Sooo....
Coming back on the various tests we did. OK, these aren't blind tests, but well... nearly. I let all filters in their identical casings and brought amp, jumpers and filters to my friend Gilles without a word. he had no clue on what was done nor what to expect, except one filter had a knot to be identified. I said: Just try it over the coming days, take your time, compare with your Class D and Onkyo, and come back
I just gave hints about the different combos and way to plug them (jumpers and filters). I didn't change the bias. Nor have I yet tried the ACA Mini vs my own VFET. I wasn't there when he did his tests, I just had a quick listening at the begining - making sure all was right - and a few days later to enjoy the result. LS are Klipsch RP-8000F, very well suited to this amp. Source was high res through Eversolo Z8, quite decent DAC BTW.
First of all, whatever the filters and jumpers, the ACA Mini is a fabulous little amp! It is marvelous, not just cute, but really a non expensive amp that sounds great. Probably the best option vs Class D for beginners! An absolutely great addition!
Now to the results. I suspect jumper position is system related, but to our amazement, shame on us "distorders", jumpers out was this time prefered. Sounded for whatever reason more real, closer. Strangely, "in" sounded a tad more holographic, something normaly associated to H2 but well. Flattering but sounded for whatever reason less realistic, less homogeneous. So quite the opposite from what I would have expected as engineer, but Gilles is musician and has also good ears...
More important now, as anyone can play easily with the jumpers: to the filters.
I love MJ's SMPS filters and they never brought any negative results. I built over a dozen of them and always have 2 extra "ready to try".
Well, this time the result is more mitigated. Yes, one filter indeed cleaned the music message and brought more clarity and details. But it robbed some dynamics at Gilles according to him. Something I hadn't experienced so far, or experienced sometimes when daisy chaining 2 of them as I did for my own VFET before moving back (and back then Gilles was also there with me when I did all this). BTW, daisy chaining here didn't brought any improvement on that, but that could be expected at least probably.
So yes, the usual SMPS filter brings some positive despite the VG "oversized low ripple SMPS", but is overall not really a great addition here due to the pereived dynamic loss. So rather mitigated.
On the contrary, the simple Murata coil between the SMPS and ACA Mini sounded.... as the ACA Mini should have sounded IMHO. Just RIGHT. First off, it did a lot of cleaning, more than MJ's Filters. Secondly, it didn't rob any of the dynamics. Nor anything else. In short: it is a worthwhile non expensive addition without any negative (to our ears). You could build that filter in a small casing the way I did it and just plug it in and out for quick comparisons. Or just think on how to include it directly. IMHO, the ACA Mini kit from the store could greatly benefit from this additional part in the future - it costs just a couple of bucks and really sounds as if it was made for it.
Gilles tried daisy chaining MJ's filter and the Murata coil filter (credits go here to MJ aswell anyway!) and that brought no additional benefits... nor really any negative BTW. It is like the coil filter dominated the overall sonic signature and dynamic. We had only one coil, so we didn't try daisy chaining the "coilfilters", but given the expected results on the paper not sure it would bring more... but who knows?
At the end of the day Gilles loved the ACA Mini so much he is... likely to have one in a very short time. Or perhaps even 2 as they come cheap and it could be fun experimenting (various) biamping configurations.
The big SMPS we have allow us to expriment aswell in this area (one SMPS + 1 coil filter for 2 amps... or doubling it all) before finalising the end set up.
There is also one thing I wish one day to experiment: additional conventional filtering upstream (CLC) or even linear PS. In a perfect world at Gilles we would keep everything of the ACA Mini including its sonic signature (very close to my VFET on that point) but say with (even) more speed or slam, immediacy, without loss in fluidity though. That would then be a giant killer... well, it probably is anyway already with efficient LS!
Bottom line: try that Murata coil ASAP! Unless you like it very relaxed with the std SMPS filter, you are likely to be very surprised. Either way, once again with SMPS filtering is the way forward.
I hope this helps
Claude (and Gilles of course)
Coming back on the various tests we did. OK, these aren't blind tests, but well... nearly. I let all filters in their identical casings and brought amp, jumpers and filters to my friend Gilles without a word. he had no clue on what was done nor what to expect, except one filter had a knot to be identified. I said: Just try it over the coming days, take your time, compare with your Class D and Onkyo, and come back
I just gave hints about the different combos and way to plug them (jumpers and filters). I didn't change the bias. Nor have I yet tried the ACA Mini vs my own VFET. I wasn't there when he did his tests, I just had a quick listening at the begining - making sure all was right - and a few days later to enjoy the result. LS are Klipsch RP-8000F, very well suited to this amp. Source was high res through Eversolo Z8, quite decent DAC BTW.
First of all, whatever the filters and jumpers, the ACA Mini is a fabulous little amp! It is marvelous, not just cute, but really a non expensive amp that sounds great. Probably the best option vs Class D for beginners! An absolutely great addition!
Now to the results. I suspect jumper position is system related, but to our amazement, shame on us "distorders", jumpers out was this time prefered. Sounded for whatever reason more real, closer. Strangely, "in" sounded a tad more holographic, something normaly associated to H2 but well. Flattering but sounded for whatever reason less realistic, less homogeneous. So quite the opposite from what I would have expected as engineer, but Gilles is musician and has also good ears...
More important now, as anyone can play easily with the jumpers: to the filters.
I love MJ's SMPS filters and they never brought any negative results. I built over a dozen of them and always have 2 extra "ready to try".
Well, this time the result is more mitigated. Yes, one filter indeed cleaned the music message and brought more clarity and details. But it robbed some dynamics at Gilles according to him. Something I hadn't experienced so far, or experienced sometimes when daisy chaining 2 of them as I did for my own VFET before moving back (and back then Gilles was also there with me when I did all this). BTW, daisy chaining here didn't brought any improvement on that, but that could be expected at least probably.
So yes, the usual SMPS filter brings some positive despite the VG "oversized low ripple SMPS", but is overall not really a great addition here due to the pereived dynamic loss. So rather mitigated.
On the contrary, the simple Murata coil between the SMPS and ACA Mini sounded.... as the ACA Mini should have sounded IMHO. Just RIGHT. First off, it did a lot of cleaning, more than MJ's Filters. Secondly, it didn't rob any of the dynamics. Nor anything else. In short: it is a worthwhile non expensive addition without any negative (to our ears). You could build that filter in a small casing the way I did it and just plug it in and out for quick comparisons. Or just think on how to include it directly. IMHO, the ACA Mini kit from the store could greatly benefit from this additional part in the future - it costs just a couple of bucks and really sounds as if it was made for it.
Gilles tried daisy chaining MJ's filter and the Murata coil filter (credits go here to MJ aswell anyway!) and that brought no additional benefits... nor really any negative BTW. It is like the coil filter dominated the overall sonic signature and dynamic. We had only one coil, so we didn't try daisy chaining the "coilfilters", but given the expected results on the paper not sure it would bring more... but who knows?
At the end of the day Gilles loved the ACA Mini so much he is... likely to have one in a very short time. Or perhaps even 2 as they come cheap and it could be fun experimenting (various) biamping configurations.
The big SMPS we have allow us to expriment aswell in this area (one SMPS + 1 coil filter for 2 amps... or doubling it all) before finalising the end set up.
There is also one thing I wish one day to experiment: additional conventional filtering upstream (CLC) or even linear PS. In a perfect world at Gilles we would keep everything of the ACA Mini including its sonic signature (very close to my VFET on that point) but say with (even) more speed or slam, immediacy, without loss in fluidity though. That would then be a giant killer... well, it probably is anyway already with efficient LS!
Bottom line: try that Murata coil ASAP! Unless you like it very relaxed with the std SMPS filter, you are likely to be very surprised. Either way, once again with SMPS filtering is the way forward.
I hope this helps
Claude (and Gilles of course)
Claude,
I found MJ's AmyAlice filter to do a better job with the ACP+ and ACA Mini, over the PO89ZB filter.
Though I might have to give the Murata 1433428C a try.
I found MJ's AmyAlice filter to do a better job with the ACP+ and ACA Mini, over the PO89ZB filter.
Though I might have to give the Murata 1433428C a try.
If anybody is trying to put together an LTSPICE simulation, the unwanted but oh-so-very-real parasitics of the Murata inductor were stated in post #849:
However, you are on your own when seeking the oh-so-very-real parasitic values of the two "C5" electrolytic capacitors, one per channel. I don't know them.
_
However, you are on your own when seeking the oh-so-very-real parasitic values of the two "C5" electrolytic capacitors, one per channel. I don't know them.
_
Attachments
Thanks a lot for your reply, Papa!
My guess is with linear PS you have the choice between say CRC or CLC "variations" and it might be more a matter of taste and cost, whereas with SMPS there is nothing to beat an L in the line to get read of high frequency noise. Is that correct?
Thanks a lot to Mark for this wonderfuly simple idea (before reading his post, I wanted to use the very same part, but in a far more complicated way!)... and to carry on now helping with the scientifical explanation / model behind.
And thanks a lot to 6L6 for the build guide and... his kindness. Why? No the background - Why did build this amp whereas I have enough amps at home, wife and kid wondering what that fragile amp could do at ours, plus I have my very own amp to build? Because of Jim / 6L6. He got me (re) started ato this wonderfull hobby after 15y and his enthousisam to tease and help future people is just great.
Inspired by this I build a dozen of Mark's excellent SMPS filter... mainly to try them at friend's, for them. I built the ACA Mini because it is cute, fun, easy, non expensive... and easy to carry at friend's to get THEM started on Papa's wonderfull kits. And Gilles wants now firm 1 or 2 ACA Mini, and another friend is already standing queue, willing to test it at his.
For this tests / loan to friends, not handy with the english language / forum, SMPS + external coil in casing (always fun to play with and without it) + ACA Mini is just perfect and good enough.
For Gilles, and a more ultimate ACA Mini (more power is not needed here, we are after a bit more speed and slam... and that only, without changing anything else to the sonic signature), I have something different in mind re PS and that also goes back to my initial (and more complex) idea of using that coil. More a bit later, time permitting, as I would be delighted to have your POV on 2 possible options and don't have your skills modelling PS systems.
This is trully a fabulous forum, so educative, so fun, with so many kind people to help... everyone should build one of the projects here, be it for the fun of it, not even mentioning the sonic qualities and excellent VFM!
Claude
My guess is with linear PS you have the choice between say CRC or CLC "variations" and it might be more a matter of taste and cost, whereas with SMPS there is nothing to beat an L in the line to get read of high frequency noise. Is that correct?
Thanks a lot to Mark for this wonderfuly simple idea (before reading his post, I wanted to use the very same part, but in a far more complicated way!)... and to carry on now helping with the scientifical explanation / model behind.
And thanks a lot to 6L6 for the build guide and... his kindness. Why? No the background - Why did build this amp whereas I have enough amps at home, wife and kid wondering what that fragile amp could do at ours, plus I have my very own amp to build? Because of Jim / 6L6. He got me (re) started ato this wonderfull hobby after 15y and his enthousisam to tease and help future people is just great.
Inspired by this I build a dozen of Mark's excellent SMPS filter... mainly to try them at friend's, for them. I built the ACA Mini because it is cute, fun, easy, non expensive... and easy to carry at friend's to get THEM started on Papa's wonderfull kits. And Gilles wants now firm 1 or 2 ACA Mini, and another friend is already standing queue, willing to test it at his.
For this tests / loan to friends, not handy with the english language / forum, SMPS + external coil in casing (always fun to play with and without it) + ACA Mini is just perfect and good enough.
For Gilles, and a more ultimate ACA Mini (more power is not needed here, we are after a bit more speed and slam... and that only, without changing anything else to the sonic signature), I have something different in mind re PS and that also goes back to my initial (and more complex) idea of using that coil. More a bit later, time permitting, as I would be delighted to have your POV on 2 possible options and don't have your skills modelling PS systems.
This is trully a fabulous forum, so educative, so fun, with so many kind people to help... everyone should build one of the projects here, be it for the fun of it, not even mentioning the sonic qualities and excellent VFM!
Claude
Ok, general question to you experienced guys... my brain is a bit tired today and can't work out between 2 models LOL (plus I might have a bias on one of them, habits habits...)
Say you have a decent big SMPS with sadly an unknown output impedance (and I presume these devices have some caps at the output).
Say you want to drive any power amp directly, that is without its PS section (no caps, no rectifier, no filter). Obviously you don't have its impedance either, let's go for common values / ball park?
Now inbetween the SMPS and the amp you want to put some filtering device, obviously including some caps.
The question is: doing this, are you better off going for a filtering device corresponding to LCR... or to RLC ?
In either case the values of L, R and C would be the same.
If you really need some values, let's take L = 330uH , R=0.3R, C= 15000uF and assume near perfect components.
Can anyone give me a hint on the differences / prefered option between LRC or RLC on a pure filtering basis?
My poor brain says "LRC = L doing some mild HF filtering (TBD!) - followed by a classical RC first order filter" vs " RLC = Useless R to reduce just voltage - followed by classical second order (steeper / better than 1st order) LC filtering", so the latter would possibly win re filtering unless the coil is doing a great job, which is possible. And as in both cases the amp sees the same C there is no fear re PS output impedance .... but there must for sure be more to it as placing the coil where I just did it with ACA Mini (and not where I wanted it initialy, thanks to MJ's great hint!) proved very effective...
Sorry for the stupid question, bit tired tonight, this is part of a bigger scheme for the ACA Mini, having an answer / confirmation of that simple question would reduce all the options currently flying in my head... for a bigger scheme!
Many thanks
Claude
PS: I am old shool (no sim software) and with an old nearly useless brain (eg my differential equation solving is very very rusty) - Ok, I admit I am very lazy as well here to work it out, but that's Deep Purple's fault ;-)
Say you have a decent big SMPS with sadly an unknown output impedance (and I presume these devices have some caps at the output).
Say you want to drive any power amp directly, that is without its PS section (no caps, no rectifier, no filter). Obviously you don't have its impedance either, let's go for common values / ball park?
Now inbetween the SMPS and the amp you want to put some filtering device, obviously including some caps.
The question is: doing this, are you better off going for a filtering device corresponding to LCR... or to RLC ?
In either case the values of L, R and C would be the same.
If you really need some values, let's take L = 330uH , R=0.3R, C= 15000uF and assume near perfect components.
Can anyone give me a hint on the differences / prefered option between LRC or RLC on a pure filtering basis?
My poor brain says "LRC = L doing some mild HF filtering (TBD!) - followed by a classical RC first order filter" vs " RLC = Useless R to reduce just voltage - followed by classical second order (steeper / better than 1st order) LC filtering", so the latter would possibly win re filtering unless the coil is doing a great job, which is possible. And as in both cases the amp sees the same C there is no fear re PS output impedance .... but there must for sure be more to it as placing the coil where I just did it with ACA Mini (and not where I wanted it initialy, thanks to MJ's great hint!) proved very effective...
Sorry for the stupid question, bit tired tonight, this is part of a bigger scheme for the ACA Mini, having an answer / confirmation of that simple question would reduce all the options currently flying in my head... for a bigger scheme!
Many thanks
Claude
PS: I am old shool (no sim software) and with an old nearly useless brain (eg my differential equation solving is very very rusty) - Ok, I admit I am very lazy as well here to work it out, but that's Deep Purple's fault ;-)
Last edited:
I have a dumb question. I think I know the answer but will seek your op[inion to be sure. If the 330uH Murata is used between the SMPS and supply is any kind of diode needed across the coil as with a relay, to mitigate kickback? Or, unlike a relay, will no dode be needed because the coil will remain under load when the power has been shut off? I'm thinking the latter but would like to be sure. Thanks in advance for any insight.
Not sure I understand your post correctly, but just as a side note to make sure we get the same picture...
The coil is in series with the +, not in parallel between plus and neutral. Eg. no current flowing = no tension at the coil.
Obviously at shut off the coil will for a very limited time per nature fight against this change of status (24V vs 0V) and try to compensate (it won't but will try to) , but it is a ("very" in our case) transient matter. Same at power on...
I hope this helps - although you perhaps got it right first time anyway and if so sorry for the interference
Claude
The coil is in series with the +, not in parallel between plus and neutral. Eg. no current flowing = no tension at the coil.
Obviously at shut off the coil will for a very limited time per nature fight against this change of status (24V vs 0V) and try to compensate (it won't but will try to) , but it is a ("very" in our case) transient matter. Same at power on...
I hope this helps - although you perhaps got it right first time anyway and if so sorry for the interference
Claude
As said normal operational transient, so I would say normal.
At least at ours it didn't cause any side effect, os we didn't bother at all. Sale as caps discharging really...
Just as a side note, one may think that shutting off the PS (not the switch on the ACA Mini obviously, but the SMPS power) would cause the coil to discharge and whatever distorted tone to "come out". Well, if so forget the coil, our (indeed uprated) SMPS despite being unplugged.... lasts for several dozen of seconds and the ACA Mini keeps playing undistorted for a while, and then distords womewhat shortly and funnily. Not that I would do that everytime, but at least we tried once without willing to LOL.
I would say the biggest time constant in our amp so far has been... from the SMPS side at switch off. Unsuspected by us until we found out.
BUT... back to topic, anyone to answer my question re LRC vs RLC? ;-)
At least at ours it didn't cause any side effect, os we didn't bother at all. Sale as caps discharging really...
Just as a side note, one may think that shutting off the PS (not the switch on the ACA Mini obviously, but the SMPS power) would cause the coil to discharge and whatever distorted tone to "come out". Well, if so forget the coil, our (indeed uprated) SMPS despite being unplugged.... lasts for several dozen of seconds and the ACA Mini keeps playing undistorted for a while, and then distords womewhat shortly and funnily. Not that I would do that everytime, but at least we tried once without willing to LOL.
I would say the biggest time constant in our amp so far has been... from the SMPS side at switch off. Unsuspected by us until we found out.
BUT... back to topic, anyone to answer my question re LRC vs RLC? ;-)