> but I am afraid If I forgot charger for a few days when no playing
You misunderstood what I said.
I listen to music for 10% of the time at most.
And for the 90% of the time, the charger is always on.
The charger (designed for NiMH) has a charging rate of C/40 when fully charged.
And I have not experienced any negative effect for this continuous, non-stop, trickle charge at C/40.
But of course in the end you decide what you want to build.
There are many modern ICs for charging NiMH automatically.
I am sure you will find something easy and suitable.
Patrick
You misunderstood what I said.
I listen to music for 10% of the time at most.
And for the 90% of the time, the charger is always on.
The charger (designed for NiMH) has a charging rate of C/40 when fully charged.
And I have not experienced any negative effect for this continuous, non-stop, trickle charge at C/40.
But of course in the end you decide what you want to build.
There are many modern ICs for charging NiMH automatically.
I am sure you will find something easy and suitable.
Patrick
Patrick, next project should be a hassle-free, plug and forget LiFePo4 charging scheme. I've got a shortlist of a few ICs but I trust your skills way more! 
Regarding filters, have you guys seen this line from Murata?
I bought a few from the NFL series (LC filters) to try them out on my Xen 9022.
Regarding filters, have you guys seen this line from Murata?
I bought a few from the NFL series (LC filters) to try them out on my Xen 9022.
I'm about to do the same/similar thing but
relay's "no power" = batteries -> Cen O/P stage,
relays energised = charger turned on, and batteries -> charger
Pushbutton led voltage meter (test) and low voltage auto charging switchover
System here works maybe 18hrs a day sometimes, so need bigger batteries but no real design problems so far - not sure yet if bigger Nicad/NiMh pack or 18V 2.4Ah SLA and as different charging routine required, but determined to keep it as simple as possible.
Also been unsuccessful with Li-ion batteries but they do work extremely well in the power tools so I guess it's just my ignorance again.
relay's "no power" = batteries -> Cen O/P stage,
relays energised = charger turned on, and batteries -> charger
Pushbutton led voltage meter (test) and low voltage auto charging switchover
System here works maybe 18hrs a day sometimes, so need bigger batteries but no real design problems so far - not sure yet if bigger Nicad/NiMh pack or 18V 2.4Ah SLA and as different charging routine required, but determined to keep it as simple as possible.
Also been unsuccessful with Li-ion batteries but they do work extremely well in the power tools so I guess it's just my ignorance again.
SEN without matching jfets
I've been playing around trying to design a better servo to avoid the high order harmonic distortion of my last one and also compensate for my badly matched jfets.
This is what I've come up with (they are all op-amps, not LM323s).
The servo for the DAC bias current has an active low pass filter with cutoff about 0.1Hz and stop band below -200dB at audio frequency - should hopefully prevent it injecting high order harmonics.
The trick with compensating for mismatched jfets seems to be to use a floating servo which preserves the current conveyor path and so avoids introducing distortion.
The low value resistor is so the opamp doesn't need to swing all the way to the bottom rail.
Spice indicates 3rd harmonic is about -145dB at the 250R Riv shown and about -130dB at 510R.
The passive filter on the output of the floating servo maybe needs a bit more work. It might take too long to settle.
I think it should work with jfets unbalanced either way around.
I've been playing around trying to design a better servo to avoid the high order harmonic distortion of my last one and also compensate for my badly matched jfets.
This is what I've come up with (they are all op-amps, not LM323s).
The servo for the DAC bias current has an active low pass filter with cutoff about 0.1Hz and stop band below -200dB at audio frequency - should hopefully prevent it injecting high order harmonics.
The trick with compensating for mismatched jfets seems to be to use a floating servo which preserves the current conveyor path and so avoids introducing distortion.
The low value resistor is so the opamp doesn't need to swing all the way to the bottom rail.
Spice indicates 3rd harmonic is about -145dB at the 250R Riv shown and about -130dB at 510R.
The passive filter on the output of the floating servo maybe needs a bit more work. It might take too long to settle.
I think it should work with jfets unbalanced either way around.
Last edited:
One thing to remember is that for real high performance servo's, the used opamps have to be very high (audio) performance also.
Many people know that the servo 'works' at very low frequencies but don't forget that they are required to work as a perfect integrator up to the highest audio frequency.
So basically all that you would want for an opamp for the signal path also is valid for the servo opamps!
jan
Many people know that the servo 'works' at very low frequencies but don't forget that they are required to work as a perfect integrator up to the highest audio frequency.
So basically all that you would want for an opamp for the signal path also is valid for the servo opamps!
jan
I think the ones connected to the audio signal need to be good. I'm not sure if I can get away with cheap ones for the active low pass filter after the integrator. The integrator should compensate for any DC offset and the output of the low pass filter should be -200dB so hopefully it won't matter if there is a little distortion from those op amps.
Something to measure to find out perhaps.
Is there a way I can use spice to tell me if it's going to oscillate? The frequency response doesn't seem to have any nasty peaks in it.
Sometimes Bode plots lie, so transient response simulations are really the safest.
I should qualify my statement a bit: if there is sufficient loop gain and two close poles, you will see ringing. But an integrator followed by a little passive lowpass with a different tau may be a stable solution. This also filters out some of the broadband opamp noise, which appears at the output of a Sallen-Key lowpass.
If you get a complete schematic with real opamp parts I could take a look at it in sim, although I'm a little busy right now.
Brad
I should qualify my statement a bit: if there is sufficient loop gain and two close poles, you will see ringing. But an integrator followed by a little passive lowpass with a different tau may be a stable solution. This also filters out some of the broadband opamp noise, which appears at the output of a Sallen-Key lowpass.
If you get a complete schematic with real opamp parts I could take a look at it in sim, although I'm a little busy right now.
Brad
Patrick: ANR26650M1B lifepo4 batteries like The Shaman is using are 2500mah vs probably something like 150-200mah from your 8.6-9v nimh, so given the minimum 600 cycles before it doesnt reach their high performance specification (not lifetime, they will last a lot longer) its not an issue.
Take however long your batteries last in a cycle (20hrs is a number I remember you mentioning I think?), multiply that by 12-16 (240-320hrs/10-13.33 days 24hrs a day) and multiply x 600 (144000-192000hrs/6000-8000 days/16.4383-21.9178 years non-stop) still seem like an issue? thats before they start to show their age. basically I expect battery technology to have been revolutionised again several times by then. dont let them ho below 2.8-3v though, they wont explode like some lithium cells can, but they can age quicker, though nowhere near as badly as nimh do if you deep discharge them.
Take however long your batteries last in a cycle (20hrs is a number I remember you mentioning I think?), multiply that by 12-16 (240-320hrs/10-13.33 days 24hrs a day) and multiply x 600 (144000-192000hrs/6000-8000 days/16.4383-21.9178 years non-stop) still seem like an issue? thats before they start to show their age. basically I expect battery technology to have been revolutionised again several times by then. dont let them ho below 2.8-3v though, they wont explode like some lithium cells can, but they can age quicker, though nowhere near as badly as nimh do if you deep discharge them.
600 cycles is 600 times you play music. If you play 3 times a week, that is in theory 4 years, not 16~24.
But do feel fee to use what you think is appropriate.
I personally do not see any issue with NiMH.
They will do 2000 cycles without too much trouble, have no memory effects, and will take trickle charge of C/40 continuously.
I have not used LiFePo4 myself, but a friend used them a lot for his bike lamps.
And he said one should expect to change them every 2 years.
I have not changed my NiMHs for 5.5 years now.
Unless one can prove that there is a performance advantage, like lower internal resistance, or noise, or whatever else.
Patrick
But do feel fee to use what you think is appropriate.
I personally do not see any issue with NiMH.
They will do 2000 cycles without too much trouble, have no memory effects, and will take trickle charge of C/40 continuously.
I have not used LiFePo4 myself, but a friend used them a lot for his bike lamps.
And he said one should expect to change them every 2 years.
I have not changed my NiMHs for 5.5 years now.
Unless one can prove that there is a performance advantage, like lower internal resistance, or noise, or whatever else.
Patrick
Patrick, they are referring to full discharge cycles - which are in fact not recommended.
Normal charge cycles like the one you describe are in the order of 1000s... Check here for example.
Their lifetime is considered higher than that of NiMh.
It mustn't be a coincidence that the automobile industry is moving to said batteries for new electric vehicles (I've read about Renault, GM and even Porsche, among others).
Not sure about noise but I think their internal resistance is a lot lower than other types of batteries up to very high frequencies. I recall reading a paper with such measurements @electrochem.org but I can't seem to find it now.
Anyhow, there are plenty of references to the advantages of LiFePO4s in the literature (e.g. Ref.1, Ref.2)
A less scientific test can be found here.
Also here's a quick reference from "Battery University":
Patrick, I can send you three of these batteries and a charger to test them (objectively and subjectively
) and you send them back when you're done.
Last edited: