Uh, sorry, I've thought you were asking about thick film... obviously Mouser say nothing about it.
Thick film resistors vs thin films is similar to carbon film vs metal film, IMHO.
Both thick and carbon films have usually higher TC and, if you trust this sort of things, fuller sound; both also are less detailed and precise than thin films/metalfilms.
Siva, I really like (and am a little jealous) of your approach. Having intact stereo pairs - containing various component sets - has to be the most valid platform for comparisons. Though sometimes when I look at my collection I think I must be either nuts or greedy, I hope to be able to duplicate that setting at some point with V1.2, V1.3, FE RC an FE Final.
I'm not clear on the changes between 1.3 and 1.4. Can you fill us in?
Also, I'm curious about how you rate/value/prioritize/combine... simulation, lab/scope readings and listening tests. Are you finding good correlation in the results?
I'm not clear on the changes between 1.3 and 1.4. Can you fill us in?
Also, I'm curious about how you rate/value/prioritize/combine... simulation, lab/scope readings and listening tests. Are you finding good correlation in the results?
I intend to change the transformers fairly soon to either R core or encapsulated types, either way they will be lower rated, probably a maximum of 160VA. Lower value fuses will obviously be needed but I don't think i'll bother with inrush limiting or soft start facility with such small transformers.
(Please bear with my delayed/missing responses - I'm down with a cold/flu, so will probably be slow/missing for the next week or so.)
V1.3 and V1.4 differ mainly in slight layout changes and component flexibility. The layout changes are incremental improvements, such as reduced parasitic inductances with multi-filar routing for some capacitors (C4, C22, C7, C23). For several of the components (electrolytics and film caps mainly), alternate pad spacing (pitches) have been provided in V1.4, for instance to allow 2.5, 5 or 7.5mm pitch caps at C7. V1.4 can also accommodate some larger parts, for instance 15mm pitch box-type caps at C4, and 16mm diameter electrolytics at C1, C2 and C9. Some resistor locations can accommodate longer and/or wider resistors. Some capacitors (IIRC C17, C18, C19, C20), which are no longer in use in any BoM/variant, have been completely dropped to create space.
I do use simulation to get an idea of which direction I'd like the tweaks to proceed in, but since there are no official good working public Spice models of either the LM318 or the LM3886, I still use unofficial or substitute models (e.g. LM1875 instead on LM3886). It has certainly helped to get some insight on stability; but for audible sonics, there is no substitute to careful auditioning. My measurement capabilities are still rudimentary, but I'll try to remedy that, going forward.
Broadly, the results have been consistent with the simulations so far. There have been some mishaps (the instability in the MiniRef 3886 didn't show up in simulation and had to be solved manually by trial-and-error on the prototype board), but also some excellent predictions from simulation (MiniRef 1875 is virtually exactly identical between the simulation schematic and fully-functional first-pass physical boards, for which I'd give most of the credit to the excellent unofficial LM1875 Spice model from Pedja).
V1.4 sounds extremely user friendly and flexible. Great for at home trials and tweaks. Maybe if we can get like 500 MyRefs builders to bulk email TI we can force them to publish a new model. 😀
Thanks for the info and take care of your health first and foremost. We will still be here doing the same things when you get back.
Thanks for the info and take care of your health first and foremost. We will still be here doing the same things when you get back.
Three comments to make about my current setup, and that's enough.
Dario, the MF resistors I used for R104/204 are Vishay/Dale CMF. I am not aware of any sonic issues with these parts. They have a pretty good reputation. In fact, some modders use them to replace cheap Xicon and such. I doubt that they contribute at all to amplifier harshness.
I completely disagree with your comments about the TC at C13. This is the only time I will say such a thing. I think you are incorrect to suggest it is the cause of brightness or an edge to the sound of the FE. The TC capacitor sounded perfectly balanced and accurate with the Rev C, and not because the Rev C was masking anything. It sounds just as good, perhaps better, with the FE if it is properly tuned. If my current compensation configuration is causing distortion, it is the most beautiful distortion I have ever heard. If the TC cap is producing brightness, I cannot hear it. In my opinion, the TC is a reference part and the single component that is NOT subject to trial and error. I played that game too long with too many caps and too much expense. The TC is easily superior to every other cap I tried, including Mundorf. To simplify it: if your amp doesn't sound good with the TC in it, it is the amplifier at fault, not the cap.
Bob, regarding soundstage: last night I had my first session listening to a pair of matched monoblocks with the 33/15 configuration. I can't say if it was the best I've ever heard, but, if not, it was close enough. It was certainly the best I've ever heard amplifiers of any price or type in my living room with these speakers. I won't say that the speakers disappeared because they're not quite that good, but the walls of the room definitely disappeared. The soundstage was solid and coherent from several feet outside of each speaker to a complete fill of the center space. Images were steady, properly proportioned, and incredibly three dimensional with plenty of air around them. Inner details of the music, not false details from excessive brightness, and the spatial relationships of one instrument to another were easily audible. What most impressed me was the wall behind the speakers dropping away and the sound emanating from a very deep, silent space behind them, stretching back beyond the physical boundaries of my living room. I listened to the Analogue Productions SACD of "Wish You Were Here" along with some Aimee Mann, some classical, and, of course, Beth Orton. I did not want it to stop. Unlike some systems, it seemed that the longer I listened, the better it sounded.
I know now that these amplifiers, conceived by Mauro, refined by Dario, given suitable source material and a decent pair of speakers in a good room, are capable of producing a well defined three dimensional image of a sound event with complete tonal accuracy.
Peace,
Tom E
Dario, the MF resistors I used for R104/204 are Vishay/Dale CMF. I am not aware of any sonic issues with these parts. They have a pretty good reputation. In fact, some modders use them to replace cheap Xicon and such. I doubt that they contribute at all to amplifier harshness.
I completely disagree with your comments about the TC at C13. This is the only time I will say such a thing. I think you are incorrect to suggest it is the cause of brightness or an edge to the sound of the FE. The TC capacitor sounded perfectly balanced and accurate with the Rev C, and not because the Rev C was masking anything. It sounds just as good, perhaps better, with the FE if it is properly tuned. If my current compensation configuration is causing distortion, it is the most beautiful distortion I have ever heard. If the TC cap is producing brightness, I cannot hear it. In my opinion, the TC is a reference part and the single component that is NOT subject to trial and error. I played that game too long with too many caps and too much expense. The TC is easily superior to every other cap I tried, including Mundorf. To simplify it: if your amp doesn't sound good with the TC in it, it is the amplifier at fault, not the cap.
Bob, regarding soundstage: last night I had my first session listening to a pair of matched monoblocks with the 33/15 configuration. I can't say if it was the best I've ever heard, but, if not, it was close enough. It was certainly the best I've ever heard amplifiers of any price or type in my living room with these speakers. I won't say that the speakers disappeared because they're not quite that good, but the walls of the room definitely disappeared. The soundstage was solid and coherent from several feet outside of each speaker to a complete fill of the center space. Images were steady, properly proportioned, and incredibly three dimensional with plenty of air around them. Inner details of the music, not false details from excessive brightness, and the spatial relationships of one instrument to another were easily audible. What most impressed me was the wall behind the speakers dropping away and the sound emanating from a very deep, silent space behind them, stretching back beyond the physical boundaries of my living room. I listened to the Analogue Productions SACD of "Wish You Were Here" along with some Aimee Mann, some classical, and, of course, Beth Orton. I did not want it to stop. Unlike some systems, it seemed that the longer I listened, the better it sounded.
I know now that these amplifiers, conceived by Mauro, refined by Dario, given suitable source material and a decent pair of speakers in a good room, are capable of producing a well defined three dimensional image of a sound event with complete tonal accuracy.
Peace,
Tom E
CMFs... they were one of the first I've tried there and the first I've discarded as they sounded weird and unbalanced (towards mid and high freq) it that position, preferring the much better and much more neutral KOAs indicated on the 'On a budget' BOM.
Usually CMF fare much better, apart detail, and are not harsh but on that position they're not a good choice.
Not a single part can be considered NOT subject to trial and error.
Sometime an excellent part in a particular position it's not the right one.
And, BTW, I've suggested a simple test to exclude or confirm compensation contribution to the problem you're hearing...
If that test would have excluded compensation not necessarily the cuplript would have been TCs...
And also TCs role could be easily verified DC coupling the amp, if the DC-coupled amp has the same harshness TCs would have been excluded as the source of the problem.
But it seem that every simple test that could possibly contradict your presuppositions is a no-no... 😉
I appreciate all your comments, but this part stands out to me. As you say it depends on the program material, but sometimes that depth of stage is simply stunning. I have never heard better.
I was laughing to myself this morning as the first 30 seconds of the FE from a cold start really sounds awful. 30 to 45 minutes in they really open up and mellow. I'm still hearing improvements after months of use. I suspect some of that may be due temp and atmospheric changes, but I'm convinced there is still some component blending at every extended listening session.
Addendum:
My previous post implied TC is the only appropriate cap for C13. Of course that's not true. Such things are very dependent on personal taste, system, etc. Reading through these pages clearly shows that each builder's selection process works differently. The TC is not a perfect cap because there isn't one. If anything, it might be a little too laidback for some tastes. It doesn't really have the excitement factor that some listeners seek. There isn't much WOW there, just layer upon layer of honestly musical tonality. In fact, if one casually compared it to some bright, sparkly thing, he might very well prefer the b, s part.
For sure, Bob, these amps are not switch on and enjoy. It doesn't usually take long, but there is certainly some warm up required before that immersive quality manifests itself. Could just as easily be speakers or source, too, or even your "ears" relaxing. I've left mine on overnight, and it doesn't seem to make much difference from a cold start with a single hour's playing time, but then my ears and brain aren't usually as perceptive first thing in the morning so I could be wrong. Might be an interesting experiment. You could leave one monoblock on all night and start cold with the other, see how long it takes for them to match.
Peace,
Tom E
My previous post implied TC is the only appropriate cap for C13. Of course that's not true. Such things are very dependent on personal taste, system, etc. Reading through these pages clearly shows that each builder's selection process works differently. The TC is not a perfect cap because there isn't one. If anything, it might be a little too laidback for some tastes. It doesn't really have the excitement factor that some listeners seek. There isn't much WOW there, just layer upon layer of honestly musical tonality. In fact, if one casually compared it to some bright, sparkly thing, he might very well prefer the b, s part.
For sure, Bob, these amps are not switch on and enjoy. It doesn't usually take long, but there is certainly some warm up required before that immersive quality manifests itself. Could just as easily be speakers or source, too, or even your "ears" relaxing. I've left mine on overnight, and it doesn't seem to make much difference from a cold start with a single hour's playing time, but then my ears and brain aren't usually as perceptive first thing in the morning so I could be wrong. Might be an interesting experiment. You could leave one monoblock on all night and start cold with the other, see how long it takes for them to match.
Peace,
Tom E
Soft start
Hello,
about the soft-start, I do think that it's useful to avoid useless stress to the transformers, rectifiers and 10000uF caps. I'm going to implement this soft-start design (just to keep the family feeling 😀):
soft-start by Mauro Penasa (in italian)
I do not know if it has been ever reported here on diyaudio. If someone is interested, I may make a short summary in english of it.
Daniele
Hello,
about the soft-start, I do think that it's useful to avoid useless stress to the transformers, rectifiers and 10000uF caps. I'm going to implement this soft-start design (just to keep the family feeling 😀):
soft-start by Mauro Penasa (in italian)
I do not know if it has been ever reported here on diyaudio. If someone is interested, I may make a short summary in english of it.
Daniele
PLEASE! I am both curious and interested in anything Mauro has thought about and tried.
post1510 link to Mauro soft start is a direct on line device that has ALL components connected to Mains Voltages.
In normal circumstances discussion, of all these direct to mains devices, is banned from this Forum.
But we are not in "normal" mode.
The Forum is permitting some discussion in some circumstances.
In normal circumstances discussion, of all these direct to mains devices, is banned from this Forum.
But we are not in "normal" mode.
The Forum is permitting some discussion in some circumstances.
Absolutely... on Mauro project page there's clear disclaimer and the mandatory suggestion to enclose in a plastic/insulating case the whole soft start circuit.
(Image extract from Costruire HiFi magazine #116)
Attachments
Here I go again. Has it been established a soft start is necessary? We went through a lot of this on the MyRef Integrated thread and a soft start was eventually added. You can read about it around this post. Though the SS does provide an element of protection, I've built amps with 125 VA, 200 VA and 300 VA dual transformers without even a CL-60 and have never experienced any problems. I may be pushing my luck, and I'm not trying discourage the use of a soft start - but how significant is it's performance with this design.
I remember seeing a photo of the components of Dario's eventual final build and I believe a SS was included. Sound like a worthy topic of discussion for a few posts.
I remember seeing a photo of the components of Dario's eventual final build and I believe a SS was included. Sound like a worthy topic of discussion for a few posts.
Attachments
Andrew,
For clarity of my understanding, is the reason why this type of device would be banned because of the risks of working at mains voltage? If not, can you explain the reason why it might be banned?
Personally, I haven't built a project with a soft start so far, so like Bob, I am also curious about the guidelines for when it would be recommended. I would have thought that for relatively low power and relatively low supply capacitance as in the 40W and 10 kuF per rail of the My Ref series, soft start would not be necessary. That said, I have no physics or even engineering guidelines that lead me to this conclusion, just opinions of other builders that smaller amps and supply caps don't require it. Can anyone offer a supported guideline?
Jac
Hello Dario,
I am really interested in your version of the My_ref and would like to the answer to a few questions concerning the LM318 and forgive me if they have been answered before but the collection of your threads and Mauro's are really huge and hard to go through 🙂...
1- Why are these two voltage increase in the op-amp supply considered an improvement? What effect do they have on the op-amp performance?
2- Can the op-amp power supply section (Lm317,zener diode, BC639) be replaced by a simple zener shunt regulator (resistor+zener+capacitor) if the amplifier is fed by a regulated power supply?
3- If the input capacitor value be increased to 4.7uF, how will it affect the amplifier performance?
4- Is a 7W output resistor (0.47 ohm) more suitable than a 5W one?
Best regards.
It is not my decision.
It is not up for discussion.
The Forum already has the Rule banning discussion of direct to mains powering.
I had another sentence but deleted it.
Andrew, the confusion stems from the term "direct to mains". There has to be some distinction as the normal AC inlet is a direct to mains device/connection. There is a definition gap here somewhere. Can you explain or direct us to a place on the forum that can help us understand?
Last edited:
Quite all opamps works better if powered with higher rails and on the original My_Ref thread has been noted that the LM318 is sensitive to rails voltage.
The My_Evo (a closed design evolution of the My_Ref) uses higher rails.
The plain zener shunt can be used, like in the original My_Ref but you would loose the advantages of the new regulator:
- decoupling the LM318 from the LM3886 power
- a low impedance supply that can be used without caps to ground.
It will work but the LM318 input would be biased differently from what Penasa planned.
a 5W output resistor is perfectly adequate but a 7W one will not harm.
I have already raised the existing "Rule" issue with the Forum.
They won't enter into a discussion.
I cannot say more.