
here are results:

the lower lines green & red show the voltage on the caps

the upper lines green & green ! show the current in the nominal 0.1R resistor

You can see how the snubber smooths out the resonanes in the voltage on the caps.

But more significantly the upper plots show how much the current in the rails is also disturbed when unsnubbed & clean when snubbed
 

Attachments

  • PSU bypass.jpg
    PSU bypass.jpg
    100.5 KB · Views: 244
Here are the results from the similar test when I modelled the FB cct with caps.

You can see that the voltages ( below ) are disturbed and cured just as before.

But perhaps significantly the current in this much higher impedance circuit is disturbed very little. Tiny little red blib around 1Mhz - honest it is there but has got lost in the upload

So I think the bypass cap here in the FB circuit will not have anything like kind of an impact on the sound when compared with the bypass power rails.

Technically it will still be a good idea to include a snubber cct ( 0.1R worked well with 4.7uF ) but it might be hard to hear any difference
 

Attachments

  • FB cap bypass.jpg
    FB cap bypass.jpg
    96.8 KB · Views: 240
Last edited:
diyAudio Chief Moderator
Joined 2002
Paid Member
Mikelm thanks for the analytical presentation. Now what happens on the rails decoupling if the power source (PSU) shows the output impedance of a capacitor multiplier that has been proposed as a positive add on? Any adapted recommendation in this case?
 

Attachments

  • Typical_heavy_Cmult.png
    Typical_heavy_Cmult.png
    17.8 KB · Views: 229
  • Cmult_Zo.png
    Cmult_Zo.png
    12.7 KB · Views: 223
I just spent about 4 hours trying to find, calculate, estimate, guess real world values.

It's surprising how few capacitor impedance graphs there are out there.

my presented results were using "guestimates" Now after my excertions my conclusions remain pretty much the same but I think some of the cap inductances guesses were a bit on the optimistic side.
 
Mikelm thanks for the analytical presentation. Now what happens on the rails decoupling if the power source (PSU) shows the output impedance of a capacitor multiplier that has been proposed as a positive add on? Any adapted recommendation in this case?

Well, if you use snubbers as suggested on by pass film caps your cap bank should be stable.

If your regulators are also stable - which hopefully should not be difficult with no FB and 10,000uF in parallel then I think it should be OK.

Sonny,

Do you think this circuit will substantially benefit from regulators ?

After thinking about how you have protected your i/p devices with the CFPs & Jfet cascodes on one side and high perfomance CCS's on the other I am begining to think that this cct might be reasonably imune to regular audio frequency / signal induced noise.

My choke regulators with fused outputs provide excellent regulation against mains / diode VHF RH garbage etc but the highish o/p Z means they probably sing along to the music more than many supplies - yet your amps sound very good with them.

What are your thoughts on this topic ?
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.