Dario, I think you understood very badly what I wrote.
the fixed LDR I used it on R12 and stop.
on the pump I have to place Z-foils (47K x2 and 22K x2). I have no intention of using the fixed LDRs on the pump.
However I am very convinced that on the entry R (R12) the performance of the LDR is higher than the Z-foil.
Having clarified this, I do not base my conclusions on the datasheets but on what I know from prolonged listening. I have quite difficult tastes and I have to be convinced of what I feel.
So, do you agree?
Oops! George, you're right!
Sorry Giovanni...everytime I read about those LDRs used as fixed resistors I get my eyes veiled...
Last edited:
You're right Giovanni, sorry.
I've already said it in my previous post.
Giovanni... what I've said it's not simply on the datasheet, it's electronic common sense.
Increased noise and distortion are quite certain.
Now I trust your ears and I'm sure it will 'sound' better and/or more pleasant, I've no doubts about it.
But better and/or more pleasant is often way different from transparent and the latter should be our goal.
Exactly George. On the pump, as you know, I put the Susumu RS and certainly I preferred them to the susumu RG. I'm waiting for the Z-foil that I'll try as soon as possible. Something tells me that they will go even better than RS.
Giovanni
Ok!!
now it's all clear.
I agree. On the pump the ldr is not suitable
I repeat that I appreciate the LDR just because it is more transparent and the details come to light much more easily ..... not because it is more pleasant
Giovanni
Last edited:
Giovanni,
I know how difficult is to understand it. I discovered it by chance after many years.
The only way to say DUT A is more transparent than DUT B is to compare them to a copper wire and in most cases this is simply not possible.
But I have had the opportunity to do it comparing resistors for R11.
R11 can be replaced by a wire jumper without problems (apart, maybe, a small noise increment in the audio ground) and the sound you hear... on a whole another level! So much more open, free!
How it is possible? Let replace the wire with several resistors!
Ouch! how terribly colored are all those resistor!
It's in this way that I've first noticed how terrible are the KOA MF, overbloated by bass.
Bass, yes, because main coloration is in the bass for most resistors.
Why I've ended-up using a RN55? While it was a bit more colored (slightly warmer) than others (but more neutral than most) it was the only one which retained the most amount of details and harmonic content heard via the plain wire.
Later comparing resistors in other positions, particularly RN55 Vs Z foils I've heard pretty much the same differences between RN55 and wire in R11 so, while not a direct comparison, Z foils are IMHO our best bet for one of the most NEUTRAL resistor.
Precison wire wound resistors, like Rhopoint Econistor could be even better.
All of them are passive components, they don't need power, they will vary their value according temperature (one of the main source of distortion in resistor) ORDERS OF MAGNITUDE less than LDRs.
Even if they were a bit more 'transparent' I would NEVER choose them for their many disadvantages.
But we live in a (almost) free world.
LDRs will NOT lengthen signal paths. The SIGNAL path will be shorter. Circuit path? Yea I suppose because we are adding more power supply components but the last time we designed to a circuit path?
Having power lines with much less than one milliamp "flying" over anything at about 2VDC (zero hertz) will make no difference. You are talking about EMI somehow translating itself into noise in the signal traces and of course the opamp will eliminate this as it will have manifested itself as common mode noise and if anything at all somehow injected itself into the power supply this also would be rejected by the opamp.
LDRs do change in value with heat. This doesn't matter. Pick any resistor in the signal and remove it and replace it will a resistor of a different value but similar value and there will be no change. None. Replace a 5k resistor with 5k1 and tell me you hear or even measure a difference. This does not add noise. This does not add distortion. The changes in LDR value are not the same action as is happening in Johnson noise although they are susceptible to Johnson noise just like all other resistors.
Low distortion is probably a good thing. Its also what I have measured. I use an AP Distortion analyzer and an HP Spectrum Analyzer. There are people out there using LDRs incorrectly and getting ridiculous distortion numbers. I use LDRs and generally get .004% distortion. This is usually considered LOW. This will also give the user the LDR sound they are looking for.
Using LDRs correctly is important and there is a small window of "correct use." I dont like the sound of LDRs above 12k and I dont trust them for use above 22k. A lot of people like the sound above 12k but my preference is 6k and below. They do become veiled in higher resistances in my opinion and I dont bother with them in those values, nor have I measured them to see if the problem in higher resistances is noise or distortion, could be either or both. I do measure in the values I value.
I do understand he is NOT using them in the current pump and I completely agree that with how finicky a Howland Current pump can be that LDRs would be likely the very worst resistor anyone could use. Luckily only part of the MyRef is a current pump and there is still room for LDRs and improvement in sound.
Good thing I have a room full of equipment.
Here is a link to a youtube I just made of a resistor replacer at 1k ohm with 1V of 1kHz going through it. This thing is in the wide open, not protected from noise, with long flying leads.. A veritable audio nightmare! Lets see how much noise we can find in this LDR.
YouTube
A long time ago I made a long video of me testing a Lighter Note which is using 4, count em FOUR, LDRs for quadruple the awful audio. The only measurements on this video that will matter at all to this conversation are the first measurements and the last. Both are the distortion measurements. The rest anyone can feel free to watch but it will all be in regards to channel matching, channel crosstalk, etc. I can vary the input voltage, frequency and LDR resistance to get different distortion numbers. This video shows distortion starting at around .004% and up to about .04%. I can increase input voltage to about 13V and get I think .6% if I remember right and I can get down to about .0024% as well. This is what I am talking about. Correct use of the LDRs can give fantastic results and luckily the best results are found with input voltage levels we are likely to use.
YouTube
Having power lines with much less than one milliamp "flying" over anything at about 2VDC (zero hertz) will make no difference. You are talking about EMI somehow translating itself into noise in the signal traces and of course the opamp will eliminate this as it will have manifested itself as common mode noise and if anything at all somehow injected itself into the power supply this also would be rejected by the opamp.
LDRs do change in value with heat. This doesn't matter. Pick any resistor in the signal and remove it and replace it will a resistor of a different value but similar value and there will be no change. None. Replace a 5k resistor with 5k1 and tell me you hear or even measure a difference. This does not add noise. This does not add distortion. The changes in LDR value are not the same action as is happening in Johnson noise although they are susceptible to Johnson noise just like all other resistors.
Low distortion is probably a good thing. Its also what I have measured. I use an AP Distortion analyzer and an HP Spectrum Analyzer. There are people out there using LDRs incorrectly and getting ridiculous distortion numbers. I use LDRs and generally get .004% distortion. This is usually considered LOW. This will also give the user the LDR sound they are looking for.
Using LDRs correctly is important and there is a small window of "correct use." I dont like the sound of LDRs above 12k and I dont trust them for use above 22k. A lot of people like the sound above 12k but my preference is 6k and below. They do become veiled in higher resistances in my opinion and I dont bother with them in those values, nor have I measured them to see if the problem in higher resistances is noise or distortion, could be either or both. I do measure in the values I value.
I do understand he is NOT using them in the current pump and I completely agree that with how finicky a Howland Current pump can be that LDRs would be likely the very worst resistor anyone could use. Luckily only part of the MyRef is a current pump and there is still room for LDRs and improvement in sound.
Good thing I have a room full of equipment.
Here is a link to a youtube I just made of a resistor replacer at 1k ohm with 1V of 1kHz going through it. This thing is in the wide open, not protected from noise, with long flying leads.. A veritable audio nightmare! Lets see how much noise we can find in this LDR.
YouTube
A long time ago I made a long video of me testing a Lighter Note which is using 4, count em FOUR, LDRs for quadruple the awful audio. The only measurements on this video that will matter at all to this conversation are the first measurements and the last. Both are the distortion measurements. The rest anyone can feel free to watch but it will all be in regards to channel matching, channel crosstalk, etc. I can vary the input voltage, frequency and LDR resistance to get different distortion numbers. This video shows distortion starting at around .004% and up to about .04%. I can increase input voltage to about 13V and get I think .6% if I remember right and I can get down to about .0024% as well. This is what I am talking about. Correct use of the LDRs can give fantastic results and luckily the best results are found with input voltage levels we are likely to use.
YouTube
Last edited:
Dario, I share everything you have written.
in R12 the resistance must not handle any power and it can not be subjected to thermal stress ... and it is better in that position the LDR than the z-foil.
For example. placing it on R7 (feedback R) did not have an improvement effect and I put the z-foil back, which does its job well. I had the experience and now I know with absolute certainty that on the pump they do not do anything to us or even feedback. I tried everything and everything as hypothesized .... but you have to try (you .... is generic).
The overall performance of my FE has reached a very high level ..... much higher than the starting condition and seems never to end up improving .... each step was an improvement. At the base of this possibility of performance increase is the quality of a brilliant project born from the mind of Mauro Penasa.The architecture of MyRef is an architecture made to preserve the original signal to the maximum and Mauro, we know, that he had clear ideas how to get it ... we only collect the fruits.
Giovanni
On the R3 discussion, has anyone tried the Bourns power resistors? Basically the same specs as the Caddock...
PWR221T-30-R330F Bourns | Mouser
How much current actually needs to be dissipated at R3? Ohmite also has a thin film option that can handle up to 10W.
TNP10SHR330JE Ohmite | Mouser
PWR221T-30-R330F Bourns | Mouser
How much current actually needs to be dissipated at R3? Ohmite also has a thin film option that can handle up to 10W.
TNP10SHR330JE Ohmite | Mouser
Last edited:
well said!! I add only one thing: first try an LDR resistance and then judge with the appropriate caveats that probably have had a good explanation at this point.
Giovanni
Hi JW,
I haven't tried the Bourns in an audio signal position, but the way that thick film resistors are made means it's unlikely to be better than Caddock. That said, the only way to know for sure is to try them.
The Ohmite is an interesting find that I wasn't aware. Thanks for bringing it to my attention. I am a little wary because their data sheet specifies they use either a Ni-Cr (a typical thin film resistive material) or Ru-O (which is a typical thick film resistive material). The TCR for the 0.33/0.47R values is 100 ppm/C which is better than Caddock and not as good as others, but might be worth trying.
To answer the question, with an 8 Ohm speaker, R3 is at about 2.5W, so having 4 or 5 W would cover it. In this case, like the Caddock, a heat sink is needed. Also note that, like the Caddock, they start de-rating the power from 25 deg C. Many thin film resistors start de-rating power at 75 deg C. Just something to keep an eye on when sizing a heat sink.
Jac
Last edited:
Dario R11 experiments reminds me of a question I have been curious about. I hope those of you who have a better understanding of the circuit will be able to answer it.
As I understand it, R11 is a "ground breaker", that is, a resistor between audio ground and power supply ground that helps keep power supply noise from the the signal by floating the audio above ground. The strange thing to me is that we connect the negative speaker terminal to power supply ground and run all of the return current through R11. Am I getting this right?
I am wondering if it makes sense to connect the speaker negative to audio ground and use a more traditional ground breaker circuit? For example, back to back diodes with a small capacitor in parallel to isolate the power supply noise from the audio signal.
Jac
As already suggested, don't forget to try speakers crossovers resistors, like Mundorf Supremes, Duelund Graphite or Jantzen Superres. They are rated > 10W, non-inductive and are meant (on paper, at least) for keeping the signal clear, natural and uncolored.
They are expensive, but only 2 of them are needed...
I encourage all experimentation, but a couple of things to be aware of with these. The Jantzen and Dueland have a minimum value of 0R47, so evo A isn't available. The Jantzen in under 1R has a TCR of 500 ppm/C and the Dueland doesn't spec TCR other than a negative TCR. The Mundorf has a very good TCR of 50 ppm/C, but has up to 75 uH inductance. That is fairly low for a wirewound, but much higher than other technologies. Dueland doesn't give much technical information, but it seems like they are essentially a carbon composition resistor. I wonder what they are doing differently if they are not colored like most carbon comp resistors?
Oh yes, these things are huge.
Jac
Last edited:
And their new yardstick, not yet in the webstore.
Texas Components Corporation
Both the tx221z and CSNG should easily be better than a Quad, at a price, of course.
Jac
Last edited: