My stereo 300D's are nearing completion. Upon reading a few of the posts here, I decided to // a 1uF MKP across the 47uF electro. I honestly don't know if it is any better because I hadn't spent enough time listening to the amp before the mod. Regardless, it sounds incredible... but it sounded incredible before too. None the less, it can't hurt.
Greg says that the DC stability is quite good and can be maintained below ~20mV with the link in place... but I'm sure you guys already know this.
Too bad Greg isn't on DIYAudio any more. His amp is excellent and he is a very patient and helpful guru.
Greg says that the DC stability is quite good and can be maintained below ~20mV with the link in place... but I'm sure you guys already know this.
Too bad Greg isn't on DIYAudio any more. His amp is excellent and he is a very patient and helpful guru.
KLE and others, instead of using a DC servo, why not do as Francis/Rellum suggests:
... Then I desoldered/resoldered one input transistor of each pair, so as to couple them thermically, face to face. This greatly enhenced the offset behaviour, +/-20mV max when hot, and is no more an issue with my speakers.
Greg makes a similar suggestion and says that you can then use VR2 to minimise DC offset to +-20mV:
I might point out that the DC stability of the amp is still good in DC mode especially if the lid is on and a small cup is siliconed over Q1Q2 and Q3Q4 pairs.
A cup can be made from a ball point cap cut off. The transistors only dissipate 3.75mW each so heat is not an issue, just enclosing them for the same environment. Then a link across C19 (on the GB300D or C9 on the GB150D) sets it to DC coupled feedback and you can trim the offset with VR2 when it's warm. Drift should be < +/-20mV.
I have tried DC servos on my pre in the past and I can say for sure that the DC servo tends to cut off both ends of the spectrum, both bass and highs.
I'll try Gregs idea and post feedback.
... Then I desoldered/resoldered one input transistor of each pair, so as to couple them thermically, face to face. This greatly enhenced the offset behaviour, +/-20mV max when hot, and is no more an issue with my speakers.
Greg makes a similar suggestion and says that you can then use VR2 to minimise DC offset to +-20mV:
I might point out that the DC stability of the amp is still good in DC mode especially if the lid is on and a small cup is siliconed over Q1Q2 and Q3Q4 pairs.
A cup can be made from a ball point cap cut off. The transistors only dissipate 3.75mW each so heat is not an issue, just enclosing them for the same environment. Then a link across C19 (on the GB300D or C9 on the GB150D) sets it to DC coupled feedback and you can trim the offset with VR2 when it's warm. Drift should be < +/-20mV.
I have tried DC servos on my pre in the past and I can say for sure that the DC servo tends to cut off both ends of the spectrum, both bass and highs.
I'll try Gregs idea and post feedback.
Re: Re: NFB network & DC servo
Hi KLe,
in the annex you find the schematic of my DC-servo which works just fine. It was mounted on a small lab-type PCB together with the +/- 15 VDC regulators. If you have a SKA stereo version, one +/- 15 VDC regulator is enough for both DC-servos. Power consumption is in the mW region.
I noticed that the output of my OP97 lingers around 0.4 to 0.6 VDC.
The SKA DC output offset stays below 500 uV, regardless of the temp of the heatsinks.
ciao.
KLe said:
Thankyou stoeffle/rellum
DC servo sounds really good, but, removing the electro or shorting it and thermally coupling the input transistor pair, sounds even easier.
stoeffle, yes, please send me the DC servo circuit diagram.
No feedback electro, let's see ...
1. Closed loop gain to DC
2. No capacitor non-linearities
3. No degradation of sonics with time
4. No plopp! noise (which could degradate speaker driver performance over time)
Hi KLe,
in the annex you find the schematic of my DC-servo which works just fine. It was mounted on a small lab-type PCB together with the +/- 15 VDC regulators. If you have a SKA stereo version, one +/- 15 VDC regulator is enough for both DC-servos. Power consumption is in the mW region.
I noticed that the output of my OP97 lingers around 0.4 to 0.6 VDC.
The SKA DC output offset stays below 500 uV, regardless of the temp of the heatsinks.
ciao.
Attachments
DC servo and sonic implications
Hi Dr. H,
yes, definitely, thermally coupling the two LTP input pairs is of great benefit to DC performance. I mounted my Q1/Q2 & Q3/Q4 on longer leads and placed them with the flat sides face-to-face, even with a small amount of heat spreading paste in between, then strapped a tye-wrap around them tight. To avoid shorts between the rather twisted leads, I used a short piece of thin silicone isolating hose on the middle leg.
To the DC-servo influencing the sound spectrum:
if the time constant of the integrator is made too small, we for sure will have the servo trying to partially counter-act against very low-frequency signals. We also would notice group-delay changes at the lower end of the audio spectrum.
At the upper end, the servo has no influence whatsoever, as it is a dead DC voltage through a high ohmic resistor (in my case 180k), which acts as a DC current source into the inverting input node of the SKA, correcting for the DC offset at the amp's output. In case you sensed changes in the higher end of the spectrum, I suppose that the reason must be sought somewhere else. My SKA300D delivers a full & solid bass and very high defined highs.
But as the romans already stated: degustibus non est disputandum. (for non-romans: taste is not a discussable matter).
ciao.
Dr.H said:
Hi Dr. H,
yes, definitely, thermally coupling the two LTP input pairs is of great benefit to DC performance. I mounted my Q1/Q2 & Q3/Q4 on longer leads and placed them with the flat sides face-to-face, even with a small amount of heat spreading paste in between, then strapped a tye-wrap around them tight. To avoid shorts between the rather twisted leads, I used a short piece of thin silicone isolating hose on the middle leg.
To the DC-servo influencing the sound spectrum:
if the time constant of the integrator is made too small, we for sure will have the servo trying to partially counter-act against very low-frequency signals. We also would notice group-delay changes at the lower end of the audio spectrum.
At the upper end, the servo has no influence whatsoever, as it is a dead DC voltage through a high ohmic resistor (in my case 180k), which acts as a DC current source into the inverting input node of the SKA, correcting for the DC offset at the amp's output. In case you sensed changes in the higher end of the spectrum, I suppose that the reason must be sought somewhere else. My SKA300D delivers a full & solid bass and very high defined highs.
But as the romans already stated: degustibus non est disputandum. (for non-romans: taste is not a discussable matter).
ciao.
Hey guys,
Out of curiosity... How do the SKA amps compare to others? ie. AKSA 55W, Sym A Sym, Aussie, etc... I have only had the pleasure of listening to a GB300D as my first exposure to a high resolution system and it is incredible! In fact, I am contemplating building nothing other than SKA's for the remainder of my DIY career yet at the same time i know that if I build nothing else I will know nothing else. But is it worth building something else if I already have the best? I am just trying to spur conversation here.. Devils' advocate. I am not an audiophile.. Yet... but I am young still
Out of curiosity... How do the SKA amps compare to others? ie. AKSA 55W, Sym A Sym, Aussie, etc... I have only had the pleasure of listening to a GB300D as my first exposure to a high resolution system and it is incredible! In fact, I am contemplating building nothing other than SKA's for the remainder of my DIY career yet at the same time i know that if I build nothing else I will know nothing else. But is it worth building something else if I already have the best? I am just trying to spur conversation here.. Devils' advocate. I am not an audiophile.. Yet... but I am young still
Spartacus said:Sorry to tell you this guys ..... but going without the FB cap results in a clear improvement. Arrgh, more stuff to fiddle with.
As for the sound, well I can hear more music and less electronics.
iihay said:No tweaks for me this weekend, I cooked a channel earlier with a stray probe whilst rebiasing
The air was blue I can tell you!
Iain
Spartacus
That's excellent news ... have you done this already?
Iain
That's not good news, hope that's there isn't too much damage and its a quick rebuild ...
Like I said in the ska forum...bigeyes:
I had a muse cap bypassed by .5uf nichicon best quality
I removed it and there was no difference in the bass or high frequency except...
The only difference was the decay of notes, it was more detailed you could hear the microphones distortions of old recordings just bellow the high frequency : like 1k - 3khz around that.
But the amp has already lot of details... so I would not suggest doing any DC coupled without a dc servo. you dont want 28 X amplified DC current
just unplugging one input and you can blow the fuses with dc coupled. Also you will need to adjust it every time you change your source or components. You will need to thermally link the q1-q2, and q3-q4 for around +- 5mv dc output... With the cap in you can get +-0.1 mv of dc !!!!!
The sound itself doesnt change at all with dc couple... so just put a good cap there and don't worry about it and biaised people who say a capacitor in the feedback is bad...
Really , just getting a better cd player will completely change your sound and is a way better modification
I had a muse cap bypassed by .5uf nichicon best quality
I removed it and there was no difference in the bass or high frequency except...
The only difference was the decay of notes, it was more detailed you could hear the microphones distortions of old recordings just bellow the high frequency : like 1k - 3khz around that.
But the amp has already lot of details... so I would not suggest doing any DC coupled without a dc servo. you dont want 28 X amplified DC current
just unplugging one input and you can blow the fuses with dc coupled. Also you will need to adjust it every time you change your source or components. You will need to thermally link the q1-q2, and q3-q4 for around +- 5mv dc output... With the cap in you can get +-0.1 mv of dc !!!!!
The sound itself doesnt change at all with dc couple... so just put a good cap there and don't worry about it and biaised people who say a capacitor in the feedback is bad...
Really , just getting a better cd player will completely change your sound and is a way better modification
KLe said:
Sorry I missed this originally. Yes, I've done the mod, and heard a clearly beneficial improvement. Maybe I'm biased, but quite why I would be, I have no idea.
As Gab says, you need to take care over DC offset. My SKA is relay protected at the output.
Also, a really expensive CD player WILL make a bigger difference. Strange, but true.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.