[SOLVED] Help with project

[GH Crash]

I need help in determining the best approach for a lithium ion capacitor, LIC, (also know as a hybrid capacitor) low-voltage protection circuit. Can you suggest an approach, or a simple circuit design to address the requirement of this circuit?

Background: Hybrid (lithium ion) capacitors have a minimum voltage limit. Discharging them beyond that point, or storing them below their minimum voltage will damage or destroy the capacitor. The circuit (device) needed would be placed between the capacitor and the load and would interrupt the current flow when the capacitor voltage dropped to its minimum allowable voltage. The circuit will be used in environments where size and weight are a major concern.

Circuit requirements

1. The circuit shall not reduce the maximum voltage and/or amperage available at the load.
2. The circuit must be capable of handling currents up to 3 amps.
3. The circuit should operate satisfactorily at input voltages between one and six volts DC.
4. The completed circuit board needs to be small, preferably under 15mm by 15mm square.
5. The completed circuit must be light, Under 0.75 grams max weight, under 0.5gshould be the goal.
6. Items, or conditions, not directly mentioned in the requirements are left up to the designer's discretion.

Additional information (information that may be helpful in designing a solution.):

1. The LIC typically used will be between 10F and 100F capacitance and 5.0 max volts.
2. The load will typically be a DC coreless motor of under ten watts and 3.7 volts.
3. It is desirable that there be absolutely no current flow once the LIC reaches its minimum voltage limit but quiescent current flow through circuit components may be unavoidable.
   

I would appreciate any help you can give. A simple circuit diagram may be the best way to explain your thinking. I try to answer any and all questions concerning this circuit design request.

George

 

[KlausST]

Hi,

This sounds as if the requirements are similar to LiIon protection.
There are ready to buy modules, ready to buy ICs, and already available schematics.
Please try to read through datasheets, application notes .... no need to understand every detail, just to get familiar how they solved it, what is good and what needs to be adjusted.

Some comments to your informations:
* "the best approach" needs some more information .. in which regard "the best". For the one it may be the lowest cost, availability, some technical items.
* "not reduce voltage" is impossible. Every piece of wire, every tiny current will reduce the voltage. You should better say (in value with unit) what voltage drop you are willing to allow.
* "satisfactory .. down to 1V". What das satisfactory mean? Are there any new/other limits? 1V without a circuit to boost this voltage any pass transistor will have problems to stay low ohmic for low voltage drop. Also you talk about "input" voltage. Is it the battery voltage or the charger voltage, or any other voltage?
* "absolutely no current flow" is impossible. There will be stray currents ...and there may be a circuit to "measure the voltage" .. which needs some minimum current. What current is still O.K. in OFF state? Is it 100uA? 1uA? 1nA? Less? I understand" the less the better, but where is the upper limit?

I miss exact threshold levels and the tolerance of them. Are there critical limits, like at LiIon where the upper cell voltage should be kept within 2%.

And I miss any information about charging. How is it handled? Does the charging current flow through this (to design) circuit, or is there an extra path?

Klaus

 

[barry]

I think there is a fundamental problem here, as Klaus alluded to: getting a circuit to operate at 1V.

Li Ion ICs generally work down only to 2.5V.

A conventional transistor (BJT) used as a pass element will require a lot of current to drive it (300mA); that’s probably not acceptable for your application. There are no MOSFETs (that i’m aware of) that will work down to 1V- this would require a circuit to step up the 1V to a suitable level to drive the gate. And I don’t think there are any relays that will work at 1V.

So, one solution I can see right now would be an opamp or comparator + reference + MOSFET + charge pump voltage booster.

 

[GH Crash]

Hi,

This sounds as if the requirements are similar to LiIon protection.
There are ready to buy modules, ready to buy ICs, and already available schematics.
Please try to read through datasheets, application notes .... no need to understand every detail, just to get familiar how they solved it, what is good and what needs to be adjusted.

Some comments to your informations:
* "the best approach" needs some more information .. in which regard "the best". For the one it may be the lowest cost, availability, some technical items.
* "not reduce voltage" is impossible. Every piece of wire, every tiny current will reduce the voltage. You should better say (in value with unit) what voltage drop you are willing to allow.
* "satisfactory .. down to 1V". What das satisfactory mean? Are there any new/other limits? 1V without a circuit to boost this voltage any pass transistor will have problems to stay low ohmic for low voltage drop. Also you talk about "input" voltage. Is it the battery voltage or the charger voltage, or any other voltage?
* "absolutely no current flow" is impossible. There will be stray currents ...and there may be a circuit to "measure the voltage" .. which needs some minimum current. What current is still O.K. in OFF state? Is it 100uA? 1uA? 1nA? Less? I understand" the less the better, but where is the upper limit?

I miss exact threshold levels and the tolerance of them. Are there critical limits, like at LiIon where the upper cell voltage should be kept within 2%.

And I miss any information about charging. How is it handled? Does the charging current flow through this (to design) circuit, or is there an extra path?

Klaus

--- Updated Aug 7, 2022 ---

Hi,

This sounds as if the requirements are similar to LiIon protection.
There are ready to buy modules, ready to buy ICs, and already available schematics.
Please try to read through datasheets, application notes .... no need to understand every detail, just to get familiar how they solved it, what is good and what needs to be adjusted.

Some comments to your informations:
* "the best approach" needs some more information .. in which regard "the best". For the one it may be the lowest cost, availability, some technical items.
* "not reduce voltage" is impossible. Every piece of wire, every tiny current will reduce the voltage. You should better say (in value with unit) what voltage drop you are willing to allow.
* "satisfactory .. down to 1V". What das satisfactory mean? Are there any new/other limits? 1V without a circuit to boost this voltage any pass transistor will have problems to stay low ohmic for low voltage drop. Also you talk about "input" voltage. Is it the battery voltage or the charger voltage, or any other voltage?
* "absolutely no current flow" is impossible. There will be stray currents ...and there may be a circuit to "measure the voltage" .. which needs some minimum current. What current is still O.K. in OFF state? Is it 100uA? 1uA? 1nA? Less? I understand" the less the better, but where is the upper limit?

I miss exact threshold levels and the tolerance of them. Are there critical limits, like at LiIon where the upper cell voltage should be kept within 2%.

And I miss any information about charging. How is it handled? Does the charging current flow through this (to design) circuit, or is there an extra path?

Klaus

I will try to answer each of your points, hopefully without stepping on your toes.
Similar to LiIon protection: Yes, I guess, but the Liion protection I'm aware of protects against over charging. Do you know of a specific IC or circuit that would protect against over discharging? Can you give me an idea of the IC or circuit you are thinking of?
Data sheets: I hear you. I read them and interpenetrate them as best as I can. If you can suggest a particular IC to use, I will read and read its data sheet. I promise.
Best approach: The best approach is the one that best satisfies the stated requirements. Items like time. costs, availability, accident prevention among others are not mentioned in the requirements
Not reduce voltage: Your are right of course. Hopefully you understand, without it being stated that normal circuit loses are excepted. A source to load voltage drop of less than 0.05 volts is acceptable.
No current: I'm sorry that this is not clear to you as I tried to explain that quiescent current flow through the components may exist but that it should be kept to as low as possible. "As low as possible" means as low as reasonable possible. The higher the quiescent current, the less useful the device/circuit will be in prevent over discharge of the capacitor.
Exact thresholds/tolerances: There are no exact thresholds or tolerances. LIC data sheets usually just state that capacitor should not be discharged below a certain voltage. That minimum voltage varies between different brands and manufactures. Usually it is around 2.3 volts.
Charging: Upper capacitor voltages are immaterial to the design of this circuit as the charging of the capacitor is done through another device with a separate lead.

Hopefully this answered your questions and comments sufficiently. If not, hit me again with more questions.

George

--- Updated Aug 7, 2022 ---

I think there is a fundamental problem here, as Klaus alluded to: getting a circuit to operate at 1V.

Li Ion ICs generally work down only to 2.5V.

A conventional transistor (BJT) used as a pass element will require a lot of current to drive it (300mA); that’s probably not acceptable for your application. There are no MOSFETs (that i’m aware of) that will work down to 1V- this would require a circuit to step up the 1V to a suitable level to drive the gate. And I don’t think there are any relays that will work at 1V.

So, one solution I can see right now would be an opamp or comparator + reference + MOSFET + charge pump voltage booster.

OK, Barry. I'm listening.
Rohm's BU48xx series of IC have a minimum detection voltage of 0.9v. that is where I came up with the 1.0 volts minimum. I need (think I need) chip that will work down to at least 1.5 volts.

Opamp? Do you mean a operational amplifier. It sounds like a comparator + reference + MOSFET + charge pump voltage booster circuit would be difficult to get it to meet the size and weight requirements of the project. Can you point me to such a circuit so that I can study how it is laid out?

Thanks for you help.

George

 

[KlausST]

Hi,

LiIon protection circuit:
I have no IC in mind, just read about a couple of them in the forum. I guess mis of them have under discharge protection inside.
I don't want to do part selection nor the datasheet / application note "reading" for others.
But I'm quite sure IC manufacturers provide overview and parametric search for battery protection ICs.

Just came into my mind:
Some time ago I bought a cheap LiIon protection board for playing around. It surely had under discharge protection.
Don't have it by hand now to see what IC it uses.

Voltage drop.
Thanks for the info. There is nothing like a "commonly used value" some are satisfied with 0.5V .... some want just 10mV.

No current:
Yes, you were very clear. But: no number --> no calculations --> no focussed circuit design.
If you can't give a target value ... no one of use can give either.
So basically it's just a delay. Sooner or later you will have to decide.
Let's say we provide a circuit ... and it has 123uA quiescent current ... at the latest then you have to decide.
Means for us: we maybe have to design multiple circuits until it meets your expectation.
Any "over the thumb" value is better than none.
Maybe tell us two values: a "nice to have" and one as the "worst limit you can live with"

I have no experience with those cells. I guess they have a lower healthy voltage limit, and they will have self discharge...so use these numbers to do some calculations maybe in the way: after "shut down" (decide the threshold) you want to store them for a week without charging until the voltage enters the "unhealthy" region. Just as an example...

Klaus

 

[GH Crash]

Hi,

LiIon protection circuit:
I have no IC in mind, just read about a couple of them in the forum. I guess mis of them have under discharge protection inside.
I don't want to do part selection nor the datasheet / application note "reading" for others.
But I'm quite sure IC manufacturers provide overview and parametric search for battery protection ICs.

Just came into my mind:
Some time ago I bought a cheap LiIon protection board for playing around. It surely had under discharge protection.
Don't have it by hand now to see what IC it uses.

Voltage drop.
Thanks for the info. There is nothing like a "commonly used value" some are satisfied with 0.5V .... some want just 10mV.

No current:
Yes, you were very clear. But: no number --> no calculations --> no focussed circuit design.
If you can't give a target value ... no one of use can give either.
So basically it's just a delay. Sooner or later you will have to decide.
Let's say we provide a circuit ... and it has 123uA quiescent current ... at the latest then you have to decide.
Means for us: we maybe have to design multiple circuits until it meets your expectation.
Any "over the thumb" value is better than none.
Maybe tell us two values: a "nice to have" and one as the "worst limit you can live with"

I have no experience with those cells. I guess they have a lower healthy voltage limit, and they will have self discharge...so use these numbers to do some calculations maybe in the way: after "shut down" (decide the threshold) you want to store them for a week without charging until the voltage enters the "unhealthy" region. Just as an example...

Klaus

Hi again Klaus. It seems like only you and I are exchanging ideas. That makes it a sort of public PM.

I've done a quick, hour long or so, search of Li Ion protection boards but I've yet to find one that sounds like it will do what I want, All I've seen, so far, are for lithium ion or lithium polymer batteries and the high/low protection points are set and based on the characteristic of those batteries. I will make a more extensive search.

My problem, and where you can help, is that I'm not knowledgeable enough what is a reasonable voltage drop. Also, wouldn't a reasonable voltage drop be different for different approaches to the problem?

The same sort of problem exists for quiescent current. I'm not knowledgeable enough what would be reasonable. If I have to pick a number, less than 10uA.

"Means for us: we maybe have to design multiple circuits until it meets your expectation." I don't want anybody to design the circuit. I asking for ideas of a circuit design that might work. Just give me a starting point and I will try to do the actual circuit design and the selection of the ICs.

Let me try a different approach. Let me propose a circuit (one that doesn't work but I don't know why) and let the forum tell me how to make it work or suggest a better approach.

You have been a great help. Much more than I have expressed. My background is oil production and I know to talk oil field, the information/items that need to established to start and complete a project. I do not know "electronics speak." You have helped me understand better what is important in designing an electrical circuit and how to communicate that information to the people that can help me.

Let's call this project un-doable because of the lack of clear project requirements and specific design parameters.

George

 

[KlausST]

Hi,

I've done a quick, hour long or so, search of Li Ion protection boards but I've yet to find one that sounds like it will do what I want,

Sorry, it seems I was not clear in this. I didn´t ask you to find a perfectly suitable IC. I asked you to read some documents to see how the LiIon solutions work. To find out what suits you and what not. Just to have more context to talk about.

I'm not knowledgeable enough what is a reasonable voltage drop.

You already answered it in your previous post. I´m fine with this information.

less than 10uA.

Fine, it´s a starting point.

I asking for ideas of a circuit design that might work. Just give me a starting point and I will try to do the actual circuit design and the selection of the ICs.

This is why I suggeested to read the datasheets / application notes.
Just out of curiosity I do a search on my own. "Battery protection IC" at farnell.
Within 1 minute I have a list of 192 items

• > Battery Protector ICs

Without any further selection I opened the datsheet of the first one. AP9101C.

https://4donline.ihs.com/images/VipMasterIC/IC/DIOD/DIOD-S-A0007229744/DIOD-S-A0007229744-1.pdf?hkey=6D3A4C79FDBF58556ACFDE234799DDF0

Opening the datasheet it shows on the first page some features including "over-discharge detection".
On page 2 there is a schematic how it is meant to operate.
And on page 3 there is a sketch how it internally works.

Isn´t this what you are asking for?
I didn´t go deeper .... what parts of the schematics can be used for your application...
Maybe it has alot more than you need.

It also gives you values of operating current (which is 5.5uA max) operating voltage (1.5V ... 5.5V).
And it gives a brief descripion how it works, timing diagrams and so on.

For me - without much knowledge about battery protection - it gives a lot of information. Within 2minutes.
Let´s talk about this device. What do you think is good and what do you think you is not that good and needs to be adjusted?

Let's call this project un-doable

I´m shocked now.
You write a long post ... and at the very end you say: Let´s stop here.


Klaus

 

[GH Crash]

Not stop. maybe, Just start talking circuit specifics in a new or separate posts..

Your first link was helpful. I'm in the US and had only done general searches through Google using variations of "Li ion protection" You searched and electronics supplier. I hadn't don that in the past. I will now.

I do read the data sheets, once I know what type of ICs I should be looking at. Your example appears to be a good starting point. One question, what are Q1 and Q2? What IC component? I will continue to investigate the use Ap9101C.

I did designed circuit that used a voltage detect IC. It was similar to the open drain application circuit in the data sheet except the micro controller was replaced with a N-channel MOSFET. The circuit didn't work because the motor rpm varied high low in a cyclic fashion with a perion of around a cycle/sec.

 

[KlausST]

Hi,

Q1 and Q2 are external MOSFETS as power switches.
(I wonder why you ask, because you already used an N-CH MOSFET and your appended datasheet clearly uses the same symbol)

The problem with voltage detectors like BU48xx is that they don´t have (enough) hysteresis.
As soon as you draw a current form a batterry the voltage will drop a bit. Let´s say just 50mV. Now if the threshold is at 2.500V and the battery voltage is let´s say 2.510V then the circuit switches ON. but immedialtely - due to the current - the voltage goes down to 2.460V (50mV drop) it switches OFF. This may happen several times per millisecond.
Thus you need a circuit that includes a hysteresis that is bigger than the worst case drop.
As example: OFF below 2.500V and ON above 2.600V.
In your case the thresholds can be far apart, like: OFF below 2.00V and ON above 3.5V (after some charging).
--> You have to decide.

Klaus

 

[GH Crash]

Q1 and Q2 are external MOSFETS as power switches.
(I wonder why you ask, because you already used an N-CH MOSFET and your appended datasheet clearly uses the same symbol)

That is what I assumed but I asked because there are n-ch and p-ch MOSFET, aren't there? And n-ch appear to be scarce in the US right now.
I understand the what you say about the hysteresis thing. The on again, off again action is not an issue as long as the final off voltage is not below the set point, +/- 2% leeway.

What i want to know about the circuit in the attachments is what is causing the motor speed to increase and decrease in a cyclic manner The motor rpm varies up and down from around 60% to 100% continually even when the input voltage is well above the low set point.. Can you give any insight as to why that is occurring?

What if I replaced the voltage detector with the AP911C?

I have another circuit under development that uses the voltage detect BU48xx to provide the enable for a LDO voltage regulator. The problem there is the the voltage regulator is getting hot even at currents well within the specifications of the voltage regulator. Do you think that use of the AP9101 would benefit that circuit? (Yes/no answer please. If yes, I will post it as a separate question.

Thanks much.
George

 

[KlausST]

Hi,

Hysteresis: you complained about the non stable motor RPM, and I guess this is a side effect of ON/OFF/ON/OFF cycling.
You my find it is not an issue ... but it is bad circuit design and may cause bad EMI effects.
There are regulations regarding EMI/EMC/safety ... and so on ... one has to comply with.
Also one has to care about power supply stability, this means to add a bigger bulk capacitor in parallel a fast ceramics capacitor.

Additionally your big capacutors around the gate are counter productive. It makes the MOSFET gate signal slow, thus the mosfet is in the analog region and also causes a lot of dissipated heat, maybe even kills the MOSFET.

I can't comment about the AP9xxx because I didn't read the datasheet yet. As written it's just a random choice.

Klaus

 

[GH Crash]

I wondered if hysteresis might be the cause of the cyclic motor rpm but the motor rpm cycle seem, to me to be too slow for it to be caused by hysteresis. How does one go about removing hysteresis from a circuit?

I was told by someone who claimed to know circuit design that the 0.1uF capacitor was needed for noise filtering.

"add a bigger bulk capacitor in parallel a fast ceramics capacitor." You lost me there, Klaus.

"I can't comment about the AP9xxx because I didn't read the datasheet yet." You read enough that you thought it a suitable example. I'm not asking specifically about the AP9101, I'm asking about the possible suitability of that type of IC. (My fault for not making that clear.) I've read the datasheet. Although I do not fully understand it, it sounds like an IC similar to the AP9101 (might) work. I don't know if it might work better than the BU48xx or not.

Klaus, you have been a great help. The circuit is still not functioning but I know a different way to approach the problem and items that should be considered in the circuit design. Give me a couple of days to adsorb what I have been exposed to.

George

 

[KlausST]

Hi,

I did not see your circuit malfunctioning, I have no measurements, I just have your input about the failure.
The cyclic ON/OFF indeed us faster than what you desribe ... thus I talk about sude effects, they can be as slow as you describe.
But I guess it's a combination of several circuit faults.

You should not vremove hysteresis, one should add hysteresis.

Filters and hysteresis are different functions ... one has to combine them meaningfully for clean circuit functionality.

You read enough that you thought it a suitable example.

Again: I picked it as random example, just to show you that there is a lot of information inside the documents.
You asked for a schematic, there it is. Not saying that it is suitable. Just as first discussion base.
I don't need to read it, it's your job, to get familiar with some working principles.

If I had to choose a suitable device, I'd use a parametric search.

I'm asking about the possible suitability of that type of IC

Yes, I think a dedicated battery protection IC may be suitable as battery protection ;-)

Although I do not fully understand ........ Give me a couple of days

Take your time. Rome was not built in a single day.

Klaus

 

[GH Crash]

Klaus,

One Two last questions. Hysteresis is the phenomenon in which the value of a physical property lags behind changes in the effect causing it. Is that a reasonable definition of what you are talking about? How do I, could I, add hysteresis to the circuit?

George

 

[KlausST]

Hi,

Hysteresis: wikipedia: chapter: Electronic Circuits

Hysteresis - Wikipedia

en.m.wikipedia.org

Basically there are two different trip points for ON and OFF respectively, as explained already.

Klaus

 

[GH Crash]

I have interpenetrated what Klaus has said to mean that hysteresis and the use of the voltage detect chip may be the cause of the problem and that the use of a different type of IC, a battery protection circuit may be a better approach. Is that basically correct, Klaus? After reading, and re-reading, the data sheet for the AP9101C lithium ion/polymer protection chip, I think that it should work for in a lithium ion capacitor circuit.

I do have one question. The AP9101C chip has both high and low voltage protection capabilities, I will not be using the high voltage protection feature. What should I do with the FET gate pin for the high voltage interrupt, CO? Can it be left un-connected? The attached file shows a typical application circuit for the AP9101.

One other question. The typical application drawing has the two MOSFETs, Q1 and Q2, as mirror images of each other. Is there a significance in that?

 

[KlausST]

Hi,

hat hysteresis and the use of the voltage detect chip may be the cause of the problem

How to explain again? Not hysteresis is the problem, but the lack of hysteresis is the problem in your design.

Still you stick on the AP9101 ... while I explained that this is just the first hit of a coarse Farnell search.
I never meant this IC for your application, as explained it was just meant to find some informations, schematics, ideas.

Using this IC ... with handstands ... to make it suitable for your application was not my intention.

And - also again - I did not read the datasheet. I just had a quick view on it.

I strongly recommend to do a search on your own to find an IC that does what you want.

Klaus

 

[GH Crash]

I understand what you are saying about Hysteresis and the IC. The problem was in my phrasing ins the summation of what you said.

No handstands are required. I heard you the first time. I reviewed many Lipo protection IC data sheets. It just happened that your course search yielded an IC that was (could be) suitable to my needs. I found no other IC that came any closer to meeting my needs. Those IC designed to protect one cell Lipo batteries all had high voltage protection included Therefore my question about what to do with the high voltage protection pin is applicable regardless of the IC chosen.

Klaus, before I even considered starting this thread, I search extensively for a chip, or means, to protect a lithium ion capacitor from being over discharge, discharged below it minimum recommended voltage. My lack of an electronics background means that I don't always know where to search or what to search for. Your comment about using a battery protection IC lead me to search where I hadn't known to search.

I'm not asking you to make anything suitable. I will say that again, I'm not asking you to make anything suitable.
I will do the making. Can you just give some generalized answers?

You keep telling me to read the datasheets. I do that. I do that before I ask. Yes, most of them contain typical application drawings. None of my projects are not typical applications. They are unique, they are non-typical so a typical application is often of limited help. Sometime you might consider, you can look at the datasheet and it wont tell you how to use that IC in the environment where you want ot use it. You can read the specification sheet on gasoline but you still won't know how to distill or formulate it. A specification (data sheet) for a specific aircraft wont tell you how to make the plane or how to fly the plane.

Let's put the datasheet thing to rest. And on to the questions I ask.
1. Should I connect the CO lead to anything if I'm not going to use it? The data sheet does not cover that question.
2. Is there any significance to the two MOSFET being mirror images of each other? The AP9101 makes no mention of Q1 andQ2.



Now back to the question I aks

 

[GH Crash]

It appears unlikely that there will be any further discussion on this thread. Let's call it quits?

 

[KlausST]

Hi,

it´s on you to press the [mark as solved] button.

Klaus