Help with diode form and selection...(OT)

[abn]

I'm trying to wire in a permanent magnet gear reduction starter. I would like to power the whole thing with one fat cable. I tried to jumper the starter motor terminal and the starter activation solenoid..but I found out that the back EMF from the permanent magnet motor energizes the starter solenoid and causes it to stay engaged with the flywheel.

I'm assuming that as the motor becomes a generator and sends current via the jumper wire to the solenoid that polarity is reversed...so an in line diode should solve my problems right???

I pulled apart some old alternators to get the diodes...but they conduct through the chassis and a heat sink which makes them hard to connect into my simple jumper wire system. Can someone help me select a diode or compact assembly that I can maybe solder or connect in line with the jumper wire and shrink tube over for a simple but robust install?

Thanks for any help!!

[CCWKen]

Whuuuuuut?

What's the application?

What leads on the solenoid are you jumping? S to B+? You need a PB or ignition switch between S and B+. When the switch breaks contact, there's no way for the starter to backfeed the solenoid. You have a wiring problem.

[J Tiers]

yes, back EMF polarity is opposing current, that is how motors work.

I'm not totally clear on what you are doing, on what sort of motor........

Yes, a diode (heavy duty rectifier) would do the job, likely, but I'd add a parallel resistor to avoid an unlimited voltage generated... that could "pop" the diode.

You want a diode rated for sufficient current to run the starter.... Almost all of the suitable ones will probably be inconvenient to mount., it goes with the territory on high current. There are a few with a plastic and metal flanged package that have screw terminals, but they are less common now.

CCWKEN..... his issue is that a PM motor generates a back EMF after power is disconnected, enough to hold the solenoid in too long, while a series starter motor won't usually give enough to hold the solenoid in. This is because he is NOT using the high current contacts in teh solenoid....

Thing of it is, the high current switch and solenoid are the same in most standard starters, so its "on" until the solenoid magnet drops out, at which time its "off" ... and the problem doesn't occur.

Seems like it might be much easier to use a solenoid with a switch in it for the high current, if you can.

[darryl]

The inline diode will work, but it needs to be a very high current one, and it will get hot even after only a few seconds. Because there is a voltage drop through the diode, it will also reduce power to the motor, especially when it's cold and the amp draw is higher. I would be tempted to go with a diode rated for at least twice the maximum current the starter will draw (worst case scenario, winter conditions probably) and mount it on a heatsink or just a block of aluminum. The aluminum block will soak up the heat from the diode and keep it within operational limits for a short time, so under normal circumstances a standard finned heatsink might not be required.
My thinking is it would be better to run a high current switch to handle the motor current, and have that switch and the solenoid controlled from a start switch, whatever that is in your case. Might be a part of the ignition switch as usual, or a separate push button switch. This means the motor wire and the solenoid wire have to be separable at the motor, and an extra wire run. I think you'd be happier doing it this way, even though it's more wire and parts to start with. The first time you make it through a borderline start situation, you'll be glad you didn't use the diode.

[Evan]

If you can find a few dead ATX computer power supplies you can salvage usually at least one dual schottky diode from each one. Usual rating is about 20 amps per diode at 40 volts. The unit is a dual diode in the same package and they can be paralled to give 40 amps per package. About 4 of these in parallel should give a surge rating over 1000 amps. Schottky diodes have very low forward voltage drop.

Try a small local computer store like mine. I'm keeping my grandsons busy by having them scrap out the parts in about 50 dead power supplies.

[abn]

Maybe I should just explain what I'm trying to do...there has to be an existing circuit for it.

I've got a trunk mounted battery and a Ford type solenoid also in the trunk...I'm trying to get the simplest, cleanest installation where where all power cables are fuse protected, except the starter cable, and it is only hot when the key is turned to start. To avoid a constant hot battery cable running the length of the car. This was easy with the original 4.5" old style starter with one wire and a Bendix to engage the starter with the flywheel, but now I'm using a new PMGR "mini" starter for header clearance...this requires about 30amps to operate a separate solenoid on the starter itself that engages the starter pinion with the flywheel. I tried a jumper as stated in my original post but ended up with the back emf problem...now I'm using a separate relay at the front of the car, triggered by the ignition swith, to power the solenoid on the starter itself and trigger the remote solenoid in the trunk. I would like to be able to get rid of the second circuit and just have one trigger wire going to the trunk and one wire going from the trunk mounted solenoid to the starter area, and a diode protected jumper to the solenoid on the starter itself. Darn now it sounds even more complicated.

[abn]

What exactly is a Shottky anyway? I just threw away a dead ATX power supply, I still have a selection of dead and untested commercial power supplies and some Tripp-Lite battery backup power supplies. I was hoping for something sort of OEM or at least with a Digikey number in case it fails later. Someone on corner-carvers.com is supposed to have a circuit for what I'm trying to do maybe I'll ditch the diode idea.

[darryl]

A schottky diode differs from silicon diodes in the construction and materials of the reverse blocking conduction path. This gives it a lower forward voltage drop and speeds up it's action, important in low voltage output and high frequency switching power supplies. The lower forward voltage drop is useful when the voltage source is low, such as only 12 volts in a vehicle, and where currents will be high. The schottky diode will heat less than the standard type silicon diode because less power is lost in the diode due to the lower voltage drop. The high speed operation aspect of a schottky diode is of no value in this potential application, but it doesn't matter.

For the starter- if you want only a single wire going to the starter, the diode will have to be inline with the starter current, not the solenoid. If it's inline with the solenoid, it won't block voltage from the motor from reaching the solenoid, but it will reduce that voltage by a half a volt or so if using a schottky diode, or by about a volt if using a standard diode. That diode will also reduce the voltage from the ignition switch to the solenoid, so if the battery starts going dead, the solenoid will start clattering a little earlier.
In any event, using a diode inline with the motor is going to require two wires leading to the starter anyway, whether they come from the trunk or from an area inside the engine compartment where the diode can be mounted. If it's in the engine compartment, only one line need come from the battery to that point, but then it must branch to send two wires to the starter. I don't think it's any simpler than using the relays and solenoids as you already are, and you'd be introducing a voltage loss to the starter with the diode. You'll always need two wires to the starter either way, so it would be better for the diode to not be in the hot engine compartment. The only way around the extra wiring is if you put the diode right on the starter itself, which is going to be a hot zone, and a wet zone, which the diode won't like and nor will any extra connections you have to make there.

[J Tiers]

<font face="Verdana, Arial" size="2">Originally posted by darryl:
If it's inline with the solenoid, it won't block voltage from the motor from reaching the solenoid, but it will reduce that voltage by a half a volt or so if using a schottky diode, or by about a volt if using a standard diode. </font>

More likely, at the high currents, it would be almost a volt for the schottky, and nearly a volt and a half for the standard rectifier. They have resistive losses as well as the forward voltage issue. Even more if teh rectifiers you use are marginal for the current.

I'd bite the bullet and use the "extra" relay. more juice to the starter, more of a "big hammer" solution, which is appropriate for that type useage.

[winchman]

Here's what you need to do.

Put the standard solenoid in the trunk next to the battery.
Run the big cable from it to the big terminal on the starter.
Connect the wire from the ignition switch "start" terminal to the solenoid on the starter as it would normally be.
Run a jumper wire from the solenoid mounted on the starter (or from the ignition switch "start" terminal) back to the coil terminal on the solenoid in the trunk.

When you turn the key to "start", the solenoid on the starter will close. At the same time, the solenoid in the trunk will close, making the big cable "hot". When the engine has started and the key is moved off the "start" position, the solenoid on the starter and the solenoid in the trunk will open.

You won't need a diode, all the small wires will be fuse-protected, and the big wire will be disconnected from the battery by the solenoid in the trunk.

The jumper to the solenoid in the trunk can be pretty small (probably 16ga.) since those stand-alone solenoids only draw about three amps.

The small wire could be secured to the larger cable with short pieces of shrink tubing of the same color so it would hardly show.

Roger



[This message has been edited by winchman (edited 07-29-2005).]

[jimmstruk]

Yes!!! exactly what winchman said.Jim

[Paul Alciatore]

If you still want to use a diode:

The stud mounted diodes are that way for a reason. A high current combined with the forward voltage drop in the diode winds up creating a lot of heat inside the diode. It has to be disappated or the diode will melt or vaporize. 100 Amps x 0.7 Volts = 70 Watts and that's a lot of heat inside a small package. A heat sink is really needed at these current levels. All diodes over 5 or 10 amps absolutely need and will have some provision for a heat sink.

Most such diodes come with insulating washers for mounting them. A thin mica, teflon, or other such insulating material washer is used under the diode to insulate it's mounting surface from the heat sink. Sone are even aluminum with a black anodized coating. There is also a shoulder washer, usually nylon or other high temperature plastic, that is used under the nut on the other side to insulate it from the heatsink and ensure that the stud does not touch it inside the hole. The hole is drilled larger than the stud size to allow for the shoulder washer's diameter. If the diodes do not come with these washers, you can buy them separately from electronic supply places.

You should get some electronic heat sink compound to coat both sides of the washer under the base of the diode for maximum heat transfer.

These washers will allow the diode to be completely isolated from the heat sink.

Paul A.

[CCWKen]

Use the solenoid? I think that's what I said in the beginning. ???

In-line diode? If this is the type of starter I'm thinking about, it will draw 100A+. A regular starter can draw 200A+.

[J Tiers]

<font face="Verdana, Arial" size="2">Originally posted by CCWKen:
Use the solenoid? I think that's what I said in the beginning. ???

In-line diode? If this is the type of starter I'm thinking about, it will draw 100A+. A regular starter can draw 200A+. </font>

According to my meter, about 400A or more for a 4 cyl at -25 F, meter was wiggling so much I don't know how accurate it was (compression peaked it).

[darryl]

Right you are, J Tiers, the voltage drop across the diode will rise as current through it rises. Your figures sound about right. Still better to use schottky diode instead of regular diode, but still not good anyway to use a diode in this application.