The thing Im most proud of so far/ switchin bike with relays

[auraslip]

I've wanted to do this for a year or so, and finally got around to it on my newest build.

All activated by a single keyswitch!

3 cell logs, 2 temp sensors (one for motor and one for controller), and on back lit clock. The two button pcb beneath the clock is to set the clock.

Box was made from balsa wood.
Had to use a pcb board from radio shack because the pins on the 12v relays looked a little bot flimsy.

Bottom is plastic sheeting from online metals.com
I'm a little worried about water proofing here. The hole isn't small enough to be sealed by the tube. Need to figure something out.
The top is to be made from clear plastic from lowes.

Already a rats nest and only half the battery pack is installed

Here is a huge tip which I wish I could follow, but make the wiring harnesses last so you won't have extra wiring to deal with!
I swear though there is like 1000ft of wiring on this bike. It makes me really realize how complex the wiring must be in automobiles. I had a temp job at a place that specilized in making harnesses for military aircraft. And that was it. They had hundreds of people on the floor meticulously assembling the harnesses on giant boards with the proper layout pre-printed on paper tacked to the boards. Wow, blows my mind. You could wire a residential block with the wiring in one plane.

Ebikes are easy. Ebikes with lights, chargers, and electronics are a whole 'nother ball game

 

[DAND214]

I like it!

How big is it?

Dan

 

[auraslip]

6"x7"x3"

 

[auraslip]

First problem. The controller enable line is shorted! I've traced the short back to relay.

http://www.mouser.com/Search/ProductDetail.aspx?R=V23105A5503A201virtualkey65520000virtualkey655-V23105A5503A201

The spec sheet says the relay can handle up to 200vdc, so I don't know what could be wrong?

Could the spark caused by the connection be the cause?

 

[texaspyro]

They had hundreds of people on the floor meticulously assembling the harnesses on giant boards with the proper layout pre-printed on paper tacked to the boards.

All women, I bet... dudes can't do that...

 

[hjns]

Nice work. But it is BIG. Can't you reduce the size a bit? And how did you plan for all the wires?

I am trying to do plan for a full rewire of my bike, in such a way that I can easily troubleshoot and/or replace a component if necessary. Planning for all the wires is now really a challenge for me....

 

[Tom L]

First problem. The controller enable line is shorted! I've traced the short back to relay.

http://www.mouser.com/Search/ProductDetail.aspx?R=V23105A5503A201virtualkey65520000virtualkey655-V23105A5503A201

The spec sheet says the relay can handle up to 200vdc, so I don't know what could be wrong?

Could the spark caused by the connection be the cause?

How much current is the relay switching? Looking at the data sheet it's only good for breaking 0.55A @ 200VDC. If it's more than that, then yes, you could have welded the contacts.

What's more is that the the UL rating for switching 200V is only 0.15A (inferring from text and 30W load line on the graph).

 

[auraslip]

Thanks Tom. I'm not sure what the load is on these controllers 5v bus. It should be low, but perhaps the initial charging of the caps is enough to weld it shut.


The box really can't be made that much smaller or you won't be able to fit the components.

 

[Tom L]

If it's only 5VDC you are switching then the relay is good for breaking 3 Amps.

 

[auraslip]

I'm switching 22s so 90v fully charged.

 

[Tom L]

Only good for 0.8A @ 90V. So if you have any sort of inrush current charging capacitors then you're in trouble.

 

[auraslip]

The caps are very small on the 5v buss, but I guess the in rush current is more than .8a. Certainly enough for a spark when plugged in.
It's odd because a standard 12v 5a keyswitch held up for a year in the function. Much beefier looking though.

I guess I need to find a bigger 12v relay! Radioshack has one for $5. Hopefully it'll work.

 

[Tom L]

You just said you were switching 90v not 5v. I'm confused.

 

[nonlineartom]

mounting your LiPo width ways, This means the frame box is atleast 175mm wide right? interfere with pedalling at all, feel like you are straddling a whale?

 

[auraslip]

You just said you were switching 90v not 5v. I'm confused.

The relay switches the enable line on the controller. The main leads are plugged in, but the controller won't function with out power to the voltage regulator that powers the 5v and 12v buss. I've decided this was much easier than manually plugging in the main power leads and using a precharge relay every time.

mounting your LiPo width ways, This means the frame box is atleast 175mm wide right? interfere with pedalling at all, feel like you are straddling a whale?

box is 6" wide. I have an extra wide BB from sickbikeparts. It allows the pedals to clear. It's not perfect, but you can pedal it just fine.

 

[nonlineartom]

That could store a LOT of Lipo that way, ever considered storing them length ways as groups of 3, any advantage?

 

[auraslip]

Storing them on their side helps a lot with wiring them up. Makes it much easier to service them too.

 

[auraslip]

Got a new 10a relay from radio shack. It's worked for a dozen cycles now even though it's only rated at 28vdc. I'm still gonna carry the jumper in my tool kit in case it fails. thanks for the help everyone.

 

[auraslip]

Well that didn't last long. Guess I need a beefier relay. What a PITA. Can anyone recommend one?

 

[Tom L]

There's something wrong. Your first relay should have been more than adequate if you are only switching the controller's enable line.

 

[auraslip]

actually I think it was only a 5a relay....
the spark is pretty significant when I manually make the connection, so I'm sure it could be welding it closed.

I'm thinking now maybe I could have a high value resistor always in place to across the enable line. enough to keep the caps charged but not enough to power up the bike. this is how some people recommend a precharge resistor should be done. how would I calculate the value of the resistor I needed?

 

[auraslip]

Ok, I'm looking at this stuff trying to figure it out and I don't know where to start.

Here is the diagram. So the first cap gets charged immediately. At 90v with zero resistance it sees 90a for a split second. That's why the relays are frying.

So, what I'm thinking is that I need to supply enough current to this setup so that the caps can stay charged but not enough for the 7805 to function. The data sheet says the 7805 quisenct current is 5ma. So lets say we want 10ma to account for that standby current and a little extra to keep the caps charged. So ohms law says (90v/.010a)there should be a 9000k ohm or resistance (including the base and switched resistors) to drop the current down to 10ma. Then I should be able to use a a tiny relay.

Thoughts?

 

[auraslip]

OK, so the resistor thing isn't going to work! I tried it with a 10k resistor, and the resistor got very hot and the setup was stilling putting out 2.5v on the 5v rail. Forgot ohms law to calculate power draw across the resistor. Almost 1w!

So I tried a 20k resistor. Which is half a watt of power across it still! And there was a big spark!

I may need a time delay relay!

edit: actually I guess I'll build something like this using a 555

 

[amberwolf]

If you are only switching the enable line, there should be no spark, as the caps will already be charged. :?

If you are switching the main power, you need a relay capable of the full current of the whole system under max load.

 

[auraslip]

If you are only switching the enable line, there should be no spark, as the caps will already be charged. :?

you'd think that.... but in fact there is quite a big spark. The PSU on this board draws a bit of power, and the caps aren't tiny.