[Frankencar]

So I'm not sure if there is a thread for this already...I didn't find it. I was wondering what/if people use to measure motor RPM? Seems like it would be useful, especially given that many conversions done use the existing manual transmission, I would think it would be helpful for maintaining rpm when shifting. Or is this simply not an issue? Any thoughts would be appreciated.

And yes, sorry if it is an easy answer. I'm not a car guy but I'm building an EV anyway!

 

[cbliss]

On of my favorite devices can be installed to show RPM if you can put a permanent magnet on the motor shaft somewhere. That would be Cycle Analyst from www.eBikes.ca It will show just about everything else you need, except if you use the speed input for rpm, you would loose some of its capability. For cars, you would need to make sure you order the high voltage version. I have a couple of them.

 

[Yukon_Shane]

Alot of the folks here who have a strong electronics background seem to recommend using an industrial proximity sensor to measure RPM; however, I think that this "Net Gain" speed sensor is probably the most common/off the shelf product: http://www.ngcontrols.com/speedsensor.html

Shane

 

[cruisin]

So I'm not sure if there is a thread for this already...I didn't find it. I was wondering what/if people use to measure motor RPM? Seems like it would be useful, especially given that many conversions done use the existing manual transmission, I would think it would be helpful for maintaining rpm when shifting. Or is this simply not an issue? Any thoughts would be appreciated.

And yes, sorry if it is an easy answer. I'm not a car guy but I'm building an EV anyway!

The AC-50 motor and controller comes with the spyglass that provides RPM, Amp, Volts, etc. at no extra charge.

 

[Frankencar]

Alot of the folks here who have a strong electronics background seem to recommend using an industrial proximity sensor to measure RPM; however, I think that this "Net Gain" speed sensor is probably the most common/off the shelf product: http://www.ngcontrols.com/speedsensor.html

Shane

That looks like an awesome device. I like it. Is it compatible with the FB1-4001? From what I see, my motor has a tailshaft of 0.750", according to KTA, but the smallest "exciter ring" is 0.875"

 

[Yukon_Shane]

That looks like an awesome device. I like it. Is it compatible with the FB1-4001? From what I see, my motor has a tailshaft of 0.750", according to KTA, but the smallest "exciter ring" is 0.875"

I'm not sure. My understanding is that the product isn't actually made by "Net Gain" so I wouldn't be suprised if they had a verison for the FB1-4001. If not I'm sure you could adapt it. I have one on my Warp 9 and it seems to have a fair amount of adjustability.

 

[gdirwin]

I needed 2 speed sensors:

I use the Netgain sensor to feed my Zilla over-speed detection (note there are different magnet patterns on the shaft - Zilla needed a pure 50% duty cycle so magnets were wider).

For the car tach, I kept the OEM/Toyota sensor, which created signals from a 36 tooth wheel (with 2 teeth removed) which was attached to the gas engine motor shaft. I mounted the sensor into my aluminum transmission to motor mount, and welded the toothed wheel to my motor adaptor/shaft.

 

[azdeltawye]

I needed 2 speed sensors:
....

I needed two sensors also for my 2001 VW Jetta w/ CAN bus:

One sensor for the motor controller (Soliton1) which gives two pulses per turn and uses a standard ABS wheel sensor with a couple bolt heads on the motor fan.

The other sensor was for the OEM tachometer which gives 60-2 (60 tooth wheel with two teeth missing) pulse train to the engine computer.

 

[mizlplix]

I wanted an "old school" Sun tach on my steering column, so I got one and set it to 4 cylinder mode. (that uses 2 pulses per rev.) I now need to make a simple aluminum hub for the 7/8" aux shaft on my AC50 that has two 3/8" set screws in it to give a halls effect sensor something to read.

In theory, it is quite simple compared to other guys problems making the existing in-car tach work...

Miz

 

[Gozer]

Miz

Do the set screws need to be magnetic?

Similar challange

Thanks

George

 

[mizlplix]

No. Just anything a magnet will attach to. An aluminum hub is perfect and two poles of steel. A wheel speed sensor or a crankshaft angle sensor will work to signal the tach.

 

[cbliss]

I wanted an "old school" Sun tach on my steering column, so I got one and set it to 4 cylinder mode. (that uses 2 pulses per rev.) I now need to make a simple aluminum hub for the 7/8" aux shaft on my AC50 that has two 3/8" set screws in it to give a halls effect sensor something to read.

In theory, it is quite simple compared to other guys problems making the existing in-car tach work...

Miz

I tried hooking up my Sun Tach to hall sensors and it didn't work. It seems like it should and I would like to know what I did wrong. 12V+/- to power hall, and the switched 'ground' from hall (hall low) to the gage. I am pretty sure I tried it with a pull up resistor.

 

[mizlplix]

Sorry if I made it sound so simple, but I skipped some things.

If you have an OLD generation 1 SUN tach that requires a huge box sending unit, Sorry, I dont know how to do them.

If you have the second generation SUN (or any other) tach, that reads distributor dwell pulses, I can.

They use a grounding type signal, not an electrical pulse like a sensor generates.. They can not directly read a Halls effect sensor. The signal must be converted.

Years ago I did one for a diesel engine using a GM HEI module as an interface. The crank angle sensor triggered the GM module and the tach read the module like a normal system. (MSD also has 8910 & 8920 interface modules to do this, but they are pricy.)

I can not say but some variation of this same set up MIGHT make the older SUN transmitter work too.

EDIT: This link shows the two type of tachs and how they differ in usage. http://www.msdignition.com/page.aspx?id=3300 Current VS voltage type triggers. It might help you some.


Miz

 

[dtbaker]

the super low tech solution is to just use stickers or grease pencil to mark 'redline' speeds on the speedo. easy to calculate using tranny gear ratios and tire size.

 

[Dave Koller]

And of course my 2 cents

http://www.diyelectriccar.com/forums/showpost.php?p=151836&postcount=37

 

[m38mike]

I mounted a 2" Westach tach in my EV. They sold me a magnetic sensor and a plastic collar for my output shaft. They were easy to install and worked perfectly right out of the box. The plastic collar has 2 magnets imbedded in it. The metal plate behind the pickup sensor helps to cut down the EMI from the motor. This is on the small shaft on a WarP11.

 

[notailpipe]

How hard would it be to convert one signal to another?

For example, I'm using the RPM sensor that Manzanita sells because it was specifically designed to work for the Zilla for overspeed warning. What would be the easiest way to take this signal and convert it to a suitable signal that the original tachometer takes? i.e. electrically instead of mechanically (two sensors). I stupidly did not keep my original RPM sensor from the distributor.

 

[TigerNut]

How hard would it be to convert one signal to another?

For example, I'm using the RPM sensor that Manzanita sells because it was specifically designed to work for the Zilla for overspeed warning. What would be the easiest way to take this signal and convert it to a suitable signal that the original tachometer takes? i.e. electrically instead of mechanically (two sensors). I stupidly did not keep my original RPM sensor from the distributor.

Your OEM tach probably expects one pulse per ignition event. On a four-stroke engine you get half as many ignition events per revolution as you have cylinders, eg. on a four-cylinder you get two pulses per revolution.

Depending on how the Manzanita sensor is made (and what the Zilla expects) you may only get one pulse per revolution. While it's easy to divide a signal down (i.e. turn two pulses per revolution into one pulse per rev), it's tricky to go the other way, especially if the incoming pulse train is narrow (which it probably is, if it's generated by a magnet or Hall sensor) and if the RPM range is broad (100 to 5000 RPM is a span of 50 to 1, very broad).

So, while it's nice to have a trick electronic solution, it's pragmatic and a lot easier to avoid unexpected issues, to have two sensors, each optimized for the task at hand.

The other option is to create a single sensor that puts out two pulses per rev to drive your tach, and then divide it down to drive the Zilla.


Note that if your sensor is generating two pulses per rev, then using that to drive your tach might be as simple as wiring in a buffer transistor (to keep the tach input from contaminating the Zilla's input expectations) and a pullup resistor.

Hope that helps... if you can provide more details on what your system has then I can give you more specific pointers.

 

[ElectriCar]

Here's a thread I did after installing an aftermarket tach. Lots of details.

 

[notailpipe]

Your OEM tach probably expects one pulse per ignition event. On a four-stroke engine you get half as many ignition events per revolution as you have cylinders, eg. on a four-cylinder you get two pulses per revolution.

Depending on how the Manzanita sensor is made (and what the Zilla expects) you may only get one pulse per revolution. While it's easy to divide a signal down (i.e. turn two pulses per revolution into one pulse per rev), it's tricky to go the other way, especially if the incoming pulse train is narrow (which it probably is, if it's generated by a magnet or Hall sensor) and if the RPM range is broad (100 to 5000 RPM is a span of 50 to 1, very broad).

So, while it's nice to have a trick electronic solution, it's pragmatic and a lot easier to avoid unexpected issues, to have two sensors, each optimized for the task at hand.

The other option is to create a single sensor that puts out two pulses per rev to drive your tach, and then divide it down to drive the Zilla.


Note that if your sensor is generating two pulses per rev, then using that to drive your tach might be as simple as wiring in a buffer transistor (to keep the tach input from contaminating the Zilla's input expectations) and a pullup resistor.

Hope that helps... if you can provide more details on what your system has then I can give you more specific pointers.

Update: the 2171S RPM sensor made for the Zilla is an open-collector output which provides 4 pulses per motor revolution. Although the manual says specifically "four pulses per motor revolution at a 50% duty cycle," which I'm not sure how a 50% duty-cycle makes any sense. I would think four pulses per revolution would mean each full up/dn pulse takes a 1/4 of the period.

Anyway, then my CRX manual says the tachometer displays 100rpm for every 200 pulses per minute from the distributor. So I interpret that as consistent with what you were saying that most 4-cylinder cars are 2 pulses per revolution.

So it sounds like I can just make a divide-by-two circuit and drive the tachometer with that. I will need to research that a bit more, probably can easily find an existing divide-by-two driver circuit without reinventing the wheel. Maybe even on this forum!

Also, the two physical sensors for each application doesn't work so well for me. One reason I got the 2171S was that it completely covers the end tail shaft of my motor and I liked that protection, so there isn't an easy way to mount any other proximity sensor.