Author Topic: DIY Dynamic Wheel Balancer, Mark 2  (Read 33747 times)

0 Members and 1 Guest are viewing this topic.

Offline pmbrunelleTopic starter

  • Regular Contributor
  • *
  • Posts: 183
  • Country: ca
  • lost mech. guy
DIY Dynamic Wheel Balancer, Mark 2
« on: November 12, 2015, 01:54:32 am »
A not so long time ago, in June of 2013, three classmates and I began our senior year mechanical engineering project. We chose to do a wheel balancer, mostly because it involves a fair number of topics, and because it's cool 8)

How it works: The machine rotates a wheel on its shaft. Two force sensors, one per shaft support, measure the vibration forces created by the unbalanced wheel. Knowing the vibration forces at the shaft supports as a function of shaft angle, and knowing the dimensions of the wheel, the correct combination of counterweights can be placed on the wheel, thus giving a smooth ride.

Along the way, I asked for help regarding some subsystems, and I didn't even share what it was for! Silly me.
https://www.eevblog.com/forum/beginners/mounting-screw-through-inductor/
https://www.eevblog.com/forum/beginners/overundervoltage-protection-for-adc/

So we presented our incomplete machine to our professors in April 2014, and apparently we did well enough to graduate, woo-hoo!
The balancing machine, just days before the presentation:


Our gadget was fairly crude, with no microcontroller. We directly read the vibrations from the force sensors with an oscilloscope (BNC sockets on the front panel), and we could manually calculate how much counterweight would need to be added, and where it needed to be added to balance the wheel. In general, it worked quite poorly...

So during my free time in my post-school life, I worked on improving the balancer by doing the following things:
  • Replaced the crude 90 VDC thyristor motor drive (with 120 Hz + harmonics galore) with a nice high frequency (claimed 16 kHz) PWM drive, which didn't transmit all this 120 Hz vibration crap to the load cells (which pretty much destroyed the signal-to-noise ratio).
  • Reduced the dynamic braking resistance value, for quicker stops after running.
  • Ordered a new aluminium laser-cut toothed wheel, so that the optical shaft encoder would work as designed.
  • Added an AVR microcontroller with a user interface to facilitate use...
So after doing all these things, I had debugged the balancer as much as I could with the current design:

A skilled person could balance (quite well actually) one wheel in 30 minutes.

Yet after all this work, we still noted 31 deficiencies with the balancer. It would be impossible to fix them all on the current machine.
A semi-clean-sheet redesign, with all the lessons learned from the original balancer would be needed to correct the deficiencies.

The original machine was dubbed Mark 1, and then the Mark 2 wheel balancer project was kicked off!

General Mark 2 goals:
  • Converge faster to a balanced wheel with less iterations
  • Be less picky about the machine being leveled
  • To save $$$, reuse components from Mark 1 where the compromise is not too severe
  • Reduce the parts count, by a lot
  • Use more generic parts where possible, to avoid problems with spare parts availability in the future
  • Reduce number of manual adjustments/trimming type operations at assembly
  • Provide a fun and stimulating leisure activity; learn something!
So anyway, in this thread I expect to chronicle the development of the Mark 2 wheel balancer as it progresses ;) I got started on the debounce circuit for the future wheel cover switch: https://www.eevblog.com/forum/beginners/transistor-schmitt-trigger/

I've attached the a copy of the code that runs the Mark 1 balancer today. Totally not portable, but it's my first embedded programming project. Actually, first programming project on anything...
 

Offline Alexei.Polkhanov

  • Frequent Contributor
  • **
  • Posts: 684
  • Country: ca
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #1 on: November 12, 2015, 02:05:39 am »
"A skilled person could balance (quite well actually) one wheel in 30 minutes."
- 30 Minutes???!!! That sounds like way to much time to balance a single wheel. That is 2 hours per car! You'll be out of business in a day if you do it for living.

I just checked - commercial shop wheel balancer spec: "Cycle time: 6 - 9 seconds (avg.)".
« Last Edit: November 12, 2015, 02:10:04 am by Alexei.Polkhanov »
 

Offline pmbrunelleTopic starter

  • Regular Contributor
  • *
  • Posts: 183
  • Country: ca
  • lost mech. guy
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #2 on: November 12, 2015, 02:18:27 am »
I never was satisfied with the typical $10 flat rate per wheel service...

Actually, I would think that proper balancing (competent operator, multiple iterations, not just one cycle and call it close enough) for a car would be worth $200 if done right, but there's really no market for that. Besides, it's much more fun to DIY!

Don't worry, I have a day job! This is just hobby stuff.
 

Offline retrolefty

  • Super Contributor
  • ***
  • Posts: 1648
  • Country: us
  • measurement changes behavior
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #3 on: November 12, 2015, 03:55:12 am »
I never was satisfied with the typical $10 flat rate per wheel service...

Actually, I would think that proper balancing (competent operator, multiple iterations, not just one cycle and call it close enough) for a car would be worth $200 if done right, but there's really no market for that. Besides, it's much more fun to DIY!

Don't worry, I have a day job! This is just hobby stuff.

 Well your welcome here to share and document your passion. We already have volt-nuts, time-nuts, and lots of unattributed nuts, so happy to have a balance-nut added to the mix.  :-+
 

Offline rs20

  • Super Contributor
  • ***
  • Posts: 2317
  • Country: au
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #4 on: November 12, 2015, 05:18:07 am »
Very cool, I always thought this was an interesting concept, and vaguely thought about implementing it. Cool to see one in action!
 

Offline pmbrunelleTopic starter

  • Regular Contributor
  • *
  • Posts: 183
  • Country: ca
  • lost mech. guy
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #5 on: November 18, 2015, 05:34:29 pm »
So I will be re-using the Mk1 DC motor drive, which works fine. I have the KBWS-25D.
http://www.kbelectronics.com/Variable_Speed_DC_Drives/DC_Drives_Chassis.html

It can output the same 0-90 VDC to the electric motor, but it can accept 115/230 VAC at 50/60 Hz. Quite versatile, if I wanted to move elsewhere (not that I want to).

I need a single 5V DC rail for the Mk2 electronics. Current is still unknown, because I haven't done the power budget of all components, but lets say around 500 mA.

Previously, I would have done something traditional like this for a power supply of lets say 500 mA:

1. Transformer with dual primaries.
2. Bridge rectifier.
3. Electrolytic capacitor.
4. (sometimes) follow with LC or RC low-pass to reduce the ripple seen at the regulator.
5. Linear regulator, such as 7805.
6. Probably some 100 uF electrolytic to keep the regulator from oscillating.

Then, if I wanted to switch between 115 and 230 V operation, I would have to change the transformer winding configuration from series to parallel or vice versa. If you would forget to put the primaries in series for 230 V, but plugged the thing into 230 V, you could break something.

So, in the spirit of learning more (and designing the Mk2 balancer to be more robust that its predecessor), it would be time to graduate to something that changed over seamlessly without user intervention (just like the DC motor drive), such as a switching power supply.

However, unlike the traditional sort of cookbook power supply I have posted above, there does not seem to be a similarly standard cookbook recipe for switching power supplies.

There seems to be a fairly bewildering choice of switching regulator chips, and it's not all the same format at the three-terminal linear regulators.

Where to start  :-//
 

Offline jeroen74

  • Frequent Contributor
  • **
  • Posts: 396
  • Country: nl
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #6 on: November 18, 2015, 11:01:57 pm »
 

Offline ajb

  • Super Contributor
  • ***
  • Posts: 2582
  • Country: us
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #7 on: November 19, 2015, 04:23:45 am »
Yeah, unless you have very unusual needs or specifically want the experience, there's really nothing to be gained by rolling your own line input switcher for a one-off device.  The RECOM RAC series is good (note that you'll need to provide external filtering/protection on the input side), and of course there are dozens of options in small enclosed or open frame PSUs that will do the job as well.
« Last Edit: November 19, 2015, 04:26:25 am by ajb »
 

Offline pmbrunelleTopic starter

  • Regular Contributor
  • *
  • Posts: 183
  • Country: ca
  • lost mech. guy
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #8 on: November 19, 2015, 05:17:24 am »
I came across that exact "power brick" while searching through digikey. A similar pre-packaged solution will be the backup plan if the DIY power supply is a failure.

Ultimately, the goal of this is not just to balance tires, but it's to learn about electronics along the way. This is just for fun, so there are no strict project deadlines to respect. Delays due to subsystems holding up the entire project are OK. If I don't roll my own power supply, I won't get that learning experience.

On the other hand, a fully discrete switching regulator would be too much for a beginner like myself to handle.

So, I want to do a switching regulator based around a "jellybean" IC. In the world of switching regulators, the MC34063 seems to fill the role of the most standard jellybean IC (kind of a 7805 equivalent). So I will see if I can do something with that.
Oh cool, Dave has a video on that  :)


I like 8-DIP packages. The 0.1" pin pitch is about as fine as I want to deal with, considering my 40mil trace width/40mil spacing PCB fabrication process, which includes a Sharpie marker...

Lets see what the 5 VDC power budget is looking like right now. Some items are taken from part datasheets, others are guesstimated for now. Probably I will update this list as things become more stable.

1. 20 mA - microcontroller, 28-DIP AVR, probably ATmega168
2. 20 mA - main shaft quadrature encoder: two SOT-23 Hall-effect open-collector sensors, assuming both sensors are sinking a 1k pullup simultaneously
3. 5 mA - PWM output to control the speed of the DC motor drive
4. 80 mA - DPDT relay coil current, NO contact supplies power to DC motor drive, NC contact connects braking resistance in parallel with DC motor armature.
5. 2 mA - wheel cover status switch hardware debouncer
6. 50 mA - flat laser, used to paint the inside of the wheel with the "angular origin" of the balancer, indicating where the weight should be placed
7. 100 mA - two pre-amplified load cells. I have no idea, this is not specified, I had best order these and measure them.
8. 43 mA - OLED character display
9. 80 mA - two optical encoders (for coarse/fine) for data entry/menu interface
10. 20 mA - two SPDT switch (for Back/Next) debouncers for data entry/menu interface

Total: 420 mA

Multiply by 1.5 factor because surely I missed something and/or something will go wrong:

Power supply minimum target current capacity: 630 mA

Well, at least now I have a rough idea of what is needed...
 

Online CatalinaWOW

  • Super Contributor
  • ***
  • Posts: 5173
  • Country: us
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #9 on: November 19, 2015, 05:39:08 am »
I've often wondered why iterations are required.  If SNR is sufficient you should be able to compute exact location and best approximations in standard size weights immediately.  If SNR is not sufficient, why would the result improve on second iteration.  Only explanation that makes any sense to me is nonlinearity in the sensors, which should be possible to eliminate, possibly with software only if the data can be captured before any mixing occurs.  Maybe the non-linearity is saturation due to inadequate dynamic range.  Back in the day of expensive A/D converters that might have made sense, but now that high resolution A/D is both fast and cheap that should not be an issue.
 

Offline rs20

  • Super Contributor
  • ***
  • Posts: 2317
  • Country: au
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #10 on: November 19, 2015, 05:54:28 am »
I've often wondered why iterations are required.  If SNR is sufficient you should be able to compute exact location and best approximations in standard size weights immediately.  If SNR is not sufficient, why would the result improve on second iteration.  Only explanation that makes any sense to me is nonlinearity in the sensors, which should be possible to eliminate, possibly with software only if the data can be captured before any mixing occurs.  Maybe the non-linearity is saturation due to inadequate dynamic range.  Back in the day of expensive A/D converters that might have made sense, but now that high resolution A/D is both fast and cheap that should not be an issue.

There are so many things to take care of: absolute rotational accuracy w.r.t. position of the wheel, any stretching/bending of components causes by wheel shake that throw out measurements; not to mention the human factor involved with fitting the counterweight (how do you specify to a human exactly where to place them?). No, it's far easier to just have an approximate device, and then rely on the power of the Newton-Raphson method to get you to an arbitrary degree of precision.
 

Offline pmbrunelleTopic starter

  • Regular Contributor
  • *
  • Posts: 183
  • Country: ca
  • lost mech. guy
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #11 on: November 19, 2015, 06:59:00 am »
I've often wondered why iterations are required.  If SNR is sufficient you should be able to compute exact location and best approximations in standard size weights immediately.  If SNR is not sufficient, why would the result improve on second iteration.  Only explanation that makes any sense to me is nonlinearity in the sensors, which should be possible to eliminate, possibly with software only if the data can be captured before any mixing occurs.  Maybe the non-linearity is saturation due to inadequate dynamic range.  Back in the day of expensive A/D converters that might have made sense, but now that high resolution A/D is both fast and cheap that should not be an issue.
Good questions.

So in general, you want to have the highest possible shaft RPM, because this will generate the strongest vibration forces. However, if the shaft spins too fast, then the springy shaft and wheel assembly's resonant vibration mode can be excited. Therefore, to avoid carnage, we want to stay somewhat below the shaft and wheel assembly's first resonant peak.

On the Mk1 balancer, the first whipping resonance is probably a little above 1000 RPM. So I think our 500 RPM DC gearmotor was a good choice. The pros want low RPM, to reduce the acceleration/deceleration times. But the cycle time was a non-consideration for this design.

Then, I thought that at maximum RPM, the load cells should be sized so that the typical smallest imbalance we would want to measure would generate a near full-scale output, for the best possible SNR. So we bought 3.4 lbf load cells.

Now, to avoid overloading the sensitive 3.4 lbf load cells with heavily imbalanced wheels, then the RPM would have to be reduced.

The idea was to slowly turn up the RPM until you reached full-scale on the load cells, then balance what you can. On the second iteration, with the partially balanced wheel, you should need to set the RPM higher than last time to hit full-scale vibration on the load cells. Eventually, you would balance the wheel well enough to reach maximum RPM without a load cell overload.

In practice, the sensitive 3.4 lbf load cells that were supposed to have a stellar SNR due to their sensitivity were a complete failure. The bottleneck in the system is not the load cells, not in the instrumentation amplifiers; neither is the 10-bit microcontroller ADC the limiting factor here.

The primary issue is the mechanical friction hysteresis in the link system. With about 50 lbs of wheel weight hanging from the end of the shaft, it can take a fraction of a pound to move the shaft supports (in grey) before the load cells will see any of the vibration force.

As a side note, because the load cells measure compression only, the load cells need to be DC-biased to 1.7 lbf with adjustable preload springs, because the vibration is an AC signal. This 1.7 lbf DC-biasing thing is finicky and largely swamped by the friction hysteresis, which is really a major PITA.

I can't do much about the mechanical friction hysteresis, but I can give the Mk2 balancer 10 lbf load cells. So the idea is that for a given amount of imbalance, the RPM would have to be higher to generate full-scale vibrations on the load cells. So now that the vibration forces will be 3x stronger (except for the very last iterations), the vibrations should dominate over the friction hysteresis, thus improving the overall system SNR... Faster convergence with fewer iterations is the hope here! And less pain with the DC biasing...

The Mk1 also has a mechanical resonance issue. Essentially, it has two shaft supports, each support supported by 5 links. The 6th DOF is restrained by the load cell. Each shaft support contains a double-row self-aligning bearing, which must act as a perfect frictionless ball-and-socket joint if the vibration force equilibrium equations are to be solved.

What happens during operation above 350 RPM is that the shaft moves around within the play that is present in the bearings. So, we get a resonance at that speed, and we can't even run the machine to its 500 RPM maximum design speed. The cure for looseness in  bearings is to preload them. But, in this case, if we preload the double-row self-aligning bearings, they cease to be remotely similar to frictionless ball-and-socket joints, so that's a non-starter.

In addition to the 10 lbf load cells, the Mk2 will use a 4-link system which holds an all-moving spring-preloaded assembly. Now, bog-standard 6207-ZZ bearings with a tight spring preload (to take up the slack in the bearings) can be used. With this topology, any binding that happens in the bearings is internal to the all-moving assembly, so the load cells won't care.
« Last Edit: November 19, 2015, 07:04:55 am by pmbrunelle »
 

Online CatalinaWOW

  • Super Contributor
  • ***
  • Posts: 5173
  • Country: us
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #12 on: November 19, 2015, 04:39:37 pm »
Why have linkages at all?  Those will always have free play that results in hysteresis.  Just use spring supports and measure the strain in the springs.  This is the approach used in the balance machines that I have seen (these are intended to balance much smaller objects, think of things like the rotor of an electric motor.) 

The springs can be high stiffness so that there is little actual motion, and in two parts, one of which supports a small part of the load, but has a much higher strain at a given extension, making measurement of the strain much easier.

Balancing automotive wheels is a strange business.  I know it helps but in many balance machines, including yours, the wheel is free, with a totally different geometry than when it is carrying the load of the vehicle.  It is easy to imagine cases where a tire perfectly balanced on the machine will bounce on the road.  Some balance machines may improve on this by applying drive and braking forces with a drive roller on the tire, but I haven't really noticed an on the road better result.

 

Offline pmbrunelleTopic starter

  • Regular Contributor
  • *
  • Posts: 183
  • Country: ca
  • lost mech. guy
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #13 on: November 20, 2015, 07:07:53 am »
See below the pictures of a balancer I found online. Essentially, it's kind of like my Mk2 4-link setup with dual load cells, except that it replaces the 4 links with a flat shear plate as a sort vertical support. Very neat, probably very low hysteresis. Low parts count, less adjustments...

If I wanted to go away from a rod-end link setup, that's the way I would go. The nice thing about the rod-ends is that "good-enough" connections to the blue steel tube can be easily made. The whole setup ends up fairly well located in space. I also have a rough idea of how to attach the fixed ends of the links to the plywood box, from experience with the Mk1. At the moment, I don't have an easy solution for how to mate the blue steel tube (without distorting it) to the flat shear plate, especially considering that I am not (yet) a welder. Also I have no idea how I would support the flat shear plate in the air, though this would not be impossible to figure out.

Here is our (unproven) hypothesis with a rod-end link setup:
Once loaded with the weight of a wheel, for very small movements, the rod-end ball will rock (maybe roll is the right word) within the rod-end housing, rather than rotate and slide, thus acting as a frictionless pivot.

In practice, some of our rod ends were pretty terrible; we had to spin them in a drill to get them to "wear-in", but some remained tight even after that. Some rod-ends had no slack, and therefore no possibility for rocking action. So I think that if I go with a 4-link setup with the Mk2, I should be able to cherry-pick the 8 best (most slack) rod ends from the 20 rod-ends that are present in the Mk1.

To keep the scope of this project limited, I will only attempt to correct vibration caused by inertia. But in practice, if you have a fairly round quality wheel and tire, and you correct only the inertia, there should be minimal vibration sensed in the car. All four wheels (16x7 aluminium wheel with 205/50R16 tire) on my dad's car where balanced on the Mk1 balancer, and the result seems excellent (there is a large chance of observer bias here, however). Probably if you have an out-of-round wheel, you're better off trying to fix the root cause rather than trying to patch it with inertial correction.

I have considered that the machine could have a secondary purpose as a tire lathe, which would be used to make a tire true again for instance after a flat-spotting incident.
« Last Edit: November 20, 2015, 07:20:25 am by pmbrunelle »
 

Offline pmbrunelleTopic starter

  • Regular Contributor
  • *
  • Posts: 183
  • Country: ca
  • lost mech. guy
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #14 on: November 24, 2015, 06:24:05 am »
So I was thinking that I would use a 230 VAC nominal input transformer, with 18 VAC output, and 20 VA. Then, with a bridge-rectifier + capacitor combo, there should be some 8~40 VDC available over lets say a 108 to 264 VAC range.

This DC output range would suffice for an LM2675-5.0 buck converter:
http://www.ti.com/lit/ds/symlink/lm2675.pdf

The "suggested application" they mention has 20 mV ripple. This is probably too much for the load cells and 10-bit ADC. Even though the load cells and ADC are ratiometric, and will be powered by the same analog voltage, the ratiometric noise rejection probably doesn't work a darn for 260 kHz switching noise...

I am not sure of the approach to take regarding the ripple. If I increase the output capacitor, then would it cause stability issues? Effectively, an "ON" pulse where the switch is closed (and an associated amount of charge) will cause less voltage change for a larger output capacitor, and thus less ripple. But then it seems like the control has less authority over the output voltage, so it may begin to  chase its tail?

I am in over my head when it comes to the math of feedback loops  :-[

Maybe I should tack on another low-pass LC, with a cutoff below the switching frequency? Presumably, for effectiveness at the switching frequency and harmonics, the L would need to be a ferrite?

Or should I use the buck to bring the voltage to say 9 V and then post-regulate with a 7805. But since the 7805 is probably crap at rejecting the switching noise, then I should just stick with passive filtering and forget about a linear regulator?
 

Offline tggzzz

  • Super Contributor
  • ***
  • Posts: 19284
  • Country: gb
  • Numbers, not adjectives
    • Having fun doing more, with less
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #15 on: November 24, 2015, 09:19:37 am »
Yet after all this work, we still noted 31 deficiencies with the balancer. It would be impossible to fix them all on the current machine.
A semi-clean-sheet redesign, with all the lessons learned from the original balancer would be needed to correct the deficiencies.

Well done, and that list will impress potential employers: don't forget to mention it during interviews!
There are lies, damned lies, statistics - and ADC/DAC specs.
Glider pilot's aphorism: "there is no substitute for span". Retort: "There is a substitute: skill+imagination. But you can buy span".
Having fun doing more, with less
 

Offline pmbrunelleTopic starter

  • Regular Contributor
  • *
  • Posts: 183
  • Country: ca
  • lost mech. guy
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #16 on: November 25, 2015, 03:20:16 am »
Maybe "31" sounds bad, but I'm actually quite happy with how the Mark 1 balancer turned out for a first effort.

As for the buck regulator, low ripple is pretty much the highest priority. This is so the load cells and ADC can have clean readings. So, I'm somewhat planning on some 22000 uF / 1 mH style monster combo.

With a large output capacitor, maybe the startup will be an issue. The LM2675 datasheet claims current-limiting, so I'll see how it likes to charge a big capacitor on startup... As for the microcontroller, I can set the AVR to wait a certain number of clock cycles while the power supply stabilizes. This doesn't really seem at all sensible, but for a one-off hobby project that is not too cost-optimized, this is OK.

I will have to make a preliminary PCB layout of the entire electronic board. Then, when I have a rough idea of what the power supply section will look like, I will prototype the power supply section only. Then, on the power supply section prototype board, I can try different combinations of inductors/capacitors, and see what happens.

Once I converge on a design on the prototype board, then I can transfer it as an atomic unit back to the main PCB.

So I'll need to make myself some sort of load resistor kit with banana plugs, so I can design the power supply. I will also need some sort of square-wave MOSFET switch apparatus, so I can make a load that repeatedly switches on and off. Then, I can observe the power supply's transient response on my (non-storage) oscilloscope.

Or I could look into purchasing a pre-built electronic load. To be evaluated.
« Last Edit: November 25, 2015, 03:26:44 am by pmbrunelle »
 

Offline pmbrunelleTopic starter

  • Regular Contributor
  • *
  • Posts: 183
  • Country: ca
  • lost mech. guy
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #17 on: November 25, 2015, 03:37:06 am »
On a second thought, as per the suggestions on the previous page, I may buy pre-built power supply bricks.

However, I would not install them as-is into the Mk2 balancer. Rather, the purpose would be to benchmark these power supplies, and see how the pros do it, because right now I'm a little too clueless  :-//
 

Offline tggzzz

  • Super Contributor
  • ***
  • Posts: 19284
  • Country: gb
  • Numbers, not adjectives
    • Having fun doing more, with less
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #18 on: November 25, 2015, 09:34:25 am »
On a second thought, as per the suggestions on the previous page, I may buy pre-built power supply bricks.

A big consideration in engineering and business is to understand where you are adding value. Before PSUs became easily purchasable off-the-shelf items, a custom PSU was often designed by the least experienced person on the team more-or-less as a training exercise.

Unsurprisingly, PSUs were very unreliable. If that occurred nowadays, it would be a case of subtracting value.
There are lies, damned lies, statistics - and ADC/DAC specs.
Glider pilot's aphorism: "there is no substitute for span". Retort: "There is a substitute: skill+imagination. But you can buy span".
Having fun doing more, with less
 

Offline jeroen74

  • Frequent Contributor
  • **
  • Posts: 396
  • Country: nl
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #19 on: November 25, 2015, 11:32:31 am »
Use the off-the-shelf power supply to get 6V, then drop it to 5V with a linear regulator and feed all the analogue stuff with it. You could insert a simple LC filter between the switcher output and the linear regulator. If you use the LM2675, use ceramic caps (on the output) that have very low ESR; that will reduce ripple too.

22000uF caps are way, way overkill and are likely not effective LM2675 at all at these frequencies.

Also, a simple filter on the ADC input should block that 260KHz.
 

Offline pmbrunelleTopic starter

  • Regular Contributor
  • *
  • Posts: 183
  • Country: ca
  • lost mech. guy
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #20 on: December 03, 2015, 12:38:58 pm »
So what was I thinking, doing a roll-my-own switching regulator (headaches) before doing my own linear regulator? One must learn to walk before learning to run. Furthermore, having a main PCB with more discrete parts just looks cool, kind of like GK's computer(s).

It seens that even with my desired 120/240 VAC input voltage range, a linear regulator would be feasible. As long as you can dump enough heat... The goal is 5V @ 630 mA. So with a large enough heatsink (not that expensive actually, less than $5 from digikey without forced air cooling) and a beefy TO-247 pass transistor, it should work. According to the SOA chart of the TIP142 datasheet, there will be no secondary breakdown issues.

My main uncertainty with this design is to predict how the transformer output voltage will droop under load at half its design voltage. So I will be having to test this supply with a variac.

I am looking for a 240 VAC right-angled PCB-mounted power switch. To my surprise, this seems to be unobtainium. Flying leads from the PCB to a panel-mount power switch are possible, but this kind of defeats the cleanliness/assembly advantages of having the wiring printed on the board...
 

Offline pmbrunelleTopic starter

  • Regular Contributor
  • *
  • Posts: 183
  • Country: ca
  • lost mech. guy
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #21 on: December 10, 2015, 04:36:52 am »
So apparently 250 VAC PCB-mount toggle switches do exist on DK/Mouser. They're just not classified as such... 125 VAC/6 A (resistive) in the search engine, but then only when you dig into the datasheets, you can find an alternate rating of 250 VAC/3 A (resistive) :-+

In Mk1, there were 8 custom-made PCBs. This was to permit immediate testing/debugging of individual modules without having to wait for the whole project to be finished. This was to manage the risks better, to increase our chances of presenting the project in time to our professors. However, this scheme resulted in a huge amount of wiring between modules (though neatly tucked away in wire trays).

In Mk2, I want to consolidate as much as possible on a single vertically-mounted PCB behind the control panel. So since I want to get the power supply on this vertical PCB, I will have to abandon the idea of the heat-wasting TO-247 monster power supply and its 56 VA transformer. My father has experienced "large" thru-hole transformers tearing off (he learned the hard way) from vertical PCBs during transport/handling.

So OK, I will buy a pre-built SMPS power brick from the likes of CUI, TDK-Lambda, etc, for the weight-savings, so I can have something installed on the vertical PCB without having it rip off. To deal with the switching ripple, I will post-filter and then post-linear-regulate. Simple  :)

Except that these 10 W power bricks have a 20 A inrush |O I do not find the NTC inrush current limiters to be a robust solution, because if you power cycle the unit rapidly (faster than the thermal time constant of the NTC), then you have no inrush protection.

Is a 20 A inrush OK on a 3 A switch? I lack the experience to know these things. I think that the switch would never have to break 20 A; this massive current only happens when the contacts make contact, which is perhaps less critical. I'm also wondering if I should get a double-pole switch and wire the contacts in series, though the toggle switches are not typically characterized for this use, so I don't know if this would help anything...
 

Offline RobertBG

  • Regular Contributor
  • *
  • Posts: 126
  • Country: us
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #22 on: December 10, 2015, 05:59:21 pm »
Great project,I've learned a bit by going over your work and I like your hands on approach.  :-+

If you ever decide to head towards a marketable product and dont want to go against the big names like Hunter and Coats etc.You should look towards a portable unit aimed at racers and track use.You could drive it off of a hub adapter and use the car or possibly a DC drive powered by the car or service vehicles electrical system.

The commercial market is pretty well saturated with the big names and knockoffs but most home mechanics,hobbyist and even professional race teams still use the venerable old bubble balancer at the track.In all my years I've yet to see a shop style balancer in a race team trailer(Nascar uses dynamic machines but it's handled by Goodyear not the team) and we do quite a bit of work for a few teams.They just draw too much power and take up valuable real estate.

Anyway just something to think about,if you ever decide to take this to another level.
 

Offline pmbrunelleTopic starter

  • Regular Contributor
  • *
  • Posts: 183
  • Country: ca
  • lost mech. guy
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #23 on: December 11, 2015, 05:58:41 pm »
I can't compete head-to-head with the big names right away; maybe at the Mk6 iteration I could think about having something worthy of the big leagues... though perhaps I could make for sale a partially complete hobbyist-level kit, which would only include the special more hard-to-manufacture parts, such as the steel shaft. I expect that a revised (not as-constructed) Mk2 could be mature enough for this.

However, even for a partial kit it would seem difficult to get the BOM below $1000, which is enough to buy a good used dynamic balancer. So, for someone to buy a kit like this, it would have to be for interest/hobby reasons, and not only economic.

There is another challenge with putting a dynamic balancer in a trailer: The balancer should be installed level on a solid concrete floor. Maybe it can work in a trailer, but it's not ideal.

For a racer-type dynamic balancing solution, I think the best would be to stick (two) magnetic probes onto the knuckle, jack up the car, and spin the wheel to be balanced with the car's engine. I would go this route if I wanted to sell an affordably-priced kit.

Right now I'm just having fun with this, so I'll see as I go...
 

Offline pmbrunelleTopic starter

  • Regular Contributor
  • *
  • Posts: 183
  • Country: ca
  • lost mech. guy
Re: DIY Dynamic Wheel Balancer, Mark 2
« Reply #24 on: December 13, 2015, 08:51:53 pm »
I think I know how to deal with the SMPS power brick inrush current (probably 22 uF input capacitance or thereabouts for a 10 W supply) and the 250 VAC / 3 A power switch.

I will simply put a fixed 82 Ohm resistor in series with the SMPS input. If the SMPS draws ~120 mA @ 120 VAC, there will only be a 10 VAC drop, which isn't too bad, considering the SMPS is supposed to work all the way down to 85 VAC, albeit with a load current derating. I think that this resistance would also charge the input capacitance quickly enough, on the same order of magnitude as a mains AC half-sine, so hopefully I shouldn't have start-up issues with the DC bus ramping up too slowly. To be tested...

With the fixed resistor, I won't have to worry about power cycling the machine too fast for the NTC inrush current limiter to cool down after being powered on.

My fixed inrush resistor specifications:
  • Axial-leaded format, none of this TO-xxx rubbish where you need to add a heatsink to achieve some semblance of a power rating.
  • 400 V dielectric breakdown voltage, because when the machine is turned on, the resistor will see full mains voltage.
  • 5W continuous or better power rating; I don't want to burn my fingers if I touch the resistor, nor do I want to scorch the PCB.
  • 2~3 Joules pulse handling or better; My rough estimation for the pulse energy is ½CV2, though I'm not too sure how to precisely calculate this, as it may be affected by when the power switch is closed in the AC cycle. I would also need to know the the value of the SMPS' input capacitance, but I am not willing to buy two of them just to cut one open to see what's inside.
So I will just get a Vishay 7W 82 Ohm resistor:
http://www.vishay.com/docs/28730/acseries.pdf
No dielectric voltage rating, but the body is like 1" long...

I should get this inrush-limiting mostly right, because I will be using the power switch as the Emergency-Stop, so I can't afford to have the contacts weld together.

Technically it would be more proper to use of of those $$$ mushroom switches that is guaranteed to separate its contacts even after being welded together. But I want to keep everything clean and PCB-though-hole mounted, without having flying leads just for the E-Stop.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf