Just ran a sim with LT1431 (should behave the same as TL431) regulating at 4.1V.
Between "fully turned on, in regulation" and "fully turned off", the difference is 2mV. i.e, at 4.1V it is in regulation, and at 4.098V it's off. So... it will shut down pretty much as soon as you stop charging.
> the noise of the TLV431 at 1 mA cathode current is about 275 nV/sqrt Hz at 10 Hz.
> And 1 mA is close to the off state (as specified in the datasheet)
No, 1mA is fully on and in regulation. It's a feedback system, it controls the output as long as it's in regulation.
Between "fully turned on, in regulation" and "fully turned off", the difference is 2mV. i.e, at 4.1V it is in regulation, and at 4.098V it's off. So... it will shut down pretty much as soon as you stop charging.
> the noise of the TLV431 at 1 mA cathode current is about 275 nV/sqrt Hz at 10 Hz.
> And 1 mA is close to the off state (as specified in the datasheet)
No, 1mA is fully on and in regulation. It's a feedback system, it controls the output as long as it's in regulation.
A PSU, even very low noise is quite a different beast from an inactive BMS circuit.
Anyway, you can do the test very easily and cheaply: build the TL or TLV431 circuit, and connect it to the battery with a switch: if you can measure a difference when the switch is on (and the circuit inactive , of course), you have indeed a noise issue (and do not rely on your ears only, and when you use them, only test the no-signal background noise, that's where you have to look)
That was the idea: measuring batteries one actually can buy.
Walls @ NIST did not disclose the exact types for fear that this
would be taken as "endorsed by the US government".
I did cite him in the references.
The 2 1K resistors in front of the op amp/ADC cost you a lot of sensitivity.
that alone is abt 20 nV/rt Hz; with such resistors in series the best you
can ever see is 20nVrtHz, even with a completely noiseless source.
I think your setup could gain a lot from cheaply inserting one 20 dB AD797
in front of it.
And the 260u input coupling C is not enough. I had 20 * 10uF WIMA foil
in my 220pV/rtHz 20*ada4898 preamp. Those 200uF could not short the
noise of my 10K bias resistor through the near-0-Ohm measurement object
at very low frequencies.
Scott spotted it immediately that I had no real 1/f noise but that it
rose faster like RC.
Changing my 200uF against 10000 uF made it much better below 50 Hz.
I used 2 wet slug tantalums @ €120
, but they gave no advantagethat I could see against good quality Al electrolytics.
I'll redo these measurements soon with a better preamp.
regards, Gerhard
Last edited:
I would just add a shunt protection, using a voltage detector that turns on a fixed voltage. this is basically passive balancing. when your load can handle the charge current it is a crude but simple way. to create the series mode you can put add an optocoupler when the detector triggers, and AND all stacked optosignals to usea single fet for turnoff.
check the seiko BMS chips for more info. the BD49xx ROHM voltage detectors are only 0.55uA, have a small hysteresis and are very suitable.
check the seiko BMS chips for more info. the BD49xx ROHM voltage detectors are only 0.55uA, have a small hysteresis and are very suitable.
Hi gerhard & basreflex - thank you both for your feedback
@Gerhard:
Actually, your measurements were quite interesting to me because they are one of the few measurements I've found (if any other ...) where the noise level with a current draw is also shown. Although my own measurement equipment is not (at all, I guess) up to the spec of your preamp they correlated well with what I measured myself. Batteries can be really, really quiet not just measurement-wise ... Looking forward to seeing your upcoming measurements.
And I'll remember your tip about the AD797 - I reckon it may be good enough for my purposes - thanks for the tip
P.S.: Gerhard if you have an idea - knowing that you work with very low-noise circuitries - for an OV circuitry I'd be pleased to hear about it ...
@basreflex:
Thanks for the suggestions. I've looked at seiko's BMS chips and also the BD49xx series from ROHM, however, the ROHMs are 0.1 VDC steps which is a bit too crude in my context. I'll look a bit more on the seiko ICs - yet I can't see any noise specs in the datasheet ... ?
Anyway, I think that what I probably will do is to use a voltage reference which feeds into an opamp through a RC filter. The opamp then feeds a bipolar transistor of some kind which drains excess current. In this way the bipolar transistor may be switched entirely off when there's no over-voltage.
Cheers,
Jesper
@Gerhard:
Actually, your measurements were quite interesting to me because they are one of the few measurements I've found (if any other ...) where the noise level with a current draw is also shown. Although my own measurement equipment is not (at all, I guess) up to the spec of your preamp they correlated well with what I measured myself. Batteries can be really, really quiet not just measurement-wise ... Looking forward to seeing your upcoming measurements.
And I'll remember your tip about the AD797 - I reckon it may be good enough for my purposes - thanks for the tip
P.S.: Gerhard if you have an idea - knowing that you work with very low-noise circuitries - for an OV circuitry I'd be pleased to hear about it ...
@basreflex:
Thanks for the suggestions. I've looked at seiko's BMS chips and also the BD49xx series from ROHM, however, the ROHMs are 0.1 VDC steps which is a bit too crude in my context. I'll look a bit more on the seiko ICs - yet I can't see any noise specs in the datasheet ... ?
Anyway, I think that what I probably will do is to use a voltage reference which feeds into an opamp through a RC filter. The opamp then feeds a bipolar transistor of some kind which drains excess current. In this way the bipolar transistor may be switched entirely off when there's no over-voltage.
Cheers,
Jesper
Hi ! looking for info about batteries noise i found your very interesting lab report.
Thank you sincerely for disclosing the results
I am thinking about using this kind of battery here below ... hoping that it can be quiet enough for a little line preamp
i wonder if it can be suitable not only for powering electrical tools but also audio equipment.
A big advantage is that i would not need any regulator but just some high grade caps to buffer the output. And of course chargers are already widely available.
Your opinion would be very much welcome and appreciated.
Thanks a lot for disclosing your findings that seem to mirror some conclusions of this old work.
https://tf.nist.gov/general/pdf/1133.pdf
Last edited:
Hi ! i see. But batteries management is an issue. They must be used with some attention.
However i like a lot the 40V batteries used to power small electrical appliances.
In my case current draw will never exceed 100mA total ...
The idea is to modify the charger and place a DC socket on its case in parallel with the battery terminals.
I will almost never have the need to remove the battery ... just when it is not working anymore.
Some batteries have also a very handy indicator of level of charge
I am quite confident that they will be silent enough ... so they are an option.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.