Source for hall effect wheel???

[Stomis]

I want a tach on my diesel. Dakota digital sells a translator box to turn a hall effect signal into a gas tack signal. They advertise it as running it off the teeth on the flywheel. I have open access to mount a hall effect "wheel", I guess you would call it with the flat tabbed teeth like on an ABS reluctor wheel, where the vacuum pump would normally go in front of the injection pump.

Is there a source to buy one or should I just whip one up in CAD and look for a c&c machine shop?

 

[Fordum]

I know there is a good reason, but I have to ask. Why not use the flywheel?

My diesel genset did not come with an autostart panel, but they drilled and tapped the bellhousing and screwed in a prox sensor.

The autostart panel counts the teeth going by and uses that to determine when the engine has cranked and turns the starter off.

Then it watches for overspeed and kills the electric fuel pump if it sees it.

The head of the sensor is smaller than the body, so the whole install was done from the outside.
Just drill the hole with a hand drill, thread the hole, and screw in the sensor until it touches the flywheel teeth.
Then back it out a couple of turns and tighten the locknut.

 

[Stomis]

I know there is a good reason, but I have to ask. Why not use the flywheel?

My diesel genset did not come with an autostart panel, but they drilled and tapped the bellhousing and screwed in a prox sensor.

The autostart panel counts the teeth going by and uses that to determine when the engine has cranked and turns the starter off.

Then it watches for overspeed and kills the electric fuel pump if it sees it.

The head of the sensor is smaller than the body, so the whole install was done from the outside.
Just drill the hole with a hand drill, thread the hole, and screw in the sensor until it touches the flywheel teeth.
Then back it out a couple of turns and tighten the locknut.

Honestly the main reason is to not have a $100 mail order sensor mounted under my truck where it will undoubtedly be submerged in water and mud. Plus I need to make a cover where the vacuum pump is meant to be anyway so I figure it would be nice to incorporate it there high and dry.

 

[Fordum]

I can certainly understand the mud and water part.
But two things come to mind.
First, you do not need a special wheel. Most any old gear with teeth or even a wooden disk with bolts mounted along the edge would get the job done.

Also, you might want to look around for sources for the sensor.
Like this place for instance........

http://www.automationdirect.com/adc...ximity_Switches/8mm_Round_(AE_-z-_PEW_Series)

 

[Stomis]

Hmmm it seems like the Dakota kit has you paying for the box more so than the sensor. I don't know what it takes to make a hall effect signal work with a tach if anything but you can fine tune the Dakota box to translate to the tach.

Ill have to do more research.

 

[Blue85]

I bought one from these guys and modified it to ride on my crank pulley: http://www.diyautotune.com/catalog/trigger-wheels-c-48.html?osCsid=8d0214c651c907a7779621eabe55510f

But this is for main ignition timing. So most wheels like this will have a "missing tooth" to give absolute crank position. For tach only you don't need that, but it shouldn't hurt anything for that little signal jitter each revolution. Your box just needs to know how many pulses per rev. I think some generators use a small trigger wheel to govern speed. You might also use a magneto trigger wheel from a small engine and add teeth as needed.

For a distributor-fed fed tach on a V8, there are just 4 pulses per engine revolution. If you could get a wheel with 4 teeth, you could use an off-the-shelf tachometer with just a simple circuit to convert the sensor pulses to 12V. Then you would also have the option of using a VR sensor, which can be scored at the junkyard for next to free or maybe $10 on eBay. It's possible with a strong magnetic signal the tach could read it directly. BUT...you could also use an HEI module to do this (it has a nice zero-crossing detector built in and amplifier. The module from the ECM trucks might be even better as the signal amp and coil driver are two seperate circuits).

I have a VR sensor mounted down low and it can take mud and water all day: http://coloradok5.com/forums/showpost.php?p=2417094&postcount=55

 

[Stomis]

Hmmm good site. I'm looking at the 3in spacer as the wheel itself and just drilling two more notches in it. That would give me 4ppm set.

Although I just realized this is going to be mounted infront of my IP. So that's cam timing so I would need 8ppm right?

 

[Stomis]

Wait that's wrong. V8 pulling 4pulses off the dizzy is the same relation as the cam / IP.

 

[Blue85]

If the tach is designed to feed from a distributor/single coil, it counts every 8 pulses as 2 revs (although it's not really "counting", it's an analog frequency to voltage converter, but all the scaling works the same). 2 revs is 1 cam rev, so a cam wheel needs to make 8 pulses. But if you're using a standard tach, you also have the option of flipping the switch to "4 cylinder". That means 4 cam pulses for each engine rev and your plan works.

 

[Stomis]

OK I got you. I was trying to figure that one out in my head. This is a huge help. I'd rather build it right and know its dead on then play with a calibration box.

4 pulse will be easier to replicate using one of those spacers that site sells. How crucial is perfect spacing/gap for a tach signal?

 

[Stomis]

Looks like Im going with this

And this.

Im going to mount the spacer through the center hole and grind two more reliefs in it. Since the sensor works on the collapse that should work fine no?

 

[Fordum]

Looks like Im going with this

And this.

Im going to mount the spacer through the center hole and grind two more reliefs in it. Since the sensor works on the collapse that should work fine no?

Uhh, no.
You are probably have to grind out the original slots too. Let me see if I can find some specs.....

 

[Fordum]

OK, found it in the FAQs section of that sensor I linked to. Here is the link to that section, http://support.automationdirect.com/faq/faq_group.php?product_id=24

And here is the one you need to see.

Question: Can a proximity sensor be used to determine RPM? Answer: Yes, subject to a number of factors. It must be connected to a high-speed counter capable of reading the pulse rate. The sensors switching frequency must be higher than the rate at which the rotating target is passing the sensor. The target must be present long enough for the sensor to switch a valid signal to the counter. The 'off' portion of the target should be about twice the 'on' time.

You might want to scroll through them all. They have some good information.

 

[Fordum]

Also, maybe I better clarify some. An inductive sensor generates a high frequency field.
When a piece of metal enters that field, it loads it down and it draws more current. The electronics detects that draw, and sends a signal to the reading equipment.

After the metal leaves that field, it takes a period of time for the field to rebuild. That is what determines the max switching frequency. The larger the time between metal, the more accurate and stable the setup is.

Those I linked to show the freq in the specs. 2-3 KHZ is common. So, you need to do the math to make sure the sensor can respond fast enough.
To make the math easy, 6000 RPM is 100 RPS, or HZ. 4 teeth on that wheel makes 4 pulses per rev. If it were mounted on the crank at that RPM, then the sensor would see 400 HZ.
If the crank pulley was 8 inches across, and the wheel were mounted on a belt driven pulley that was 2 inches across, then the sensor would see 1600HZ or 1.6 KHZ.
If the sensor maxes out at 2 KHZ, then you want to make sure it sees a definite difference between metal and no-metal.

 

[Stomis]

Also, maybe I better clarify some. An inductive sensor generates a high frequency field.
When a piece of metal enters that field, it loads it down and it draws more current. The electronics detects that draw, and sends a signal to the reading equipment.

After the metal leaves that field, it takes a period of time for the field to rebuild. That is what determines the max switching frequency. The larger the time between metal, the more accurate and stable the setup is.

Those I linked to show the freq in the specs. 2-3 KHZ is common. So, you need to do the math to make sure the sensor can respond fast enough.
To make the math easy, 6000 RPM is 100 RPS, or HZ. 4 teeth on that wheel makes 4 pulses per rev. If it were mounted on the crank at that RPM, then the sensor would see 400 HZ.
If the crank pulley was 8 inches across, and the wheel were mounted on a belt driven pulley that was 2 inches across, then the sensor would see 1600HZ or 1.6 KHZ.
If the sensor maxes out at 2 KHZ, then you want to make sure it sees a definite difference between metal and no-metal.

I get what you mean. Do my math to see if my gaps and the large solid area on the wheel will jive. Im going to have to give diyautotune a call. their screw in unit is 12v and shouldnt need an amp from my reading.

 

[Stomis]

Also, maybe I better clarify some. An inductive sensor generates a high frequency field.
When a piece of metal enters that field, it loads it down and it draws more current. The electronics detects that draw, and sends a signal to the reading equipment.

After the metal leaves that field, it takes a period of time for the field to rebuild. That is what determines the max switching frequency. The larger the time between metal, the more accurate and stable the setup is.

Those I linked to show the freq in the specs. 2-3 KHZ is common. So, you need to do the math to make sure the sensor can respond fast enough.
To make the math easy, 6000 RPM is 100 RPS, or HZ. 4 teeth on that wheel makes 4 pulses per rev. If it were mounted on the crank at that RPM, then the sensor would see 400 HZ.
If the crank pulley was 8 inches across, and the wheel were mounted on a belt driven pulley that was 2 inches across, then the sensor would see 1600HZ or 1.6 KHZ.
If the sensor maxes out at 2 KHZ, then you want to make sure it sees a definite difference between metal and no-metal.

Well we're talking about mounting on the injection pump sprocket basically. The stainless adjustable sensor I showed has a max of 15khz. Tach has a 0-6000 rpm range. Like you said at 6000rpm ie 100rps is 4ppr so 400hz.

I guess the real question is what do I need for a collapsed distance to operate accurately at 400hz and below?

 

[Blue85]

I don't like the slot setup, but that wheel could still work. Just grind enough away so there are only 4 teeth sticking out (see attached). The rest should be like 1/2" or more further from the main body of the wheel.

Or you could just weld 4 little pieces of steel on as teeth. I've heard of people drilling and tapping the edge of an aluminum wheel and then threading bolts in.


This is all assuming you just want to see engine speed. The ECM must already know it, so another approach is to tap into the diagnostic bus and have some kind of display.

 

[Blue85]

I assume you want that hall effect sensor because it gives square waves inherently. The site doesn't say whether that is 5V or 12V, but a tach would probably read either. I'm also wondering if the "tooth vs. gap" thing really matters with the hall. I think Fordum and I were both assuming VR. EDIT: Yes, he said "inductive", which is basically the same thing as VR in this case.

 

[Stomis]

I may have a slight issue with where I was going to place this. Possibly may be reverting to the flywheel route. I've got to find out if the vacuum pump putting pressure on the injection pump is necessary due to slop in the from of the IP shaft.

 

[DEMON44]

What engine are you talking about? In what chassis?