The 50W version of Zen9

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Thanks VladimirK for your 50W inspiration. Although I didn't need more watts for my 101dB speaker, I still tried to learn from your 50W version. Increase bias current, bypass Alpha CS, Modulated Cascode Cap and change NFB resistor value...

Here is the list and result from the final mod: (sorry, I still touched the master's work)

(1) Reduce R8 and R4 to 1R. Increase bias current around 1.9A.
(I test several resistor values many times, this set works best in the balance of THD and high order harmonics. I prefer lowest high order harmonics than lowest THD)

(2) Increase the Cap value in Cap multiplier to 2200uf
(3) Tweak PS part cap a little, add one set 47000uF CRC filter.
(4) Increase C2 to 470uF ELNA RSK and by pass by 2uF Rel Cap TFT.
(5) Increase C7, C3 to 470uF too, also bypass it (this time Nichicon BPS)
(6) One final key step is... use Distortion Analyzer to fine tune the P1 and P2 value... it change the whole spectrum of distortion in quantity and distribution.
Modulated Cascode (P1) can change the THD and HD distribution around 1W outout. Alpha CS (P2) can change the THD and HD distribution from 6W above.

And the result:
1. I got 40uV noise when input short (better than original 100uV)
2. THD @ 1K,1W 0.002% (original 0.01%) with almost ZERO 5th or above harmonic distortion.
3. THD @ 20K,1W lower to amazing 0.05% (original is 0.19%)
4. THD @ 1K, 10W 0.015% (original is 0.1% with more High order harmonics)

Without too much time to crazy about this numeric results. I already start the listening.

Wow... it never shounds that fantastic! Bass is more solid, high is more clean but still sweet. Music density is double and 3D image is clearer...

So I write this feedback to thanks you and let you know, NP's ZV9 circuit
is really a good platform to tweak out the best sound you want...

It is great to hear about such a good result. I would only add, that listenable effects are determined by the harmonics, and some additional "substances" that are not easy to describe, like effects of electrolytics. Having accepted this, one could guess about difference between schematics with constant current consumption from PS under signal, and standard AB class schematics.
 
It is great to hear about such a good result. I would only add, that listenable effects are determined by the harmonics, and some additional "substances" that are not easy to describe, like effects of electrolytics. Having accepted this, one could guess about difference between schematics with constant current consumption from PS under signal, and standard AB class schematics.

Yes. And somehow THD is a little bit misleading too... Have 0.002% 2nd HD is totally different than having 0.001% 5th or 6th HD...

Elna RFS(Silmik) is notorious for long time break-in, so I am keep playing the amp and wait the sound for maturing. One thing is for sure, even with NEW Elna in the path, add by passing 1uF cap for Alpha CS & M. Cascode definitely help the detail from mid-high to high... It's hard to describe how much detail more there, but it is there!
 
Yes. And somehow THD is a little bit misleading too... Have 0.002% 2nd HD is totally different than having 0.001% 5th or 6th HD...

Elna RFS(Silmik) is notorious for long time break-in, so I am keep playing the amp and wait the sound for maturing. One thing is for sure, even with NEW Elna in the path, add by passing 1uF cap for Alpha CS & M. Cascode definitely help the detail from mid-high to high... It's hard to describe how much detail more there, but it is there!

I would propose, for accelerating the "maturing" or "burning in" of an amp after modifications, use VHF radio, choose some frequency between stations, where maximum noise is listenable, send this noise signal to amp's input. Use resistor as a load instead of speakers. 24 hours of such "maturing" will usually be enough for transistor amps. For new tube amps, this could take from 5 to 10 days.
 
I would propose, for accelerating the "maturing" or "burning in" of an amp after modifications, use VHF radio, choose some frequency between stations, where maximum noise is listenable, send this noise signal to amp's input. Use resistor as a load instead of speakers. 24 hours of such "maturing" will usually be enough for transistor amps. For new tube amps, this could take from 5 to 10 days.

Wonderful idea... will do.. ha ha
 
It works! After 5 days "burning in"... (I happen to have a short vocation)
It really changed a lot, transparency is the first thing I notice, then the high end
extension and bottom firmness. I am really happy with current result.

Only will do some minor resistor change, plan to use RA to replace original dale feedback resistors.
 
Recently I did several tests by changing shunts to the output cap, also measured impedance curves for electrolytics + different shunts, trying to find correlation with listening impressions. The variants were:
1) Giant 4,7uF Teflon cap (Russian K78-2 type, half a kilo weight)
2) Mundorf polypropylene 68uF 250V
3) Solen polypropylene 80uF 400V
Conclusions:
1) impedance curves do not help, I was not able to predict which variant will sound better.
2) shunt capacitance must be maximum possible, 80-100-200uF. In spite that Teflon measures better (less loss angle), 4,7uF is not enough to compete with the bad effects of electrolytics
3) shunt must pass the widest possible frequency range through it, the Low ESR electrolytic cap must only support the Low Frequency range and could be mounted out of PCB

The effects were tested for 50W Zen9, and also with Follower 99 prototype (based on RD100HHF1 transistors, matched pairs, 3,9A total current)

Not to forget, that for AB schematics the same conclusions valid for PS electrolytics
 
Recently I did several tests by changing shunts to the output cap, also measured impedance curves for electrolytics + different shunts, trying to find correlation with listening impressions. The variants were:
1) Giant 4,7uF Teflon cap (Russian K78-2 type, half a kilo weight)
2) Mundorf polypropylene 68uF 250V
3) Solen polypropylene 80uF 400V
Conclusions:
1) impedance curves do not help, I was not able to predict which variant will sound better.
2) shunt capacitance must be maximum possible, 80-100-200uF. In spite that Teflon measures better (less loss angle), 4,7uF is not enough to compete with the bad effects of electrolytics
3) shunt must pass the widest possible frequency range through it, the Low ESR electrolytic cap must only support the Low Frequency range and could be mounted out of PCB

The effects were tested for 50W Zen9, and also with Follower 99 prototype (based on RD100HHF1 transistors, matched pairs, 3,9A total current)

Not to forget, that for AB schematics the same conclusions valid for PS electrolytics

thanks for sharing... did you ever want to try nichicon BP (bi-polar) series?
It looks like my by pass value this time is not enough. But I really don't have such big value polypropylene... :eek:
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.