help needed for 50ms interrupt circuit

[fechter]

Thanks for drawing that up Rick.

If you use an ICM7556 dual timer, you don't need capacitors on the control voltage pins (fewer parts). The ICM7556 draws quite a bit less current also.
I'd suggest increasing R1 to 10k to reduce current draw. The 5v supply from a controller or CA probably doesn't have a whole lot of excess capacity.

Now I think he would like a board layout :wink:

 

[caleb7777]

I understand that and thank you, I was worried that other circuit would not perform both ways.

The grounds do not confuse me. It appears you have 2 speed input lines. Is one only ground which one goes to my trigger line?

Top to trigger line and bottom ignore?

 

[caleb7777]

ah damn.... I see. I am not used to these circuit conventions. The switch basically is the entire CA. Common will come from somewhere else on bike for simplicity.

View attachment 1

Or better yet, If I understand this is same thing:

 

[caleb7777]

This would work for the optoisolator if i went that route?

https://www.alliedelec.com/Search/ProductDetail.aspx?SKU=4310093&MPN=LH1546AD

i could replace both relays with these. Or would you recommend transistors for either?

Both relays must handle 48v up to approx 100 mA, but I don't think the ebrake line would require much current, I just don't know a number for the ebrake line on infineon.
I hope this will be a reference for everyone else though so it should be built for everyones' bike handling 100v easy. These optos are rated at 350v and 120mA so should be fine I think.

No coils to worry about either. Am I right that this would be better than using reed relays?

 

[caleb7777]

I just realized there is direct short through some of these resistors and that the trigger stays low. Is this ok on the 556? I thought the trigger had to be a pulse shorter than the desired output. The switch will remain closed for long periods. This OK?

 

[rkosiorek]

as far as i remeber it should trigger a one shot pulse regardless of how long the trigger is held low.

each LH1546AD requires a minimum of 50mA to drive the led and switch on. a pair would likely exceed the current that can be supplied by the 5V line of the controller. never mind the additional load of the timer circuit and by gosh the throttle as well.

no i figure regular plain old transistor output opto's at 15 to 20mA are gonna be pretty much the limit. that is unless an alternate source is found for the 5V supply.

rick

 

[methods]

You can do an el-cheap-o 5V regulator similar to the Infineon design to supplement the controller 5V if you need that much current.
It will be hot and inefficient as hell but it will only cost you a few bucks.
The idea is to use a linear regulator and throw a big 5W power resistor in-line with it to the main battery power.

There are tons of designs for simple 5V regulators on the web.

-methods

 

[caleb7777]

Thanks guys.

Interesting, I didn't think that this would overload the 5v out of the controller. I wish i knew what the infineon could output on its 5v bus. I was thinking of using it later to drive a headlight or something that may be in the 4.5v range. Pretty much anything that runs off 3x 1.5v batts could be hardwired to bike. I am curious what current the infineon 5v out could do.

 

[methods]

It is a very poor regulator.
Attach a 1k potentiometer and start cranking up the load.
Watch the voltage, when it drops to 4.75V you are past the max.

Measure the pot resistance and divide that into the 4.75 to calculate your max current.
Or just measure the current if you have an ammeter and a volt meter.

-methods

 

[rkosiorek]

by reading the various threads for the infineon look like one of the most common issues is how to step down the battery voltage to the +5V required. the stock system uses a bunch of power resistors and an LM317 followed by a LM7805. both are in TO220 cases. the current limit does ot seem to be with the regulator chips themselves but with the power handling ability and heat generated by the dropping resistors. anyways it looks like the limit for the 5V is about 150mA total for the entire regulator including whatever external circuitry you intend to use.

this may mean that for any type of fancy add on circuit a separate 5V logic supply MAY have to be separatetly provided. but until some one builds one, tries it and finds out for sure all of this is speculation. the debate can only be settled by experimentation. if it works fine. if something smokes, your were warned of the possibility, look for only understanding but no sympathy here. education always costs money, one way or another.

as they used to say in those WWII movies "the Smoking Lamp Is Lit"

rick

 

[caleb7777]

Great info thanks.

So one may be able mod the resistors on the infineon to provide a more powerful supply?

 

[caleb7777]

It seems I have been a fool here for a while. Originally I was going to cut the throttle directly and then I got the idea that this circuit could be used to simply pull down the ebrake line. But that won't work for anyone using regen! it only activate their regen and push current thru the motor anyways. Throttle will have to be disabled directly.

If the +5v to throttle is interrupted, it will send a throttle output of 0 yes? or no?

 

[fechter]

If the +5v to throttle is interrupted, it will send a throttle output of 0 yes? or no?

Yes. Interrupting the 5v to the throttle will do about the same thing as pulling down on the throttle line, but with less chance of frying the throttle.

It might be better to find a small 5v switching power supply (wall wart) that runs on 48v or whatever your pack voltage is.

 

[caleb7777]

Frying the throttle? by pulling down on ttrottle or by interrupting 5v+?

 

[caleb7777]

Another question,

The ground for this circuit is essentially fed from the CA. So even if I used an external 5v regulator off the battery it would protect the controller but I could still fry the Cycle Analyst as it can only handle about 100mA on that throttle override line.

Am I right in this? What would be the simplest way to use the ground as only a trigger and have no load on that line. If the common ground for the circuit came from elsewhere and only the trigger came from CA then it would be easier on it because the current would not flow thru it yes? ie: if the ground on the 555 came direct from batt, less load to CA.

I will draw it up different when I get home to show what I mean. But I don't know if I could use different ground source for reed relay 1 powering the delta switch.
I am sure there is some simple way.

 

[caleb7777]

Here are the two circuits I will try with different ground and hot source, not really different ground as whole bike shares common ground, the main difference would be the route taken. I want to keep current away from switch 1 as this is really the CA and can not handle much load.

Would this work and help?

 

[caleb7777]

It is a very poor regulator.
Attach a 1k potentiometer and start cranking up the load.
Watch the voltage, when it drops to 4.75V you are past the max.

Measure the pot resistance and divide that into the 4.75 to calculate your max current.
Or just measure the current if you have an ammeter and a volt meter.

-methods

How exactly would I do this? When I get the battrey i would like to do this just to know, it would be interesting. I am afraid of frying it though. If you can give me specific instructions I will.

Thanks!

 

[caleb7777]

I almost have this figured out guys. i just need a little more help regarding the last few posts.

 

[caleb7777]

Throttle interrupt circuit is a success!
This will be used to cut the throttle when delta relays switch over so that there is no load on controller during switch when using an automatic speed switch
This is the test board, green led is delta relay on or off and red shows momentary power either way which will be used cut throttle.

This is bernard's circuit using one 555 timer. I am not playing favorites I just don't have a 556 yet.

http://endless-sphere.com/forums/viewtopic.php?f=2&t=10209&start=0


[youtube]2H65VIhsBds[/youtube]

 

[caleb7777]

Only other thing that could be incorporated into this would be a slight delay on the delta relay side so that throttle is fully disengaged before delta relays trigger. I am guessing just 5ms 10ms max. I know this can be done simply capacitor. how would I implement it for relay 1 (top left in circuit)

Only want 5-10ms delay before relay closes

 

[caleb7777]

A delay for when the relay closes would be easy, but realistically we want it both ways, so that when power removed the throttle cut out before the realy released again hmmm ideas?

 

[rkosiorek]

A delay for when the relay closes would be easy, but realistically we want it both ways, so that when power removed the throttle cut out before the realy released again hmmm ideas?

exactly why the 556 is used with the inverter. it has 2 timers. one for switch on and the other for switch off.

i agree, it would also be nice to hane the delta/star relay close or open only after the throttle line is pulled down though. that would require a timing capacitor parallel to the relay coil and a series resistor. most likely need to figure that out experimentally though. to calculate it you would need to know the inductance of the relay coil.

remember also that the cap across the relay coil will not only delay the turn on. it will also delay the turn off. the relay will stay on until the capacitor discharges through the coil sufficiently to release the relay.

rick

 

[caleb7777]

yes
I am working on it here

http://forum.allaboutcircuits.com/showthread.php?p=141698#post141698

 

[caleb7777]

Well I got to here and I am trying to test it with the 2 timing constants at 1000x greater so that I can see operation in sec instead of ms but My power supply couldn't handle it. I gimmicked up a 120v to USB adapter for power and it doesn't work well. I will try it when i get my LM7805 so I can use a better 12v source

 

[caleb7777]

ugh.
I need more help here guys, I got the new voltage regulator LM7805 and installed it and just used a generic 12v dc supply and when I turned it on the transistor 'T3' smoked right away and that was that.
I added different capacitors and resistors to get 1000x greater timing constants so that i can see what it will do and time it out.

C3 is 20000 uF in test rig
timing constant of ~10ms with 500 ohm relay coil

R1 is 2.1 Kilo ohm
C1 is .047 F
timing constant of ~100ms for 555 timer

I am guessing that the 20,000 uF cap drew a lot of power initially and fried the transistor.
but I am guessing here.

the transistor was a 2N4403

I most likely have the capacitor relay circuit set up wrong. is there something missing from this circuit? something wrong? please help
The only purpose of the cap C3 is to delay the opening and and the closing of the relay, I don't absolutely need it but it would be very preferred. Did I not wire it proper? Do I need more resistance?