lolol yeah, that is an extra-cost accessory...if I was a scoundrel I'd propose using air dielectric Heliax for audio cables...then sell exotic gas mixtures to fill them with for big $$$...You could sell one mixture for classical, another for jazz, and fill it with a mixture of acetylene and oxygen for [fill in least favorite music here] and pray for arcing. You could claim the sonic affect was beneficial to health and then skirt the trademark by calling it Healiax! Remember you heard it here first...OK, I just threw up a little in my throat just making this shite up, how do some people sleep at night?
And for the info for those not in broadcast, we typically run 4-6 psi in 7/8" and 2-4 psi in 1-5/8" and larger Heliax and solid line. You learn to be meticulously clean and use fresh o-rings and silicone grease every time you pull EIA flanges apart, nothing like an air leak 1000' up a tower after you are back on the ground! The pop-off should be located at the highest point of the antenna /phase splitter system, keeps everything right up through the antennas dry!
Cheers,
Howie
Yes there were a bunch of processes that weren't as good as they could be and this shows up in joints that are contaminated crystalline and thin to start with and crack around the pcb hole with thermal cycling transport etc sooner or later.
Like I said the blanket resolder restores to better than new connections throughout which helps with clean operation and close channel matching.
For decades I used Savbit for the rework which effectively rinsed away existing solder reformed the intermetallic layers and left a joint of 99% new Savbit solder.
A bit costly in solder and time but that's restoration/prevention work done right such gear sounds good and never breaks down again critically important for well used pa gear and the like.
I found same process using LMP (62/36/2 Sn/Pb/Ag) or 96S sounded different nowadays I just use SnCu.
Last edited:
“I might remind some here that certain sophisticated designs, (like the Vendetta input phono stage) . . . “
John, an amplifier (any amplifier) that is sensitive to power supply changes is sub optimal.
You should not need ‘super regs’ or shunt regs to get good performance. IMV an amplifier that meets spec with three terminal regs is a good design. There are exceptions of course, like a single ended input stage, but those are a special case.
Anyway, doubt I’ll convince you of my design approach but my 2 pennies worth.
As for the solder stuff, let’s not go there. Conductivity is conductivity. . .
John, an amplifier (any amplifier) that is sensitive to power supply changes is sub optimal.
You should not need ‘super regs’ or shunt regs to get good performance. IMV an amplifier that meets spec with three terminal regs is a good design. There are exceptions of course, like a single ended input stage, but those are a special case.
Anyway, doubt I’ll convince you of my design approach but my 2 pennies worth.
As for the solder stuff, let’s not go there. Conductivity is conductivity. . .
George the average voltage of any AC signal should be zero. An AC voltage measured by the root mean squared voltage will heat up a resistive load the same as a DC voltage of the same value.
What Root Mean Squared means is you take each instantaneous value and square it. That way all the negative voltages also yield a positive squared value. Then you can average all of these values. By finally taking the square root of this result you undo the prior squaring operation to get the equivalent value.
Basically what the process does is averages the absolute value of every instantaneous reading. (Of course it is really a continuous process!)
Hi Ed
It’s a practical question.
All the Vmeters report a value of Vpeak/SQRT(2) This is referred to as Vavg >Edit:WRONG
True RMS Vmeters report a value of V peak*(2/PI()). This is referred to as Vrms >Edit:WRONG
For a sinusoidal signal Vrms=1.1107…*Vavg
George
>Edit. Correction
my apologies)
All the Vmeters report a value of V peak*(2/PI()). This is referred to as Vavg
True RMS Vmeters report a value of Vpeak/SQRT(2) . This is referred to as Vrms
It’s a practical question.
All the Vmeters report a value of Vpeak/SQRT(2) This is referred to as Vavg >Edit:WRONG
True RMS Vmeters report a value of V peak*(2/PI()). This is referred to as Vrms >Edit:WRONG
For a sinusoidal signal Vrms=1.1107…*Vavg
George
>Edit. Correction
my apologies) All the Vmeters report a value of V peak*(2/PI()). This is referred to as Vavg
True RMS Vmeters report a value of Vpeak/SQRT(2) . This is referred to as Vrms
George,
You are getting scary! For a clean sine wave the peak voltage is SQRT(2) * VRMS or the VRMS = .707 * VPeak.
Root mean square - Wikipedia
Real true RMS voltmeters often have very low high frequency limits due to the processing required.
The original RMS determination method was to use the voltage under test to heat a resistor and that was compared to another resistor that was heated by a DC voltage. When the resistors matched temperature then the DC voltage required was the AC RMS value.
You are getting scary! For a clean sine wave the peak voltage is SQRT(2) * VRMS or the VRMS = .707 * VPeak.
Root mean square - Wikipedia
Real true RMS voltmeters often have very low high frequency limits due to the processing required.
The original RMS determination method was to use the voltage under test to heat a resistor and that was compared to another resistor that was heated by a DC voltage. When the resistors matched temperature then the DC voltage required was the AC RMS value.
You think?
"I have experimented with numerous lead containing solders including Multicore 60Tin/40Lead, Multicore 63Tin/37Lead, Multicore Savbit (60Tin/38Lead/2Copper) and Multicore LMP (62Tin/38Lead/2Silver) and concluded that lead is not to my liking with Savbit sounding least bad and LMP sounding drive you out of the room bad......."
"For decades I used Savbit for the rework which effectively rinsed away existing solder reformed the intermetallic layers and left a joint of 99% new Savbit solder.
I found same process using LMP (62/36/2 Sn/Pb/Ag) or 96S sounded different nowadays I just use SnCu. "
Last edited:
Bonzai, you are wrong on this one. Sometimes, a 'better' circuit for audio has the unfortunate problem of power supply sensitivity. To fix the problem, you can change the topology to something less sensitive, or you can add a very low noise open loop buffer between the 3 terminal IC supplies and the audio circuit. I chose the buffer. My phono input circuit is famous for its very low noise (remember it is almost 40 years old) and sound quality. I still can't make anything better.
Scott,
As usual...
The article suggests using the meters near full scale. Not really a good idea for a true RMS meter when measuring a high crest factor signal like pink noise. Commonly in the RMS circuit once the crest factor gets above 3 or so the circuit will clip and give false readings. So when measuring pink noise it is often a good idea to stay around 30% or less of full scale.
Easy to test with a pink noise source an amplifier run at least 20 dB below full output and a known fixed resistive voltage divider.
Now as real music signals can have 30 dB peak to average, special considerations come into play. I know you will have a hard time realizing it but VU meters for audio are really quite limited.
As usual...
The article suggests using the meters near full scale. Not really a good idea for a true RMS meter when measuring a high crest factor signal like pink noise. Commonly in the RMS circuit once the crest factor gets above 3 or so the circuit will clip and give false readings. So when measuring pink noise it is often a good idea to stay around 30% or less of full scale.
Easy to test with a pink noise source an amplifier run at least 20 dB below full output and a known fixed resistive voltage divider.
Now as real music signals can have 30 dB peak to average, special considerations come into play. I know you will have a hard time realizing it but VU meters for audio are really quite limited.
George,
Where I think your confusion comes in from is that if we examine a sine wave from 0 to 180 degrees the average value of that does involve pi and is less than the effective power delivered from that same half sine voltage.
As power is I*E=P or E**2/R the heating effect tracks better with the Root Mean Squared principle.
I suspect you are interested in measuring an amplifiers power output, so the RMS value is more useful. As a practical matter, if you are testing using sine waves, a peak reading but RMS value displaying meter is fine. That is the way every VOM that I am familiar with does it. All it requires is a diode and a different meter scale for AC and DC.
I really don't know of any practical use for the average voltage reading of a half sine!
Where I think your confusion comes in from is that if we examine a sine wave from 0 to 180 degrees the average value of that does involve pi and is less than the effective power delivered from that same half sine voltage.
As power is I*E=P or E**2/R the heating effect tracks better with the Root Mean Squared principle.
I suspect you are interested in measuring an amplifiers power output, so the RMS value is more useful. As a practical matter, if you are testing using sine waves, a peak reading but RMS value displaying meter is fine. That is the way every VOM that I am familiar with does it. All it requires is a diode and a different meter scale for AC and DC.
I really don't know of any practical use for the average voltage reading of a half sine!
Last edited:
Our chips (used by Fluke,etc.) were 0.5% at crest factor of 6. A crest factor of 31.6 is so many sigma that I don't see it being relevant to any real world application.