No matter if one or two transformers are employed, grounds will be RF-coupled through transformer capacitance through mains wiring.
Re: Re: Daul mono, yes!
Here is a good page:
http://www.zero-distortion.com/techno/powersupply/powersi_05.htm
It discribes what most of us mean when we say dual bridge split secondaries.
Putting two Class D amps in one chassis is sub optimal. The on board ground potential of each amp will not be the same, and since the heatsinks are at the very least AC coupled to the same heatsink or chassis, you have a ground loop.
That's not to say it still can't be good, but its not ideal.
Its a budget thing I think.
I've never had monoblocks with the power grounding done with dual bridges, but will hopefully soon, also with some good caps that I haven't decided on yet.
but clearly isn't that better than having them mounted on the same heatsink?
Mike
JoshK said:
Here is a good page:
http://www.zero-distortion.com/techno/powersupply/powersi_05.htm
It discribes what most of us mean when we say dual bridge split secondaries.
Putting two Class D amps in one chassis is sub optimal. The on board ground potential of each amp will not be the same, and since the heatsinks are at the very least AC coupled to the same heatsink or chassis, you have a ground loop.
That's not to say it still can't be good, but its not ideal.
Its a budget thing I think.
I've never had monoblocks with the power grounding done with dual bridges, but will hopefully soon, also with some good caps that I haven't decided on yet.
Eva said:
but clearly isn't that better than having them mounted on the same heatsink?
Mike
When thick aluminium oxide insulators are employed, capacitance from the switching nodes to the heatsink is reduced to a few picofarads. In these circumstances, a dozen channels may be happily mounted on the same heatsink. Common-mode RF issues on wiring are far harder to solve than that.
Of course jensen 4 poles would FAR OUTDO cerafines!~
Ric, people do what they can within their means. I'm sure you're not doing all this as a "tweaker/listener" alone. Don't even tell me you're doing an auxiliary supply without a scope.
You have to realize what most commercial guys do is build monoblocks with your average low dime industrial parts... there's much bigger improvements to be had if you're building on a budget to go with high end stuff like 4 pole caps and IXYS rectifiers, with a proper layout.
Bruno himself has stated the difference to true mono isn't much, on a cost basis that means hardly worth it unless cost is no objective.
The trick is to experiment and get the layout/grounding scheme right... takes time, and some listening... right Ric?
Anyone doing a stereo amp off a single supply ought to realize the amp needs to be referenced to the input stage, which of course needs to be referenced to the source.
If you leave the front end to float anything goes. You won't get the best results! It will be sloppy and have degraded soundstage/imaging.
The modules' input stage ought to share the same ground, and that ground should be that of the source, I can't say it better than that.
Ric, people do what they can within their means. I'm sure you're not doing all this as a "tweaker/listener" alone. Don't even tell me you're doing an auxiliary supply without a scope.
You have to realize what most commercial guys do is build monoblocks with your average low dime industrial parts... there's much bigger improvements to be had if you're building on a budget to go with high end stuff like 4 pole caps and IXYS rectifiers, with a proper layout.
Bruno himself has stated the difference to true mono isn't much, on a cost basis that means hardly worth it unless cost is no objective.
The trick is to experiment and get the layout/grounding scheme right... takes time, and some listening... right Ric?
Anyone doing a stereo amp off a single supply ought to realize the amp needs to be referenced to the input stage, which of course needs to be referenced to the source.
If you leave the front end to float anything goes. You won't get the best results! It will be sloppy and have degraded soundstage/imaging.
The modules' input stage ought to share the same ground, and that ground should be that of the source, I can't say it better than that.
Eva said:
Eva,
What will help this cause of common mode issues?
Bruno has warned against Y caps I think (second hand info) because it couples it into ground.
Is this why IE's are better?
thanks
Mike
Common-mode filters are very handy, I use them even on oscilloscope probes as their effect is clearly visible on the traces.
Also, there is another issue with having several class-D amplifiers in close proximity playing similar signals without any switching synchronization: Beating frequencies.
Also, there is another issue with having several class-D amplifiers in close proximity playing similar signals without any switching synchronization: Beating frequencies.
classd4sure said:
Be Nice to Ric please!
I think what bruno said was there is an improvement with dual mono, and in his opinion it was worth while. In other words its a about how important it is to you. Opinion and value. Who'd blame Ric for going for the high end solution. And even if you do you can be respectful of his bizness cents.
I agree that the amp in a stereo single supply implemenation needs to be referenced to the input, and it is even if the signal ground is not connected to the source. That's what the instrumentation amp front end does. Commom mode rejection is quite good. It would be best if you could reference the amp ground to the source, but sometimes that causes another issue.
My reasoning for not connecting signal ground is that if you take a scope, and look at the two amps ground, they are far from a flat line. There is lots of switching noise that makes them differ. If you tie each amp signal ground to the source shield, then that noise voltage causes a noise current to flow between the shields of both right and left inputs to the source and back. (Assuming the shields are tied at the source.
You may be right, and it may be better your way. I can't try it becuase I have single ended inputs and I know that would be bad to have the HF noise current flowing in the shield which in single ended also has the signal in it.
If I recall, you have a TVC, did I say that right, tapped transformer based preamp.... Thus, you don't have to worry about that current loop thing at all. You got it good man!
I have no such luxury!
Thanks for the answer on Jensen vs Cerafin!
Warm Regards,
Mike
Eva said:
Eva,
So putting a common mode choke on the AC input is good.
X caps can't hurt, but they do nothing for common mode only transverse mode, corrent?
So, if you arn't willing to do an Xcap, you really can only gain from a single common mode choke then, correct?
Any other ways to get another 6dB per octave?
Have you heard of this "technically balanced technique" I think its called. Its were one uses basically an isolation transformer with a center tap secondary and ties that to earth.
does this make sense to you?
Thanks
Mike
Tweak on!
I hope my findings will stimulate some good listening tests. I need help. I cannot do it alone.
I am using a center tapped transformer in my amps...not dual bridges....so maybe this is why there is so much difference in going dual mono....however, my one an only experiment with using dual bridges versus one was back about 25 years ago when John Curl told a mutual friend that dual bridging was hot...so I tried it and found the opposite...the single bridge with ground to transformer sounded better....but heck, that was years ago...(maybe before some of you were born...he he)..and who knows what I would hear today using a power amp...I was using a phono stage then! At some point I will try two versus one bridge...but one thing you have to keep in mind when doing listening tests is to keep everything the same except for what you are testing...no different solder, bridges, chassis, transformers...nothing. Only this way will you have a result that has any meaning.
For instance. Let's say you want to A/B a toroidal tranny with a EI core one. They have to be the very same voltage and rating, you have to have the wires twisted the same, you have to make sure you hook up the primaries so the phase is the same (same AC potential to ground on secondaries, etc.). And then there is how you mount the transformer....how close it is to ground, if it is wobbling at all, etc. So, with everything exactly the same you might have one transformer that is sonically preferrable to another....Does that mean that either toroids or EI cores are better?...hardly. Just means you like one tranny over the other. Years ago I was making a pre regulator for a SOTA turntable (remember the flywheel, anyone?). So this was a 30 volt regulator that fed the 12 volt regulator inside the turntable which powered the motor which spun the platter with a belt. Well...I tried at least 10 different transformers (both EI and toroids) and they all made the turntable sound different....crazy....even more crazy was that the power cord to the supply also made the turntable sound different.
Both Steve McCormick and Elliot Callen (SP?) of Counterpoint have listened to many transformers. They both say the Plitron is the bomb. This is why I am using it....maybe there is a better sounding transfromer out there but I don't have time to ty everything. Sometimes I rely on the opinion of people I trust. I will be trying the super expensive transformers from Audio-consulting.ch and if they sound better, then I will be offering a more expensive version of my power amp with them.
One thing you guys may not know....is that the steel plate and steel screw mounting of toroids and diode bridges messes up the sound....you do not want steel plates on top of your transformer nor steel screws or anything magnetic....steel chasss are probably bad for the same reason. I use bronze hardware to mount my toroids and I always use toroids that have a potted center so I don't need plate mounting.
Here is an experiment for you all to try...remove the hardware on your toroids...just leave them sitting there....tell us what you hear...if you like what you hear without the steel plates and bolts then you can drill four holes inside the tranny and four holes around the outside and use tie wraps to secure the tranny down. Remember to try this one your diode bridges if you are using a steel bolt to secure to the chassis....use bronze or brass....most stainless steel hardware is magnetic. I use discrete diodes because "in the past" I have found discrete diodes sound better than their equivalent bridges I damp the diodes so they do not ring...sounds much better.
I think a lot of DIYers are cheapskates...he he....I mean, if you can build an ALL OUT amp for one tenth or less of what you would pay retail...why not spend the money? $1000 in parts for an amp that sounds better than a $25,000 amp! Yo, that's the bomb.....however, if $200 if really all you can spend...and for some that is a reality, then I can understand not wanting to get Jensens, Plitrons, go dual mono, try Blackgates, wild wire, WBTs, etc. Of course, there are those of you who do not believe that any of this makes any difference....so then why spend the money.
To me the beauty of DIY is that you can make something that can sound better than you could EVER buy. I guess I just like finding out what something can really do....you know, the hot rodder syndrome...."floor it, dude". Have fun.
I hope my findings will stimulate some good listening tests. I need help. I cannot do it alone.
I am using a center tapped transformer in my amps...not dual bridges....so maybe this is why there is so much difference in going dual mono....however, my one an only experiment with using dual bridges versus one was back about 25 years ago when John Curl told a mutual friend that dual bridging was hot...so I tried it and found the opposite...the single bridge with ground to transformer sounded better....but heck, that was years ago...(maybe before some of you were born...he he)..and who knows what I would hear today using a power amp...I was using a phono stage then! At some point I will try two versus one bridge...but one thing you have to keep in mind when doing listening tests is to keep everything the same except for what you are testing...no different solder, bridges, chassis, transformers...nothing. Only this way will you have a result that has any meaning.
For instance. Let's say you want to A/B a toroidal tranny with a EI core one. They have to be the very same voltage and rating, you have to have the wires twisted the same, you have to make sure you hook up the primaries so the phase is the same (same AC potential to ground on secondaries, etc.). And then there is how you mount the transformer....how close it is to ground, if it is wobbling at all, etc. So, with everything exactly the same you might have one transformer that is sonically preferrable to another....Does that mean that either toroids or EI cores are better?...hardly. Just means you like one tranny over the other. Years ago I was making a pre regulator for a SOTA turntable (remember the flywheel, anyone?). So this was a 30 volt regulator that fed the 12 volt regulator inside the turntable which powered the motor which spun the platter with a belt. Well...I tried at least 10 different transformers (both EI and toroids) and they all made the turntable sound different....crazy....even more crazy was that the power cord to the supply also made the turntable sound different.
Both Steve McCormick and Elliot Callen (SP?) of Counterpoint have listened to many transformers. They both say the Plitron is the bomb. This is why I am using it....maybe there is a better sounding transfromer out there but I don't have time to ty everything. Sometimes I rely on the opinion of people I trust. I will be trying the super expensive transformers from Audio-consulting.ch and if they sound better, then I will be offering a more expensive version of my power amp with them.
One thing you guys may not know....is that the steel plate and steel screw mounting of toroids and diode bridges messes up the sound....you do not want steel plates on top of your transformer nor steel screws or anything magnetic....steel chasss are probably bad for the same reason. I use bronze hardware to mount my toroids and I always use toroids that have a potted center so I don't need plate mounting.
Here is an experiment for you all to try...remove the hardware on your toroids...just leave them sitting there....tell us what you hear...if you like what you hear without the steel plates and bolts then you can drill four holes inside the tranny and four holes around the outside and use tie wraps to secure the tranny down. Remember to try this one your diode bridges if you are using a steel bolt to secure to the chassis....use bronze or brass....most stainless steel hardware is magnetic. I use discrete diodes because "in the past" I have found discrete diodes sound better than their equivalent bridges I damp the diodes so they do not ring...sounds much better.
I think a lot of DIYers are cheapskates...he he....I mean, if you can build an ALL OUT amp for one tenth or less of what you would pay retail...why not spend the money? $1000 in parts for an amp that sounds better than a $25,000 amp! Yo, that's the bomb.....however, if $200 if really all you can spend...and for some that is a reality, then I can understand not wanting to get Jensens, Plitrons, go dual mono, try Blackgates, wild wire, WBTs, etc. Of course, there are those of you who do not believe that any of this makes any difference....so then why spend the money.
To me the beauty of DIY is that you can make something that can sound better than you could EVER buy. I guess I just like finding out what something can really do....you know, the hot rodder syndrome...."floor it, dude". Have fun.
This thread IS hot! I turned the head and I missed all this most interesting talk! 🙁
Dear Ric:
First of all, thank you. 🙂 I owe you my first DIY project 😀
You seem to be using my same approach and TX's 😎
Monoblocks image like crazy. Depth, oh depth...
I use Epcos Sikorel 10000uF*4 per mono amp, wich reminds me that our mate rha61 gave me the probable answer to my HF problem (sorry Alain and Mike
) months ago. He recommends pi filtering for Sikorel, they being so fast and transparent, he said, that VHF can influence audible range somehow.
Pi filter, what's that? I said. I can't afford money or space for big air core chokes
...so I forgot the issue. Well, maybe I could use a little N30 ferrite core plus a low R to simulate the choke??
I digress, sorry...or maybe not.
Dear Eva:
Will there be sound improvement in filtering the mains ground for RF also?
Mike:
I accept all kind of comments and advices 😀
caps, bridges... Or with a "kitchen grade" teflon foil base. This way they don't couple to chassis vibes too much.
Cheers
M
Dear Ric:
First of all, thank you. 🙂 I owe you my first DIY project 😀
You seem to be using my same approach and TX's 😎
Monoblocks image like crazy. Depth, oh depth...
I use Epcos Sikorel 10000uF*4 per mono amp, wich reminds me that our mate rha61 gave me the probable answer to my HF problem (sorry Alain and Mike
) months ago. He recommends pi filtering for Sikorel, they being so fast and transparent, he said, that VHF can influence audible range somehow. Pi filter, what's that? I said. I can't afford money or space for big air core chokes
...so I forgot the issue. Well, maybe I could use a little N30 ferrite core plus a low R to simulate the choke??I digress, sorry...or maybe not.
Dear Eva:
So, it's a good thing to develop a power filter 😉
Will there be sound improvement in filtering the mains ground for RF also?
Mike:
I will begin testing my "technical balanced power filter" for influence in sound with the individual addition (from mains to units) of: Transient Voltage Suppresor (instead of VDR's); X caps on prymaries; then on secondaries, Dual Aperture Core N30 material ferrites (I think they can mannage common and differential mode noise) ; Y caps from + and - , to ground. After my listening test I hope I will end with a decent power filter to recommend.
I accept all kind of comments and advices 😀
And what about plastic screws and/or wood base?
On the run, I let all components "floating" 😀
caps, bridges... Or with a "kitchen grade" teflon foil base. This way they don't couple to chassis vibes too much. Cheers
M
Rick,
You wrote: "I have found discrete diodes sound better than their equivalent bridges I damp the diodes so they do not ring...sounds much better."
Care to elaborate on what you mean by & how you go about damping the diode?
Also does anyone have experience with Cree Schottky Diodes?
Thanks in advance.
You wrote: "I have found discrete diodes sound better than their equivalent bridges I damp the diodes so they do not ring...sounds much better."
Care to elaborate on what you mean by & how you go about damping the diode?
Also does anyone have experience with Cree Schottky Diodes?
Thanks in advance.
I use the high voltage Cree schottkies in my tube amp, they are very good.
I use MBR10100 in my UCD180 which seem ok, the higher voltage/current version is MBR2020
Anybody tried the soft recovery MSR1560's
I use MBR10100 in my UCD180 which seem ok, the higher voltage/current version is MBR2020
Anybody tried the soft recovery MSR1560's
Portlandmike said:
I'll give it a try.
Since I use my UCD's with a much higher gain than normal(180ohms gain resistor) they became prone to capacitive coupling (what degraded my soundstage)that's why I need to ground the amps at the center tap of the PSU's cap. bank and not at the inputs(tried all of the suggested grounding schemes discussed here). This would destroy my stereo image too much. Even touching the wires in the amps. would influence the soundstage( depth). Now it does not make a difference if I touch the wires or move them. Also changed my standard rectifier bridges with schottky diodes(MBR20100) and compared it with an amp. with standard rectifiers...could not hear a difference.
Well, I'll keep on trying.
t.
You wrote: "I use the high voltage Cree schottkies in my tube amp, they are very good"
Did you use any bypass caps on the Cree diodes when you used them for your tube amp? If so what value & type did you use?
Thanks
You wrote: "I use the high voltage Cree schottkies in my tube amp, they are very good"
Did you use any bypass caps on the Cree diodes when you used them for your tube amp? If so what value & type did you use?
Thanks
Yves Smolders said:
Hi guys,
I've bene searching my pants off to find the schematics mentioned above. Unfortunately (but probably understandable) Julien's email forum function is disabled.
Can someone point me to these schematics? I've already seen thepics in the Julien's gallery, but that's not helpful enough, for me at least.
Thanks!
Stevenacnj said:
I just used some high voltage 10nf ceramics across them in the tube amp, I've never compared it without them in that amp so can't really say if they made a difference🙂
I'd like to etch a psu board for my UCD180, please could anybody comment if it would be ok or is there anything I should add or change, how about a 10uH inductor between the two caps
Any snubbers etc
Thanks
Mauricio,maxlorenz said:This thread IS hot! I turned the head and I missed all this most interesting talk! 🙁
I use Epcos Sikorel 10000uF*4 per mono amp, wich reminds me that our mate rha61 gave me the probable answer to my HF problem (sorry Alain and Mike
Pi filter, what's that? I said. I can't afford money or space for big air core chokes
I digress, sorry...or maybe not.
Cheers
M
Wecome back. You are so buzy doing stuff all the time I'm amazed. A pi filter is a filter that uses a
Cap, then inductor to another cap. Make sense?
If you do that on the primary, my first choice, but no experience, then make sure you use real X caps. You might replace the inductor with a common mode choke and kill to birds with one stone attenuating both transverse and common mode noise.
Mike
Bgt said:
Remember, its only a hunch!
What do you mean by "sensistive to capacitive coupling?
Mike
Stevenacnj said:
I hear of poeple using Schotky diodes, so maybe they sound good. Schotky's shut off instantly, that is zero recovery.
Also, compared to standard diodes they have a slug more capacitance.
If they sound good, great! I'm skeptical that this would be a step in the right direction give all the hooopla over fast soft recovery.
No listening experience though.
Also, no suprise that the cap accross isn't audible. Try a series resistor accross the secondaries. Start with 1uF and 47ohms and decrease the resistance. Your resonance could be quite hi Q because of the Schotky capacitance.
Perhaps the Cree schotky's are not high capacitance?
t. said:
Looked at your layout. If your wanting to do the dual bridge, then you should add a cut seperating your +/- caps until the far right side of the layout by all the vias.
Also, don't make the trace so wide around the cap terminals. Think of forcing all the electrons under the cap terminal pad.
This is a very significant and measurable affect in switching regulators, and also a common mistake. The added (minimal) trace resistance (up until it to small to carry the current) and inductance is a filter feature in this case.
You may want to ignore this later comment in case Ric is right and center tap sounds better. Then you can just solder with solder wick to short the slit for a center tap arrangement and populate only half the diodes.
I'd also add a spot for a series cap and resistor right before (to the left) of the bridge. That would be accross the secondary inputs input pads.
Some people put caps and sometimes series RC's accross each diode leg too. Even though I don't get this method, it might be good to do if your going to go to the trouble of a board.
Also, Please consider point to point wiring. If you dig, you'll find an excellent example that ClassDforsure did. Its about as good as you get, and likely way better than a circuit board. He used Jensen 4 poles though, but the principle is the same.
Good Luck (Think about a good option for shorting the fuses too 😉
Mike
PCB question
Mikes suggestions are good but I would go further and completely separate the 2 grounds, running 2 separate wires into the single ground terminal at the amp. This makes the amp module the only place all the grounds are connected together as in implementing a star ground.
I would also suggest a quality film cap of sufficient voltage and around .1uf across each transformer output winding to take care of the switching transient caused by leakage inductance. These should be capable of handling high freq. so should be stacked film pc mount types.
You might also consider adding other pads so you can try different types of the large caps.
Roger
t. said:
Mikes suggestions are good but I would go further and completely separate the 2 grounds, running 2 separate wires into the single ground terminal at the amp. This makes the amp module the only place all the grounds are connected together as in implementing a star ground.
I would also suggest a quality film cap of sufficient voltage and around .1uf across each transformer output winding to take care of the switching transient caused by leakage inductance. These should be capable of handling high freq. so should be stacked film pc mount types.
You might also consider adding other pads so you can try different types of the large caps.
Roger
- Status
- Not open for further replies.