spinner sensitivity

[tony.silveira]

hey guys,

i have a small issue with my turbo twist 2 spinners, they seem to be way to sensitive and i'm getting what i think is called backspin.  meaning, if i spin it really hard in one direction, it registers as if it's going the other way.  is there some windows setting where i can cut the sensitivity?  i've tried the sensitivity in mame but that doesn't really seem to help.

thanks

[gamuhar]

What MAME, what game, what Windows? Did they work properly before, what MAME, game and Windows was then?

There is "mouse polling rate" in Windows set as default to 125Hz. If you think some software or new hardware drivers might have changed that value then you should check it by downloading program called "mouserate". If it's already at 125Hz don't set it any lower, and actually you may try increasing the rate, but I'd assume the real problem in that case would be somewhere else, MAME settings most likely.

I'd try temporarily moving or deleting INI file and stuff in CFG folder, then take at least three games and test it again.



In another thread about 720 Degrees controller I am stuck at similar question as yours, only I don't have a spinner yet, so if you take 720 degrees game for one of your test games you will help us both solve our problems, besides 720 Degrees is excellent game to test your spinner accuracy and robustness all in the same time.

http://forum.arcadecontrols.com/index.php?topic=117130.0



If you don't already have a dot you can just mark it with a piece of sticky tape, screw-handle is optional. The point of the test is to set up sensitivity in MAME so the dot always points the same way as character by matching their rotation 1:1. Then start making turns all in one direction and the number of turns you can make before the dot de-synchronize with the character will determine spinner's accuracy and robustness. Ideally, if it's possible to match rotations 1:1 via sensitivity settings, and if the spinner is well made, you should be able to make dozens, if not hundreds, of full circles in one direction with the dot and the character still being aligned.

[RandyT]

True "backspin" is not possible with any TurboTwist2's shipped after 03/09.  There is code in the controller to prevent any possibility of this happening.  Increasing the mouse polling rate will prevent it with earlier versions.  Regardless, there are no games which require the hard spin where this could occur, so it has no effect on actual gameplay, even with the earlier units.

Something else to keep in mind is that some "backspin" will be tied specifically to the game you are playing, meaning that if you can do it with the post 03/09 TurboTwist2, it can be done with the original spinner as well.  A good example of this is with Tempest.  A sensitivity setting of 6 will give a 1:1 resolution correlation to the actual controller shipped with the game.  But that particular ROM will show backspin if the spinner is spun hard enough.  It will do this with any spinner, including the original.  As for the sensitivity setting in MAME not having any effect, I don't believe that this is actually the case.  It may not eradicate what you have been observing, but the lower settings should improve it considerably.  Higher sensitivity settings than 6 in Tempest, for example, will exacerbate the issue inherent to this particular ROM.

As for 720, it looks like the same resolution encoder was used for that game as was used for Tempest, and being that both are Atari titles, I wouldn't hold much hope for any regular spinner being able to stay in sync.  There is a reason why that title included an extra index wheel to re-sync with the control on every rotation, and I believe it has equally to do with the initial syncing of the control position upon power up, as well as the possibility that the user could induce a "backspin" which would un-sync it in the absence of the regularly occurring index pulse.  Also, since the original was "geared" through the use of a chain mechanism, one would need to know exactly the number of pulses per revolution to get the sensitivity correct.  If this is off by even the tiniest bit, sync would be lost.

To do a proper controller of this nature, the same type of setup would need to be duplicated.  Something as simple as a magnetic reed switch in proper proximity of a small magnet attached to the ESC, or other holder fixed to the lower part of the shaft, should be able to provide a reasonably good index pulse for this purpose.  The game would also need to support a "button" for the index input, which IIRC, later versions do.

RandyT

[gamuhar]

As for 720, it looks like the same resolution encoder was used for that game as was used for Tempest, and being that both are Atari titles, I wouldn't hold much hope for any regular spinner being able to stay in sync.  There is a reason why that title included an extra index wheel to re-sync with the control on every rotation, and I believe it has equally to do with the initial syncing of the control position upon power up, as well as the possibility that the user could induce a "backspin" which would un-sync it in the absence of the regularly occurring index pulse.  Also, since the original was "geared" through the use of a chain mechanism, one would need to know exactly the number of pulses per revolution to get the sensitivity correct.  If this is off by even the tiniest bit, sync would be lost.

Sounds like a bet to me. How about I make an order for two spinners from you right now. Then you perform the test as I described above, but try as best as you can to first match rotation 1:1, and if you then make more than 10 full circles without de-synchronization you give me one spinner for free. How about it?

Although, it sounds like you'd be betting against your own spinners. Never underestimate the power of your own spinners, maaan!! I'd bet they can be synchronized perfectly with ANY game and stay in sync for hundreds and thousands of turns! You get to sell at least one spinner in any case, so do you accept the bet?

[RandyT]

No need for such silliness.  If the code in 720 checks the index after the first time the control is calibrated, and I am quite sure it does, then it does so for the reasons I described above.  Otherwise, it's just wasting cycles.  If the game itself needs to keep checking for synchronization with it's own control, then it similarly will need to do so with any other.  Also, if the number of pulses per revolution, as defined by the gearing, is not evenly divisible as whole percentage points into the native 1200 transitions of the TT2, then it's an exercise in futility.

Relative positioning encoders <> absolute positioning encoders.  The TT2 (and every other normal spinner, mouse, trackball, etc...) is the former, and the 720 controller is much more closely related to the latter.

[gamuhar]

Silly? Oh, come on!

If, if, if. You could try it out and know for a fact. I'm asking you as a customer, ok? But, all I really need to say is: "I think you are wrong, sorry"... and argument goes on and we still don't know anything for sure, which is more silly?

Only a Sith deals in absolutes, 720 Degrees encoder is relative like spinners, trackballs and mice.

[molton]

... I'm asking you as a customer, ok? But, all I really need to say is: "I think you are wrong, sorry"... and argument goes on and we still don't know anything for sure, which is more silly?...

What is this?  Are you threatening RandyT to give you free stuff or you'll act like you know what you are talking about?  

I for one love groovy game gear stuff and want Randy to stay in business and don't appreciate this kind of silliness either, if he listened to your demands he'd have to give free ---shoe--- to every idiot on the internet.

I've experienced weirdness with the TT2's in mame with the sensitivity on 100% too, but I was never not able to correct the problem by lowering the sensitivity.  the TT2 has at least double the resolution of any spinner emulated in mame anyway, so expect weirdness if everything is set to 100%  I definitely have new ones and have only been playing around with them for a little, I don't mess with that 720 game though, my computer/mame acts really strange with that specific game.

[gamuhar]

RandyT,

I took old Apple iMac's ball mouse, you know those transparent ones with only one button...



I used fresh install of MAME 0.134 and "spinner" is actually default "driver" settings, so all I had to do is set sensitivity to 85% (by trial end error) and it worked... worked well enough for me to now be quite certain your spinners could do it well enough to actually play the game seriously.

How well it worked? With 85% sensitivity I could turn it more than 20 full circles, but slowly. If I turned fast it would come shorter and shorter every two or three turns. If I set sensitivity to 86% it would work better with faster speeds, but not as good as with slow. You also have to take into account my poor setup. Every time I turn the wheel (my finger is from down below where the ball goes) it all gets shaken, more so the faster I try to spin, so I believe these results really do mean your spinners ought to do much better.

In any case I forgot to mention 'turning speed' before, it's without a doubt very important factor in this whole story, not only due to more stress on equipment, but also due to such limitations as is polling rate of Operating System, size/number of notches in encoder discs, and poling rate of the game itself. At some point every spinner type device will have some maximum turn rate of their encoder disc relative to Operating System and each specific game, so the test and calibration should be performed on each game separately and with some "normal maximum" speed you would use in the game, not maximum possible speed you could turn your particular spinner.


I really want to buy couple of spinners, some big and wide ones. What do you think about this modification and could you make something like that for me on one of the spinners, for free? It only needs to be good enough so it doesn't fall apart as easy as if I made it with a screw, and you can probably come up with something even better:

[gamuhar]

... I'm asking you as a customer, ok? But, all I really need to say is: "I think you are wrong, sorry"... and argument goes on and we still don't know anything for sure, which is more silly?...

What is this?  Are you threatening RandyT to give you free stuff or you'll act like you know what you are talking about?  

Hahaha! That is indeed funny, touché!

I for one love groovy game gear stuff and want Randy to stay in business and don't appreciate this kind of silliness either, if he listened to your demands he'd have to give free ---shoe--- to every idiot on the internet.

I've experienced weirdness with the TT2's in mame with the sensitivity on 100% too, but I was never not able to correct the problem by lowering the sensitivity.  the TT2 has at least double the resolution of any spinner emulated in mame anyway, so expect weirdness if everything is set to 100%

I'm not so bad, once you get to know me. Why are you taking offense anyway? He knows I'm only being cheeky, I don't think he took any offense himself, he just got all "grown up" on me probably because he's too busy lately. In any case, I love anyone who relates himself with an imagery of a wizard, especially if they make joysticks and spinners for a living. How cool is that?!

[tony.silveira]

RandyT, as always, a hot bed of information!

you mentioned spinners dated 3/09 wouldn't show this behavious.  anything that can be done to "fix" pre 09 spinners?  soem kind of bios flash for the board is it physical?

i'll give tempest a go tonight with it set at 6.

i do love me some groovygamegear!

here's a thought, let's start a list of spinner games with the optimal settings for a TT2.  I'll start us off:

tempest - 6 (ok, i cheated...)

[nick3092]

here's a thought, let's start a list of spinner games with the optimal settings for a TT2.  I'll start us off:

tempest - 6 (ok, i cheated...)

This comes up about once every 6 months.  You can calculate this with a simple math formula found with many of the spinner counts on the wiki:

http://wiki.arcadecontrols.com/wiki/Spinner_Turn_Count

The TT2 has a 1200 turn count.  Form there, you can do the math for any game listed.

[RandyT]

...if he listened to your demands he'd have to give free ---shoe--- to every idiot on the internet.

Well stated.  We wouldn't be around long if we participated in shenanigans, so I steer clear whenever possible.

After scrounging up the hardware schematic for the original controller, it seems the chain gearing of the controller is only for the damping effect.  It appears, contrary to my earlier assumption, to be a direct drive device, so given the information in the wiki, the proper sensitivity setting should be 12%.  12% of 1200 = 144.

Of course, this still doesn't change the fact that there needs to be an index signal generated by the control to maintain synchronization in the event that the player exceeds the ROMs ability to process the decoding.  If the player can do it in Tempest, there's a very high likelihood it can happen in 720 as well.  The indexing wheel is there, and used by the code during play, for a reason.  That being stated, if one doesn't intentionally try to knock it out of sync with a spin that is faster than whatever that threshold might be, then it should, in theory, remain in sync, once calibrated.  Much depends upon the way the control is handled by MAME here as well.  If the output from the spinner is averaged, or interpolated in any way, all bets are off.

Gamuhar:  Yes, I can put something together, knob-wise, for you to approximate the control as you have it shown.  With the AccuTwist option, reasonable damping could also be achieved.  One could even drill and tap (or install a bushing in) the 5" steering wheel and install a joystick shaft.  However, the radius of the wheel or knob to the point of the extension will have a direct impact on the playability of the game.  Too short, and it will be hard to control.  Too long, and it will feel less responsive than the original.  As the knob needs to be 1:1 for positioning purposes, this is not something which sensitivity twiddling will be able to overcome.  If you are serious about this, and willing to compensate me for the time, parts and effort involved in the endeavor, I can do what you want.  It could be simple and inexpensive, or something which will be closer to the original, and likely more complex and costly.  Just email me with the specifics.

you mentioned spinners dated 3/09 wouldn't show this behavious.  anything that can be done to "fix" pre 09 spinners?  soem kind of bios flash for the board is it physical?

If the board you have is the current surface mount version, we can replace the old chip with one containing the post 03/09 code for a nominal fee.  Email me for details.  Otherwise, upping the poll rate in the OS achieves the same end result.

RandyT

[gamuhar]

We wouldn't be around long if we participated in shenanigans, so I steer clear whenever possible.

All grown up I see... you'll lose your marbles like that, can you even catch your own shadow anymore? Shenanigans!? You must be running out of magic dust, it's a challenge and adventure! It's what fairies fart and what dreams are made of.

Of course, this still doesn't change the fact that there needs to be an index signal generated by the control to maintain synchronization in the event that the player exceeds the ROMs ability to process the decoding.  If the player can do it in Tempest, there's a very high likelihood it can happen in 720 as well.

How well can Tempest do it, how do you measure? Is position in Tempest also "absolute" similar to 720 Degrees?

The indexing wheel is there, and used by the code during play, for a reason.  That being stated, if one doesn't intentionally try to knock it out of sync with a spin that is faster than whatever that threshold might be, then it should, in theory, remain in sync, once calibrated.  Much depends upon the way the control is handled by MAME here as well.  If the output from the spinner is averaged, or interpolated in any way, all bets are off.

It's not averaged, it seems to work well enough in either "spinner", or "real" 'driver mode' where you can map calibration input to a button and so calibrate during the game if necessary.

Gamuhar:  Yes, I can put something together, knob-wise, for you to approximate the control as you have it shown.  With the AccuTwist option, reasonable damping could also be achieved.  One could even drill and tap (or install a bushing in) the 5" steering wheel and install a joystick shaft.  However, the radius of the wheel or knob to the point of the extension will have a direct impact on the playability of the game.  Too short, and it will be hard to control.  Too long, and it will feel less responsive than the original.  As the knob needs to be 1:1 for positioning purposes, this is not something which sensitivity twiddling will be able to overcome.  If you are serious about this, and willing to compensate me for the time, parts and effort involved in the endeavor, I can do what you want.  It could be simple and inexpensive, or something which will be closer to the original, and likely more complex and costly.  Just email me with the specifics.

Now we're talking! I'm for the simplest (cheapest) solution that produces "playable" results and doesn't fall apart after 15 minutes, for the prototype and proof of concept kind of thing. I am not willing to compensate you for it, unless you manage to convince me better. I think I am already sacrificing a spinner and you get to have all the fun with experimenting. I'd just let you do it because you probably can find more suitable parts and would know better what and how to combine.

720 Degrees is very good game, and like Star Wars not many people get to play it today due to necessity to have authentic controller to play it properly. It's a shame, and so there is market potentially for this mod. I was hoping you will see interest in it for your own knowledge and possibility to make and sell more of these modifications. So, while I'm not willing to compensate you for testing and prototyping, I'd like to share profits if it turns out this thing works well and you get other people asking you to make them one these too. We could work together, I'll pay for half of everything then, and do half or as much as I can, and share profits accordingly, and I'll give 70% of my profits to UNICEF that we can write off as your donation so you get some tax returns or whatever benefits that can give you. I'd kind of work for your company and also provide half of the investment, sounds like an adventure?

[Xiaou2]

Gamuhur, you really have no clue... and so you are just becoming a troll.

 If you dont want something that will break in 15 min, that actually plays the game without losing any calibration.. and plays the game properly... its not going to be a quick and cheap hack.

 As for Starwars, you can play it with a mouse, or an analog joystick "ok".  The problem is that it does not feel as good as a true yoke.   Having the correct feel, means a lot of engineering, and a lot of expensive development, and a high price tag in parts & assembly.

 720's controller is the same as Starwars, but worse.  The game in unplayable without having an accurate dual encoder system in place.   But even if you could get it to work without that... you still cant just hack a handle onto a spinner.   It will never feel right. Never control well.  And will probably tear itself apart in a very short time period.

 Theres are countless reasons why a 720 controller is like 10 inches deep, and contains a boatload of parts.  Everything is made to specifications that allow the design the correct control, player comfort, player safety, and lasting durability.

 One time I made a spinner with a 1/4" shaft.  Didnt feel right at all.  The smaller shaft spinners were easier to spin, spun faster, and just felt better.  Theres a lot of science in mechanical design to consider.  Which is why most arcade controllers went through extensive testing and revisions, before hitting the arcade floor.

[gamuhar]

Gamuhur, you really have no clue... and so you are just becoming a troll.

You mean like last time (http://forum.arcadecontrols.com/index.php?topic=117130.0) when you attacked me with your assumptions and when it turned out you were wrong you just went away without so much as saying "sorry" for your childish insult?

If you dont want something that will break in 15 min, that actually plays the game without losing any calibration.. and plays the game properly... its not going to be a quick and cheap hack.

Hey, kiddo. Let Mom and Dad talk for a minute, will ya? I'm talking to Randy here, and I said I want "proof of concept". You make yourself whatever and however you like, stop bothering me.

As for Starwars, you can play it with a mouse, or an analog joystick "ok".  The problem is that it does not feel as good as a true yoke.   Having the correct feel, means a lot of engineering, and a lot of expensive development, and a high price tag in parts & assembly.

 720's controller is the same as Starwars, but worse.  The game in unplayable without having an accurate dual encoder system in place.   But even if you could get it to work without that... you still cant just hack a handle onto a spinner.   It will never feel right. Never control well.  And will probably tear itself apart in a very short time period.

Go play Star Wars with a mouse then and get of my back, goblin.

Theres are countless reasons why a 720 controller is like 10 inches deep, and contains a boatload of parts.  Everything is made to specifications that allow the design the correct control, player comfort, player safety, and lasting durability.

 One time I made a spinner with a 1/4" shaft.  Didnt feel right at all.  The smaller shaft spinners were easier to spin, spun faster, and just felt better.  Theres a lot of science in mechanical design to consider.  Which is why most arcade controllers went through extensive testing and revisions, before hitting the arcade floor.

You are boring me. Go watch some TV and do whatever you like, I prefer to try the things that have not yet been done.

[tony.silveira]

Post count:

2,983 - Xiaou2

20 - Gamuhar

who's the parent and who's the kid saying he wants free stuff?   

[gamuhar]

Heh. Then read his post and see if that solves your problem.

[CheffoJeffo]

We all know who Gamahar is, right ?

Banned under at three other nicks that I know of. Spammer of BYOAC,  MW and KLOV (he is the guy who is the reason for the new vouching/donating rules over at KLOV).

All around general nitwit troll.

[ark_ader]

We all know who Gamahar is, right ?

Banned under at three other nicks that I know of. Spammer of BYOAC,  MW and KLOV (he is the guy who is the reason for the new vouching/donating rules over at KLOV).

All around general nitwit troll.

Don't forget Mameworld.   

If you don't mind me getting slightly back on topic and hijacking this thread.....

I have a mouse hack that I want to change sensitivity. 

I printed out some spindle wheel templates, but if I want more sensitivity I would be looking at something with less spokes right?