LVC - Low Voltage Cutoff

[safe]

LVC - Low Voltage Cutoff

This messageboard is one part "chat room" and another part "library". A good library needs to have searching abilities and titles that make sense. I've seen threads that discussed strategies for dealing with LVC situations called "Lipo with its clothes off" of all things.

Bob Mcree's LVC circuit is another thing that was found under the title "The care and feeding of a123-based packs...".

So enough with the past sloppy behavior... I'm now creating a CLEARLY titled thread that will condense all the things related to the LVC. From now on folks try to make titles that match some topic so that people can scan the titles and have some idea what they might get out of the thread.

It's been like a library of books... with no titles...

It's called 'stream of conciousness'.

I don't mind things being a "chat room" and things flowing along a 'stream of conciousness' once the basic theme is set, but you should stay within the topic at least. When you stray from the topic that's when you spawn a new thread and start over. That way you refine a topic rather than turn it into spagetti. (many threads have been turned into absolute junk because they were badly fragmented... becoming useless)

 

[safe]

Bob Mcree's LVC Circuit

This uses the braking wire in some controllers to turn off the controller once a cell drops low.

 

[safe]

Dime Sized LVC

http://www.bphobbies.com/view.asp?id=V936613&pid=EF-LVD-1T

This one only monitors the voltage and does not act to prevent damage to occur. It's strictly a signal to stop, not a direct prevention.

 

[Toorbough ULL-Zeveigh]

It's called 'stream of conciousness'.
And it's only by not self censoring, by not constantly worrying about straying off topic do you get that seemingly unrelated spark of an idea that takes off into a firestorm.

It's what made someone like Reid a valuable contributor.

It's called 'stream of conciousness'.

I don't mind things being a "chat room" and things flowing along a 'stream of conciousness' once the basic theme is set, but you should stay within the topic at least. When you stray from the topic that's when you spawn a new thread and start over. That way you refine a topic rather than turn it into spagetti. (many threads have been turned into absolute junk because they were badly fragmented... becoming useless)

Hey that's life, no one's punching a clock here.
Given a choice I'd rather have the ideas on the table & format them to whatever suits you best (as you are doing now).
Instead of someone hold back their idea, thinking what they have to contribute might not be relevant or appropriate for a given thread & not worth starting it's own.

 

[safe]

Analysis of Bob Mcree's Circuit

Going through the exercise of creating an Op Amp based simple circuit to compare voltages has allowed me to understand (better) the value of buying prebuilt components that can do the same thing. (makes things alot easier) From what I can tell this circuit has two parts.

Part One:
There seems to be a part that does the voltage comparison to what is likely 2.5 volts or so and it acts in it's own little loop.

Part Two:
The second part uses an opto device that reads the current flow in the first voltage comparison loop and creates an actual output signal.

My GUESS as to why you do this is to prevent leakage.

Is this correct?

Is the main reason for setting things up this way (with the opto half of the circuit) is so that you can cut leakage to zero?

Also, precisely what brand names (product numbers) are we dealing with here?

 

[TylerDurden]

I'm now creating a CLEARLY titled thread that will condense all the things related to the LVC.

http://images.google.com/images?hl=en&q=lvc&gbv=2

 

[Toorbough ULL-Zeveigh]

You laff, but one of those actually yielded something relevant.
One that I wanted to post earlier but held back, & it wasn't just Lebanon Valley College.
Thanx TD for the opportunity to correct that.

Ya see how that stream thing (SOC) works, safe?

http://homepages.paradise.net.nz/bhabbott/lvw.html

 

[safe]

Ya see how that stream thing (SOC) works, safe?

I downloaded the pdf for the TL431 and I think I understand that it's basically the very same comparator circuit I was trying to build elsewhere, but it's "pre configured" to be 2.5 volts. I'm okay with that.

My question remains about WHY in the Bob Mcree circuit he decides to add the extra layer of the opto current reader to isolate the voltage measurement from the output signal.

Is it all about reducing leakage?

Otherwise you could just buy these TL431's and take the output directly for use in either an LED or use it for some other purpose like pulling down a throttle signal like my Armature Current Limiting circuit. I could simply add this as an extension of my ACL circuit so that the throttle is being pulled down for either reasons of heat (current sensor) or LVC. (the TL431's)

 

[EMF]

Safe- your heart is in the right place, but you are out of line with this thread. IMHO, I think it's time the admin(s) gave you some guidance. A lot of these threads you have created lately, to me at least, are really spamming up the forums. New people coming in are going to get frustrated.

For instance it's not up to you or me how these forums are run. What I read in this thread is you are more concerned in how we are storing information. This is a valid concern and should have been placed as a topic in general discussion where we could all form a consensus as what is best to do, or to do nothing.

It's not up to any one of us to dictate to the rest of the community how things are done here.

You yourself just created a thread with two subjects, which is exactly what yu are bitching about.

Just my 2 cents

 

[safe]

LifeBatts BMS

Keeping the "theme" intact this thread is supposed to catalog the BMS techniques that are out there. If someone wanted a "jumping off point" to learn about LVC systems this thread is supposed to act as a central repository for doing that. LifeBatts is in the process of presenting their "latest and greatest" and here is the circuit diagram:

http://endless-sphere.com/forums/viewtopic.php?f=14&t=3345

A good read on that link too...

 

[fechter]

Yes, our threads have a tendency to meander a bit.
I like spawning new ideas, but I don't like how trashed some of the threads get.

One thing I've been working on is a way to collect all the 'good' posts from a thread or threads and condense them into a single new one without disturbing the origninal posts. I'd like to create a few of these in the EV Basics section, where only a few people have permission to post.

In fact, it would be great if someone with lots of spare time and an eye for good information would volunteer to help. It would create a good resource for newbies and guys with CRS (can't remember shit) like me.

Safe, the reason for the opto isolaters is the controller is referenced to the battery negative. As you go up the battery , the voltage of the comparator outputs are offset by the cells under it, so a low really isn't low anymore. The opto couplers allow all the comparators outputs to be referenced to the same potential.

 

[EMF]

One thing I've been working on is a way to collect all the 'good' posts from a thread or threads and condense them into a single new one without disturbing the origninal posts. I'd like to create a few of these in the EV Basics section, where only a few people have permission to post.

I wonder if it would help to create a new "Brainstorming" subforum and let people theorise about concepts in there? You could even take it a step further and take the theories that are valid and then let people carry these to a Design and Development subforum. Could be fun and even exciting.

At least when you went in there, you knew what you were getting into. If you don't like that sort of thing, then you can easily avoid it, by simply staying out. If you like to work on such matters, then you can do so in a special environment created just for that.

If we don't do something, we could end up with a mess. With the new subforum, you can catch this thing early and start moving these hypothetical threads there.

Don't get me wrong, I think these type discussions are important, but I think it is even more important to isolate them from the rest of the catagories. As Safe says 90% of this stuff never pans out, so this can dilute an information base. But, if the 10 percent does, we don't want to miss out on that either.

At least it's not mixed in with the hands on or day to day problems we like to solve too.

In a nutshell, I'd like to keep exsisting products and their associate problems/challenges separated from hypothetical issues. Just my 2 cents.

 

[safe]

In a nutshell, I'd like to keep exsisting products and their associate problems/challenges separated from hypothetical issues. Just my 2 cents.

On this type of thread you would essentially be directing questions towards a preexisting example. So for example, the Bob Mcree LVC uses optocouplers and some are arguing that there's a way to turn them off when not in use.

Looking at the circuit it's a valid question to ask how that happens.

So how does the LVC prevent wear and tear on the optocoupler since they wear out if left on all the time?

So is that chip in the middle involved? What is it? What's the TCSxxxxx circuit, it's just a black box in the circuit diagram. More specifically I just don't see any outside connections that would turn the optocouplers off, so that means that within a few years they will be junk because optocouplers don't last that long and wear out when used.

Optocouplers are supposed to be at 50% power after 5 years of continual use... (so they might begin to malfunction after a year)

Maybe this is the "simplified" LVC circuit and not the real one that would be used? (parts are missing?)

Even a comparator (diferential amplifier) could prevent the current to flow to the optocoupler, but it would require it's own power... this doesn't show applied power, is there some circuit that can provide voltage detection WITHOUT supplied power? (I'd like to learn about that! )

 

[GGoodrum]

We really do need a "safe" haven...

The TC54 voltage detector chip used in our LVC circuits (including the new BMS...), comes in a variety of "flavors", from 2.1V up to 4.0V. These detectors are typically used in laptop power supplies, in order to trigger a shutdown/hibernation if the battrey voltage goes too low. With the a123 LVC boards, I used the 2.7V version, and for the LiFeBatt and other LiFePO4-based cells, the 2.1V chip is used. In any case, the way it works is that the output remains at a logic-high state as long as the input voltage is above the threshold voltage (i.e -- 2.1V...). When the input voltage dips below the threshold, the output goes low, which then causes the LED in the opto-coupler to turn on. The TC54 is an extremely low power device, drawing only about 1 uA. That is less than the self-discharge "leakage" of any cell type I've seen, which is why we can leave the LVC circuits connected to the cells all the time.

The opto outputs are all ganged together, in a logical "OR" configuration. This combined signal is then connected to the ebrake signal and the controller ground. If any one cell dips below the threshold, the ebrake signal is pulled low, which causes the controller to cut the throttle,which removes the load. In actual operation, the first time an LVC circuit trips it is usually while under a long steady load, like going up a hill. The power is cut, but only for a split-second, and then it returns.If you keep the load on, it will do it again, almost right away. At that point, if you back off the throttle, you can actually go another couple of miles. When you get to the point that the pack is just about dead, even half-throttle will cause the power to oscillate, at about a 2Hz rate. I did this the other day, testing a 16-cell LiFeBatt pack, and when the oscillation started occuring, the WattsUp said I had used 9,85 Ah one time, and 9.78Ah the second time I tried it.

In the new BMS, we also now have the option of using the LVC opto output to control two 4110 FETs in order to actively cut the negative pack lead connection.

-- Gary

 

[Ypedal]

With the FET option, will the BMS auto-reset or will it need to be " Un-Plugged and reconnected " ?

 

[eP]

One thing I've been working on is a way to collect all the 'good' posts from a thread or threads and condense them into a single new one without disturbing the origninal posts. I'd like to create a few of these in the EV Basics section, where only a few people have permission to post.

I wonder if it would help to create a new "Brainstorming" subforum and let people theorise about concepts in there? You could even take it a step further and take the theories that are valid and then let people carry these to a Design and Development subforum. Could be fun and even exciting.

It's bright idea IMHO.
You should create a poll with that idea i'm sure.
Maybe should have been created also subforum: Most Viewed &Replied for 60+ replies and 2000+ views ?
That way such threads could be available in one place and could be possible to view them sorted by last post. As a result they would be easier available to wider audience and more dedicated to continue them.
Actually very narrow member's percentage checking what was hot I suppose.

Regards

 

[safe]

Thank You, Thank You, Thank You

Now I get it.

Somehow in the process of people being so fascinated with the optocoupling aspect of the LVC I missed the real gem in it which was this Voltage Detector component. From one of the pdf's:

GENERAL DESCRIPTION

The TC54 Series are CMOS voltage detectors, suited especially for battery-powered applications because of their extremely low 1mA operating current and small surface mount packaging. Each part is laser trimmed to the desired threshold voltage which can be specified from 2.1V to 6.0V, in 0.1V steps. The device includes a comparator, low-current high precision reference, laser-trimmed divider, hysteresis circuit and output driver. The TC54 is available with either an opendrain or complementary output stage.

See this is the kind of stuff that the common observer isn't getting. I'm really proving the value of FEEDBACK on these things... when you just post an item and then let it sit there no one is going to be able learn the quirks and know how to find the odd components.

http://digikey.com/scripts/DkSearch/dksus.dll?Detail?name=TC54VN2902ECB71CT-ND

 

[safe]

When you get to the point that the pack is just about dead, even half-throttle will cause the power to oscillate, at about a 2Hz rate. I did this the other day, testing a 16-cell LiFeBatt pack...

What I'm trying to design is a LVC that integrates into my Armature Current Limiting circuit. After all, if the act of pulling too many amps can overheat the motor it also tends to produce sag in the cells, so there is a convergence of interests in combining the two. The effect at runtime when the cells begin to get low would be to simply lower the current limit. The bike would not cut out but would instead "moderate itself" by producing less power. Rather than having to "throttle twiddle" you could rest assured that the throttle will keep pulling down all the way until at some point the throttle has no effect.

I like these Voltage Detection components...

 

[GGoodrum]

With the FET option, will the BMS auto-reset or will it need to be " Un-Plugged and reconnected " ?

It will auto-reset, just like the regular LVC function that uses the ebrake signal.

 

[safe]

More Questions

With comparators the big fear is high "common mode voltage" and the distortion that it has on the comparator functioning. This forces you to buy expensive stuff that is built to deal with it. (expensive for me is $5 each) My first question is if these Voltage Detectors are properly designed to deal with the 36v - 72v range of "common mode voltage" that they would have to deal with?

Should we fear the "common-mode rejection ratio" or something like it:

http://en.wikipedia.org/wiki/Common-mode_rejection_ratio

Second question is about SPICE models... where are they, are they even possible to be found?

 

[safe]

Some Information

STMicroelectronics has these for 65 cents each. The logic and block diagrams are below. My question remains... going by the block diagram there's a comparator inside one of these. If comparators are susceptable to the "common-mode rejection ratio" (see above) then wouldn't this be too?

These should be fine until you put cells in series and the "common mode voltage" increases... then the internal comparator would begin to become destabilized. The thing is that you might not know it... Hmmmm... I'm hoping someone far smarter than I am can explain this...

Does the fact that those two gates exist in the circuit change things so that the "common mode voltage" is factored out?

(I suppose I could recreate this block diagram as a circuit and see what happens)

http://digikey.com/scripts/DkSearch/dksus.dll?Detail?name=497-6181-1-ND

 

[safe]

Question

Just want to remind folks that there is a pending question here...

My question remains... going by the block diagram there's a comparator inside one of these. If comparators are susceptable to the "common-mode rejection ratio" (see above) then wouldn't this be too?

How can these Voltage Detectors get away with this?

 

[GGoodrum]

Question

Just want to remind folks that there is a pending question here...

My question remains... going by the block diagram there's a comparator inside one of these. If comparators are susceptable to the "common-mode rejection ratio" (see above) then wouldn't this be too?

How can these Voltage Detectors get away with this?

All I know is these TC54 chips work really well, and draw next to nothing. I'm not sure what your issue is, but there certainly is not a problem connecting the lvc circuits in series. Each circuit works completely independent of the others.

-- Gary

 

[curious]

safe-
I think you now confused everybody (and possibly yourself ). What the hell common mode rejection has to do with a single input voltage detector ?
When suggesting the idea could you please draw a simple schematic to illustrate it. I bet in 90% of the cases the schematic will have an answer for you right away.

 

[safe]

safe-
I think you now confused everybody (and possibly yourself ). What the hell common mode rejection has to do with a single input voltage detector ?
When suggesting the idea could you please draw a simple schematic to illustrate it. I bet in 90% of the cases the schematic will have an answer for you right away.

I know that the more expensive comparators can do a good job under very high "common mode voltage" situations (like 200 volts) but normally you have to pay extra for that kind of capability just in the comparator alone.

These Voltage Detection units seem to be a combination of a comparator and some FET's all in one and some how they seem to have figured out how to get it all to work.

I was just wondering "why"... if they really do work correctly and are cheap that's great.