opus101 Posted April 6, 2018 Share Posted April 6, 2018 13 hours ago, barrows said: You are making yourself look really bad here. My only disgreement with your points @barrows - this has already been done. Link to comment
opus101 Posted April 6, 2018 Share Posted April 6, 2018 19 minutes ago, firedog said: He’s simply saying he - and others - can design an SMPS that is quieter and more linear than what those same people regard as audiophile level linear power supplies. I don't see anywhere he's talking about linearity of power supplies, seems to me meaningless as power supplies don't amplify signals. He took issue with the terming of 'traditional' mains-frequency transformer power supplies as 'linear' ones that's all. Link to comment
opus101 Posted April 6, 2018 Share Posted April 6, 2018 1 minute ago, Ralf11 said: ??? To those with an engineering background, its been obvious for quite some time that @GUTB doesn't have one. But that fact is quite irrelevant - what's much more relevant is his apparent imperviousness to correction. Link to comment
opus101 Posted April 6, 2018 Share Posted April 6, 2018 1 hour ago, monteverdi said: There is a general idea that toroids are the best transformers. There is no 'best transformer' just like there is no best vehicle. It all depends what you're trying to achieve in engineering terms. Toroids are best for efficiency (low losses) and radiated flux and have wide bandwidth (low leakage flux). But in audio we don't want high bandwidth power transformers, even though we want output transformers to be wide bandwidth enough for audio. Link to comment
opus101 Posted April 9, 2018 Share Posted April 9, 2018 1 hour ago, monteverdi said: Devialet's ADH is not just a class A input stage (as more or less all class D amps) followed by class D but they claim the class A stage generates the full voltage for the speaker output and 4 class D amps work in parallel to generate the necessary currant. My understanding of Devialet is they're rejigging the old Quad concept of 'current dumping' where instead of the 'current dumper' being classB, its class D. So 'hybrid' seems a perfectly reasonable term for it to me. 'Hybrid' is only about the output stage, not about the preceding signal stages (which in many classD amps with opamps, are classAB) Link to comment
opus101 Posted September 2, 2019 Share Posted September 2, 2019 20 minutes ago, March Audio said: When you say "better" what do you mean or like to see? The LM4652 is an exceptionally good buffer. If you feel that a component with 0.00003% THD + Noise (-130dB), IMD of 0.00005%, voltage noise density of 2.7nV etc is somehow lacking, then you need to explain why. Of course the measurements of the LM4562 aren't in doubt, what makes it a questionable choice subjectively for an input stage is its susceptibility to RF. Check out the thread(s) on DIYA where the device is shown to rectify DECT (home portable phone) transmissions. Link to comment
opus101 Posted September 2, 2019 Share Posted September 2, 2019 10 minutes ago, March Audio said: Any amp should have an input RF filter which makes the point moot. Mine do. The point is only moot if the FR of the RF filter demonstrates the received energy is truly negligible at the frequency of interest. Do you have the measured FR of your filter up to (say) 2.5GHz ? Link to comment
opus101 Posted September 2, 2019 Share Posted September 2, 2019 10 minutes ago, March Audio said: Not measured to 2 Ghz but SPICE shows the following. If you're wanting to rely on SPICE rather than measurements then you'll need to show the circuit you modelled to be at all convincing. What parasitics did you take into account? How did your model account for the particular layout config on your PCB? Link to comment
opus101 Posted September 2, 2019 Share Posted September 2, 2019 Just now, March Audio said: I havent seen the thread you refer to but seriously, an amp getting upset by no more than 10 mW of RF from a dect handset? As I said, they need to take a look at what they are doing with the design. Basically your point is unconvincing. Any non RF hardened op amp can suffer from problems with RF and I have taken measures to minimise the potential for problems. A good question for other vendors perhaps? If you're at all curious then you'll go look at the thread (I think there's more than one where this particular issue comes up) and no, its not an 'amp getting upset by no more than 10mW'. That's just caricaturing - I am fairly sure some those involved (Pavel Macura was one I remember) were already more knowledgeable than you, but happy to be shown to be wrong. I wasn't intending to convince you, hopefully potential customers will be educated to ask more pertinent questions when they're presented with graphs and numbers set out to dazzle. Link to comment
opus101 Posted September 2, 2019 Share Posted September 2, 2019 1 minute ago, March Audio said: How much RF does a DECT phone put out - Tiny. 10mW max. Quite. So the fact that the DECT 'signature' was visible at the output is itself quite remarkable. How could an opamp be designed to be so incredibly sensitive to such low levels of RF? Inquiring minds want to know. Link to comment
opus101 Posted September 2, 2019 Share Posted September 2, 2019 1 minute ago, March Audio said: Considering it is not a problem I have seen with this op amp, even with specific RF testing, then my conclusion is that it isnt sensitive to RF in the way you are claiming when normal design considerations are made. 'in the way you are claiming' - I'm just pointing to other people's findings. Are you taking issue with them? Or saying they're mistaken or perhaps deceitful? 'normal design considerations' - you're saying the DIYA guys didn't employ them? Link to comment
opus101 Posted September 2, 2019 Share Posted September 2, 2019 'Never underestimate the power of denial'. Link to comment
opus101 Posted September 2, 2019 Share Posted September 2, 2019 Let's see the results of the tests ? Also the measured response of the input LPF, not just a SPICE sim? asdf1000 1 Link to comment
opus101 Posted September 2, 2019 Share Posted September 2, 2019 Keep up the denial! Its going well so far! asdf1000 1 Link to comment
opus101 Posted September 2, 2019 Share Posted September 2, 2019 1 minute ago, March Audio said: There is no denial, I have fully acknowledged the potential for RF issues. It wouldn't be denial if you accepted it, would it? 😎 asdf1000 1 Link to comment
opus101 Posted September 2, 2019 Share Posted September 2, 2019 Just now, March Audio said: I think you are one for the ignore button. QED Link to comment
opus101 Posted September 2, 2019 Share Posted September 2, 2019 Here are Pavel Macura's findings with LM4562 : https://www.diyaudio.com/forums/analog-line-level/308415-ne5532-buffer-post5097962.html If anyone's interested in further anorak-y stuff concerning LM4562, there's also this thread on TI's support forum which might raise the odd eyebrow : https://e2e.ti.com/support/audio/f/6/t/415907 Link to comment
opus101 Posted September 2, 2019 Share Posted September 2, 2019 Keep on playing that denial card, its working a treat! Link to comment
opus101 Posted September 2, 2019 Share Posted September 2, 2019 1 minute ago, March Audio said: Why do keep saying that, it's totally non senscal? Of course it won't make any sense to you that I note your textbook denial behaviour. Coz from your point of view, there isn't any denial. But the defense mechanisms you display are textbook ones. Perhaps acquaint yourself with some of Anna Freud's work? I think there may well be free downloads. asdf1000 1 Link to comment
opus101 Posted September 2, 2019 Share Posted September 2, 2019 16 minutes ago, sandyk said: In this case the length of the cable would have had <100pf capacitance. It is when the Data sheets show a graph with Small-Signal Transient Response AV = 1, CL = 100pF This is an interesting point - the squarewave response with the stated 100pF load shows about 30% overshoot on the positive excursion, less on the negative. Not oscillation by any stretch of imagination (or denial!). asdf1000 1 Link to comment
opus101 Posted September 2, 2019 Share Posted September 2, 2019 4 minutes ago, sandyk said: ...the 30MHZ maximum operating frequency of the original DECT handsets. I had a quick look at Wikipedia on DECT, didn't see any mention of DECT going that low in frequency. In my understanding its >900MHz but you're a telecoms guy, feel free to educate me. Link to comment
opus101 Posted September 2, 2019 Share Posted September 2, 2019 2 minutes ago, sandyk said: You obviously didn't look hard enough Well it is true I only scanned the Wikipedia article. But now I've had more time to look through it, I am seeing things like : Before 1.9 GHz band was approved by the FCC in 2005, DECT could only operate in unlicensed Region 2 2.4 GHz and 900 MHz ISM bands A cite for your 30MHz claim would be useful. Link to comment
opus101 Posted September 2, 2019 Share Posted September 2, 2019 I see no reference to DECT there. Handphones of course existed prior to the establishment of DECT standards. Link to comment
opus101 Posted September 2, 2019 Share Posted September 2, 2019 4 minutes ago, sandyk said: It should have been obvious that an LM4562 wouldn't have been able to demodulate at frequencies of 900MHZ !!! Clearly this document's going to come in above your pay-grade but others might find it relevant : http://www.ti.com/lit/an/snoa497b/snoa497b.pdf Link to comment
opus101 Posted September 2, 2019 Share Posted September 2, 2019 Just now, sandyk said: You can continue your silly arguments without me !!! For some reason I can't quite believe that claim. kumakuma 1 Link to comment
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now