Have not visited thus thread for a while:
It appears to be Groundhog Day!!!
(or as the late Yogi Berra said it's "déjà vu all over again")
Or as my (now 21year old) son's childhood book says "If you give a mouse a cookie, he will probably ask for a glass of milk...."
Anyhow
Tom:
Regarding the That driver PCB - what does it require for a power supply? I see a three pin terminal block in middle of PCB. My system plan is that Mod86 will be fed from the tape out of my preamp (just downstream of phono pre and source selector switch) and thus access all my music sources but drive different speakers from some tube gear connected to the pre.
A minimalist box with an attenuator and a bit of gain is just what I need, but viewing your website it's not clear what one needs to complete the package once the board is assembled.
My apologies to all for (once again) interrupting the , umm, discourse
PS - really enjoying the amp for last 7 months
It appears to be Groundhog Day!!!
(or as the late Yogi Berra said it's "déjà vu all over again")
Or as my (now 21year old) son's childhood book says "If you give a mouse a cookie, he will probably ask for a glass of milk...."
Anyhow
Tom:
Regarding the That driver PCB - what does it require for a power supply? I see a three pin terminal block in middle of PCB. My system plan is that Mod86 will be fed from the tape out of my preamp (just downstream of phono pre and source selector switch) and thus access all my music sources but drive different speakers from some tube gear connected to the pre.
A minimalist box with an attenuator and a bit of gain is just what I need, but viewing your website it's not clear what one needs to complete the package once the board is assembled.
My apologies to all for (once again) interrupting the , umm, discourse
PS - really enjoying the amp for last 7 months
Modulus-86: Composite amplifier achieving <0.0004 % THD+N.
You really do not have to have all your ducks lined up to your satisfaction to do the test. Google is your friend.
http://www.analog.com/media/en/technical-documentation/application-notes/236037846AN_843.pdf
Or
Image from http://www.winmls.com/Applications/Wizards/Pictures/ImpedanceMiniJackConnection.jpg.
Or
http://www.raymondaudio.nl/files/jig.pdf
These are open information.
If someone really gets off there butt and work, and come up with different data, then we can figure out why and what is reasonable.
You really do not have to have all your ducks lined up to your satisfaction to do the test. Google is your friend.
http://www.analog.com/media/en/technical-documentation/application-notes/236037846AN_843.pdf
Or
Image from http://www.winmls.com/Applications/Wizards/Pictures/ImpedanceMiniJackConnection.jpg.
Or
http://www.raymondaudio.nl/files/jig.pdf
These are open information.
If someone really gets off there butt and work, and come up with different data, then we can figure out why and what is reasonable.
Last edited:
Regarding "characteristic impedance" - does it EVER change for a specific cable, whether with frequency or with any other external parameter?
I always hear it as a specific value, such as 75 ohms. If this value changes for a cable, are you really measuring characteristic impedance?
I always hear it as a specific value, such as 75 ohms. If this value changes for a cable, are you really measuring characteristic impedance?
OK, so you have (sort of) covered the first two points on the list that tomchr asked for, so how about the other two?
Just to rewind. A normal interconnect has R, L and C at the lengths we are talking about. It's not a loudspeaker, its not a transmission line. Measure it on an accurate LCR meter and SIM.
From the data that I have seen, it is constant within a frequency range, not infinite. In most communications, if your application is in that range, you assume it is constant.
Takes (again sort of) care of part of point 3, but what is the scale of the Y axis?
And how about point 4?
Why is it so difficult to simply provide the information we keep asking for?
Bill, I've described how I used to conduct EMI/RFI tests. I'm sorry it doesn't meet your standards but I still stand by my claim that any gear silent in close proximity (1m or less) to this contraption will pass ANY RFI/EMI test.
'inaudible' would be at least 10dB below (cos impulsive) the noise of a 150R resistor ie -133dBV - 10dB = -143dBV measured on a BBC type PPM with CCIR 468 weighting. I think this type of (noise) measurement is now formalised as an ITU BS 😛
(Yus gurus please xcus my glossing over loadsa stuff about measuring impulsive noise but this was our (Calrec) practical definition of 'inaudible' ... subject to caveats about operating level bla bla)
These 'measurements' I took a Very long time ago. 😀 but I think with somewhat more rigour than at least one other on this thread.
I'm sorry, I have no RF field strength measurements of the Calrec CB Radio 🙁
I suppose the modern day equivalent of my junior toy would be a mobile phone placed on top of the DUT.
But enough of my nasty crude & childish tests 😱
Can you tell us how YOU conduct RFI tests and what standards you will attempt to meet? If these standards are available to us unwashed masses, I would appreciate a link.
Presumably, you have loadsa $$$ so the gear & chamber to conduct these formal tests is small change.
Last edited:
No, it changes a whole lot, maybe two orders of magnitude in the audio band.What we often see is the simplified formula for Radio Frequency Characteristic Impedance. This short formula works well as the frequency approaches 1 Megahertz. But in the audio band you need to measure 4 parameters at each frequency in question.
See my link:
http://www.diyaudio.com/forums/vend...r-achieving-0-0004-thd-n-102.html#post4463440
Actually, no information is provided at all. The links provided general information for how to obtain impedance measurements. No description of which impedance was actually measured in case of Soong's measurements of "cable impedance" (meaningless term).
I am always willing to answer specific and relevant questions.
Tom
While I do enjoy the discussion of the THAT Driver here, I figured it would be better served in a separate thread. I just launched a separate thread for the THAT Driver: http://www.diyaudio.com/forums/vend...-differential-driver-preamp-0-000021-thd.html
You need to provide the input signals and a power supply in the range of ±18 to ±40 V that can source at least 25 mA.
Awesome! I'm glad you like it.
Tom
I'm sorry, seems my humour was misinterpreted. I don't have such a contraption as you do, but assumed any generator of a suitable spark would work. Hence the Jacobs ladder, which I do have the parts to make
I have 5 kids and no money, but I do have access to friends who are RF consultants and DO have the right equipment to test DC to GHz. Plying them with beer would allow some testing and some numbers to be generated. Can't see any reason for contention there.
Really not sure what I said to upset you but was unintentional. I'm just interested in some numbers and am willing to make the effort in getting them.
I no longer have a Calrec CB radio 😡
I made it circa 1980 to investigate RFI in microphones & broadcast mixing gear. Everyone else hated it cos it was impossible to do ANY testing while it was running 😱
As a beach bum, I'm unlikely to build another this Millenium. Just googled 'Jacob's ladder'. Should be a good substitute 🙂
__________________________________________
But if your friends have access to a RF anechoic & test gear, can you get them to test the radiated field at various distances from a mobile phone trying to send a SMS message? All frequencies please.
Any model as long as its specified. I think an old GSM phone is best cos it's impulsive & hence nastier & broader spectrum. 😱
Dunno what dis new fangled 3G stuff does. Any mobile phone gurus lurking?
This would allow us unwashed masses to put some numbers to our junior gear (mobile phone) measurements 🙂
The RF anechoic will be a Faraday cage so the phone will attempt to put out its maximum power.
PS No offense taken.
PPS AC electric motor, spark plug, car ignition, rod aerial and the use of these as a Radio Transmitter were all inventions of N. Tesla. I've often been compared to Tesla .. but alas, I think more because of his crazy guy reputation in his last days than from any inventive genius on my part. 😱
Last edited:
The early GSM (used up to 2G, I believe) is the worst. It's not so much the RF, rather it's the 200 Hz burst packet rate. The RF will cause a DC offset shift in the input stage of your opamp. That's usually not a big deal, but when it does so 200 times a second, you get the familiar brrrtt!! noise. Highly annoying.
CDMA (2G in the US) didn't have this issue as it was spread spectrum, so essentially wideband noise.
I'm not 100 % up on the cell phone standards. I was in a pigeon hole delivering high-performance clocks for the cell phone base stations and not really involved with the standards, so others in the know feel free to correct. I do seem to recall that 3G and onwards is spread spectrum.
A Jacob's Ladder should be a nice wideband (and rather lethal!) noise source.
Tom
Perhaps I should apologize (again?) to Tom for the off-topicness here, but this bugs me (and I feel like this "Duty Calls" guy: https://xkcd.com/386/).
http://www.audiosystemsgroup.com/TransLines-LowFreq.pdf
That Figure 1 has this description:
"Characteristic Impedance of Typical Cable at Audio Frequencies."
Overall, that file looks like a bunch of good info, but I'd certainly double-check it before relying on any of it because that writer is MIS-USING the phrase "characteristic impedance."
Here's a couple of sentences from The Fount Of All Human Knowledge (okay, perhaps it's not The Final Word, but I trust it more than the writer of that PDF file):
It's clear that that Figure 1 measurement is simply the capacitive reactance between the two conductors of a short (for audio frequencies, less than many thousands of feet) cable. The X and Y axes are correctly labeled frequency and impedance, and this is indeed a measure of impedance (that due to capacitance between cable conductors goes down with increasing frequency), but it's NOT a measure of CHARACTERISTIC impedance.
This is the reason I asked the question (perhaps it was baiting, but I was being honest). I had always understood characteristic impedance to be a fixed value, and now that I check (sources other than that PDF), indeed it is.
Once again (as I recall someone else saying recently on DIYAudio, most likely earlier in this thread), CHARACTERISTIC impedance is meaningless when the length of the cable is significantly below the wavelength of the frequency/signal in question.
Okay, I'm looking at Figure 1 of the PDF file in that post, whose URL is:
http://www.audiosystemsgroup.com/TransLines-LowFreq.pdf
That Figure 1 has this description:
"Characteristic Impedance of Typical Cable at Audio Frequencies."
Overall, that file looks like a bunch of good info, but I'd certainly double-check it before relying on any of it because that writer is MIS-USING the phrase "characteristic impedance."
Here's a couple of sentences from The Fount Of All Human Knowledge (okay, perhaps it's not The Final Word, but I trust it more than the writer of that PDF file):
https://en.wikipedia.org/wiki/Characteristic_impedance
It's clear that that Figure 1 measurement is simply the capacitive reactance between the two conductors of a short (for audio frequencies, less than many thousands of feet) cable. The X and Y axes are correctly labeled frequency and impedance, and this is indeed a measure of impedance (that due to capacitance between cable conductors goes down with increasing frequency), but it's NOT a measure of CHARACTERISTIC impedance.
This is the reason I asked the question (perhaps it was baiting, but I was being honest). I had always understood characteristic impedance to be a fixed value, and now that I check (sources other than that PDF), indeed it is.
Once again (as I recall someone else saying recently on DIYAudio, most likely earlier in this thread), CHARACTERISTIC impedance is meaningless when the length of the cable is significantly below the wavelength of the frequency/signal in question.
I have an 8W luggable orbitel phone from 1993ish. High gain antenna so can really pump the power. But I need to build an adaptor as it used 5V Sim cards.
Tom is right its the 217Hz frame rate in GSM that causes the problem, specifically call setup (or SMS) as no power control is used. 3G is spread spectrum 4G is OFDM.
Spectrally GSM has to stick close to its 200KHz channel bandwidth so it's remarkably narrow band.
EDIT: A museum in oxford has Marconi's original spark gap transmitter apparatus that was used for the first trials on salisbury plain. Oddly they don't fire it up any more.
Tom is right its the 217Hz frame rate in GSM that causes the problem, specifically call setup (or SMS) as no power control is used. 3G is spread spectrum 4G is OFDM.
Spectrally GSM has to stick close to its 200KHz channel bandwidth so it's remarkably narrow band.
EDIT: A museum in oxford has Marconi's original spark gap transmitter apparatus that was used for the first trials on salisbury plain. Oddly they don't fire it up any more.
Last edited:
I'm not sure "meaningless" is the right word. "Irrelevant" would be a better term. I do fundamentally agree with you, though.
I'm glad you arrived at the correct conclusion. At audio frequencies, you can certainly measure the inductance and capacitance of the cable. For most practical cables, source impedances, and load impedances, these parameters are of little consequences, however. The better cable manufacturers specify (concept!) the capacitance and inductance of their cables. This allows you to easily run a simulation or do a little math to see what the impact of a certain length of that cable will be in your application.
Tom