excess noise in potentiometers
The problems with potentiometers are mainly due to the uncertainty of the wiper contact(s). Obviously John and team will have selected the best available.
Brad
By and large true, but if excess noise is there with current, the signal will be accompanied by a little additional noise on the peaks, a sort of noise modulation. I suppose one could select parts for the application based on excess noise arising from a steady current flow, knowing that this will be very worst-case.
The problems with potentiometers are mainly due to the uncertainty of the wiper contact(s). Obviously John and team will have selected the best available.
Brad
Yes, that's right and it can be seen as widening at the baseline of spectral lines. I'm just encouraging Dan to do an order of magnitude calculation assuming worst case (deposited carbon).
Hum ;-)
On my point of view, no way to create concerns, where they have no reason for being.
voltage regulators
The regulators work well for the task, and the final design does have a servo which will assure that the output voltage is stable for that first stage. The use of a JFET as a current source* to develop the reference voltage across a bypassed resistor for the pass JFET is a tried and true way to get low noise. Let's also bear in mind that the upstream voltages are likely relatively quiet as well.
One thing I about which I'm curious: why is the voltage divider for the cascode parts just naked resistors, with no bypassing? Was there a deterioration in the sound otherwise? The effect of such capacitors would be small, but with the pullup and pulldown resistors the divider noise will modulate the voltage across those pullup/down resistors. In the more detailed schematic they are 249 ohms, hence one can work out what this modulation will be and refer it to input and output. It's a small effect but ought to be measurable and maybe even audible.**
*note that an NFET could be used for both positive and negative reference voltages.
** Note that as mentioned the use of the floating recirculating current generator will also drastically reduce the voltage divider noise modulation, although its benefits are primarily to elminate the pullup/down resistors' thermal noise.
The regulators work well for the task, and the final design does have a servo which will assure that the output voltage is stable for that first stage. The use of a JFET as a current source* to develop the reference voltage across a bypassed resistor for the pass JFET is a tried and true way to get low noise. Let's also bear in mind that the upstream voltages are likely relatively quiet as well.
One thing I about which I'm curious: why is the voltage divider for the cascode parts just naked resistors, with no bypassing? Was there a deterioration in the sound otherwise? The effect of such capacitors would be small, but with the pullup and pulldown resistors the divider noise will modulate the voltage across those pullup/down resistors. In the more detailed schematic they are 249 ohms, hence one can work out what this modulation will be and refer it to input and output. It's a small effect but ought to be measurable and maybe even audible.**
*note that an NFET could be used for both positive and negative reference voltages.
** Note that as mentioned the use of the floating recirculating current generator will also drastically reduce the voltage divider noise modulation, although its benefits are primarily to elminate the pullup/down resistors' thermal noise.
yes, that is exactly what i read about.
Some points of interests ---
pcb thru hole.... it is the edge/corner of the hole that can cause problems... making the edge thick as the trace isnt always accomplished at some fab houses.
pots.... the wiper has a contact problem with some designs/materials. the best controls use gold plated contact points and have a double wiper for reliable contact.
THx-RNMarsh
Last edited:
I recently looked on hydrogenaudio for jitter listening tests - seems to be a real lack of recent ones given the press jitter gets
digital jitter simulation processing of good high rez source should be easy, should be able to verify gross >100 ns levels found audible in peer reviewed publication - or push those thresholds way down if jitter really is lots more audible
the async sample rate conversion principles are well documented, PC horsepower cheap even if the processing likely still can't be done in realtime that's no limitation - you want saved files for ABX anyway
playback could be on as low jitter hardware as you like - although below 1 nanosecond appears to be common even in consumer grade audio DAC
so where are all the positive jitter listening tests?
digital jitter simulation processing of good high rez source should be easy, should be able to verify gross >100 ns levels found audible in peer reviewed publication - or push those thresholds way down if jitter really is lots more audible
the async sample rate conversion principles are well documented, PC horsepower cheap even if the processing likely still can't be done in realtime that's no limitation - you want saved files for ABX anyway
playback could be on as low jitter hardware as you like - although below 1 nanosecond appears to be common even in consumer grade audio DAC
so where are all the positive jitter listening tests?
Last edited:
It will be fine for straight file copying to CD, I hope, and I do have excellent ADC and DAC now. Not willing to spend a lot for a new one (pro or semi-pro) for what is a legacy product for me. BUT, which do you recommend instead?
With the best equipment today for mastering in HD...... and direct playback with excellent DAC.... it sounds the best to date. The Cloud is next to see what happens there but very convenient storage for a planet roamer.
The good news is today my adopted family for 15 years made it back from their flattened village at the epi-center of the quake in Nepal to their apartment in Kathmandu.... that must have been a trek..... all is OK there but still after shocks. Close call.
THx-RNMarsh
Last edited:
ES,
I think you will find with any type of platting that when you plate over a sharp corner the platting will always be thinner at the junction of perpendicular surfaces. This has nothing to do with the type of coating , it is just a simple phenomena of surface tension when platting or coating anything at that point. The only way around that would be to have a small radius at the corner, no way around that I can think of and I don't think anyone is going to radius the drilled holes on a pcb unless it was a critical device.
Sy or one of the people who deals with this on critical applications can answer whether this is even a real problem or not.
I think you will find with any type of platting that when you plate over a sharp corner the platting will always be thinner at the junction of perpendicular surfaces. This has nothing to do with the type of coating , it is just a simple phenomena of surface tension when platting or coating anything at that point. The only way around that would be to have a small radius at the corner, no way around that I can think of and I don't think anyone is going to radius the drilled holes on a pcb unless it was a critical device.
Sy or one of the people who deals with this on critical applications can answer whether this is even a real problem or not.
Hi Richard,
-Chris
Wouldn't this increase the tension? You would really want to match the thermal expansion for the PCB material to relieve the tension. Either that, or do not plate the holes through and use component leads or wire soldered on each side. The wire is perhaps more flexible.
-Chris
Afair the "gross >100ns levels" were established for random jitter which only contributes to the noise level.
Non random jitter was audible in some cases down to ~10-20ns with music samples according to Benjamin/Gannon.
Without doing measurements Katz et al did controlled listening tests using the Grimm clock for an evaluation printed in the Tape Op magazine, so that might give an estimate for real world applications.
there are tonnes written about jitter from almost any part of a digital system..... far more than i will ever need to know beyond finding the major differences between listening to CD and HD downloads;
here is one...AES E-Library Theoretical and Audible Effects of Jitter on Digital Audio Quality
THx-RNMarsh
THx-RNMarsh
here is one...AES E-Library Theoretical and Audible Effects of Jitter on Digital Audio Quality
THx-RNMarsh
THx-RNMarsh
Last edited:
There is no reason a priori for there to be any more or less jitter based on resolution, 44.1/16 can be downloaded to a custom DAC signal chain with reclocking and buffering.
yes lots about numbers not so much about audibility
many calculated limits, some few with some comparison to psychoacoustic noise in quiet thresholds with 120 dB SPL 20 kHz source
but not verified in listening tests - demonstrated audibility really is the bottom line
correlated jitter may be audible to 10s of nanoseconds - but even that's nowhere near where the jitter argument is today
we have people arguing over picoseconds - where's the demonstrations of audibility in controlled Listening tests
many calculated limits, some few with some comparison to psychoacoustic noise in quiet thresholds with 120 dB SPL 20 kHz source
but not verified in listening tests - demonstrated audibility really is the bottom line
correlated jitter may be audible to 10s of nanoseconds - but even that's nowhere near where the jitter argument is today
we have people arguing over picoseconds - where's the demonstrations of audibility in controlled Listening tests
Last edited:
Most of the pots I've looked into already have 5 independent wipers. I don't know about gold plating, wouldn't it just promptly rub off?
- Status
- Not open for further replies.