Looking for drill testing jig ideas

[Stuey]

I've been trying to think up a benchtop testing jig I can use to compare the power output of cordless drills and maybe drivers as well.

It would have to be a benchtop system since I want to be able to hook up meters to record things like current draw and runtime.

I'm thinking I can set up a bearing-supported 3/8" shaft (drill rod?) that would be chucked into drills and onto which different weights can be clamped on, but I'm unsure as to how that can be easily done. I don't have the equipment to machine 2-piece cylindrical weights in a range of sizes.

(The different weights would be to simulate different loads.)

Maybe machinable bore locking collars would do the trick? http://www.mcmaster.com/#3374k12/=pv2ase Although I doubt my Craftsman 10" drill press would be up to the task of machining the holes.

If I use larger drill rod I can then use shaft couplers with short 3/8" and 1/2" rods that can be chucked into different sized drills, but this might affect the ability to do direct comparisons.

Alternatively, I'm thinking I could use pulleys and belting to create a jig where a drill slowly lifts certain weights from the floor via high tensile fishing line or steel wire. This wouldn't allow for runtime tests though.

Ideally the entire contraption could be placed within 80/20 frame with lockable door and polycarbonate shielding.

Any ideas for a different approach, or how to make this work? This is a down-the-road project so no urgency.

 

[A_Pmech]

I would build a Prony Brake using a band brake lining and drum from a go-kart kit. Such as:

Northern Tool go-kart band brake kit

 

[Stuey]

I would build a Prony Brake using a band brake lining and drum from a go-kart kit. Such as:

Northern Tool go-kart band brake kit

Thanks for the suggestion! I did consider creating a friction brake system, but would be concerned about repeatability between adjustments.

 

[A_Pmech]

Thanks for the suggestion! I did consider creating a friction brake system, but would be concerned about repeatability between adjustments.

Measure the torque on the brake arm with a small hanging scale. Adjust the band tightness to the same value each test.

By incrementally tightening the friction band and taking both a shaft speed and torque reading at each step you can plot a power curve for the drill.

 

[Stuey]

I appreciate the idea, but I'm not too convinced this would be the best type of setup for what I'm looking to do. It has the potential for multiple complications and sources of error. Additionally, spinning a 4-inch drum might set the minimal load too high.

Separately, do you have any recommendations for measuring shaft speed? The best idea I could think of is a handheld tachometer, but something more specific for smaller shafts might be more suitable. Some of the models I've been looking at require very large patches of anti-reflective tape to work well.

 

[zkling]

I'm not sure how crazy you want to get with this but, what I would do....

Make a setup that has a shaft supported in bushings, yet allows axial movement. On the end of the shaft mount a friction disk setup, like a clutch works. Have it spring loaded. If you purchase accurate springs, then all you will have to measure is displacement. Otherwise a force transducer would be needed. Then also measure shaft speed via an inductive pick-up.

Or you could just buy a torque transducer, but $$$$.

You are going to need a data acq box, possibly an amp and some software (download) for analysis.

This would really be the best for a drill test as it would replicate axial and torsional loadings just like a drill would see in use.

 

[signcrafter]

I appreciate the idea, but I'm not too convinced this would be the best type of setup for what I'm looking to do. It has the potential for multiple complications and sources of error. Additionally, spinning a 4-inch drum might set the minimal load too high.

Separately, do you have any recommendations for measuring shaft speed? The best idea I could think of is a handheld tachometer, but something more specific for smaller shafts might be more suitable. Some of the models I've been looking at require very large patches of anti-reflective tape to work well.

Don't use a non contact tachometer, use one with a wheel that the shaft will spin.

 

[Stuey]

I'm not sure how crazy you want to get with this but, what I would do....

Make a setup that has a shaft supported in bushings, yet allows axial movement. On the end of the shaft mount a friction disk setup, like a clutch works. Have it spring loaded. If you purchase accurate springs, then all you will have to measure is displacement. Otherwise a force transducer would be needed. Then also measure shaft speed via an inductive pick-up.

Or you could just buy a torque transducer, but $$$$.

You are going to need a data acq box, possibly an amp and some software (download) for analysis.

This would really be the best for a drill test as it would replicate axial and torsional loadings just like a drill would see in use.

I think I get what you're going for.

I'll have to look into a friction disc setup a bit more. I believe I get the concept, but cannot yet visualize how it would all go together.

With the inductive pick-up, I suppose I would have to craft a clamp-on encoder of some kind? Maybe a plastic cog with steel screws embedded in the teeth? Although I suppose at up to 2000 RPM an all-metal part would be best but I don't have the means to machine one.

That might still work well for a side project.

I have a datalogging multimeter on order, an oscilloscope on the bench, and have been trying to get an Arduino to work with a Labview trial for data acquisition.

My immediate focus wouldn't be on axial loading as much as purely radial as radial loads would more effectively increase loading on drill motors.

Don't use a non contact tachometer, use one with a wheel that the shaft will spin.

Ah, makes sense, thanks! The ones with wheels I've seen seem to be set up for linear motion, but I can always convert linear to rotational data.

 

[zkling]

You don't necessarily have to have a piece on the shaft for the non contact tach.

I wouldn't waste you money on a datalogging multimeter. Need to step up to an actual data analysis capture box. Something like a cheaper version of the multi channel NI boxes. Then you can plug in a large multitude of sensors.

You don't want to use a contact tach for something like this.

My immediate focus wouldn't be on axial loading as much as purely radial as radial loads would more effectively increase loading on drill motors.

Then you are really not testing the drill how it would be used in service.
Your best bet time, money and patience ahead would be to purchase a small dyno that people use for testing motors. Basically all you really need for what you want to do is a dyno like AP_Mech suggested.

 

[signcrafter]

Ah, makes sense, thanks! The ones with wheels I've seen seem to be set up for linear motion, but I can always convert linear to rotational data.

We used to use one for setting shaft speeds on the machines at the mill. Here is one like we used to use, http://www.amazon.com/dp/B006C2DMZU/?tag=atomicindus08-20

 

[Stuey]

You don't necessarily have to have a piece on the shaft for the non contact tach.

I wouldn't waste you money on a datalogging multimeter. Need to step up to an actual data analysis capture box. Something like a cheaper version of the multi channel NI boxes. Then you can plug in a large multitude of sensors.

You don't want to use a contact tach for something like this.



Then you are really not testing the drill how it would be used in service.
Your best bet time, money and patience ahead would be to purchase a small dyno that people use for testing motors. Basically all you really need for what you want to do is a dyno like AP_Mech suggested.

The apparatus I have in mind is not for testing how a drill would be used in service. I want to measure current draw over time and comparative runtime using range of arbitrary but repeatable loads.

But I do get what you're saying. I'll look into buying or building a small dyno as you and AP_Mech described.

With the NI data acquisition box, you mean something like http://sine.ni.com/nips/cds/view/p/lang/en/nid/201986 ?

Thanks again. This is new territory for me and everyone has given me a lot of good things to look into.

We used to use one for setting shaft speeds on the machines at the mill. Here is one like we used to use, http://www.amazon.com/dp/B006C2DMZU/?tag=atomicindus08-20

Thanks, I'll look into it!

 

[zkling]

With the NI data acquisition box, you mean something like http://sine.ni.com/nips/cds/view/p/lang/en/nid/201986 ?

Yes, call NI up and talk to their technical department. They are very helpful. Note they are not the only ones in the data aqu business, but just the ones I'm most familiar with. They are not known to be cheap, but they make good products and really stand behind them.

Here is a good read to help familiarize yourself.

http://www.ni.com/data-acquisition/what-is/

 

[signcrafter]

You don't want to use a contact tach for something like this.

Any particular reason why? We used a contact tachometer to set shaft and motor and belt speeds all the time on machines that had different sections and you had to get the speeds matched up right. I would think his experiment would be very similar and could just rig up the tach to touch part of the shaft of whatever he builds.

 

[Spudland_Dave]

Why not make a water brake...like a real dyno?
Water pump with a valve to block flow..good way to apply consistent load..

 

[Stuey]

Yes, call NI up and talk to their technical department. They are very helpful. Note they are not the only ones in the data aqu business, but just the ones I'm most familiar with. They are not known to be cheap, but they make good products and really stand behind them.

Here is a good read to help familiarize yourself.

http://www.ni.com/data-acquisition/what-is/

Thank you for the suggestion and link.

Short term I intend to play around using an Arduino and Labview trial for basic data acquisition and was going to re-evaluate my hardware needs down the road. It's been a long time since I did any data acquisition. As an undergrad we used a lot of different analog and digital sensors and software packages, but a lot of them were black box designs in all regards.

I have also been looking at torque testers and rundown adapters to help give true torque comparisons, but that would be too large of an expenditure for right now.

The frustrating part is that many of the companies that make these tools don't have the greatest websites. I found myself looking at Mark-10's TT02, buts its max torque is way too low for power tools (100 in-lbs).

Any particular reason why? We used a contact tachometer to set shaft and motor and belt speeds all the time on machines that had different sections and you had to get the speeds matched up right. I would think his experiment would be very similar and could just rig up the tach to touch part of the shaft of whatever he builds.

I'll keep your advice in mind as well, thank you!

Why not make a water brake...like a real dyno?
Water pump with a valve to block flow..good way to apply consistent load..

I considered something along these lines, but I would like to avoid a tank of water if I can. Knowing me I would let it sit for two months and then wonder why the whole place smelled so bad.

 

[Spudland_Dave]

I considered something along these lines, but I would like to avoid a tank of water if I can. Knowing me I would let it sit for two months and then wonder why the whole place smelled so bad.

Good point...Would using some AntiFreeze in there suffice to prevent microbial growth? or heck...cup of Bleach?

 

[Stuey]

Good point...Would using some AntiFreeze in there suffice to prevent microbial growth? or heck...cup of Bleach?

There are a couple of additives I could use, but ultimately I need solutions that are easy to set up and put away. Otherwise they will never get used or never get put away.

I don't have enough storage space for things that never get used and not enough bench space for too many things that never get put away.

 

[A_Pmech]

Stuey,

I think I see what you want to do now. You want to measure battery life at a given load.

For a very steady load, consider building a simple eddy current absorber from a couple of cast iron discs and some hand wound electromagnets. Then, build a simple circuit to measure a load cell value and adjust the absorber field current to match it to the load set point.

In a nutshell, that's how I'm going to build the engine dyno for my motorcycle project. Only, it will have to soak up about 400 HP continuously and will be water cooled.

Edit:

A less elegant solution is using a 3-phase motor as an eddy current absorber. At the power levels you're talking about, a cheap used 5 to 7.5 HP motor would do the trick. Inject DC into two of the field coils and voila! Variable DC brake.

The only downside is the heat is going into the rotor of the motor. It is possible to heat the rotor to the point that the lamination insulation breaks down. (Over 450F)

An ODP (open drip proof) motor would allow you to blow air across the rotor with a blower, increasing the potential continuous load capability over a TEFC motor.

 

[Stuey]

Stuey,

I think I see what you want to do now. You want to measure battery life at a given load.

For a very steady load, consider building a simple eddy current absorber from a couple of cast iron discs and some hand wound electromagnets. Then, build a simple circuit to measure a load cell value and adjust the absorber field current to match it to the load set point.

In a nutshell, that's how I'm going to build the engine dyno for my motorcycle project. Only, it will have to soak up about 400 HP continuously and will be water cooled.

Edit:

A less elegant solution is using a 3-phase motor as an eddy current absorber. At the power levels you're talking about, a cheap used 5 to 7.5 HP motor would do the trick. Inject DC into two of the field coils and voila! Variable DC brake.

The only downside is the heat is going into the rotor of the motor. It is possible to heat the rotor to the point that the lamination insulation breaks down. (Over 450F)

An ODP (open drip proof) motor would allow you to blow air across the rotor with a blower, increasing the potential continuous load capability over a TEFC motor.

Well, I had an idea of what I wanted to do, but you guys have given me a lot of food for thought and shown me better ways of approaching the problem. At the least I now have additional keywords and information from which springboard and do more research.

The basic idea is the same - I want to do continuous motor load simulation. The aim is for runtime testing and measurements that can only come from continuous loading.

I'm now circling back and warming up to your idea.

What about a hysteresis brake (e.g. http://www.magtrol.com/brakesandclutches/hysteresis_brakes.html) or possibly 2 different sizes of brakes for different classes of drills? I might give them a call next week with some questions.

As I understand it, I could set up a test jig with a shaft coupler connecting a drill (with drill rod chucked in) to the brake, possibly with a bearing for support, and could use a basic 90V DC power supply with current control to adjust the braking torque and simulated load.

Potentially, if I increase power to the brake, to where it stalls out a cordless drill, I could also do more accurate Brand A vs. Brand B maximum torque comparisons as well. This would be easier, less expensive, and possibly more informative than using a torque gauge with rundown adapters. Maybe I could write a program so the brake power increases over time to simulate fastener-driving or hole-drilling more precisely.

Later on I could add inline components to the setup to get quantitative data, as shown here http://www.ogura-clutch.com/products/industrial/applications/load-simulation-machine.html , although inline shaft torque transducers or gauges look a bit too pricey.

It looks like a brake and rotation sensor would allow for me to measure RPM stability as the simulated load changes. I'm not sure what benefit this would bring outside of personal interest. I still need to read up on what could/should be used on a shaft to be read by the inductive sensor.

Edit: I've been reading up, and suppose that if/when I'm ready to start making quantitative power measurements I could then look into a dyno rather than adding individual components to a brake. Unfortunately, the brand I'm looking at doesn't make dynos at the torque range I'd need (maybe 800 in-lbs max to test 650 - 750 in-lbs rated drills).

 

[zkling]

OK, from your 1st post I got a bit difference impression of what you wanted to do. Agree with AP, after your last long post my mind went directly to an variable eddy current brake setup as well.

Long term you may want to look at a universal multi channel data aqu setup though. It will give you lots of flexiability in measuring, testing and monitoring. For example with the proper cells, you could design your own torque gauge like the one you posted. Data acq is very similar to multimeters. In that there are really only a few things you can measure, but more so how you apply those measurements and how you manipulate the data set to get what you want. You could do anything from torque measurements to dynamic balancing with accelerometers and force transducers.

Good point...Would using some AntiFreeze in there suffice to prevent microbial growth? or heck...cup of Bleach?

Yes, if you cut distilled water with antifreeze coolant it will prevent growth. Same idea is used in tig coolers to prevent growth.

Any particular reason why? We used a contact tachometer to set shaft and motor and belt speeds all the time on machines that had different sections and you had to get the speeds matched up right. I would think his experiment would be very similar and could just rig up the tach to touch part of the shaft of whatever he builds.

Efficiency, accuracy, and for what I thought what he wanted to do data logging vs time. Initially what I thought he wanted to do was plot power curves for the drills. Thus if he could input two channels simultaneously; speed (rpm) and torque both as a function of time, he could do a calculation on the data set to obtain power curves for certain drills, load cases and so on and so forth.

Those hand held devices are designed more for a reference check at a steady state.

 

[Stuey]

OK, from your 1st post I got a bit difference impression of what you wanted to do. Agree with AP, after your last long post my mind went directly to an variable eddy current brake setup as well.

Long term you may want to look at a universal multi channel data aqu setup though. It will give you lots of flexiability in measuring, testing and monitoring. For example with the proper cells, you could design your own torque gauge like the one you posted. Data acq is very similar to multimeters. In that there are really only a few things you can measure, but more so how you apply those measurements and how you manipulate the data set to get what you want. You could do anything from torque measurements to dynamic balancing with accelerometers and force transducers.



Yes, if you cut distilled water with antifreeze coolant it will prevent growth. Same idea is used in tig coolers to prevent growth.



Efficiency, accuracy, and for what I thought what he wanted to do data logging vs time. Initially what I thought he wanted to do was plot power curves for the drills. Thus if he could input two channels simultaneously; speed (rpm) and torque both as a function of time, he could do a calculation on the data set to obtain power curves for certain drills, load cases and so on and so forth.

Those hand held devices are designed more for a reference check at a steady state.

May I ask what the benefit is to go with an eddy current brake vs. a magnetic air gap brake? I was looking over at Magtrol literature and there was a note that the eddy current brakes have torque dependency on speed, while the magnetic air gap brakes do not. The magnetic air gap brakes seem to go up to the torque range I'm interested in, and might provide more accurate results since not all drills provide similar speeds.

Thanks again, and for your patience. Before today I hadn't realized that there were off-the-shelf products that could simulate motor loads.

 

[justanengineer]

JMHO, but if I were to build a dyno it would be hydraulic (oil, not water) as its much more versatile and cheaper to build. When we need to do some true torture tests its always on a hydraulic brake for speed, durability, and repeatability.

 

[zkling]

May I ask what the benefit is to go with an eddy current brake vs. a magnetic air gap brake? I was looking over at Magtrol literature and there was a note that the eddy current brakes have torque dependency on speed, while the magnetic air gap brakes do not. The magnetic air gap brakes seem to go up to the torque range I'm interested in, and might provide more accurate results since not all drills provide similar speeds.

Thanks again, and for your patience. Before today I hadn't realized that there were off-the-shelf products that could simulate motor loads.

Primarily cost, easier to DIY and durability. The major benefit of the hysteresis brakes is the high RPM stability and accuracy. Hence you see them quite a bit in test labs. Eddy current brakes are used in some vehicles and industrial applications.

JMHO, but if I were to build a dyno it would be hydraulic (oil, not water) as its much more versatile and cheaper to build. When we need to do some true torture tests its always on a hydraulic brake for speed, durability, and repeatability.

Hum, that may not be a bad idea. I wonder what type of pump could be re purposed for the small power motors he would be testing. I think I remember some karting guys that were building these

 

[Morphious]

I may be missing something here, but would not a small DC generator make a good variable load? A selection of different resistor values gives you a selectable load and you could easily measure the voltage/current at the load resistor (if the power dissipated there is of value in your application).

Roger

 

[Stuey]

Soooo... I think I know which magnetic brakes would be suitable for my uses.

But... inspired by some of your recommendations, I'm thinking that a torque transducer or torque and speed sensor might be useful in comparing torque and peak power values for different drills.

The problem: all of the options that might work cost mega big bucks. Realistically, maybe $3K-$4.5K since I would need one that can handle up to 100 Nm (800 in-lbs and change).

Alternatively...

Measure the torque on the brake arm with a small hanging scale. Adjust the band tightness to the same value each test.

By incrementally tightening the friction band and taking both a shaft speed and torque reading at each step you can plot a power curve for the drill.

How could the torque be measured when using a magnetic brake? A torque wrench that can handle up to 65 ft-lbs of torque or so?

Since the brake's torque is independent of speed, I can turn the shaft using a torque wrench at say 5 RPM and the torque reading should be the same as when a drill turns it at 1500 RPM.

Presuming that the brake output delivers constant resistive torque at a given current, would I be able to measure the braking torque using a torque wrench to create a performance curve a la http://www.magtrol.com/drawings/hbhc_dwgs/curves/hb-750-2curve.pdf ?

If Drill A stalls when 0.150 Amps is delivered to the brake, and Drill B stalls when 0.200 Amps is delivered to the brake, I can say Drill B can deliver more torque than Drill A as it can handle the higher load.

So a brake by itself will still allow me to test for runtime and relative torque/power performance.

If a performance characteristic curve would allow me to estimate quantitative torque values, I could then use those values as a direct comparative measure and also with RPM for quantitative relative power output comparison.

I would be happy if I can get torque readings with reasonably certainty, and would only use the readings in a comparative manner. That is, I don't intend to do independent testing for other companies, and I don't wish for very high precision. ±5% accuracy would be good enough, and if there is ever a need for higher precision I could look into a commercial dynamometer or torque transducer.

I have tried doing a little more homework on the matter, but there's not a lot of info out there.

 

[Stuey]

*facepalm*

So I decided to revisit this idea...

and realized that the brakes I was looking at are rated in oz in.

1 oz in = 1/16 in-lb

In other words, this company's highest-rated brake can provide less than 315 in-lbs of torque even when powered passed its optimal power range.

Back to the drawing board. I'd rather not have to mess around with massive and pricey air- or water-cooled brakes.