Slip Ring

[luiHS]

 
Hi.

Does anyone have experience with the Slip Ring components?
I am designing a Ferris wheel for a scale model, and I need to connect supply and data voltage for lighting effects with digital LEDs type APA102, the ferris wheel will be rotating with a stepper motor or a modified servo. 3 wires, Mass, + 5v and Data are needed.

The only solution I have found is to use these Slip Ring, but it is a component that I have never used before and I don't know how they work and what problems they can give, if they need any maintenance and what durability they have. The signals are low voltage and low current (5v, 600mA).

The ferris wheel is just a drawing, it will has its axis, and spokes, coupled to an electric motor step by step, and with optical sensors to know the position of the baskets. The lighting system is much more complex than what is seen in the drawing, the spokes and baskets will install APA102 digital LEDs, and the circumference of the ferris wheel will be transparent resin with internal LED lighting, and we will try to illuminate it with Laser. It is possible that each basket also installs a sensor to detect when there is a ball in the basket.

Regards.

PD: discarded wireless systems, because that would force to put batteries in the ferris wheel, and replace them every so often, that solution is not viable.

 

[Marco]

No idea about cheap slip ring reliability, but alternatively you could use magnetic coil power transfer and hide a micro in the wheel to drive the leds.

[jonovid]

used video cassette recorders have a head drum with two sets of magnetic coils power transfer & data
if u use switched-mode power at say 1or 2 KHz head drum coils maybe made to transfer your power ?   

[luiHS]

 
Thanks for your suggestions.

The magnetic coupling with two coils, is something that I thought also a long time ago. The problem is that it is not only about transferring power supply, it is also necessary to transfer data signals.

The microcontroller cannot be in the ferris wheel, because the effects of lights are interactive and depend on signals external to the ferris wheel. So I can't put a magnetic coupling just for supply voltage.

The use of two coils, one for supply and one for data, I think it is much more complicated than using a Slip Ring, it is also about making it as simple and cheap as possible.

[beanflying]

I have a few of the cheap ones left over from a project I could take some close up pics of if you like tomorrow (post midnight here). Providing you are not hanging a lot of weight on them Ferris Wheel speeds will be fine.

I didn't look at if they can be disassembled but if they can be then I will pop the top on them for fun too.

[radar_macgyver]

Don't know if this applies to the cheap ones you can get on ebay, but the general rule for slip rings is you don't mount both halves rigidly. Only one side is rigidly mounted, the other side is allowed to have some play to take up misalignment between the rotational axes. If this isn't done, the internal bearings will wear out, quickly followed by the brush assemblies.

https://www.moflon.com/technical_install.html

[james_s]

In a pinch you could build your own using carbon brushes from small DC motors and rings made of copper tubing or etched on copperclad. The only issue you might have is on the data line for the LEDs, slip rings can be electrically noisy, but given the low speed and low current it might end up working fine.

[MarkF]

If you want to send both power and data, you might consider how Model Railroad DCC controllers work.
(Maybe even use DCC controllers.  The N-scale ones are very small.  https://www.rakuten.com/shop/hobbyking/product/9950000003-0/ (18x12x3mm))
The power can be extracted just by rectifying the waveform and filtering.
The signal is encoded in the pulse width.
This could be sent via magnetic coupled coils.



I have no ideas on how you intend to decode the signal without a MCU in the ferris wheel (with or without slip-rings or magnetic coils).

[luiHS]

I have no ideas on how you intend to decode the signal without a MCU in the ferris wheel (with or without slip-rings or magnetic coils).

The LEDs are digital APA102, each LED decodes the serial data signal, plus a clock signal, exactly the same as when connected with cables or PCBs. I said a data signal before, but the clock signal was also missing, in total 4 wires, two for power and two for data.

I know the DCC for model railroad, its an option, I will check what is the easier and cheaper.

[MarkF]

I have no ideas on how you intend to decode the signal without a MCU in the ferris wheel (with or without slip-rings or magnetic coils).

The LEDs are digital APA102, each LED decodes the serial data signal, plus a clock signal, exactly the same as when connected with cables or PCBs. I said a data signal before, but the clock signal was also missing, in total 4 wires, two for power and two for data.

I know the DCC for model railroad, its an option, I will check what is the easier and cheaper.

If you had a set of coils on each side, you could have the clock on one and the data on the other.
The clock could be rectified to generate the power.  If you can get enough current from it???

[daqq]

Does anyone have experience with the Slip Ring components?

We use some really expensive slip rings for one of our devices and the only issues we've had was when the material started corroding when exposed to something nasty. Completely our fault, can't blame the rings for that. Basically, a laminate structure was not cured properly and started emitting something nasty.

What I've heard of from the cheaper ones are issues with higher current (sparks and such), very short disconnects. And of course life time.

[beanflying]

Two minute look at the http://www.chinagoco.cn/c1098.html SRC-22-06A brought off evilbay.

It resisted my spudgers so I used a little more persuasion 

Front and back are magnetic so there appears to be bearings front and rear I suspect the reason for the 250-300RPM limit is more about balance and contact bounce as the plastics are not perfect. So take the dimensions as a guide more than gospel.

The remaining bits would be great to make into a custom 3D print and much more compact than the full outside housing. Guisstimate on current 100mA and strictly sub 100VDC IMO regardless of claims.

EDIT pulled out the soldering scope and the contacts are better than I thought dish shaped washers either side of the contact wiper.

Re EDIT turns out the two halves are not glued and yes it did go back together afterwards 

[richard.cs]

I am dubious about serial data without error correction being passed over low-cost sliprings. I'm sure it will work most of the time, but there's a high chance that contact bounce will causes glitches in the data, or stray clock edges that misalign everything by a few bits and I suspect it will be quite visible. A fast update rate may help in that glitches then get overwritten and most errors appear as brief flashes rather than LEDs that get stuck in the wrong state until you next want to change them. Really you'll just have to try it and see.

Then best solution is probably to have two microcontrollers, one on the ferris wheel controlling the LEDs and the other outside doing the user interface. That then allows you to create a robust, noise-resistant protocol over a slipring between the two microcontrollers. Even without any protocol effort microcontrollers are likely to be more robust as they will have UARTs which do majority voting internally and ignore a lot of glitches.

[Psi]

If you use a slip ring with double the number of connections you need you can parallel them up in pairs.
This will reduce any issues if one should glitch.

You can also add some capacitors on each side to smooth things out should the ring glitch.
For data connections you can still add a cap, you just have to check that it's small enough not to corrupt your data.
For slow speed signals this is easy to do e.g. (9600 baud).

[beanflying]

All the chat about a glitch here due to a bounce here or there we should remember this is for a Model Ferris wheel not a Rocket headed for Mars 

Rotational speed is likely to be sub 5RPM so the chances of bounce are low? Twisted pairs and shielding would help as would a Low baud rate and maybe still leave room for some error correction.

[Psi]

i wouldn't be worried about glitches especially if you parallel 2 contacts

[TomS_]

Slip rings with two contacts seem to be good enough for gyroscopes in military aircraft:

[radar_macgyver]

i wouldn't be worried about glitches especially if you parallel 2 contacts

^ This. Commercial units usually have two contacts per wire to reduce the probability of random bounces occurring on both.

[beanflying]

i wouldn't be worried about glitches especially if you parallel 2 contacts

^ This. Commercial units usually have two contacts per wire to reduce the probability of random bounces occurring on both.

Refer to the last photo on my previous post there is two wipers per contact.

[soldar]

How about using a phone cord swivel?

https://www.amazon.co.uk/Hasme-Swivel-Twist-Detangler-Telephone-Eliminator/dp/B073XS85FP

[m3vuv]

why not use slip rigs and brush block from a scrap car alernator?,cheap ready made solution.

[CatalinaWOW]

Slip rings can carry high bandwidth and low amplitude signals.  If you put enough time, money and energy into it.  Other than the ones used for telephone line cords I have no experience with low cost rings.  Most of the companies i worked with only provided custom assemblies so the costs are stunning.

All of the good slip ring assemblies I am aware of use multiple brushes on each ring to reduce the impact of contact bounce, and even just resistance variations across the contact.  Some have as many as eight brushes. 

To the extent that you can making the impedance of the driven circuit as high as possible consistent with bandwidth helps reduce that contact resistance impact.

It isn't terribly difficult to make your own slip rings for low speed applications like this.  Lay out a PWB with one or more circular rings.  You can go double sided though single side works too.  Stack these disks on a shaft with a non-conductive spacer.  This will be the rotating side.  You have to run wiring from the rings along the axle to your destination.

The "brushes" can be salvaged from a number of sources, including defunct DVM (from the selector switch), fuel gauges, or the emi shields from cabinets.  Remember to use more than one per signal with all brushes on a signal paralleled.

If you have access to a lathe for machining a fascinating variant is to copy the power rings on the ISS that connect the rotating solar panels to the non rotating station.  Rather than a brush the rotating and stationary part are connected by a hoop of conducting material that is slightly compressed in a track somewhat like the races on a bearing.  They use three of these hoops spaced 120 apart, and obviously have some mechanism to keep them at the 120 degree spots.  Minor variations in hoop diameter would have them running into each other otherwise.

[luiHS]

 
I also thought about making my own rings with PCB board and brushes. The PCB has no problem, I design it and the Chinese make it for me, but the brush system is more complex. It is not about manufacturing a unit using some components taken from scrapping parts, because once I have designed this product I will have to manufacture it in a certain quantity, therefore I cannot depend on scrapping parts.

I suppose that the brushes with their spring and their charcoals can be bought from the Chinese, the case is if it is worth designing and assembling my own slip ring, instead of using a commercial one. Slip rings in the Chinese cost between USD 5 and 16 per unit, and are ready to use.

In the attached image, someone already did something similar to what I need, his ferris wheel is for a model of trains, mine is not for that application, but it is similar. Theirs is simpler because it only transfers power for simple LEDs, in my case they are APA102 digital LEDs and I also need data and clock signals, in total 4 signals, but if I duplicate data and clock to avoid problems due to bad contacts in the brushes, I already need a minimum of 6 contacts.

The use of the system to couple two annular coils inductively also complicates the design to do so with some frequency. Unless these annular coils I can also do them with PCBs, in that case the solution would be quite good, because in a single PCB I could put three annular coils that would fit with the other three annular coils on the ferris wheel PCB. Maybe try it.

 

[luiHS]

 
Looking for information, what I found interesting for the motor to use with the ferris wheel, are these PCB stepper motor. I do not know if there is something commercial, but the design does not seem complicated, I understand that there are two round PCBs, one with the printed coils and the other with permanent magnets, it does not seem very complicated to do, although if there are already facts, much better.

Another thing is the power they may have, if it will be enough to move my ferris wheel.
Anyone have experience with this kind of stepper motors, is it a commercial product that can be bought?
https://hackaday.com/2018/03/28/a-brushless-motor-on-a-pcb-made-from-pcb/
 
 

[beanflying]

Given the rotational speed you will likely want KISS and use a DC brushed motor for maximum torque at slow speed anything from a simple relay to a FET/PWM from a micro will do you. Either that or you will need at least a double reduction and more control gear if you go brushless as their torque on startup is not as good.

If you strip a CD drive some of those will have a PCB type brushless and some gears but I wouldn't go down that path because of the control complexity.