any idea to build a throttle control that limit battery voltage drop to 41V ?

[mitch_781]

Hi Lanchon,

Sorry but I have a last question regarding your circuit (I think it would be difficult to find more question on a simple circuit like you proposed ), I understand it will regulate but I don't undertstand how the throttle works now.

The output of the comparator can be only Vdd(5V) or 0V, this output is linked to the throttle input controller so how works the hand throttle if the output is always Vdd (5V) or 0V at the throttle input controller ? (the hall effect sensor has no effect now ?)

There is maybe something I did wrong in the circuit : Is the Vdd linked to the positive power input of the comparator ?

Regarding the componants I can order them now, it is only the measure of the Vdd that I will do in September (or maybe you want to know Vdd voltage before choosing the comparator)

Find below the only picture I have of the controller, It is a picture of all the parts I received. (there is nothing at all written on the controller.)

Also don't worry I will leave you in peace for next two weeks beacause I will not have internet access.

Thanks again for your help

[Lanchon]

hi,

> The output of the comparator can be only Vdd(5V) or 0V
> maybe something I did wrong in the circuit

you did nothing wrong. the output of the comparator is either ground or high impedance, not ground or Vdd. that's called an open collector (bipolar) or open drain (mos) output.

> you want to know Vdd voltage before choosing the comparator

definitely.

post when you get back, have a nice time!

[Lanchon]

that's exactly like the old 36V controller I got in 2006, probably has a similar circuit. when you measure Vdd do it with the throttle plugged in, if possible. you might find it's something like 4.3V that way, and about 5V when the throttle's not connected.

[mitch_781]

Hello,

It's a long time since my last post but I have more time now to work on the project of limiting voltage drop for small batteries.

I measured the voltage of Vdd on the throttle it is 4.36v (throttle connected) and on full charge the voltage does not move, it is always 4.36v.

Next step is :
to check if a 1K resistor has an influence on the throttle
Check if voltage of ground of resistor and ground of battery is the same

Before I will post an updated circuit with the values I will use, to check if everything is ok.

Thank you for your help Lanchon with your last posts.

[mitch_781]

Circuit update :

[biohazardman]

Circuit update :

Would it not be allot easier to get a bit more battery or replace it with something that will not have the voltage sag so you would not have the need for more electronics.  That does seem to be the problem does it not?

[mitch_781]

Hello, it is true but first I have already spent 350$ for nimh batteries (because it is more safe than Lipo and more green than nicd),
 second I didn't expect that the manufacturer TENERGY (allbattery.com) will send non conform batteries less than 3Ah instead of 4.2Ah indicated.
So I prefer to stop purchasing batteries and loose money because I want to have only 4Km of range with my bike.

Also even with large batteries,i would have the cutting problem at end of use.

I think if the circuit is working it is very usefull for everybody who has cutting problem with a small battery (or end of use), and it is very simple only a few resistors,a comparator and one circuit you can do yourself.

I should be able to do the last 2 check next week end and after I'll start to look for a comparator and do the circuit.   

[biohazardman]

Hello, it is true but first I have already spent 350$ for nimh batteries (because it is more safe than Lipo and more green than nicd),
 second I didn't expect that the manufacturer TENERGY (allbattery.com) will send non conform batteries less than 3Ah instead of 4.2Ah indicated.
So I prefer to stop purchasing batteries and loose money because I want to have only 4Km of range with my bike.

Also even with large batteries,i would have the cutting problem at end of use.

I think if the circuit is working it is very usefull for everybody who has cutting problem with a small battery (or end of use), and it is very simple only a few resistors,a comparator and one circuit you can do yourself.

I should be able to do the last 2 check next week end and after I'll start to look for a comparator and do the circuit.   

I do understand that the batteries are spendy for sure.  They can be heavy also.  I spent a hundred dollars more than you and got some LifePo4s was a hard thing to spend more on batteries than the cost of the bike but the twelve pound 36V FoxPower pack has been very good and gives me about 20 miles per charge.  Currently have near 2K miles on it.  Hope you are able to get things to work correctly with the small and light battery pack you have.

[Lanchon]

hi! long time no see!

please post your results of:

>Next step is :
>to check if a 1K resistor (IN SERIES ON THE SIGNAL WIRE) has an influence on the throttle
>Check if voltage of ground of *THROTTLE* and ground of battery is the same (OR MOSTLY THE SAME)

after that let me select an open collector (or drain) comparator for you, or have you already got one? only then we can size the rest of the components.

[mitch_781]

Hi Lanchon,

Yes a long time but I have more time for the moment to work on it.

I confirm, the resistor had no influence on the throttle (or I did not noticed it).

and the level of the ground of throttle is the same of the ground of battery (0V).

So everything seems to be ok to build the circuit.

I don't have the comparator but I will try to by one to a local dealer in case I have question to ask. Thank you if you can help me to choose one. Maybe if you know a very standard or very used comparator it will be easier for the dealer to get one.

See you

[Lanchon]

I looked through your proposed retailer. this is what I recommend:

http://radiospares-fr.rs-online.com/web/search/searchBrowseAction.html?method=getProduct&R=0402885

it's a single rail-to-rail 4us open drain comparator supporting up to 30mA sinking.

it's a bit expensive @ 3.74 euro, you could save up to 3 euro going for a less ideal device, but I guess 3 euro is no issue for a prototype. the other options are dual:

http://radiospares-fr.rs-online.com/web/search/searchBrowseAction.html?method=getProduct&R=0649475
http://radiospares-fr.rs-online.com/web/search/searchBrowseAction.html?method=getProduct&R=5343009
http://radiospares-fr.rs-online.com/web/search/searchBrowseAction.html?method=getProduct&R=0810239
http://radiospares-fr.rs-online.com/web/search/searchBrowseAction.html?method=getProduct&R=3564412

please verify that you're willing to go with the MCP6546-I/P and able to find it.

[mitch_781]

Hello Lanchon,

I found all the componants I need, I bought the TLC393CP because they didn't have the others in the shop.

I have one more question before I begin to build the schematic.

Are you sure that the IN- is connected to battery / 15 and the IN+ connected to the middle of the variable resistor ?

Because if we want to ground the throttle command when the battery is to low, it should be the opposite ?

if IN- < IN+ ==> output is 5V and we would like to have 0V I think to ground the throttle input ?

Do you agree ? (I put the schematic of my proposition)

Thank you if you can help me again.
Have a nice day

[Lanchon]

yes, I agree. you drew it wrong before, check out my original post: "use a 2 resistor divider to bring the batt voltage down to the 0-Vdd range (under all circumstances) and hook the middle tap to the positive comparator input. divide Vdd using a 15-25 turn trimpot (best if impedance at the tap is similar to the tap of the other divider when trimmed to the right cutoff point) and hook it to the neg input."

the assumption is that:
IN+ > IN- => output is high impedance
IN+ < IN- => output is low

however, this is just an assumption. in the datasheet this is not specified and there are numerous errors. for example, fig 19 and fig 21 both have a triangle oscillator at the left, and they both use the same circuit except for input polarity. needless to say, one of them does not work. fig 20 uses a strange comparator symbol, where the circle (depicting inversion) is drawn on the positive input... yeah, right.

I wouldn't bother to verify the assumption, but if you want to, check out the LM393 datasheet; this is the device the TLC393 tries improve upon.

now that you decided on a particular comparator, I'll tell you the values of the rest of the components soon. I'm sorry I can't do it right now. don't start building until then.

for the actual build, I recommend a perf board (check wikipedia), or dead bug style (wikipedia) and gluing the components with a hot plastic glue gun. you can use small gauge thermoplastic insulators to use resistors directly as interconnects. once you're sure everything works really fine, you can hot-glue it all for sturdiness. use an IC socket if you use a perf board. actually, it's such a simple circuit that you can really build it any way you want.

[mitch_781]

Hello Lanchon,

Sorry if I took a particular comparator, I thought all of them were possible with no influence on the shematic.

I have every thing ready to solder on a plastic plate according the last schematic I did.

 I don't understand the datasheet at all. I wait for your next post if you think there is modifications to do I will follow you.

Thank you in advance for your great work.
Have a nice day

[Lanchon]

here's the schematic and parts. two circuits, one has a led power-limiting indicator. the led should have a monostable to keep it fully on during limiting, but it doesn't and instead it will be pulse width modulated (brightness is "proportional to the amount of limiting"), so I don't like it very much. your choice.

several optional resistors marked as "zero ohms" beg explanation. R3 & R5 would increase the precision of R4, but that's not necessary. R7 helps protect the output of the comparator form static discharges, I'd include it. R8 helps prevent a latch-up, it's being a bit paranoid, but I'd include it if it's comfortable to do it during the building (ie: if I have some space for the resistor and easy access to it).

again: use a socket for the IC if you can.

happy building!