Oh I was interested in seeing the results of your typology which you have been touting so much... Of course a stock versus yours analysis would be the only way to present your value added.
You can likely lower your PCB costs by using Chinese producers such as pcbway. Their product is good. And cheap!
Sorry to bump in Boggit,
but if you don't mind, I would like you to measure the AC idle consumption of your SMPS1200 and SMPS1200+NC500 when you will have time
Thank you
JC
Yes you have said that. Did you ever disclose what it is that you are actually selling? If you did then I missed it. What you have done is come to this thread touting the benefits of changing the input buffer configuration for the NC500 modules, comparing the stock buffer (and inferring that this is the same as the NC1200 buffer) with that on the NC400 and with your own design and those by Burson and others. You went so far as to brag about the lower noise floor of your design versus the stock configuration with AP readings. I would naturally assume that you also examined the linearity of your design versus the stock one.
The stock one was really just developed as a basic means of OEM's to test drive the amps. But then there was interest in using it for finished designs, so they decided to make it part of the catalog. I really can't see any more OEM's besides AI and Boggit using it unless they just want to make clones of each others amps in different boxes. Pretty boring if you ask me.
Whatever the application, were one to avoid using the recommended LM4562 based input buffer configuration (Fig 1 of the datasheet) one must have a rationale for doing so. Perhaps it is achieving a lower noise floor but then we dismissed that as being a sensible reason given the very low noise floor you observed on the stock module once you did a proper test. Perhaps it is a different linearity signature to introduce or remove some distortion for that "house" sound.
You chose not to implement the recommended buffer in your design. I'm wondering how you made that decision when you didn't even make the basic measurements for comparison. Or maybe you just don't know except different must be better. So I am asking you how rigorous your design philosophy is. Were it not simply random application of another buffer typology I'd expect you to have done the tests.
It's okay to respond by saying you are just winging it.
Boggit's get out of jail card in this is that he isn't (yet) touting one configuration over another. He is providing a platform where the purchaser can try and decide for themselves. Where he might get into 'hot water' would be if he started strongly advertising the other op amps as recommended upgrades when he can't provide an explanation of why that might be so.
The LM4562 is an extraordinary op amp. Self and others are full of praise for it. It's not at all a sloppy starting point.
I'm not on here to divulge every aspect of my design philosophy. I think I've shared enough information already. After all Boggit's idea for his new input buffer came from information I shared on this thread. The NC500 isn't part of the DIY catalog. If you want to DIY it, you must buy a finished product from an OEM, and mod it yourself. If you feel the standard Hypex buffer used in both the AI Atsah 500 and Boggit's Nord amp meets your needs, then perfect, it will save you the trouble of modding.
SGK: I am not sure I understand your need for pestering bavmike any further on details of what he is doing, and what the measurement differences might be? Hypex themselves notes that OEMs can design an input buffer to suit their own needs and sound preferences, the idea that the LM4562 is the "best" or "only" way to do it is ridiculous.
Mr. Putzeys own Ncore amplifiers (the Mola Molas) do not use the LM4562, as he chooses to implement his discrete buffer stage for his premium product line.
Every different input stage is going to have slightly different sound, and clearly bavmike is tuning the sound to his preferences, if you have a personal need to understand more about this, and the measurements which (might) explain these differences, I suggest you get your hands on some NC-500 OEM modules, and do the work yourself.
Mr. Putzeys own Ncore amplifiers (the Mola Molas) do not use the LM4562, as he chooses to implement his discrete buffer stage for his premium product line.
Every different input stage is going to have slightly different sound, and clearly bavmike is tuning the sound to his preferences, if you have a personal need to understand more about this, and the measurements which (might) explain these differences, I suggest you get your hands on some NC-500 OEM modules, and do the work yourself.
SKG I could see where this was going a couple of days ago. Bavmike is a thoroughly descent chap.
Our input buffer board was designed by my brother a high level professional Electronics design engineer and is based on Hypex's own Schematic provided by them to us. It has been tested and improves the already excellent sound produced by the standard LM4562 based board. It has been bought to market in a short time scale and is a reasonable price. I would suggest with all your apparent talent you come up with and produce a product at a reasonable cost that we can all use and bask in its glory. I'm sure we will all be immensely thankful. One word of warning there are many, many considerations in such an undertaking and is not for the faint hearted I look forward to you results.
Our input buffer board was designed by my brother a high level professional Electronics design engineer and is based on Hypex's own Schematic provided by them to us. It has been tested and improves the already excellent sound produced by the standard LM4562 based board. It has been bought to market in a short time scale and is a reasonable price. I would suggest with all your apparent talent you come up with and produce a product at a reasonable cost that we can all use and bask in its glory. I'm sure we will all be immensely thankful. One word of warning there are many, many considerations in such an undertaking and is not for the faint hearted I look forward to you results.
I'm not asking bavmike to divulge his circuit. He came here touting a better design with a lower noise floor. I'm asking if he ever also measured linearity in order to defend his claims. That's it. I take it the answer is "no". That will serve as a benchmark for my (now low) confidence that his claims have a solid foundation.
Frankly, and I'm just laying it out like I see it, anyone that doesn't have Bruno's level of expertise and credentials will have a hard time besting the exemplary performance the LM4562 offers as an instrumentation amplifier in the UcD's/NC's front-end. Any attempts to DIY a front-end without having the required expertise and equipment is a fruitless exercise, all you'll accomplish is pushing the NC500 into the MyFi realm.
That Bruno has opted to offer an all discrete approach on the NC400 and Mola-Mola amplifiers in my opinion can have two reasons or a combination thereof.
Either he wasn't happy with the performance of the LM4562 and figured he could do better, knowing Bruno personally that's a valid reason. Or the discrete approach allowed for a better integration of the front-end with the nCore modulator giving the designer more control over critical circuit parameters.
I don't believe however it is solely a cost-cutting exercise, not on premium products like the NC400 and Mola-Mola amplifiers.
We are not expecting you to reveal any of your secret sauce (that Bruno does is to his credit, but not everyone is a Bruno). But if you advocate one approach over another, we are of course interested in knowing what you base your recommedations on, and what informs your choice.
But if you feel there are even better approaches, it would be interesting to hear what you base your views on.
Hey Boggit, you are to be commended for your efforts. I hope your venture is very successful.
There are several reasons why I have been of focussing on the input buffer design options. One is that one of my favourite brands, Theta Digital, has NC1200 based products (Dreadnaught and Prometheus). They appear to use the stock input buffer on the NC1200. I had always wanted to see/hear a shootout between the Prometheus and the Kaluga with a view that it would be a shootout between the supplies. But it appears there's another variable in the mix.
Secondly, I have been working on a balanced input for an amp I am building. It uses Self's ultra low noise typology of quad input buffers and four differential amplifiers with quad versions of the very highly praised LM4562 op amp, the LME49740. So input buffer design, including discrete designs, is flavour of the month with me.
So when someone begins to tout massive improvements in input buffer design I am naturally intrigued as to whether they've made the noise and linearity measurements to support these claims.
There are several reasons why I have been of focussing on the input buffer design options. One is that one of my favourite brands, Theta Digital, has NC1200 based products (Dreadnaught and Prometheus). They appear to use the stock input buffer on the NC1200. I had always wanted to see/hear a shootout between the Prometheus and the Kaluga with a view that it would be a shootout between the supplies. But it appears there's another variable in the mix.
Secondly, I have been working on a balanced input for an amp I am building. It uses Self's ultra low noise typology of quad input buffers and four differential amplifiers with quad versions of the very highly praised LM4562 op amp, the LME49740. So input buffer design, including discrete designs, is flavour of the month with me.
So when someone begins to tout massive improvements in input buffer design I am naturally intrigued as to whether they've made the noise and linearity measurements to support these claims.
- Status
- Not open for further replies.