X-Plane 11 6DOF motion platform

@mariano68
That would be great to have your experience with the 2x32s. It would be far simpler for me if I had got those instead of the 2x60s with all the battery malarkey The motors I have draw 25A under load. So for others who follow us, this is an important design consideration both for cost and complexity.

I am continuing to pursue ideas for the Sabertooth 2x60. I sent an email to Dimension Engineering again:

My email:
Rather using 2 x 12V @ 30Ah lead acid batteries for each Sabertooth 2x60 driving 2 x 24v 450W DC motors (max current 25A fully loaded), could I use 10 x 2.7V 500 Farad super capacitors connected in series for a total capacity of 50 Farads and a rating of 27V? I have attached an example from eBay. These capacitors can start cars and so are capable of 100A+ but would they be suitable for the motor regenerative current? If they are then I wouldn't have to buy 6 x 12V @ 30Ah lead acid batteries saving weight, cost and safety.

DE reply:
You can try it, but I would be surprised if it worked.
If your max current per motor does not exceed 25A, then you may be able to use a smaller battery than 30Ah. Something like 12Ah might work.

So it looks like I will be getting 6 x 12V @33Ah batteries

Can you not just do as I did and use 1 Battery ?

@SilentChill
Initially, Dimension Engineering said I needed 30Ah for each Sabertooth 2x60. In their last reply they said 12Ah 'might work'. The actual difference in cost between a reasonable 12-15Ah battery and a 30Ah one is only $20 or so but there is a big difference in their working life. The smaller capacity ones have an expected life span of 5 years under good (read that low usage) conditions. Hence it pays to spend a bit more for the 30Ah ones. The cheapest suitable batteries I can source locally are around $60 each for 33Ah which works out to $120 for each Sabertooth or $360 for all three. Remember I have 24V 470W motors. The advantage of this approach is I only have to balance two power supplies (connected in parallel) and the batteries connected in series.

The other alternative, which you have proved works on your rig, is to connect all the supplies in parallel and put on a bigger battery for all the Sabertooths. The cost of 2 x 100Ah batteries (low to reasonable quality) is around $300. The advantage of this configuration is the simplicity of all the supplies in parallel but the disadvantage is getting all the supplies to equally supply current to the Sabertooths.

For both arrangements, there is still the problem of overcharging the battery because the simple LED power supplies do not have current limiting or charge sensing circuits.

1. As the second option is working for you, what brand and capacity is the battery you used?
2. I know you have said before that you simply set the output voltage of each supply to 13.8V so how has the battery behaved with this usage?
3. Lastly, when you power up the platform do you always have the supplies and battery connected when you are using it or some other arrangement?

I have a 680a 75ah battery. Got it from a scrap yard for 10 quid out of an Audi A6.

With the voltage at 13.8 it's only trickle charging and with the back current it's all been good battery holds voltage between 12.4 to 12.8 depending how long it's been switched off

Over 8 months of heavy use and no problem with the battery or PSU's.

I turn the PSU's on and then I turn the battery on . And when turning off I turn battery off and then the PSU's off.

Probably not the perfect way to do it but it works for me and works well.

@SilentChill
Thank you so much for replying. That is very valuable information to me. I will heed your experience and save myself some money

I have now come to a milestone of sorts - I have finished all my ordering of parts. Tomorrow, I pick up the batteries then I will have to wait for all the other parts as they arrive in the mail. Upon reflection the most critical component it seems is the motor: 12V or 24V, speed 3500-4500RPM and the power rating 270W-470W. This determines the drivers and batteries (what type, how many, cost), material for the top platform (weight) etc etc. So my advice to others is read through all the build threads with those aspects in mind. While theory and engineering play their part, so does the proof of a working system and sometimes those two aspects are at odds with one another.

Hi,

I added my 3D file to my thread
https://www.xsimulator.net/community/threads/6dof-sim-linear-actuator-diy-sabertooth-arduino.9392/

@baykah
Thanks so much for remembering and sharing
I will check it out tomorrow.

I have finally settled on my interpretation of how the 6DOF is going to work (updated 23JUL17@19:11):

Edit September: I have discovered that the Sabertooths respond to the Arduinos better if they are powered up first. The Arduinos send out an auto-baud command to the Sabertooth when the Arduino powers up. This means that the Sabertooth must be powered on first before the Arduino. So the turn on/off sequence will need to be changed. Also I don't think the Arduinos will need to be powered externally from the 12V source. They seem quite happy being powered by the USB hub. Once the Arduinos are powered-on they command the actuators to go the the middle of their travel. So the platform always starts from the same position.

The power up sequence is now:
1. Power supplies
2. Batteries
3. Powered hub
4. Gaming pc (otherwise the pc will power the hub on its own)

@SixDegreesOfFlight
ln another thread you asked about why I suggested using a cap bank, I should have been specific and said electrolytic capacitors.
Capacitors is a category of components and come in different flavors, polypropylene, mica, AC, DC etc
To me a super cap is more a battery than a capacitor, it is constructed the same as optima car batteries with a porous seperator between plates while regular caps are constructed with a dielectric seperator between the plates.

The biggest difference in using either a electrolytic capacitors and super capacitors is in the way they charge up.
A electro cap will pretty much charge up with unlimited current till it reaches the supply voltage.
A 2.1v super cap will charge up really fast say 0-1.6v and then slower till 2.1v, it is that slower charging rate at close to 2.1v that makes them not ideal for absorbing the back current from motors as they just can't accept the big charge. Hmmm just like charging a lead,nicad etc, battery, did l say they look like a battery in a electrolytic cap case.
Think of it as, 2 graphs electrolytic charging as a linear curve while the supercap charging as a logarithmic curve.

You do need to realize why DE are recommending a battery in parallel with your sabertooth?
It has to do with using a Switching mode power supply, they don't like to be backfed.
It is not so much the current that it doesn't like, it is the voltage rise seen at its output that shuts it down due to over voltage.
You need to buffer the output so something absorbs the current so the output doesn't rise too much to cause shutdown.
This is where a battery, resistor or electro cap bank comes in where it can absorb the current keeping the output voltage within range.

@OZHEAT
Thanks for that additional information.

DE seem quite reluctant to give firm advice about what the minimum requirements are for a Sabertooth 2x60. I would have thought that a few examples would help users make the right choices when it comes to dealing with the regenerative current. I am getting the impression that the engineering for this aspect is more of a 'black art' rather than applying a formula or at least some 'rule of thumb'. I have been reluctant to use lead-acid batteries for obvious reasons but in the end I bought them so at least that part of the mechanism is somewhat taken care of. It will be interesting to put a CRO across the battery once my platform is finished and is in use, so I can really see the dynamics of voltage noise. Thanks for your input, hopefully it will help others who take up the 6DOF challenge

Small update to the cost spreadsheet and a slight change to the schematic to remove the need for an additional 5V power supply for the Unos. I have also included the weight of the steel and seat in the spreadsheet which will give some indication of the weight of the top frame

sorry for kinda half post before, footy was on.
During the footy I thought 2 things to bring up.
1. A electro cap bank stores a lot of energy and is dangerous, it would be safer for diyers to just use sealed lead acid batteries.
2 super caps would work, add a few extra to say 30v worth per string. The costs would be you probably need more than 1 string.

I'm not really surprised DE will give defiant requirements.
It will all depend on what your motors,load inertia is during largest reversal, you need to absorb that energy so the smps doesn't shutdown.

using 12A sla batteries you can keep adding more in to your bank until it doesn't shutdown on max reversals.
monitor the voltage and you will see

@SixDegreesOfFlight , Are there three independent circuits in your schemat ? Will you have three independent power supplies? Did I understand correctly your scheme?

@Grizus
Yes that's right. If you are following the kind of build led by @SilentChill & @wannabeaflyer2 then you put all the power supplies in parallel, otherwise you split them up into three modules like @baykah (at least that is what it looks like from his photos). @SilentChill has shared his ideas about his approach in many replies and you have to consider that he has a fully working, proven system.

I suppose my concern is trying to balance the six power supplies in parallel. Inevitably, there will always be one or two supplies doing the 'heavy lifting' while the others cruise. It apparently, doesn't seem to matter too much seeing as his platform has been working fine for 8 months

So my thoughts at this stage is three separate modules each with two psu & two batteries driving a Sabertooth 2x60.

Hey Six, I was reading up on supplies in general and if their output sags as their current output goes up, they will apparently share the load fairly well since they kind of self-level. Supplies that don't sag but hold their rated/set output right up until they crowbar out don't share the load so well since they will vary at least slightly in output voltage and the one with the higher voltage will supply the most current. So how well paralleling power supplies works depends a lot on the characteristics of the supplies used.

I don't have much high power experience but lots of regular circuit experience so am wondering if the power supplies don't play well together if a small high power resistor could be put in series with each supply between the supplies and where the outputs get tied together. This could force the supply into running where the "output" voltage sags (due to voltage drop across the resistor) and that might let them share the load better. Again, this is for the problem child kind that otherwise don't share load well and would come at the expense of throwing away power at the added power resistors.

@Zed
Conventional wisdom would say that we shouldn't tie outputs together, but where necessary use diodes to prevent one output driving another. In practice we can use the valuable experience of builders like @SilentChill who has done exactly this - tie six supplies together and let the supplies sort themselves out. This approach many not work with some types of supplies but Craig has shown it does work reliably and well. I am following his advice and tying two supplies/batteries together and distributing them to be closer to each pair of actuators. My thinking is to reduce wiring, reduce electrical noise and provide some isolation between these very high current sources.

Pretty scary dealing with 470W motors, high current supplies and fast moving mechanics

I have no doubt that SilentChill's setup works reliably. I was just saying what I found online on how some supplies will share load well and some won't - depends on the characteristics of the supply. If you get the same supplies as SilentChill used then I would think you should be good to go too. But you even repeated my point in your reply - that not all supplies share load well. Some are made to share load and some aren't.

Diodes don't really help with current sharing. They prevent a supply that fails with a shorted output from being the dump for all the current other supplies can still deliver.

Good luck, though!

Yep, anything that has the power to physically throw you around should be treated with respect.