Unlikely, as I said it is hard to get the BOM on parts for a DAC up over $100 unless you are doing an exotic solution with DSP's, FPGA's, etc. It's also likely a discrete regulator is cheaper than one of the better IC's.
Tear downs are expensive so in a boutique market you don't see many, but you can pay for one on just about anything (BTW there is no hiding anything from an expert tear down outfit). You can hide code but you can always get an estimate of the basic cost of the IC's. Did you ever take one of your modded eval cards (one that you are happy with) and add up the 1000's price on the parts?
EDIT - I have to take back what I said about Luxman, the curved PC traces comment is nonsense so I don't know what to make of them these days.
Last edited:
I used to work with a Midlander, so I have been exposed to a bit of unaccented English, Really! (Recently met a Liverpudlian and he was a bit surprised that I could even understand his speech, as it was..)
Do have a local plumber who is from Cambridge! (Townie, not a collegiate fellow.)
Of course I also am mean enough to poke fun at accented Brits to ask them if they are from Australia. (I did surprise one fellow when I mentioned I recognized his accent, he assumed I meant some variety of Great Britain English, and was amused when I got it right as "Midlands.")
Then there was the fellow who at work everyone thought he had a nice classy British accent, you know the London city one!
So why does the BBC have to train their news presenters?
(For humor I have a complete set of the Harry Potter movies in the UK versions. Some interesting differences from the U.S. releases.)
How dare you to go against one famous audio story? Electrons reflecting on sharp corners and creating extra noise, destroying the fine texture of music. Clear loss of micro details, as proved by the wife-in-the-kitchen comments, while slicing the onions.
Then you are making a mistake as usual. Already told you I use IC regulators for clocks, PLLs, and digital circuitry, etc. Usually I use IC regulators for opamp circuits. For some analog audio loads I may choose a regulator that results in sound I prefer. That might be an AD797 opamp buffer for AVCC, for one example. If you don't like it too bad, nobody cares.
This is well known by anyone and everyone dealing with HF/high speed, fast Tr ciircuits --->
A right angle in a trace can cause more radiation. The capacitance increases in the region of the corner,
and the characteristic impedance changes. This impedance change causes reflections.
• Avoid right-angle bends in a trace and try to route them at least with two 45° corners. To minimize any
impedance change, the best routing would be a round bend (see Figure 13
http://www.ti.com/lit/an/scaa082a/scaa082a.pdf
THx-RNMarsh
A right angle in a trace can cause more radiation. The capacitance increases in the region of the corner,
and the characteristic impedance changes. This impedance change causes reflections.
• Avoid right-angle bends in a trace and try to route them at least with two 45° corners. To minimize any
impedance change, the best routing would be a round bend (see Figure 13
http://www.ti.com/lit/an/scaa082a/scaa082a.pdf
THx-RNMarsh
Last edited:
I'm sure you can hear the sound difference between a LT3045, a TPSA7A4700, and a AD797 buffer. I'm also sure that what you hear can be extrapolated to the potential customers of your/Jam's DAC. No need to provide any proof, nobody cares of what you think you hear.
It doesn’t make a product a penny more expensive though, so that cannot be the justification for a high price.
Hans
Absolutely nothing to do with audio. They discovered a new quantum state of matter at 0.001 degree Kelvin and 10M PSI it must have something to do with the sound of wires.
It has everything to do with audio as we use ADC/DAC in audio. Their pcb layout can affect the sound.
Dont listen to these guys here ... Its apparently not in their realm of expertise. listen to TI.
THx-RNMarsh
http://www.ti.com/lit/an/scaa082a/scaa082a.pdf
Last edited:
Of course, I don't use sharp corners in the controlled impedance traces I am doing when playing with the black magic at many GHz on a PCB, or dealing with ultra fast logic or clock trees. It is amazing how many things can be done out of the PCB copper geometries, without a single active or passive device or any soldering. Even for 45 degrees traces there are certain layout rules that minimize the impedance discontinuities. Unfortunately the Rogers prototypes are still too expensive to extend my fun at 40 GHz...
Matters as much at audio frequencies as the skin effect in wires (ouch, now jneutron will jump in and tell about his experience with Faraday induced variations in wire impedance).
Take a look at some IC layouts, rounded corners are a PITA to step and waste space, they are not relevant until multiple GHz. You are simply wrong on this one move on.
Nope. Timing issues can be directly attributed to pcb design including reflections from 90 degree corners. Well below GHz freq.
Especially critical is the clock signal --->You may also notice what appear to be “shelves” or “steps” in the clock edges. This is caused by reflections of the clock signal as it travels along a trace ...."
THx-RNMarsh
Last edited:
Welcome John. Especially return signal paths thru ground planes. You had a really good layout person who understood and applied it successfully, too.
-Richard
Last edited:
Posting claims out of such evaluation would be a contribution to his own mission here.
OK Everyone here raise your hand who has worked with precision complex measurements at microwave freqs? Like to know the name here. Must be at primary or secondary standards lab level. Microwave pcb software user also accepted.
One (me) and others?
Its to help with the lack of credibility issue here.
THx-RNMarsh
One (me) and others?
Its to help with the lack of credibility issue here.
THx-RNMarsh
Last edited:
The effect is minimal unless we are talking several GHz+.
A colleague of mine designed a PCB which had to have a 90 degree angle in an RF output (1.95 GHz) due to layout and physical constraints and it wasn't a problem at all, confirmed by VNA and 2nd and 3rd harmonic measurements of the output in a test chamber. Tested at Intertek, Cetecom, and TUV.
Video: Putting High Speed PCB Design To The Test | Hackaday
It has nothing to do with audio. In fact, I could go put 8 right angles in my 1 Gbps LVDS pairs on the board I'm working on now and I feel confident that it would still work.
Last edited:
- Status
- Not open for further replies.