Real Instrument Panels

[Howard_Casto]

Ok so I mentioned this to RamJet, but I thought I might post it here in case some of you are gear heads and can add some info. 

I was searching the web last night and ran across this usb adaptor (actually it's a usb to serial port dongle)  for bmw instrument clusters for racing sims.  Apparently the whole cluster runs on a serial line?  Who would have guessed? The problem with that of course is it's bmw.  So if you want to pay 300-600 bucks for a rpm/speedo for your fake car knock yourself out. 

My question is are there any other cars that use a serial based cluster?  On ebay right now you can get NOS or gently used clusters for as little as 30 bucks.  It's just they are usually from a Grand Prix, Taurus or Saturn.  If we could find a serial cluster that's affordable I'd be willing to do some of the heavy lifting in terms of figuring out the serial protocol. 

Of course there is always the option of using an analog cluster, but then you need a avr with a buttload of analog pins, so you've essentially added 50-75 bucks to the cost of the project. 

[mcseforsale]

You need to research the CAN bus, which most euro OBDII have been switched to. 

http://en.wikipedia.org/wiki/CAN_bus

AJ

[Howard_Casto]

Thanks for the link, I'll look into this. 

Would the CAN bus even be relevant for American panels though?  The strange thing about cars is American cars generally get exported to the UK, but not many UK cars get sent back over to the US  (except bmw of course).  So for it to be cost effective for everybody on byoac we'd probably have to use a rather common brand of car, like a Ford, Chevy, Toyota, ect....  Also of course I want a gauge in mph, not kph. 

[mcseforsale]

Yeah, it's a standard.  Basically, like LAN standards are (such as IEEE stuff).  And, since most auto manufacturers take bids for their OBDII compliant devices, this bus is usually used as part of, if not most, of the communicado between different subsystems. 

For instance, Hyundai/Kia use and entire Bosch computer management system and some of the parts are probably identical to parts in a Corvette. 

AJ

Thanks for the link, I'll look into this. 

Would the CAN bus even be relevant for American panels though?  The strange thing about cars is American cars generally get exported to the UK, but not many UK cars get sent back over to the US  (except bmw of course).  So for it to be cost effective for everybody on byoac we'd probably have to use a rather common brand of car, like a Ford, Chevy, Toyota, ect....  Also of course I want a gauge in mph, not kph. 

[RamjetR]

Hi all,

Actually I missed this thread until after replying to you Howard. Does it need to be serial? There are plenty project documents on Can-bus sheilds for arduino. Sending and receiving through the library would be alot easier then trying to fake the signals if they were in native Can-Bus.

Some links

http://www.seeedstudio.com/wiki/CAN-BUS_Shield

https://www.sparkfun.com/products/10039



Also with ODBII interfaces, it is now possible to use X-Sim to receive and actually generate ODBII signals for real onboard diagnostic data and dashboard display with a suitable Bluetooth adaptor for $15.

On the subject of X-Sim, it's component Dash Gauge design has brilliant output for additional LCD screens if you want to make a dash for your simulator.

A brief summary of my experience making my dashboard;

It comes down to cost. How much your willing to spend on dashboard you'll likely never look at. I've spent perhaps $135AUD per dashboard to display RPM, Shift, Gear and speed in addition to fuel level, temperature, tire and track warnings...

Symprojects.com do a great range of devices to either hack an old analogue gauge or make your own new ones. I own the SPI-D, Pro-Shift and Gi-Pro Gear indicator. The Symdash program works well for all PC racing titles which support telemetry output. And it reads memory locations of some that don't.

I've written a basic interface for outputting the speedo and gear info from Model 2 games to my dashboard but that is a side project which was a by product of working on Model 2 outputs last year. You can see any of them working in my Youtube channel (see sig below).

As for putting a real dashboard in? It's a lot of work, and probably something you'd take on only if you really really want the illusion of a real car dashboard.

If I were to do the dashboard installation all over again with my cabinets and simulator setup. Buy the cheapest 7" Android with Wifi you can buy and download the simdash DashMeterPRO https://play.google.com/store/apps/details?id=com.sensadigit.dashmeterforiracing&feature=search_result#?t=W251bGwsMSwyLDEsImNvbS5zZW5zYWRpZ2l0LmRhc2htZXRlcmZvcmlyYWNpbmciXQ..

It will work better, and with a little skinning can look like the dashboard of any car your driving... plus it's got a touchscreen! I can get a 7" Android 4Gb +Wifi which will do this for $79AUD and thats alot cheaper than the $135 I spend on symprojects gear...

Thats my experience with simulator dashboards though...

[Fursphere]

oops, double post.

[Fursphere]

I was thinking of doing this to my cab.  I wouldn't be directly related to game play, but at least it wouldn't be a static sticker....




Apparently this guy figured it out...



And this guy...

[Howard_Casto]

I think the cost really depends upon what kind of cluster you use. 

There are three Saturn clusters on ebay right now for 30 bucks with free shipping.  Many might say things like "that doesn't go past 100 mph, ect." but it would be virtually free to scan the display and print out one that goes higher.   

Your real cost is the interface. 

Certainly a serial connection isn't a requirement, but the gauges have to at least be digital. 

Here's my train of thought on this one.....  The avr route is certainly a good one, but the more complex the signal you have to send the more custom programming you have to do which makes it offputting for other people (trying to get a system in place for everybody, not just something for a personal project).  Store bought AVRs certainly make things easier, but as you pointed out, it really ups the cost of the project. 

So a parallel port interface would be cheapest.  I know you are thinking that's outdated tech ect, but they do make usb adaptors like so:

http://www.ebay.com/itm/150938418139?hlp=false

Yeah that costs a whole three bucks. 

Even if the entire harness isn't serial, so long as the individual instruments are, it should be more than sufficient.  I know for a lot of speedos/odos you simply "flicker" a pulse to the gage.. the faster the pulse, the higher the value.  So essentially you need one data line per gage. 

The parallel port has 8, so that's enough for:
Speedometer
Odometer
Rev Light
Temp Gage
Fuel Gage
Left Turn
Right Turn

And one bonus.  Or if you want a simpler setup you can just do the speedo/odo and a rev light and have 5 outputs left over for vr buttons and a start button. 

Of course all the gages have to operate at low voltages.  If it's a 12v signal system, that complicates the build to the point of where symprojects stuff seems more cost effective.

Also if you want a shift number indicator that isn't enough outputs, but I never got that anyway.  Can't you tell what gear you are in by looking when you shift?


So what I'm really needing is a source of cheap instrument clusters that everybody thinks looks nice and have low voltage digital controls.   I'm not very knowledgeable in the auto parts department so I can't look at a random dash on ebay and say "yup that will work". 


If we go for this very generic implementation I can probably make things universal with a simple calibration routine (adjust this setting until your rpm reads 10x100 ect...)

[Fursphere]

So..  for 12v, I would think the PC power supply is more than capable of providing enough to drive the gauges (maybe not a shift light... but maybe with a capacitor?)

And for digital guages...

http://www.summitracing.com/search/department/gauges-accessories/part-type/gauges-digital

Race cars generally don't have clusters - they have individual gauges.  Summit is expensive - probably find them cheaper elsewhere.

[Howard_Casto]

I think you are confused as to what I was asking. 

You linked to gauges with digital DISPLAYS.  I mean a gauge with a digital INTERFACE (the gauge display can still be analog or you can go with digital, your choice).

Also by low voltage, again, I mean the voltage to the interface, not the lights/motors/ect on the cluster.  In other words the signal to control the gauges needs to be in the range of around 5v.  Some of the older ones probably work on a 12v system so they wouldn't work without relays and/or transistors. 
 

Yeah I know that race cars generally have individual gauges, but you've just upped the cost of your project by a factor of 4 if you go that route.  I guess you guys are thinking GT, I'm thinking outrun.    Anyway, most cars have individual gauges in their clusters, you'd just have to pull/cut them out and maybe put a trim ring around them.  Of course if you can find an aftermarket "racer" replacement it would work with the interface as well. 

[BadMouth]

At one point in time, most gauges were just 12v voltmeters and the sensors were just variable resistors that reacted to heat or pressure.
If you could find a cheap source of 5v voltmeters, you could toss them behind one of the overlays they sell on ebay or print your own.
I would think you'd be able to adjust them by pulsing the voltage.  Analog adjustment of voltage from a PC is a PITA as far as I've been able to tell.

I think the app RamjetR linked to looks pretty sweet.

[Fursphere]

I think you are confused as to what I was asking. 

You linked to gauges with digital DISPLAYS.  I mean a gauge with a digital INTERFACE (the gauge display can still be analog or you can go with digital, your choice).

I knew what you meant - I guess I was just putting a higher dollar amount on the project than you were.     I've got ~$1,000 into my driving cabinet already, another $100 wouldn't be horrible.  (same for my standup cabinet actually).

So are you just doing memory scans to identify the variable for the speed (and other things), then translating that to the gauge (system) basically?  Or do some of the simulators have that information output (in some format) already?

[deathrow]

This is all way over my head but its looks awesome!

well worth the expenditure. its got to increase the value of the cab?

[Howard_Casto]

Fair enough fursphere.  The way I look at it is if the device is useful, spend as much as needed.  On the other hand if it's useless (like this decidedly is) spend next to nothing and save your money for the important stuff. 

BadMouth:  I looked into the app RamJet linked to and it's entirely possible that I'm missing something, but there are multiple versions of it for each sim engine, around 6 in total.  Each one is 7-8 bucks as well.  So by the time that you buy all of them you might as well look into real gauges as you are entering their price range.  I do like the concept, but perhaps that particular app isn't the best one. 

Also on the client end of things I'm going to have to write something possibly for the model 2 stuff and most definitely for outrun 2006.  X-sim looks like the ideal solution software-wise, but I'm not sure if it can handle all the hackery I do to get things working.  I"ll have to look into it, but regardless we can probably make whatever physical interface x-sim compatable as well so you can use it for my software and x-sim. 

5v voltmeters might work I'll look into it, but I think they are decidedly analog.  Something that expects a digital signal (in the case of speedos a square wave pulse) is easy to calibrate... not sure about something that doesn't. 

[Howard_Casto]

Ok in the realm of vu meters I checked my favorite "crappy Chinese product" website they've got a variety of vu meters. 

I liked this one myself

http://dx.com/p/as-current-flow-vu-meter-500ua-7057

2.84 shipped... you rally can't get much cheaper than that... this seems even too cheap for me.    They are typically odd shaped or small.  You'd just about have to rip it apart, put a bigger needle on it and make a new face.

If a VU meter is used though, it might be necessary to stick to a serial port as opposed to the parallel port when means you can only have one dial per port without the need for a avr.  It has to do with the timing really. 

With a serial connection I can set the baud rate and I have a rock solid timer regardless of the pc you are running it on.  I can modify the pulse rate by just sending different patterns of data

Example:

sending 0x0x0x0x0x1x1x1x1x1 would be a rather slow pulse... it's off for a while then on for a while
sending 0x1x0x1  would be a rapid pulse, it's constantly toggling between 0 and 1

With a parallel interface I don't really have that functionality.  I can do a timer on the software end, but I don't know if a pc is up to the task of pulsing things with that level of accuracy. 

There are gpio boards floating around the net (think usb parallel port) which might do the trick, but they start at 10 bucks and don't have a lot of pins, so an avr might be better. 

We could pick a very cheap avr and I could write some code for it.  I know you can get a teensy for under 20 and I've had pretty good luck with those.  16 I/o pins as well. 

The other option of course is one of the existing hobby avrs.  The ledwiz has power level settings for the leds which would work for a vu meter. There are only 48 degrees of change though, so it wouldn't be terribly accurate.   

I should also mention that if you are just wanting to do a secondary pc monitor mamehooker can handle that right now.  I've got a display file for turbo and it tracks the speed really well.  So long as you are willing to make enough image files you can lay that out however you want. 

Any more ideas on the subject?

[BadMouth]

I was looking at cheap 5v analog meters on fleabay earlier.
The needle on them looked like it would be very difficult to swap out.
The post it was attached to was recessed and the needle had a few 90 degree bends to bring it out to the face of the meter.
Looks like these might be made the same way.  Still useable, but maybe more of a hassle than anticipated.

Not sure if they have enough range of motion to pass for a speedo or tach.

[Howard_Casto]

Yeah my thoughts as well. 

I would think that with that wave of useless usb gadgets we had a few years back somebody would have made a usb tachometer than we could just hack.  Hmm.... maybe I need to check dream cheeky... I know I figured out the protocol for their missile launchers in an afternoon. 

[BadMouth]

http://www.meterpointer.com/

Not sure if they sell to the public, but at least we can learn what the parts are called.


EDIT: Not sure if 18 degree steps is too harsh, but these tiny buggers are interesting:
http://www.ebay.com/itm/4PCS-Micro-Stepping-Motor-6-11mm-2-Phase-4-Spools-/170979672182?_trksid=p2047675.m1851&_trkparms=aid%3D222002%26algo%3DSIC.FIT%26ao%3D1%26asc%3D163%26meid%3D6269876152546623129%26pid%3D100005%26prg%3D1088%26rk%3D2%26sd%3D170993951598%26

[Howard_Casto]

I don't think I've ever seen a two wire stepper before.  Not sure how that would work. 


In terms of steps if you want it really accurate you'd need one step per mph/rpm.


I actually found some of the stepper motors for gauges online but they were around 30 bucks a pop... the needles were around 10.  Then then you'd need some sort of face plate ect...

On the other hand, buy a cheap instrument cluster for 30 bucks and you've got the needles and steppers for a whole cluster. 

So we are back to that again. 

[Howard_Casto]

Bah!  I hate it when I can't sleep, especially when I'm not feeling well. 

Anyway since the gauges in a modern car use stepper motors apparently, I thought I'd take a look into the most common one we would all have readily available. 

http://www.codeproject.com/Articles/16715/Controlling-Floppy-Drive-Stepper-Motor-via-Paralle 

Now all of this is pretty straight forward until you look at the delay he has to use in his code:

Code: [Select]

System.Threading.Thread.Sleep(50); // Delay
PortAccess.Output(888, 0);
System.Threading.Thread.Sleep(50); // Delay

50 Milliseconds per half step, which means 100 ms per step.  That doesn't sound like a lot but 1000 ms = 1 true second and each mph/kph would be one step for our purposes. 
So if you slam on the brakes and go from 150 to a full stop that's 150x100=15000 ms, which is 15 seconds!

I know that different steppers operate quicker but damn, if this is the acceptable delay I'm not sure what is best. 

I looked into CAN bus btw.  It's a pseudo dead end.  Unless you want to simulate the car's entire can computer (I'm not doing that) you really need the interface box to go along with the instrument cluster.  That puts the cost waaaaay back up to the point of where another solution would make more sense. 

[wetwaterwalker]

I know this would add to the cost of the build, but does anyone know what kind of interface that the guy with the ford gauges is using? in his video, the gauges were definitely smooth.  What about using something like a pacdrive with an older set of gauges? A twelve volt power supply from the computer and then pulse the grounds with the power supply? Is this possible? I'm just scratching the surface of all this stuff so I'm really shooting in the dark, please let any of the higher ups with more knowledge correct  me if I am completely off track.

[BadMouth]

I know this would add to the cost of the build, but does anyone know what kind of interface that the guy with the ford gauges is using? in his video, the gauges were definitely smooth.  What about using something like a pacdrive with an older set of gauges? A twelve volt power supply from the computer and then pulse the grounds with the power supply? Is this possible? I'm just scratching the surface of all this stuff so I'm really shooting in the dark, please let any of the higher ups with more knowledge correct  me if I am completely off track.

the title of the video is "Gear Indicator Using Arduino/TFT Module"

I need to get into the Arduino (or generic version of it) someday.
Seems there's a way to do whatever you want as far as using it as an interface.

[wetwaterwalker]

I need to get into the Arduino (or generic version of it) someday.
Seems there's a way to do whatever you want as far as using it as an interface.

Yeah, I seen some guys program them to calculate gas mileage on there cars before.  Something about splicing into the fuel injectors.

I don't think it's all that hard but I've never really messed with it.

[Howard_Casto]

AVR's are actually quite easy to work with, it's just the more pins you need the greater the expense.  You can get a basic teensy (arduino ripoff) for as little as 20 bucks.  It only has about 20 useable pins though.  If the gauges use pulses to control them that's more than enough, but if it's something more complex you are probably going to run out of pins real quick. 

One of those 7 segment displays for example, well that's 8 pins right there. 

[Howard_Casto]

I thought I would resurrect this based upon something I ran across. 

If you want to go RamJet's route and use a secondary display this is how I'd go about it:

http://emerythacks.blogspot.com/2013/04/connecting-ipad-retina-lcd-to-pc.html

Retina displays have a very high viewing angle and although finding a display-port capable video card would be a pain in the butt, it'd probably be worth it. 

[Howard_Casto]

Ok let's bump this Necro thread and get some new ideas rolling. 

TwistedSymphony mentioned that he has played around with this stuff a little more than me, so maybe he can chime in. 

After a year of looking the options I've ran into are basically the ones we all suggested to each other at the start:

1.  Hook up real instrument panels via a AVR
2.  Use a secondary display and fake it.
3.  Spend a lot on sim-centric hardware that is essentially a turn-key avr solution.

Most of us are either converting a sega cab or using a sega cab as the basis for our design.  That makes our options in terms of the dimensions rather limited.  Motorcycle gauges are around the same size, or you can of course roll your own by pulling the stepped motor and needle out of a large speedometer. 

I also ran across this article a few days ago:

http://hackaday.com/2014/04/14/a-7-touchscreen-tv-remote-control-from-scratch/#more-119852

What does this have to do with gages?  Well if you are going the secondary display route, the vendor mentioned in the article sells all kinds of displays for well under 20 bucks.  They would need an avr to control them (I'm a bit lost as to exactly how these work), but you could get a few smaller screens that fit just behind a dial cut-out on your dash. 

I'm still up for suggestions.  Honestly I wish we could just find a good supplier of real speedometer stepper motors for a reasonable price (under 5 bucks a pop).  If that were the case then I could write avr code/software just for those particular steppers and we could all use the same ones. 

[vandale]

How about squeezing a 7 inch tablet inside the dash, lay the gauge bezel over the top and fake it. I am sure somebody can drum up an android app that connects via wifi or usb and fakes the needles and allows the gauges to be placed per x,y coordinates.

Simple dual core tabs are $50 new.

Or if we are limited to PC based app then how about a 7 inch asus, acer, msi netbook. Easy to hide, no fancy interface etc.

Or a small LCD with external power but has a video board with vga etc, no like a laptop panel that does not.

Chur

[SegaOutrun]

they make usb monitors that arent too expensive. 16" for 90$ isnt bad. You could get a 2 decent sized dials in there and imagine it would be easier than making it work on android or ios. 

http://www.ebay.com/itm/AOC-15-6-Widescreen-Flat-Panel-USB-Powered-Portable-LED-Monitor-Piano-Black-/191117636450?pt=Computer_Monitors&hash=item2c7f7fb362

[Howard_Casto]

It's not bad, but I'm not sure it would fit. 

I mean here's a Daytona cp for example:



Like most control panels the shaft of the wheel is in the center and the fake gages are positioned around it. Of course behind the cp is all the mechanics for the wheel. 

So unless you can figure out how to cut a hole in a monitor I'm not sure it would work.  A person would probably need a series of small displays, thus the article I linked to (they sell them as small as 3.5 inches). 

I don't have my cp yet... still designing that part.....  But I'm sitting here holding things up and unless the monitor is really tiny I can't see integrating one without the cp looking funny.

I tried the tablet route, but the only way I could hook it up to gain any amount of flexibility in terms of what I could display was via those virtual desktop apps.  Those lagged like hell unfortunately.  But if anyone can suggest an app that would fit the bill that would work on low end tablets (or else it would defeat the purpose) I'll try it out. 

[Generic Eric]

Like most control panels the shaft of the wheel is in the center and the fake gages are positioned around it. Of course behind the cp is all the mechanics for the wheel. 
I tried the tablet route, but the only way I could hook it up to gain any amount of flexibility in terms of what I could display was via those virtual desktop apps.  Those lagged like hell unfortunately.  But if anyone can suggest an app that would fit the bill that would work on low end tablets (or else it would defeat the purpose) I'll try it out. 

I thought I saw that Google had a remote desktop?

[twistedsymphony]

After reading the older posts in this thread i had some thoughs.

1. Can bus stuff is still very new and as a result still very expensive. the benefit here is you can just do everything in software.. you can also do cool stuff like gauge sweep patterns etc.... however using older 90s era clusters will require more hardware work (like an avr) but be way cheaper.

 2. I would recommend looking for clusters from foreign cars with a strong tuner community. I'm big into the Nissan s-chassis andiknow a lot of those guys upgrade to japanese clusters meaning that the us market clusters are cheap and easy to find...i've litterally got a pile of clusters for these cars laying around... half of them were just givem to me for free because theyre worth so little peopleoften throw them out.

3. most of these machines have a very small space available for gauges so Perhapse the best thing to do is to figure out the dimensions and find a cluster that will actually fit... this might acrtually end up being more difficult than getting the electronics interfaced.

on that note i know a lot of the 90s era japanese clusters ive worked with are actually modular... meaning you can remove the individual gauges out of the cluster and they're self contained and will run without the main circuitboard... so this does offer some flexibility.

[Howard_Casto]

Ok well let me ask you this.  In the video you linked to I noticed that only two wires were hooked up.  Were you using the existing electronics of the board to achieve that or were you somehow pulsing a 4 wires stepper with only two wires?

Check it out:

http://www.ebay.com/sch/i.html?_odkw=toyota+stepper+motor&_sop=15&_osacat=0&_from=R40&_trksid=p2045573.m570.l1313&_nkw=gm+stepper+motor&_sacat=0

Ebay regularly has steppers for cars for around 5 bucks a pop, which I think is reasonable.  You can also generally get a kit of 6 steppers and led bulbs for around 20 bucks, which in all honesty is probably enough to do two cabs for our purposes.  I linked to the GM steppers because unlike most manufacturers, the pins are on the back and they have two mounting holes, meaning, in theory at least you could just drill a small hole for the needle pin, screw the stepper to the back of the plastic, and you are good to go, even in limited space situations like a sega control panel. 

I just don't know about controlling.  I mean simple I/O stuff is easy enough on a avr, I could code a teensy to act like a pac-drive or similar device, but I know steppers are timing dependant.  I also saw stepper controllers in ebay, but I think those are more for upping the voltage for a cnc machine. 

[twistedsymphony]

In that video i had 3 wires hooked up.. 12v, ground, and my fake vss pulse.... they were attached driectly to the gauge.. on those clusters the gauges mount into the cluster via 4 screws... those 4 screws is also how the gauge attaches ekectronically... its really pretty clever... the 4th post is a clean output pulse for the ecm.

[Howard_Casto]

I believe you, but that contradicts everything I've learned about stepper motors. 

It was my understanding that it was a 4 wire process, alternating the voltages on all 4 wires for precise stepping.  Last year somebody actually made a breakout board for switec steppers, which are used in most cars.  The guy even made a library for them, but the reason I ignored it was that 4 pins per gauge is pretty damn expensive. 
https://www.tindie.com/products/TheRengineer/analog-gauge-stepper-breakout-board/

If you were just doing dials it's not so bad, but if you want an over-rev light, shift position indicator, view and start lamps, ect then you aren't going to have enough pins. 

On the other hand, if these can indeed be controlled out of the cluster with one wire (ground and voltage don't count) then it's much more doable.  Shoot I bet we could get a decent, rough indicator with a ledwiz.... the brightness settings are just a timed pulse afterall.  I think there are only 48 steps though... I'd probably still have to put a avr in there
for the final product.

Something that struck me as odd though was the fact that the needles are really expensive!  Your stepper, as I mentioned, is around 5-6 bucks, with free shipping.  The needle, even a cheap plastic one, STARTS at 6 bucks and the nicer ones cost even more.  So it might still be cheaper to buy a instrument panel and part it out. 

[twistedsymphony]

Im not controlling the stepper motor direcly.... the gauge has an intergrated pcb that handles the vss conversion to something the stppermotor understands. Thats what i meant by it being a modular unit... the gauge itself contains all of its own electronics without needing anything from the larger cluster to operate.

To give you an example here is the back of a tach from a toyota pickup... you an see the 3 metal screw tabs on the edges... this is how it attaches to the cluster both electronically and physically.



I wouldnt try connecting to a motor directly even if you got the electronics tuned properly youd still need a retun spring and ive never been able to get one installed properly (not for lack of trying).

[Howard_Casto]

Sorry I didn't reply, I've been busy. 

Do you have a parts link or an ebay link to one of those?  I've started construction on my cab and I might pick some up to play with. 

[Howard_Casto]

Heh, I guess ts hasn't seen this. 

Anyway I'm slightly ill so I was stuck inside today.  I decided to give the android tablet route another try.  iDisplay seems to be the one that most people use.  I tried it on a crappy tablet and a much nicer galaxy.  The amount of lag in the program seems to greatly depend upon the processing power of the tablet. 

The galaxy performed just fine. While there was a tiny bit of lag it wasn't anything worth mentioning.  There were some quirks though, most notably that the newer tablets don't have a hardware menu button and the idisplay only lets you change settings upon startup, so you actually have to turn off the desktop app to change settings as it autoconnects by default. 

The cheaper tablet technically worked, but there was a decent amount of lag... like a second or more.  It would be fine if you just wanted to display stuff for eye candy, but if you actually wanted to read those gauges, then it wouldn't work. 

Since the "monitor" is done in software, it does take a bit of resources on your pc as well.  For outrun 2006 it was fine, but my guess is you might notice a performance hit on a high-end pc game or a very demanding emulator.  Also a severe limitation of the app is the fact that you can only hook up one tablet at a time, so forget about using a couple of small tabs for each gage. 
 
So in conclusion, if you happen to have a mid to high end tablet lying around that you aren't using, you could go this route.  Also if you have a low-end tablet and just want some motion on your dash it would be ok as well, but all-in-all considering the lack of functionality on the low end and the expense on the high end, a actual second pc monitor would probably be cheaper and perform better. 

[twistedsymphony]

Sorry I didn't reply, I've been busy. 

Do you have a parts link or an ebay link to one of those?  I've started construction on my cab and I might pick some up to play with. 

sorry I guess I missed this post.. if you're looking for some gauges to play around with then you might have a hard time finding JUST the gauges.... you'd have an easier time buying a whole cluster....

I'd say just buy whatever looks like it would fit in your cab and your budget and as long as it's not mechanical just run with it. (mechanical speedos were common up through the 80s. These are driven by a spinning cable instead of electrically, you can tell these because they have a 1/2" diameter nub for the cable hookup sticking out directly behind the needle)....

----------------

I have a S14 Silvia gauge cluster with a broken speedo (it was fully working until I accidentally borked the needle return spring attempting to remove the needle. The tach and the rest of the cluster is fully working working... if you want it I'd gladly send to to you for shipping costs.

I have a few other fully working clusters but I think most of them are mechanical... I'll have to check.

[BadMouth]

True Story

[Howard_Casto]

I'll look into it some more, thanks guys.  Eh don't be sending me stuff just yet.  The rate things are going real-world wise, I don't know if I'll get the time to finish now. 

BM, you totally did it wrong.  You are supposed to "throw them away" at your home, on eBay or at the local swap meet. 

[twistedsymphony]

BM, you totally did it wrong.  You are supposed to "throw them away" at your home, on eBay or at the local swap meet. 

EXACTLY... that's how I got my first few arcade cabinets:

I worked as a repair tech at an arcade and was instructed to truck a bunch of machines to the dump...

[BadMouth]

BM, you totally did it wrong.  You are supposed to "throw them away" at your home, on eBay or at the local swap meet. 

EXACTLY... that's how I got my first few arcade cabinets:

I worked as a repair tech at an arcade and was instructed to truck a bunch of machines to the dump...

I brought home quite a bit of what I considered saleable and listed in on ebay after the place was shuttered and I was unemployed.
I was throwing stuff into a dumpster attached to the building.  I could only get away with taking so much to my car in a day.
The stuff I brought home was retail packaged end-consumer stuff.

However, I threw away cases and cases faces, pointers, mechs, etc for OEM clusters.
Had I known that a decade later I'd be tinkering with racing arcade cabinets......

I'm still using spools of solder that came from there.

I miss having a job where people actually make stuff.

[Howard_Casto]

Yeah this modern society where almost everybody seems to have a job as a glorified middle man kind of sucks.  I'm with you on that.  There's something to be said about just making a quality product and selling it because people need it. 

[twistedsymphony]

Yeah this modern society where almost everybody seems to have a job as a glorified middle man kind of sucks.  I'm with you on that.  There's something to be said about just making a quality product and selling it because people need it. 

I can't argue with you there... I've always tried to work somewhere where I actually felt like I was doing something with a direct end product.

These days I work for an aerospace bearing manufacturer writing software... nothing I write leaves the company but my customers are essentially the engineers here so it's nice to be able to communicate face to face with the people I'm actually developing for.... though it also means that if anything goes wrong they know where to find me 

Unfortunately, aside from being familiar with a few languages none of what I do at work translates into being useful for working on stuff like MAME or other emulators.

[Howard_Casto]

Well, I always say... you don't have to be a rocket scientist to work on MAME, but fortunately we've got two or three hanging around the forums just in case. 

Literally you are like the 5th or 6th person over the years that I know of that has been on this forum that works in aerospace, even if you design software. 

[twistedsymphony]

Well, I always say... you don't have to be a rocket scientist to work on MAME, but fortunately we've got two or three hanging around the forums just in case. 

Literally you are like the 5th or 6th person over the years that I know of that has been on this forum that works in aerospace, even if you design software. 

I actually have a degree in Mechanical Engineering from Rensselaer, my programming knowledge is mostly self taught (I did take a couple of classes on it on college "for fun"  )  my electrical knowledge is 100% self-taught.

I like programming because most engineering work can be soul-crushingly tedious. In the bearing world it's like doing the same set of math problems over an over again just with different starting variables... at least with programming I get to develop new apps all the time and I'm always trying out new techniques or adding fancy new features.

As for Arcade machines... I think us engineering types just like to tinker. I'm big into car modification too and there are A LOT of aerospace engineers in that hobby as well.

[Howard_Casto]

Well my strange medical stuff has left me a bit out of sorts so I'm just now starting to get back to this stuff.  This is probably more Outrun related but....

One of the things I could work on while I was more heavily medicated was outrun radio.  While the lag is simply too great for a speedo, the iDisplay method works great for a radio, which doesn't really have to be in real time.

Check it out:

[Xiaou2]

Im not a huge fan of video display for mechanical gauges.  I really love the analog nature of real gauges.

 
 However, a compromise would be a gauge that looked nearly indistinguishable from the real deal.

 Maybe if you used a half-silvered mirror that has a Display under it.. and the gauge face behind it.   Show a black background which will become transparent.. with the Red Needle graphic.   The Needle will appear to float above the gauge face.

 Or you could just use an LCD coupled with a front face mirror at 45 degrees, to show the gauges indirectly.  At least with them being indirect, and with distortions from glass lens (as a sort of bezel)  from the gauge face,  ... it may appear less like a display...  especially if you use a good photograph(s)  rather than making some brand new looking vector / 3d model crud.

[Howard_Casto]

Yeah an lcd for the gages is a hard sell for me.  I'm just exploring all options atm.  It'd be nice if there was some sort of compromise. 

The only thing about mechanical gages that I don't like is the fact that I can't adjust the scale.  If I put a 220mph gage in there then I'm stuck with it, even when playing an arcade racer that doesn't go much beyond 100 mph.  It's not a huge deal for a tacho though... those generally are all in the same range. 

As I've been going over in the thread, the main problem with any solution is space.  Arcade dash boards are around half scale, so it's hard to find anything that'll fit period.  I found lots of nice motorcycle gages, but they use the old school "spin a wire" design so that'd be hard to implement.  Ironically my Grandma's old 99 Ford Taurus has a really nice gage cluster and it's very compact,I had to replace the speed sensor as well, so I know it's digital, but all the graphics would have to be scrapped as it only goes up to 100 mph. 

[twistedsymphony]

Ironically my Grandma's old 99 Ford Taurus has a really nice gage cluster and it's very compact,I had to replace the speed sensor as well, so I know it's digital, but all the graphics would have to be scrapped as it only goes up to 100 mph. 

it looks like the SHO trim cluster goes to 150



Honestly most non-sports cars only go up to 120 or 140... and sports cars rarely have speedos that go above 160 or 180

generally the speedo wont go much past 10 or so MPH over the cars top speed.

-------------

PRO TIP: the UK typically uses MPH on their speedometers as well and many of their speedos read at higher speeds than US Models... I swapped my USDM Nissan cluster out for a UKDM cluster from the same car because the UK cluster reads to 180 while the US cluster only went to 140.

IDEA: Clusters from Germany or Japan (or anywhere else outside of the UK or US) will read in kph... since a kilometer is shorter than a mile (~64MPH = 100KPH) their cluster will have substantially higher numbers.. sports cars with KPH speedos will generally top out over 200KPH. Since you'll be controlling the driver you can mapp the in-game MPH to read the "number" on the kph speedo.

[Howard_Casto]

Yeah I'm wondering how bad it'd be to swap out the graphics on one of these.  Obviously you could just print some out and swap it that way, but I don't know how the lighting would work...  is that stuff made out of translate or what?

[twistedsymphony]

on some clusters it's literally impossible to remove the needle... remember that 240sx cluster I mentioned with the broken needle return spring in the speedometer... yeah, I broke it trying to remove the needle.

Most cluster I've seen are plastic with the numbers and masking printed on the back and another mask and additional printing on the front.

There are companies that make EL back lit overlays for clusters for people who want to change the color and font style... and I'd doubt any of them would be willing to re-index it for you.... not to mention it's unlikely any of them even make something like that for a tarus.

------------

you could always light it from the front... I have a 99 Subaru Legacy and it does have some back-lighting but it's primarily lit front the front via 2 lights mounted in the hood in front of the gauges (but behind the glass... still part of the cluster assembly) I had done some modifications to mine and snapped a picture:



for reference this is (essentially) what it looked like before I modified it (not my cluster, since this one has a European kph Speedo):

I got rid of the car icon top center, got rid of the PRNDL lights between the gauges, added a set of polished aluminum trim rings around the gauges and swapped the originally green front lighting for "natural" uncolored lamps (I liked the soft amber tone as opposed to a harsh white that an LED would produce).

-------------

BUT this brings me back to my other point about KPH clusters... that German cluster above goes to 180KPH... there's no reason you couldn't black out the little "km/h" logo and then just change the signal to read up to 180mph instead.

[Howard_Casto]

Oh I'm not hung up on the Taurus specifically, I just thought it had a very clean "arcadey" design.  It seems like half the clusters on the markets have 50 do-dads randomly sprinkled over the dash... I like rpm's on the left, mph on the right and little else. 

I guess the smart thing to do then would be to pick a cluster that they are also selling replacement needles for on eBay.  Then, in theory at least they could be removed.  My ears are still giving me heck so I looked through a bunch of clusters ebay last night.... They had quite a few in the 20 dollar range, for various makes and models and that's with free shipping + buy it now turned on.  I guess if you wanted to get really cheap (I'm not even that cheap) you could get them for far less.  You've got to watch for pics of the back though.  I was amazed at how many mid 90's era cars still used a cable for the tacho. 

For the most part the back-lighting just lights up the tick marks primarily doesn't it?  I mean I know there are "hidden" lamps for turn signal and what-not, but I wouldn't want to change those anyway.  Maybe do like you are suggesting but with uv-reactive paint?  Then it wouldn't outright look like a spotlight is shining on it, and yet the numbers would appear to glow. 

I've been piddling around with concepts for weeks now.  I think I need to pull the trigger, pick up something to play with and see what I can come up with.

[Howard_Casto]

Check this one out:

http://www.ebay.com/itm/04-06-Volvo-S40-Instrument-Cluster-Speedometer-30728646-OEM-Used-1729-/291147987696?pt=Motors_Car_Truck_Parts_Accessories&hash=item43c9c5baf0&vxp=mtr

I would definitely want a better deal on one... I'd probably go for a junky used version, but that's the kind of cluster a person would want.  It's really simple and it looks like each portion of it is a separate module.  So you could build your dash and then put the gages in it instead of building a dash around the cluster. 

[Howard_Casto]

Ok I've got a couple of choice selections from the bargain bin, maybe you guys can chime in on any electrical concerns... I'll probably order one tonight or tomorrow. 

Here's one from a Sebring:
http://www.ebay.com/itm/96-97-98-CHRYSLER-SEBRING-SPEEDOMETER-CLUSTER-INSTRUMENT-GAUGES-142247-MILES-/310549617006?pt=Motors_Car_Truck_Parts_Accessories&hash=item484e33196e&vxp=mtr

What I like about it are the fully round dials.  I'm going to probably do a vintage styled dash.... you know wood with the gages as inserts, so that would be easier to mod as I have a template.  No cabling on the back though, so it might be a pain in the butt to interface with.

Here's a Saturn:

http://www.ebay.com/itm/95-96-SATURN-SL-SPEEDOMETER-CLUSTER-INSTRUMENT-GAUGES-140953-MILES-SSZ01-/260751533890?pt=Motors_Car_Truck_Parts_Accessories&hash=item3cb600a742&vxp=mtr

Very clean, and it has the wires this time so that's nice.  Also you'll notice the curve at the bottom is very similar to the profile of a Logitech wheel... so a person could probably just bolt it directly above if they wanted to get lazy.

Speaking of easy to implement.

http://www.ebay.com/itm/00-01-02-03-PONTIAC-GRAND-AM-SPEEDOMETER-CLUSTER-INSTRUMENT-GAUGES-DP1462-DA-/310541720506?pt=Motors_Car_Truck_Parts_Accessories&hash=item484dba9bba&vxp=mtr

Very modern, probably not for me BUT....check out the back of that thing.  It looks like each gage has a single two-wire connector unless I'm looking at it wrong.  That probably means that a 5v pwm would run anything on it, which would make things really simple electrically. 

Here's another, very 80's not for me BUT I could see it making an awesome theme to build a cab around.

http://www.ebay.com/itm/89-90-91-92-93-94-95-MAZDA-MPV-SPEEDOMETER-CLUSTER-TACH-INSTRUMENT-GAUGES-/310007038184?pt=Motors_Car_Truck_Parts_Accessories&hash=item482ddc00e8&vxp=mtr

And here is a Taurus one:

http://www.ebay.com/itm/96-99-FORD-TAURUS-MERCURY-SABLE-INSTRUMENT-SPEEDOMETER-CLUSTER-F6DF-10B885-AA-/371061356973?pt=Motors_Car_Truck_Parts_Accessories&hash=item5664faf9ad&vxp=mtr

Normally it wouldn't be worth mentioning, but see that board in the lower right on the back?  I know for a fact that's the amp/calibration board on those cars.  It converts the 5 v signals throughout the engine into 12 for the panel and has some minor adjustments as well, so again, I'll be it would be easy to interface with.

Now this dodge is my favorite:

http://www.ebay.com/itm/CL044-1998-1999-2000-2001-2002-DODGE-CONCORDE-SPEEDOMETER-CLUSTER-UK-MILES-/271443870280?pt=Motors_Car_Truck_Parts_Accessories&hash=item3f3350c248&vxp=mtr

It's kind of cheesy, somebody has obviously put white stickers on the thing, but that is the look I'm going for.. white dials with chrome trim.  I just have electrical concerns... those connectors look kind of funky. 

Obviously there are dozens upon dozens of styles out there, I'm just trying to narrow it down.  Many of the examples I posted are automatic transmission gages, so I'd want manual anyway for the gear indicator and over-rev, but at least now I know what models to search for. 

[twistedsymphony]

Sebring cluster - notice the back the speedo and tach both screw into the circuit "sheet" this would make it a prime candidate for removing the gauges from the cluster and mounting them wherever you want.... also no need to source a connector since you can just use a ring terminal

Saturn Cluster - the picture of the back is too small to get any idea of what the electronics are like or if you could remove the individual gauges, pigtails are good and it looks like it has front illumination.


Grand am cluster - I'm honestly not sure if those are connectors on the back of the cluster... the connectors might be on the top (which isn't pictured) this is a really weird shaped cluster too I would avoid it, it looks like more of a hassle than it's worth.


Mazda MPV cluster - this has a mechanical speedo so it's not an option right out of the gate regardless of it's other merits.


Tarus cluster - seems to be the smallest cluster of those listed which is good, the design doesn't look like you'd be able to remove the gauges individually so you'd have to use the cluster as a whole... Personally I hate the PRNDL shift lights but you could probably block them out.


Dodge Cluster - hard to see what's going on electronically since it's all covered up but there are real metal pins for the connectors so you could plug in with small spade connectors or solder directly. nice looking cluster if you're going for a retro look

[Howard_Casto]

I think I might give the dodge a shot... found this video:



Notice that the mileage and gear indicator are fully digital.  On the one hand that kind of sucks on the gear indicator... I'd probably have to scrap it and make a lamp array, but I could potentially display anything where the mileage goes. 

Then again that might make it really hard to interface with so I dunno. 

[twistedsymphony]

I think I might give the dodge a shot... found this video:



Notice that the mileage and gear indicator are fully digital.  On the one hand that kind of sucks on the gear indicator... I'd probably have to scrap it and make a lamp array, but I could potentially display anything where the mileage goes. 

Then again that might make it really hard to interface with so I dunno. 

chances are those displays will not be controllable via the external connections.

Mileage particularly is usually calculated from the speedometer readings and then stored in the cluster itself (most people seem to assume it's control by the computer.. it's not) ... usually with some security measures to prevent tampering too which can make it nearly impossible to play around with.

unless you desoldered the display itself from the rest of the electronics and built your own circuit to control the display directly it's not likely you're going to be custom displaying anything on it.

as for the gear indicator there will probably be discrete inputs for each gear... so there's that.

See if you can find a FSM (factory service manual) for that car online, they'll typically have wiring diagrams so you can check out how the cluster is actually wired before you drop any money.

[Howard_Casto]

Well there are only two to three chips/protocols still used in alpha-numeric displays these days, mostly because alpha numeric displays are so rare... so I was thinking it would be possible directly at the display, similar to how we used to do it back in the day with those little serial displays and the parallel port.  That sort of thing is uncertain of course, I'm just hopeful. 

I'm going to have to hold off anyway.  I could order one but I won't be doing anything anytime soon.  Got off my meds Thursday, by Saturday I felt like a dump truck had fell on my head, went all over town to find something actually open on the weekend besides the emergency room, because apparently you aren't supposed to get sick on a Saturday in this country.  They've got me on meds again, but basically said they probably wouldn't do any good.  I'm going to try and get an appointment to a ear/nose/throat specialist tomorrow. 

For whatever reason my head isn't hurting in this particular instance, so I'm wide awake but I doubt that'll last long. 

[twistedsymphony]

Well there are only two to three chips/protocols still used in alpha-numeric displays these days, mostly because alpha numeric displays are so rare... so I was thinking it would be possible directly at the display, similar to how we used to do it back in the day with those little serial displays and the parallel port.  That sort of thing is uncertain of course, I'm just hopeful. 

I'm going to have to hold off anyway.  I could order one but I won't be doing anything anytime soon.  Got off my meds Thursday, by Saturday I felt like a dump truck had fell on my head, went all over town to find something actually open on the weekend besides the emergency room, because apparently you aren't supposed to get sick on a Saturday in this country.  They've got me on meds again, but basically said they probably wouldn't do any good.  I'm going to try and get an appointment to a ear/nose/throat specialist tomorrow. 

For whatever reason my head isn't hurting in this particular instance, so I'm wide awake but I doubt that'll last long. 

Ah yes the Hitachi HD44780 protocol... I used to love playing around with those displays... I still have a few. I've even got one setup on a Raspberry Pi to NetBoot Sega NAOMI games  (info here if anyone is interested: http://capane-us.github.io/piforcetools/)

I know what you mean about not wanting to buy until you're ready to start playing around with it... buying parts too early is a bad habbit of mine, enough such that I have boxes and boxes of parts all over my garage and my office for projects where I had an idea and bought stuff and then was too busy with something else by the time it arrived so it never got used. 

In any case good luck with the medical stuff... I swear most doctors don't seem to know what they're doing. I don't usually go to the doctor unless it's something I consider serious and the last 6 or so times over as many years they just scratch their head pick a drug seemingly at random and it doesn't seem to do anything to actually improve the situation...   

If I had that kind of results at my job I'd be canned... you'd think with as many thousands of years we've been studying the human body we'd have a better grasp as to how to fix things.

[Howard_Casto]

Yeah I'm up and about (for now) and I've been looking through my arcade pile.  Dear god I've got to get rid of some of this stuff.  I've got 4 death race wheels... I'm thinking of constructing an under-dash assembly to use one of them as a optical wheel.  If I do that and make a removable hub for the FF wheel that should allow me to play offroad and pole position as well.  I've gotta go to Lowes this evening anyway, so I thought I'd look for suitable bearings while I'm there. 

Yeah the doctor's are particularly frustrating.  I think I'd be more sympathetic if they actually done something besides stare at me for a minute and get out the prescription pad.  I've been 4 times now... they haven't taken a single culture or done a single scan.  It's the equivalent of trying to fix a pc by randomly buying parts off the internet without ever bothering to open the case and look for something obvious.  I called to get a referral from my primary physician yesterday, because apparently the medical industry thinks I'm not capable of the judgment to go to a specialist when needed. They informed me that he wasn't in but they'd give him the message.  So at the rate I'm going this time next spring I might get to see someone who actually knows what they are doing.

[lamprey]

Granted it's not available yet, just a interesting twist on instrumentation. Who knows, it might be something that someone can do something cool with someday.. 

Sharp Develops Free-Form Display

http://sharp-world.com/corporate/news/140618.html

[Xiaou2]

After seeing some of the Digital displays on the Supercars... I might be kinda changing my opinion of the non mechanical speedo.

 Very cool looking graphical color changes and animations...

 But as said before...  I think with some high res photos, and a decent quality display.. one could make some really realistic looking dashes.

[Fursphere]

Sorry to bump an old thread - but did anyone ever get any of these ideas actually working in their cabs?