Two parts suppliers in NZ.
The first one doesn't have that chip.
The second one only has very expensive eval boards.
You must live in the US.
The rest of the world has problems getting dedicated chips like that.
Ordering from the US is a problem because of the very high shipping charges.
One of the reasons China does better than the US.
Leo..
Wawa:
You must live in the US.
The rest of the world has problems getting dedicated chips like that.
Ordering from the US is a problem because of the very high shipping charges.
One of the reasons China does better than the US.
Yeah, I'm in the US. Digikey has them in stock here -> http://www.digikey.com/product-search/en/integrated-circuits-ics/pmic-battery-management/2556336?k=LTC6804
Did the OP mention where they're from?
Interestingly, China is the #1 consumer of that IC at the moment. Their government is really pushing EVs and Hybrids over there.
One LTC6804 would cost me US46.86
Plus Visa/Paypal charges.
Leo..
Wawa:
One LTC6804 would cost me US46.86
Plus Visa/Paypal charges.
Leo..
Ouch. And you get the other components on your schematic for normal prices?
In the US, your 4op-amps/4optos/1mux would cost 25% of the LTC6804 by digikey prices. If you were to triple the number of components to get 12 cells (the most the LTC6804 can support), you're approaching even on cost.
And with 12 cells the amount of PCB space and assembly costs the LTC6804 would save is substantial. It uses few enough external components, a hobbyist might be able to use a breakout board even to avoid a custom PCB.
And we shouldn't ignore how much better the performance would be as well. Your schematic is unlikely to measure 12 cells within 1mV accuracy in 300us.
That balance charger was not mine.
At some stage I helped another poster iron out the wrinkles.
I designed the board with the relays.
Leo..
Ah, I see. It's not a bad schematic really. I'm just saying if you extend it to 12 cells, the LTC6804 justifies its cost. Heh...although that price is obviously not an accident.
For a hobbyist it could still make sense for 4 cells if it saves them a custom PCB and a lot of soldering.
BigBobby:
cost. Heh...although that price is obviously not an accident.
No. Freight was more than the price for the chip.
I have asked US parts suppliers in the past to add an option to send cheap/small/SMD items just by post.
At buyers risk.
I don't mind, as a hobbyist, to sometimes loose a $5 item in the post (never happened to me).
But I do mind paying $20 freight for a $5 item.
The US could have soooo much more export if they abandon those ass-covering legally safe policies.
RS components do freight-free (any value), even if it has to come from the UK (to NZ).
Leo..
Wawa:
No. Freight was more than the price for the chip.
I have asked US parts suppliers in the past to add an option to send cheap/small/SMD items just by post.
At buyers risk.
I don't mind, as a hobbyist, to sometimes loose a $5 item in the post (never happened to me).
But I do mind paying $20 freight for a $5 item.
The US could have soooo much more export if they abandon those ass-covering legally safe policies.RS components do freight-free (any value), even if it has to come from the UK (to NZ).
Yes, I understood that the freight was the issue for you. I meant the chip (which is expensive in US even at $18) does justify itself compared to the discrete solutions (as long as you don't have to pay crazy costs to ship it).
Heh...you're making me think that I should get into the business of shipping digikey component internationally by using a less ass-covering legally safe policy 
It sounds like there's a market there.
Hi everyone....i am new here. As per my knowledge all voltages measured will referenced to the same point, EI 0v.The simplest thing to do is just subtract the readings from the next highest one. This does however give you less effective resolution as you go down the string.
Yes, except for A0.
That resistor has to go between +E1 and A0. Not from E1 to ground.
A resistor there protects the pin if the Arduino is accidently off (phantom powering).
Leo..
Thanks Leo.
The older I get, the foggier I get....
From this locked thread: ADC Issues while measuring battery voltage - Project Guidance - Arduino Forum
Well the pin is not damaged for sure, since when I leave the pin open it floats random values and when I ground it with 20k, it's going to zero. I checked via multimeter, the first cell reads 1023 even though on the DMM it says 4.13V . Also, the first resistor on E1 I have not connected since I don't require a voltage divider at for the first cell, it being well below 5V at all times
Here's the simple circuit
(-ve)------([-] Li-ion battery [-])------> A0
a0 reads 1023
DMM says 4.13V atm.
Also checked with A1, A2. Same results
As said, that simple circuit could damage the pin.
Pins shouldn't have more than VCC +0.5volt on them, unless current limited.
VCC could be 0volt when the Arduino is (accidently) off.
The returned A/D value depends on TWO things.
The voltage on the analogue pin, and the reference voltage (Vref).
If the input voltage is higher or equal to Vref, the A/D will return 1023.
So check the input voltage as well as Vref.
Leo..
Wawa:
If the input voltage is higher or equal to Vref, the A/D will return 1023.
Actually, according to section 24.7 of the 328P datasheet, "0x000 represents analog ground, and 0x3FF represents the selected reference voltage minus one LSB."
That might sound like nitpicking, but when I read that it really surprised me. If you can find another microcontroller datasheet that says this, please post me a link. The ones I checked all worked the way you described.
This is a long thread and I haven't had time to read all of it so maybe this was already mentioned, but if the batteries you are measuring are the ones powering the arduino (or ATmega328), if the voltage for Vcc drops below 5V, the readings will no longer be accurate .
True.
But OP is using four cells in series, so I assume he is using that to power the (unknown) Arduino.
If the voltage is above 5volt, the readings also won't be accurate.
You can't really measure the voltage of a LiPo (stack) accurate with default Aref.
One way out is to drop every cell to ~1volt, and use 1.1volt Aref.
Leo..
I'd like to do something similar anyone had some more details?
I would say, read this thread. Ideas enough!
Hello and what about use differential amplifier ?
You just need 3x LM258 dual op-amp for 6cell battery.
10 pieces under 1USD on ebay.
Circuit below. No sw substraction needed.
http://tinyurl.com/zfp9ceq
Just a thought.
Use individual battery measuring circuits and then send them in through an RS-422 addressable bus,
well it wouldn't be as complicated as Wawa's balanced charger circuit.
Dead_Ard
