Scott,
I try not to repeat things, but anyway...
My interpretation of brain research is that the cerebrum is basically a few billion six stage FIR fuse programmed filters with millions of interconnects. (Fuse = scar tissue.) Now as these filters are formed making memories your brain tries to reuse as many filters as possible and not set new ones. As you get older so many are used up there is less capacity to learn new stuff and instead to cram it into the closest fit.
AS to charge, it is charge not an electron. When a charge enters a conductor at almost the speed of light a charge comes out the other end. Not the same electron that carries the charge. This is an issue we have never been able to successfully communicate.
Now as to a low noise preamplifier Gerhard is studious and serious and has specific goals. I built one to play and learn. Still doing it.
One of the goals is to accurately measure battery noise. George cited a nice bit that uses TWO batteries so there is no DC issue and a bit more noise to measure.
Now I built a dead bang simple low noise amplifier using 25 TSH300 amplifier as they were cheap. I have four of these cards. Gerhard proposed that the reason I was not seeing 1/F noise was that with a gain of 20 (21 actually as it was non-inverting) my AP system 2 was the actual noise limit. So I added another stage for a gain of about 5 to bring the total gain to 100. I also added a servo based on an OPA134 (have bunches) with a 1 Mohm resistor and 4.7 uF capacitor. This takes care of the DC offset that changes with source resistance due to the high input currents (relatively speaking) of the preamp. Testing this out seems to give me a bit better than 6 dB less noise than a 10 ohm resistor. Input impedance seems to be around 600 ohms at 1 MHz.
Now I did some measurements before the KD2650M1A 4 cell battery pack discharged enough for the protection circuit to open. Up for today is to improve the recharge circuit. I originally thought a low current float would be fine, but that takes too long for my patience.
The under design, construction and testing input stage as I briefly mentioned is a transformer or more specifically a magnetic amplifier. This requires some support circuitry besides a primary and secondary winding. There is a tertiary winding driven by a very slow low noise high current and passively filtered amplifier to correct for any DC on the primary. There is a quaternary winding that is used for HF bias.
Then with a dual channel amplifier I can use a multiplier on the output to get additional correlation. Should be interesting to see how low in the dirt I can go.
As to thermal effects, my LF cutoff is .2 hertz. So hopefully the room temperature doesn't drift that fast.
Now even where things are now it should be possible to measure battery noise. It seems that the sealed lead acid batteries can vary quite a bit as to noise level, so I will order in a bunch and compare them to the used ones I have about.
If I get around to it I will post a schematic of the project to date.
I am curious to see if I can get below 1 pV/rtHz.
As to the current topic of loud, my personal best was 105 dBa slow weighting at 600 feet.
Now my more important work is designing a new style cable for long loudspeaker lines.
Last edited:
No. The active electronics are a bit above 100 pV/rtHz, a simple transformer gets that to 10's and correlation should get me to 1's.
Attached picture of the inside. Note the batteries, power supply regulators +/- 2.5 on heatsink, +/- 5 no heatsink, main preamp, second stage preamp and DC servo.
Attachments
I hope you accept the invitation... you will have fun in the process of trying to figure it out with them.
And, I am sure they will treat you wonderfully.
-Richard
Last edited:
well here is a collection of 'true beleivers' gushing My review of the Altmann BYOB amplifier (Bring Your Own Battery)
My loudest public peformanve sound was in Ulm, July 1969, I was drawn to the cathedral by loud organ music. Insde, an old guy about s old as the Bible was adjustung and tuning the orgna. Afte soe time of racticing Bach's music for organs, he issued a warning that a very loud passage was coming, and he liet the organ rip. The music became almost palpable, physical, started rearranginf by internal bofy organs. The sound was deafening, yet perfectly clean and coherent, providing a perefect setting for meeting God himself. A once in a lifetime event, and I got rel lucky.
For the New Yorkers here, I was told that St. Paul's cathedral on Fifth Avenue is a replica of the Ulm cathedral. And if there's a musical instrument made in Heaven, it has to be the organ.
The end effect was that I became an affictionado if Bach's organ musical works and remain so to thiis day.
For the New Yorkers here, I was told that St. Paul's cathedral on Fifth Avenue is a replica of the Ulm cathedral. And if there's a musical instrument made in Heaven, it has to be the organ.
The end effect was that I became an affictionado if Bach's organ musical works and remain so to thiis day.
I'm an aficionado of Ulm. The word more than the city, as it seems to be the shortest word that takes so long to actually say properly.
@Gerhard: I recall the Teledyne/Crystalonics C413, even the bad grammar on the datasheet, as being one of the first low-pinchoff large-area JFETs. I think it actually got a JEDEC number as the 2N6550. Good parts, surpassed today of course.
@Gerhard: I recall the Teledyne/Crystalonics C413, even the bad grammar on the datasheet, as being one of the first low-pinchoff large-area JFETs. I think it actually got a JEDEC number as the 2N6550. Good parts, surpassed today of course.
But wrong, charge is a property not an entity, your statement has no meaning. An EM wave is not charge, who teaches your view?
Don't forget to call the LHC and Arecibo people after succeeding in this endeavour. 10pV/rtHz before "correlation" is 6 milliohm, quite a hell of a transformer that needs to be air core (nice antenna), otherwise the magnetic core Barkhausen noise will kill the performance. That's before any considerations of contact potentials, thermal (gain, noise) fluctuations, etc...
1pV/rtHz is about 4 millimeters of a 1 sq. mm copper wire, good luck with the interconnects.
Frank was well known as an obsessive task master.
The loudest thing I ever experienced was some nameless too small pub with a band that was probably a bit too well equipped for the venue, impedance mismatch.
I forgot to bring my earplugs to the Jacob Collier set on the 13th at the ill-named Hotel Cafe in Los Angeles (on the outskirts of Hollywood). But fortunately, the most demanding part of the evening was having to stand up for three hours straight. The SPLs were high but not ridiculous. When Arturo Sandoval played a guest solo he blew the roof off in his usual masterful fashion.
It was definitely the e ticket ride in town. Among other luminaries Quincy Jones was there. He got one of the very few seats, entirely appropriate. Toole was ill so couldn't make it, and I told him it may have been as well, as he doesn't like to stand all that long either.
It was definitely the e ticket ride in town. Among other luminaries Quincy Jones was there. He got one of the very few seats, entirely appropriate. Toole was ill so couldn't make it, and I told him it may have been as well, as he doesn't like to stand all that long either.
No, that's about 5 sq. mm, 20mm length for 1pV/rtHz, there.
6 feet of AWG 10 is 6 milliohm, that is 20pV/rtHz, 20 times your noise target.
Obviously we are not communicating.
Vn(RMS)= sqrt(4*KB*T*R) or sqrt(4 x 1.38e-23 x 300 x .006) or about .13 x sqrt(.006) or 10 pV
Now two transformers with a 2' primary would produce about 6 pV/rtHz or 8 pV/rtHz if added. But we are multiplying noise. So for random numbers around 6 x 6 would cluster about 11 not 36 so about 2 pV result.
Have fun Waly, it has already started. The bulk resistance is almost meaningless after a point, those guys made bespoke connectors out of large pure copper bars with the capability of applying serious clamping force for a reason.
Last edited:
I found a small company that sold me the complete work of a smps 1000w 100 euros. pcb. component link. cheap. easy to find. discrete component. PC40 toroidal. instruction for winding. absolutely everything. and smps. working after 1 year. it's free right?
- Status
- Not open for further replies.