mains filtering
with regards to a power supply for the UcD modules,
what is the done thing in terms of mains filtering. a design posted earlier by paradigm used an X-Cap across the active/neutral mains lines. other designs ive seen use a varistor.
ive also seen IEC sockets with mains filters built in.
which one (or which combination) are the go?
thanks,
ben
with regards to a power supply for the UcD modules,
what is the done thing in terms of mains filtering. a design posted earlier by paradigm used an X-Cap across the active/neutral mains lines. other designs ive seen use a varistor.
ive also seen IEC sockets with mains filters built in.
which one (or which combination) are the go?
thanks,
ben
matjans said:
why is that, if you don't mind me asking? I would think it's just the other way around...
this has been beaten to death on several forums. Search the AA cable asylum for some opinions. I've tried it both ways, and with my relatively short cable runs, it was clearly better with the long ICs than with long speaker cables.
it's simply a lot more efficient to build high grade long ICs than similar grade speaker cables that need to carry large currents. If money was no object, well, I don't really know if you could get speaker cables to sound better than the longer ICs
Peter
Optimally, the modulation scheme to be used on such a design is NOT fixed frequency, but a first order hysteresis modulator. Think of it: with a fixed frequency amp, the ripple current decreases as duty cycle is modulated. However small you may decide to make the inductor, there will always be a modulation index where the current no longer reverses mid-cycle. With a 1st-hyst-mod, you can indeed design an amp to have ZVS all the way.Jaka Racman said:One way to avoid this is to use very large ripple current in output inductor (more than twice the output current) and very large MOSFETS paralleled with some nF caps. Advantage is lossless ZVS and low on state losses and I also think parasitic diode does not come into conduction at low enough Rdson. Disadvantage is unsymetric transition time and requrement for adaptive gate drive. Is there any disadvantage of using this approach in non cost sensitive applications (besides that it works best in fixed frequency applications) ?
Adding caps to the FETs has nothing to do with ZVS, but everything with resonant transitions. From a distortion point of view this might be less than optimal.
This used to be my experience too. What I found is that using balanced interconnection with a low-impedance source can reverse the situation. If you are using unbalanced wiring and/or a highish impedance preamp output, you may indeed be better off using short interconnects.matjans said:why is that, if you don't mind me asking? I would think it's just the other way around...
Really, the wire that acts as a screen and connects chassis grounds should not double as a signal wire. This is what unbalanced connections do. >80% of "mysterious cable sound effects" come down to this.
Really, the person that first had the idea of using coax lines as audio connections should be nominated for the ignobel prize. Of course, exotic cable salesmen will disagree, because it's this technical flaw that fills their wallet.
Now that you have a power amp with a differential input (and presumably a preamp with rca outputs) , make interconnects with an rca connector on the preamp side and a male xlr on the power amp side, using balanced cable. On the rca end, tie the cold lead and the shield together to ground. Wire the xlr end as appropriate:
pin1=shield, pin2=hot (signal from preamp), pin3=cold (the one that went to ground on the preamp side). In this way, ground currents that flow from chassis to chassis do not flow on the cold wire and place no voltage across it.
You'll notice that you can live with much longer interconnects now.
Re: mains filtering
Do not use mains filters with Y capacitors installed. An Y capacitor makes a capacitive connection (several nf) from the mains lines to your chassis and hence audio ground.
So there is the audiophile designer, using transformers with a shield to remove the slightest chance of the mains polluting his system ground, then installing a schaffner inlet that puts 2.2nF straight from the mains into the chassis, simply because he heard somewhere that mains disturbances are audible. They are, more precisely when such a filter is used.
All commercially available "combined mains filters" have Y capacitors. Don't use them. If all your audio devices are connected to one wall outlet (likely), you can use such a filter there, but by no means on individual boxes.
In principle you could make a filter with only chokes (cm/dm) and an X capacitor. Fine with me, but unless you have a problem with heavily polluted mains you don't need them. The UcD will not be a source of mains pollution. If your mains are polluted by an external cause (like you have a car factory next door), a centralised filter on the outlet is more effective.
Fyi, none of my audio devices have mains filters on them.
anomalous said:with regards to a power supply for the UcD modules,
what is the done thing in terms of mains filtering. a design posted earlier by paradigm used an X-Cap across the active/neutral mains lines. other designs ive seen use a varistor.
ive also seen IEC sockets with mains filters built in.
which one (or which combination) are the go?
thanks,
ben
Do not use mains filters with Y capacitors installed. An Y capacitor makes a capacitive connection (several nf) from the mains lines to your chassis and hence audio ground.
So there is the audiophile designer, using transformers with a shield to remove the slightest chance of the mains polluting his system ground, then installing a schaffner inlet that puts 2.2nF straight from the mains into the chassis, simply because he heard somewhere that mains disturbances are audible. They are, more precisely when such a filter is used.
All commercially available "combined mains filters" have Y capacitors. Don't use them. If all your audio devices are connected to one wall outlet (likely), you can use such a filter there, but by no means on individual boxes.
In principle you could make a filter with only chokes (cm/dm) and an X capacitor. Fine with me, but unless you have a problem with heavily polluted mains you don't need them. The UcD will not be a source of mains pollution. If your mains are polluted by an external cause (like you have a car factory next door), a centralised filter on the outlet is more effective.
Fyi, none of my audio devices have mains filters on them.
Hi
Today i shorted the caps between the 5532 and the UCD stage
(my CD player has a DC servo , i use a passive preamp and i measured an offset of 131mv on channel 1 and 2mv on the channel 2 )
The result was really amazing but after 10mn , C35 behind Tr9 blew up on the channel 1
The module still works , Tr9 and Tr10 are OK
Is it because i use a +/-52v power supply ( I've already replace the 470/50 caps with 470/63 ) or is it a DC problem please ?
Alain
Today i shorted the caps between the 5532 and the UCD stage
(my CD player has a DC servo , i use a passive preamp and i measured an offset of 131mv on channel 1 and 2mv on the channel 2 )
The result was really amazing but after 10mn , C35 behind Tr9 blew up on the channel 1
The module still works , Tr9 and Tr10 are OK
Is it because i use a +/-52v power supply ( I've already replace the 470/50 caps with 470/63 ) or is it a DC problem please ?
Alain
Sounds like pumping from here. This 131mV, was that at the output of the power amp or at the preamp?rha61 said:Hi
Today i shorted the caps between the 5532 and the UCD stage
(my CD player has a DC servo , i use a passive preamp and i measured an offset of 131mv on channel 1 and 2mv on the channel 2 )
The result was really amazing but after 10mn , C35 behind Tr9 blew up on the channel 1
The module still works , Tr9 and Tr10 are OK
Is it because i use a +/-52v power supply ( I've already replace the 470/50 caps with 470/63 ) or is it a DC problem please ?
Alain
Alain,
Watch out, the NE5532 has a big disadvantage and that's the output offset voltage.
You blew up now the small 22uF capacitor. Please change this to a 22uF/63V!
Regards,
Jan-Peter
Watch out, the NE5532 has a big disadvantage and that's the output offset voltage.
You blew up now the small 22uF capacitor. Please change this to a 22uF/63V!
Regards,
Jan-Peter
Bruno Putzeys said:
Sounds like pumping from here. This 131mV, was that at the output of the power amp or at the preamp?
Jan-Peter said:Alain,
Watch out, the NE5532 has a big disadvantage and that's the output offset voltage.
You blew up now the small 22uF capacitor. Please change this to a 22uF/63V!
Regards,
Jan-Peter
Thank you for your reply
The 131mv was at the power amp output
If i have to change some caps with 63v one , is it worthwile to replace them with Silmic or BG ?
Alain
Hi Bruno
All the stuff I've built so far contains those Schaffner filters. I was under the impression that I was doing my ears a favor by doing this. Maybe time to review this topology.😱
Sorry for being a bit of topic here
Cheers
Walter
"Fyi, none of my audio devices have mains filters on them."
All the stuff I've built so far contains those Schaffner filters. I was under the impression that I was doing my ears a favor by doing this. Maybe time to review this topology.😱
Sorry for being a bit of topic here

Cheers
Walter
BG: yes. Silmic: no. I mean to say: this 22u cap does make a small difference, but I don't like the silmics.rha61 said:
Thank you for your reply
The 131mv was at the power amp output
If i have to change some caps with 63v one , is it worthwile to replace them with Silmic or BG ?
If I'm not mistaken there's a post from JP somewhere in this thread showing how to add an overvoltage protection to the amp, so that it shuts down safely in case of pumping. It basically consists of a zener diode that triggers the current sense transistors.
131mV is normally not enough to make the power supply run away but you never know.
Thank you Bruno and Jan-Peter
Do you think i cant short these coupling caps as long as i keep the 5532 ? ( i was listening at middle level )
shorting them was really worthwhile
Alain
Do you think i cant short these coupling caps as long as i keep the 5532 ? ( i was listening at middle level )
shorting them was really worthwhile
Alain
It seems wise to move to an op amp with lower input bias current.
Most people favour jfet input op amps (ad8620, opa2134) for sonic reasons, and they don't have the input current problem either.
Most people favour jfet input op amps (ad8620, opa2134) for sonic reasons, and they don't have the input current problem either.
Bruno Putzeys said:It seems wise to move to an op amp with lower input bias current.
Most people favour jfet input op amps (ad8620, opa2134) for sonic reasons, and they don't have the input current problem either.
Hi Bruno,
And what opamp do you prefer? What would be your cost-no-object choice for that opamp?
Best regards
Gertjan
I make my own op amps with discrete parts. I haven't been using IC op amps for ages. Actually I've yet to hear the AD8620 for myself. I should be getting samples next week.
Alain,
Just checked the datasheet of the Philips NE5532, Vos is typical 0.5mV and max. 5mV. The total gain is 20x suppose the both NE5532 has the opposite offsetvoltage, you can get 200mV offset max. I woudn´t take the risk.
I didn´t do it for nothing the couplings caps 😉
Regards,
Jan-Peter
www.hypex.nl
Just checked the datasheet of the Philips NE5532, Vos is typical 0.5mV and max. 5mV. The total gain is 20x suppose the both NE5532 has the opposite offsetvoltage, you can get 200mV offset max. I woudn´t take the risk.
I didn´t do it for nothing the couplings caps 😉
Regards,
Jan-Peter
www.hypex.nl
AD8620 is pretty OK. I have it in my Gainclones and in two of my headphone amps. "Spitzenklasse" I would say. (did I spell it right? = world class)
peranders said:AD8620 is pretty OK. I have it in my Gainclones and in two of my headphone amps. "Spitzenklasse" I would say. (did I spell it right? = world class)
I'll listen first and will tweak some later modules to see the difference.
Spelling? For Bruno I expect Spitzenklasse is OK but for Jan-Peter I expect Spitsenklasse to be more appropriate but I could be wrong..
Hello Bruno,
The UcD180 looks like a great product and makes most of the other switching amps look well overpriced.
Questions:
How does the amp respond to capacitive loads? I have seen 0.1uF maximum specified elsewhere although your spec states "load independant bandwidth". Are there any issues with stability into difficult loads?
Regards,
Dave
The UcD180 looks like a great product and makes most of the other switching amps look well overpriced.
Questions:
How does the amp respond to capacitive loads? I have seen 0.1uF maximum specified elsewhere although your spec states "load independant bandwidth". Are there any issues with stability into difficult loads?
Regards,
Dave
UrSv said:
I'll listen first and will tweak some later modules to see the difference.
Spelling? For Bruno I expect Spitzenklasse is OK but for Jan-Peter I expect Spitsenklasse to be more appropriate but I could be wrong..
Hi UrSv,
For both Bruno and Jan-Peter it would be "Topklasse" as they are both Dutch speakers (as myself) 🙂
Best regards
Gertjan
- Home
- Amplifiers
- Class D
- UCD180 questions