PDA

View Full Version : Unstable tacho



morrisman
12th August 2015, 07:35 PM
I bought X2 of these for lathe tachometers .

I am using a neo magnet on the spindle . The readout is unstable, it wanders about +5 and -5 rpm . I have tried running it from a regulated supply and varying the supply voltage - no effect . I also shortened the sensor head lead as it may be picking up stray interference , no effect . Also tried different magnets and positions, no effect . The LED in the sensor head is flickering as it should as the magnet passes the sensor - it has NPN written on the head, a transistor is in there ?

Any ideas . It may be that these are cheap junk from China and the unit is actually running within specs .

jhovel
12th August 2015, 07:44 PM
What are the specs for this unit?
I wonder if the battery voltage is too low. I'd try with 12V next....
Other than that, there may be a trim pot somewhere to reduce hysteresis or increase filtering.
Is it more stable at high speed than at low speed?

Stustoys
12th August 2015, 07:55 PM
What rpm are you talking about?
If its low you could try adding a magnet.

(I assume its low as a 1000rpm whos going to care about +/-5rpm?)

Stuart

simonl
12th August 2015, 08:04 PM
Hi there,

I run 2 of the same units, one in my mill and one in my bandsaw. When set up correctly they produce very reliable and stable results. I can't remember exactly the setup but when used in conjuction with one of the commonly available sensors (induction, magnetic etc), the sensors are either NPN or PNP which refers to the transistor amplified output. It should have a diagram on the case to show which lead is +ve, -ve and signal. Since the input impedence of the frequency counter unit is high, you shouldn't need a resister in line with the signal wire but you may need a small capacitor going from the signal wire to ground. A 0.1 uF cap will act as a low pass filter, only allowing lower frequencies through.

I had lots of trouble with the same unit and sensor on my mill. The VFD created lots of noise and created erratic readouts on the display. Once I installed the cap it got rid of all the HF noise and left a stable, accurate readout. You also need to make sure you are not asking the sensor to pick up signals in frequencies greater than it's maximum response frequency. It has been my experience that some of these sensors are much over rated and when they claim a frequency response up to 500Hz, sometimes they can not handle anymore than about 100Hz before giving unreliable signal.

Given your setup will produce 1 pulse per revolution, a spindle speed of 3000 rpm should produce a "square wave" of about 50Hz, well within the limits of these sensors.

Try a capacitor and go from there.

Good luck.

Simon

Michael G
12th August 2015, 08:14 PM
Mike, I understand the frustration with readouts flickering around, but remember that knowing the RPM on a machine is just data to help you work out where you should be - the actual speed you run at will depend on chatter, tool material, workpiece material, support, work piece form and a number of other factors. A job will not turn out perfectly at 350rpm but at 345 or 355rpm be an abject failure. Even with a laser tacho I get speeds jumping around by 5 or so rpm, so you are not alone.
With variable speed on a machine it is useful to have a display so that you are not starting too fast but other than that I would be adjusting conditions to suit. I can't think of a job I've done lately where it is all at the 'optimum' speeds. Most time I'm running below that although occasionally I am running slightly faster just because it seems to cut better that way. One day I hope to get to the stage where I can start the mill or lathe up, look at the spindle and decide whether the speed is right without first referring to the speed chart. One day... (I don't have continuously variable speed)

Michael

morrisman
12th August 2015, 08:32 PM
Mike, I understand the frustration with readouts flickering around, but remember that knowing the RPM on a machine is just data to help you work out where you should be - the actual speed you run at will depend on chatter, tool material, workpiece material, support, work piece form and a number of other factors. A job will not turn out perfectly at 350rpm but at 345 or 355rpm be an abject failure. Even with a laser tacho I get speeds jumping around by 5 or so rpm, so you are not alone.
With variable speed on a machine it is useful to have a display so that you are not starting too fast but other than that I would be adjusting conditions to suit. I can't think of a job I've done lately where it is all at the 'optimum' speeds. Most time I'm running below that although occasionally I am running slightly faster just because it seems to cut better that way. One day I hope to get to the stage where I can start the mill or lathe up, look at the spindle and decide whether the speed is right without first referring to the speed chart. One day... (I don't have continuously variable speed)

Michael

Hi

Yes the old time machinists mainly only had fixed speeds to play with and with experience they were able to adjust feeds and depth of cut and tool angles to suit their needs , quite a lot of that skill is intuitive to some but some also pick it up with experience . We are spoilt with today's variable VFD's and I guess one can get carried away with it but thinking about it those old timers would probably have jumped at the chance to try out a digital tacho .

KBs PensNmore
13th August 2015, 12:33 AM
Hi

Yes the old time machinists mainly only had fixed speeds to play with and with experience they were able to adjust feeds and depth of cut and tool angles to suit their needs , quite a lot of that skill is intuitive to some but some also pick it up with experience . We are spoilt with today's variable VFD's and I guess one can get carried away with it but thinking about it those old timers would probably have jumped at the chance to try out a digital tacho .

True, but they might also have said "stick that technological crap" I for one find technology to hard to keep with, I like the KISS principle.
Kryn