Projektti Mörkö - The Bogeyman - DD Wheel


#1

In Finnish language ‘Mörkö’ means Bogeyman - or something such.

From Wikipedia[color=green]
[i]Bogeyman (usually spelled boogeyman in the U.S.; also spelled bogieman or boogie man; see American and British English spelling differences), pronounced /bʊɡimæn/ or /boʊɡimæn/, is a common allusion to a mythical creature in many cultures used by adults to frighten children into good behavior.

This monster has no specific appearance, and conceptions about it can vary drastically from household to household within the same community; in many cases, he has no set appearance in the mind of an adult or child, but is simply a non-specific embodiment of terror. Parents may tell their children that if they misbehave, the bogeyman will get them.

Bogeymen may target a specific mischief—for instance, a bogeyman that punishes children who suck their thumbs—or general misbehaviour, depending on what purpose needs serving. In some cases, the bogeyman is a nickname for the Devil. Bogeyman tales vary by region. The bogeyman is usually a masculine entity, but can be any gender, or simply be androgynous.[/i][/color]

As some of you may know, I have been involved with DIY DD wheels for quite a number of years already, since mid-2014 if memory serves. See my Bruteforce 3 log here for a bit more info: http://members.iracing.com/jforum/posts/list/3491567.page#9904030

During that period, I have built quite many systems, and tested even more servos:

  1. Big and small MiGe
  2. Lenze LV
  3. AKM52G
  4. AKM53K and AKM 53G
  5. Bodnar SS2 AKM54G

Other than 5) above, and after my Argon experiments, all the above servos were powered by SimuCUBE and IONI Pro HC. The small MiGe was the only one in that list that could realize its full potential, the others almost all voltage-limited and not able to extract the full torque potential.

Some time ago I got my hands on an AKM65L (and more recently, also an AKM64L) - the idea always was to built something special using them, but they got packed on the shelf and almost forgotten - until Sam Maxwell decided to built a monstrous wheel with the AKM73L - a fair bit bigger than mine, but with 3x as much inertia.

Here is a link to Sam’s built, quite interesting approach on some of the items, check it out! http://members.iracing.com/jforum/posts/list/3575875.page

I was searching for a servo that would provide natural damping due to higher inertia, in the order of 3x or 4x what the AKM54G would provide, so as to reduce electronic damping and associated latencies.

Well, this AKM65L may just be that servo - Here shown next to the Lenze, which in itself is a rather big servo unit…but the 65L is significantly larger.

And compared with the Lenze and an AKM53G - this is going to be fun I think…

And a few more for scale…although the photos don’t do it justice.

Truly a case of David and Goliath, as who would have thought that such a small drive-controller would be able to drive this servo to very high torque-levels?

Money-shot right here - picture of the slightly longer 64L, it has a brake fitted inside the rear cover as well…:


Simucube 1 vs simucube 2
#2

Before going crazy and ripping my existing 53K from the rig, I decided to test this thing on the bench first.

It shipped with an EnDAT encoder, which we know is not supported by the IONI - but it has secondary SinCOS channels, which we know is…I had to make up an encoder cable quick, luck I had the right plugs and pins around - to easy…

So a quick and easy quality 2M CPR SinCOS setup, should feel pretty decent in iRacing when combined with the latest beta SimuCUBE FW.

Here ready to go:

Basic setup done inside Granity…

\

As well as in the Tuner for the SimuCUBE:


#3

Well, after establishing the servo to be in great working condition, I designed a new mounting plate that will allow this servo to be a bolt-in replacement for my AKM53K…

A big thanx to George Alamaras for cutting these for me, I’m very happy with the end result. I also got two wheel adapters, one for each of the 64 and 65 servos…all parts are machined from 16mm aluminium - probably overkill, but I luv rigid, can’t stand sloppy vibrating setups…

Funny though, it reminds me a lot of a Flint-mobile…

I had to brush the servo mount, as it was cut on a router, whilst the adapters were machined on a Tormach Mill…almost ready, and no comments about my color-scheme, please :slight_smile:

Here it’s fitted with the mount, as well as a test-fit of the adapter - for my first tests, I will use a green 3d-printed adapter, which I designed a few days ago. Reason is I need to tap the 70PCD M5 holes on the aluminium unit, but I need to get a new tap first. So PLA it will be for now.

But it still looks rather nice, I think - note the panic button - I have never used one before, although I have fitted them to the units I sent to Mates - but this is the first time I am deliciously scared of a servo, especially seeing it is Bathurst week in the GT3 this week…

Also note the button on the right, which is a remote power-switch for my PC.

Removed the AKM53K from the rig, here shown next to the 65L.

And fitted to the rig, how is if stuff just fits…it literally took 10 minutes to remove the old one and fit the new setup…damn, if life was just this easy…

And the OMP 310ALU bolted down, ready for some high-torque action…


#4

Well, before testing it out, I decided to upgrade my setup from the SDR-480 PSU to the SDR-960 setup - this servo will do just about a real 38NM at the 25A POS the IONI Pro HC can deliver - the servo is rated at ~62NM torque though, but in this application, we are current-limited, which is just as well.

I doubt 62NM will feel different to the 38NM any way, as those transient FFB spikes in iRacing is so narrow and sharp, and with the low 60Hz FFB update rate, there won’t be any way we will be able to feel them.

38NM is still a lot of torque, my plans is to run the GT3 cars 1:1 - thus I will have massive FFB, but zero saturation, and all the fidelity I could want, spiced up with a bit of good old mechanical damping. Life is good…

Anyway, digressing as usual - below you can see the new SDR-960 fitted, and I am also prepping a spare r1005 SimuCUBE, adding a small fan just to keep things nice and cool. As an added bonus, it comes with dual resistors to share the power-dissipation on the regeneration side, the only reason I am replacing the r1004 unit in use up to now…

Just peace of mind with the large servo, nothing else…

And fitted - quick and easy, how much fun to assemble one of these SimuCUBE systems, when compared to the Argons or IONI Cube/Discovery builds…

And a few final pics showing off the beautiful work that is the Augury case - Jesús specifically designed this one for me so I can fit the SDR-960 PSU - thank you again Augury, you guys are tops!!


#5

I will update with my first drive experience during the next few days, need to formulate what I experienced into words.

Cheers,
Beano


#6

ummm i got an akm 63m at home and i will look carefully what is the result of this experiment is, im selling it right now on ebay cause it will not fit the osw project well as it would need a massive power supply, 17 nm with 13,8 Arms @ stall 1,23 nm/A, it is fitted with a resolver too that will be another problem, i would need to fit a sincos or something similar into it, thats why i got the more simple small mige route, a motor that i think would be a dream for this is the allen bradley mpl 540d very low inertia and high torque coefficient, the thing is to find it with the suitable feedback device. It has 0.00147 kg-m2 inertia, 19,4 nm at stall and 10,5 Arms at stall.


#7

Can’t wait to hear about this absolutely mad combination :flushed:
Hopefully it doesn’t beat you to death with your own arms!!
Love it :sunglasses:

Seriously though, I think this is going to be super smooooth, we all get to find out soon!


#8

Wow, very nice project and all of your work and rig looks to be top notch indeed. :yum:


#9

The 63M could probably be fine running off of an SDR and it shouldn’t be voltage limited but it does have a relatively low torque constant which would limit its calculated Nm output to a relatively LOW 21.92Nm for it’s size. (Large Mige or 54K are better)

I just purchased a 73L, like the one Sam Maxwell Is using now but I have no idea when I will get it running with the current IONI and 48V it has a calculated top Nm of 61.87Nm…

Beanos 65L has a calculated Nm of 37.12

Should be fun, Once I get mine up there will be three of these monsters out there lol…


#10

<Hi Guys,
Now, after having used this big servo for a while, herewith a quick overview of my thoughts on the Bogeyman vs Bodnar SS2 with 54G servo.

What is it, where to buy
The latest SimuCUBE/IONI-based DD wheel is the successor to the original Argon- and IONICube 1x/4x direct-drive steering wheels.

The very first version Argon-based wheel basically was born through the efforts of Bernhard Berger, one of our German iRacers, using an open-source code and communicating via RS-485 to your PC. Michael M (MMoS) later developed a pretty nice USB HID closed-source interface, running on the STM32F Discovery development board. First versions were Discovery/Argon-based hardware solutions.

Around this time I started discussions with Granite Devices to develop a better solution for FFB wheels, and first, the IONI was born, followed shortly thereafter by a nice small daughterboard, the IONICube 1x and 4x. This solution, when paired with the 32STM Discovery Dev board, flashed with the MMoS FW, provided many hours of joy to SimRacers globally. There however were complexities in getting these up and running, as well as dealing with EMI issues.

This prompted me to have the next level discussions with Granite Devices again - Tero and his team was superb in listening to my requirements, and thus the SimiCUBE was born. More on history here: http://members.iracing.com/jforum/posts/list/3491567.page

This brings us to my latest incarnation of the SimuCBE-based DD wheel, ‘Bogeyman’ - you can view the buildlog by following the next link, it should give you a good overview of the hardware used - http://members.iracing.com/jforum/posts/list/3581690.page.

Although, unlike the Bodnar SS2, reviewed earlier, which is developed for professional use, the SimuCUBE-based DD wheels are mostly assembled and supplied by a variety of iRacing members, a few below:

Ollie at https://sim-pli.city/collections/all
Tomo at https://www.simracingbay.com/shop/
Carlos at https://www.simracingcoach.com/en/contenido/kit-augury-osw/

Components
My specific unit is comprised of:

AKM65L servo motor (Bogeyman)
SimuCUBE interface and IONI Drive Controller unit
Motor power and encoder cables
Emergency stop switch
USB cables

Specifications
A wide variety of servomotors are compatible and available for use on this system, however, for the sake of an apples-to-apples comparison vs the earlier-reviewed Bodnar SS2, I am using an AKM65L servo with 2.1M CPR. a Few of the tested options listed below:

130ST-M10010 - 20NM (small MiGe)
130ST-M15015 - 28NM (large MiGe)
Lenze MCS12H15L - 25NM

AKM52G - 16NM
AKM53G - 20NM
AKM53K - 22NM
AKM54G - 24NM
AKM65L - 38NM

Please note the servos might have a higher torque rating, however, I am listing what is possible being driven via the 48V HV PSU, as commonly used in the SimuCUBE builds. I followed the same standard for the Bodnar SS2 servos.

Alos note the Bodnar SS2 Servos are not directly compatible for use on the SimuCUBE builds, as they have SFD encoders, with proprietary protocol. You could however, replace those with either BiSS, CoSine or Incremental Encoders, which are compatible. We have a few iRacing members selling kits/upgrades for a variety of Servos, Davy Watts being the main one.

Installation
Mechanical installation onto my 8020 rig was relatively quick and easy - I had to remove my custom SS2 AKM54G servo steering mount, and replaced it with a new custom mount that accepted the much larger AKM65L servo. Other than that, I had to fab a new wheel adapter plate and use a 32MM ID Beltingonline split-bush, but the idea is very much the same as the Bodnar SS2 solution.

Electrical installation was painless, with simple plug-and-play connectors on the SimuCUBE. Nice and easy and very reliable for this application, although the ones used on the SS2 is more industrial and hardcore (You can however DIY the same solution into a SimuCUBE build, which is what I often do). The rest was connecting the USB B cable, and mains.

Setup
Setup is quick and rather easy, once you get the hang of the new Granite Devices TunerTool - this is the UI used to interact with the new SimuCUBE FW (non-MMoS) - it has pretty nice new features and filters compared to the old MMoS FW, which should allow anyone to tune to a FFB response they might like. One nice feature available here, as in the SS2 UI, is the graphical animation, showing the recommended calibration window and center position of your wheel.

The SinCOS encoder I have installed on this servo still requires the setup to go through the well-known phasing process, but it is quick and painless, and the wheel is available for use within 30 seconds after power-on.

Settings and Options
The User Interface is rather advanced, but well layed-out and specific features are available on tabs. Easy to navigate, should not take one more than a few minutes to get familiar with the general layout and operation. It is a very powerful user-interface, with quite many options to tune the final FFB behaviour to your liking. There also is a pretty good manual available that explains the basics, all that is left for you is to experiment should you wish to do so.

Some comments on these in the section below.

First impressions
The Bodnar SS2 is currently touted as the best DD FFB wheel out there - can it defend it’s title against this setup? Mmm…a tough one, please see my earlier impressions on the SS2 above.

My take on the Bogeyman - I will highlight both the good and the bad:

The good
I prefer to DIY my own solutions, as I select components, like PSU’s, based on the servo-requirements, rather than what would provide best bang-for-the-buck or best profitability. As such I can construct a quality-comparable solution to the SS2, which always was my aim. The Bodnar SS2 is a complete package, with (almost) all parts included to make it a complete DD wheel solution. In the DIY ‘OSW’ world though, I have to source each component/part - but, you can also find the ‘OSW’ as a complete package, see links earlier.

The Bogeyman wheel takes response and tracking accuracy to another level again, as the SinCOS encoder does inherently have a higher positional accuracy when compared with the SFD used in the SS2. I also find general latency to be lower, with an immediacy to input/output which is hard to describe.

As with the SS2, the IONI Controller is running at around 2million counts from the encoder (in my current build), whilst this signal is stepped down to ~65,535 counts, the maximum windows can handle. I have however successfully tested a 4M CPR BiSS encoder, mounted on a 53G servo.

Catching slides with the Ford GT3 around Motegi Twin, the BMW around Silverstone and the Mercedes around Monza and Yokohama is even easier than ever before, I suspect very much due to the low latency between the Sim and the Drive Controller. Counter-steer was intuitive and very natural, even better than the SS2. I could not fault the steering accuracy and response behaviour - even one would expect that the massive 65L servo, with higher inertia, would dampen things down, I actually found the feel to be closer to real-life than any DD wheel before, the pretty wild 38NM torque at 1:1 assisted in conveying exactly what the Sim was putting out.

Damping, Friction and Inertia worked very similar to the SS2 wheel, but the new Reconstruction Filter implemented in the IONI-side just take tings to another level - this has been my gripe with my SS2 setup, always making me think I was driving on over-inflated balloon tyres - instead of race-tyres. With the Recon-Filter, you can go from one extreme to the other, with everything in between.

I typically opt for setting 3, which gives me a pretty realistic race-rubber FFB response, whilst setting 5 feels like low-profile road tyres and setting 1, whilst providing the lowest latency, is just a bit to rough for my liking, reminding me of the older Argon-based solution with zero damping on the MMoS FW.

Overall, the best DD wheel I have tested over an extended period by a fair margin - I do believe Leo can attain similar levels of FFB behaviour from the SS2, but it will require a bit of effort on his controller and HID FW. Great potential, but that setup is currently been undone by a heavily-filtered input, even more pronounced now that I had an opportunity to compare both systems on high-end AKM servos.

Main area for improvement
Oscillation is one area where there is work to be done, not so much on the SimuCUBE FW side though, as that can be very well controlled with Damping/Friction/Inertia settings. The iRacing Simulator though, is very dated with a 60Hz FFB update rate, which can be felt once you reduce the Reconstruction Filter to 0 or 1.

To fully control oscillation at 1:1 38NM on this big servo, with Amperage running at 100%, I have to run Damping at 4.5%, Friction and Inertia at between 1.5% each. With the above, the wheel behaves itself very well, and no amount of enticing can induce oscillation.

At these settings, the wheel still feels very responsive and fast, I do not gain much in running zero settings here vs my preferred ones - but of course, having a faster Sim FFB rate would allow even more latitude (without oscillatory behaviour).

Recommendations
At this point, without changes in iRacing FFB update rate, I believe we are at the pinnacle of DD wheel technology, even spending big additional amounts won’t have a significant benefit to your experience.

This Bogeyman setup as is, is arguably the best wheel out there at this point in time, my determination is after many months of testing SS2, vs a couple months with this wheel.

With that said, building a solution such as this will take you close to the price of the SS2, so my advice still stands - always test before you buy, as it is a big investment and I am aware of more than one dissatisfied buyer, irrespective the DD wheel technology.

As always, feel free to discuss.

Cheers,
Beano


#11

When will iracing finally takes us serious and make it a real priority upgrade the 60hz FFB rate?
I have made a few topics about it on the iRacing forums and kicked david tucker a bit :wink: but it hasn’t helpen much yet.

I would like to put some more pressure on them with the whole direct drive community. More people should speak up about it and we should stand together so they finally start working on it. They don’t seem to see this is a important and high priority thing.

They used to hide behind you won’t notice it on a g25. i even proved them i can notice it on a g25 with a perfect irffb 360hz test with a 10/10 score. Because i know what to look for. but a g25 is indeed to bad to notice a difference. but even with a fanatec csw v1 its very easy to spot and a big improvement.

Their problem is their engine / dodgy coding everything is tied together which makes it hard to turn up the rate without breaking other stuff. So it will take some work. But they seem to lazy to streamline and improve the coding because they don’t see the priority.

I would say this would be a very important core thing to upgrade and they are already years behind on it.
it should have been here by now. DD Wheels are becoming mainstream now.

Lets organize a topic for this @ iracing forum and get some people behind it!


#12

Stupid question, but is the difference between the Mige motors very big?
I mean are the AKM from a other world category or is it between 10%


#13

Hi Seb,
Yeah, I’m not sure when those changes will come, if ever - we can only hope :frowning:

Perhaps one day, as you say, there are very many DD wheels out there now, things have definitely changed in this regard during the last 2 years or do.

Perhaps I will message him again, I’m reluctant to start an open debate as their are so many over there that goes into verbal dhiarrhoa mode when this topic surface. David has always been very approachable on this topic, but done in the community just reacting in a very negative fashion.

Let’s see how it goes though, will message him perhaps and get his view first, before opening it up again.

What do you reckon?

@Shorty: These expensive servos are probably 10-15% better, but st many times the price. The small and large MiGe are worth their weight in gold, they are really good in their own rights!


#14

True topics seem to go in a bad direction pretty fast by the brainless fanboys, nothing is wrong we don’t need anything better :smiley: i don’t get those people.

Yes i think its good if you pm him, could be good to let him think about it again.

remember the topic i posted about downforce vs ffb and strength build up at speed. I just drove the McLaren mp4-30 a week ago just for fun.

I really hate it a cripple no downforce f1 car :wink: but boy was i surprised, i think that its the best car with most noticeable down force increased strength when you build up speed and go trough corners at speed.

The williams fw31 generates a lot more down force and you virtually feel no increase. still very funny to me. i start to think their power steering simulation is broken / wrong on the fw31 and some more cars.

or atleast something is wrong in it but i am the crazy person :smiley:

@ beano how much improvement would a 2M sincos be over a 10.000 prr encoder with a small mige


#15

Good comments, fully agreed. I will also test the MP4-30 and see how it goes, at least the DW12 feels quite good wrt aero-loading once the speed increases - you can clearly feel the effect on the wheel.

Anyway, herewith some comments from David Tucker ages ago on iRacing, after I first arranged an Argon-based DD wheel for him…interesting comments in Test4, exactly what I am expecting to happen. It will be even better with today’s higher-resolution encoders and significantly better FW now, compared to older MMoS FW, when that test was run.


#16

Hey Beano,

Any chance of another Talk & Drive Youtube video this time with the The Bogeyman in action, I’m pretty sure at some point were gonna see you spin like your in a washing machine while holding the wheel lol.

I’ve said it before and I’ll say it again, when I grow up I wanna be just like you lol


#17

:lol:

That brought a smile to my lips - for sure I can do another one, let’s see if I can fit it in by the weekend! It’s great fun driving with Bogeyman, but it’s also a very powerful beast if you get in trouble…don’t ask me how I know, haha


#18

Te upgrade to a SinCOS on a Small Mige is a HUGE improvement. It to me probably has the biggest effect on that particular servo and it is due to the reaction speed of the Small Mige… It really smooths it out and reduced the ticking and knocking as well as the sense that it is driving you as much… In Back to back tests on my setup it made the Small Mige feel better than my Large Mige with the 10K… and the step up in feel was greater with the Small to SinCOS than the Large to SinCOS.


#19

Just mailed Jesus about that beano Augury case. Need one!

And if you ever get tired of one of your strong kollmorgens, give me a hint! :wink:


#20

they already have a new version of their case ask about that one too :wink: