Interesting, but I doubt this is correct if the heat sink is grounded and therefore also transfers heat through the bottom and top plates, if you consider the air flows (fins are somewhat acting as chimneys) and if you don't consider each fin as radiating into free air. Years ago I had the opportunity to run some heat simulations with a simulator available online for free from a now defunct Canadian company (Rho-Theta). It was available only for their own extruded profiles, but for a typical power amp heat sink, the optimum was closer to the Rod ESP recommendation.
Sorry, but that makes no sense (nonsense?). They use supercaps to get paid? Where is the logic?
I have added Samsung Li Ion 18650 somewhere in a thread in the power supply section here; no time to search it now. They can hold the candle to the large NiCds. The stronger than 1/f rise at low frequencies goes on the pre amp; I've got the boards for my new JFET amplifier on Monday; two channels with much less noise current with one common input will allow cross correlation in the Agilent 89411A; that brings the noise low enough that I can see the thermal noise of a sub-Ohm resistor. But first I take a few days off in Provence.
Then there will also be an update of the battery article.
Yes , I put the cross at "I don't care about solder mask color", but white is ugly.
( 1 IF3602 or up to 16 BF862-alikes. No AC feedback in the input which opens a new can of worms,
but also no negative input impedance at 1 MHz.)
Thanks Gerhard
I found your battery noise measurement - hadn't seen it before.
Very interesting & I would suggest a LiFePo4 26650 battery to you to solve this dilemma you posted "Having a LT3042 after a LiIon battery will probably worsen the noise. But when you need 3.3V +-5%, then you need 3.3V +-5%."
LiFePo4 are nominal 3.3V & perfect for directly driving digital devices which require 3.3V +-5%
You may then be very interested in measuring its noise when you hear the improvement direct battery power brings
Last edited:
50pV/rtHz @flat floor, pretty good indeed. Any idea what amplifiers they used? With the best op amps today I was unable to get anything under 150pV/rtHz on a good day. Which is why I'm curious about other results.
Yes, that was the the reference I was looking for but couldn't find it - thanks for posting it.
Again note this is old battery chemistry & no LiFePo4 included
NiCd measure -200dB @ 10KHz
One factor emerges in these measurements - the bigger the battery, the lower the noise. Why? Maybe because the higher capacity of the bigger cells means less chemical noise than smaller batteries when under the same load. 26650 LifePO4 are big plus they also have a larger internal surface area due to using Nano Phosphate effectively simulating a much larger battery - hence further reduced chemical noise - i would predict even lower noise (near impossible to measure) than NiCd although LIGO, CERN, etc could probably do it
Last edited:
Thank you, now I have all I need to know about supercaps. I am in your debt.
Surely a parallel supercap does not slow down anything, but may affect extreme LF noise perhaps (please be openminded). And there is no reason to say that they hamper the DAC, never has. As for what an electron does, we know that is not always as predictable as classic theory might tell us.
The Oppo comments makes no sense to me either. To complain about it is same as complaining about DIY or about having fun. Particularly odd considering it is on a DIY website. I don't agree with everything there either, but I don't allow it to bother me, may I suggest adopting a similar stance? Just saying.
BTW, supercaps are now cheap. And like adding spice, easy to try.
0.33F/5.5V are available everywhere!
Last edited:
Thank you, now I have all I need to know about supercaps. I am in your debt.
Surely a parallel supercap does not slow down anything, but may affect extreme LF noise perhaps (please be openminded). And there is no reason to say that they hamper the DAC, never has. As for what an electron does, we know that is not always as predictable as classic theory might tell us.
The Oppo comments makes no sense to me either. To complain about it is same as complaining about DIY or about having fun. Particularly odd considering it is on a DIY website. I don't agree with everything there either, but I don't allow it to bother me, may I suggest adopting a similar stance? Just saying.
BTW, supercaps are now cheap. And like adding spice, easy to try.
0.33F/5.5V are available everywhere!
What do you think happens to a linear regulator when you swamp the output with a huge cap?
It depends on the particular battery and measurement conditions (BW and LT3045 Cset cap value, for example), but I already posted a plot from an Analog Devices PDF where the LT3045 had lower noise than a 3.7V Li-Ion battery.
And no, an unregulated LiFePo4 is not a good idea for 3.3V digital unless aggressive state-of-charge management is used.
Last edited:
IIRC that was some ordinary JFET, 1 nV/rtHz class. (386 pair?) The improvement comes from cross correlation and then averaging the hell out of it. I think, Walls has described that somewhere. In the frequency scene, Walls has some demi-god status, and phase noise and frequency stability is what I'm interested into, DC noise is only ancillary science for me. But needed.
That does not mean that you must believe everything he says; his musings about current
noise don't hold. IIRC, he uses a 750R series resistor to measure current noise and that
Heisenbergs the source. But all in all, NIST Time Frequency group is a great place to
dig around. If only they had chosen file names with a meaning, and not 1234.pdf.
When measuring phase noise, that cross correlation thing does wonders for me
in the E5052B or time pod.
Last edited:
No need to invent stuff, read the reference. It clearly states that the NiCd noise is due to the lower Johnson noise of the battery internal resistance (lower for higher capacity batteries).
Of course battery type & size is important - if you have aproblem with NIST or the Hoffman measurements spell them out then
I missed your link to AD PDF - I'll have a look
Who said it was unregulated?
Yeah, now that I recall better my experiment, I don't think I averaged for 24 hours or more, since I was still getting those nasty fluctuation of likely thermal origin, so no wonder I couldn't get lower. But my gain was much higher than the NIST, 80dB if I recall correctly.
Let me know what you get, now I think you are on the right track (from paralleling 20 opamps
).Who's inventing "stuff" ?
What do you think the internal resistance results from?
You don't believe that chemical reaction noise contributes anything to the total noise measured?
Last edited:
No idea, they do say they start around 1nV. The conclusions also say the residual noise is basically due to the ESR of the battery so bigger has lower ESR, makes sense.
I suspect that was a dodgy sample used for the photos
I see that you've got your negative feedback coupled into your push-pull input-output filters:YouTube
- Status
- Not open for further replies.