1103 replies to this topic

[mjr]

Microsoft has a LOT of talented programmers that are very smart (management is suspect though and always has been) and I would think Windows 10 could optimize the cores for gaming automatically.

 

Windows (like all modern operating systems) does indeed have scheduling algorithms to optimize CPU usage, and it's pretty good at it, but an OS scheduler can only go so far because it has to be designed for the average case.  That's why all modern OS's also give you tools, such as thread priority and core affinity settings, to manually optimize particular programs that don't fit the average profile.  VP is one of those special cases.  The thing that makes VP need a little help with manual tweaking is that it's intermittently CPU-bound - most of the time, it's waiting for  GPU rendering, but between GPU cycles it wants to do flat-out CPU work.  That's a misleading combination for the OS scheduler, because the OS sees the relatively low average CPU time and thinks it means that the thread doesn't need a core to itself.  But it really does because of the intermittent bursts of CPU computation between rendering cycles.  The processor affinity settings help with this by reducing the number of competing threads that the OS scheduler has to juggle for the core it assigns to VP.

[Fusionwerks]

Hey mjr, 

can you confirm for me that DOF r3 beta is the correct version for use with the pinscape controllers? i seem to have read, and cant find it now that there where some missing things. I just want to make sure im starting off right.

 

And DOFlinx 5.20? It says i need DOF r3 FROM mjr in the changelog

[mjr]

Hey mjr, 

can you confirm for me that DOF r3 beta is the correct version for use with the pinscape controllers? i seem to have read, and cant find it now that there where some missing things. I just want to make sure im starting off right.

 

And DOFlinx 5.20? It says i need DOF r3 FROM mjr in the changelog

 

You can get my version here:  http://mjrnet.org/pi...ll-updates.html

[Fusionwerks]

Thank you... I think i know now why things were not working right. and just use the latest DOFlinx 5.20 right?

[mjr]

Thank you... I think i know now why things were not working right. and just use the latest DOFlinx 5.20 right?

 

As far as I know - I don't use DOFlinx myself so I can't give you any help there.

[DDH69]

Thank you... I think i know now why things were not working right. and just use the latest DOFlinx 5.20 right?

 

Yes

[marie]

Hi mjr,

 

This time no question but some good news (nice for a change )

 

I mentioned earlier the Wemos addressable led controller (a Teensy clone) activities over at the French side by kakou. 

Kakou modified the DOFconfig files and their Wemos-solution is compatible with the Pinscape now!

At least, until now I haven't experienced any problems during the testing in my cab, just shiny leds everywhere!

 

Good news for all Pinscape-users who want a very affordable addressable led controller

[kakou]

Just a precision, the wemos only cost 3-5€$, and it's the only thing needed (no need an extension card like the optoWS2811)

[Fusionwerks]

well besides having a temporary under-powered PC, i managed to hack my way through the software loading and DOF configuration and all my stuff works like a champ...

But... here i am with another stupid question:

I was reading previous posts on this thread regarding the cabinet button lights. I didnt wire them up on my "test cabinet", but i am ready to now. I have the virtuapin button kit that i believe has the 6.3v lamps. Im fine with them being a little dimmer by hooking them to 5v. And i would like them to react off DOF instead of always lit. So here it is...

 

Correct me if im wrong:

+5v to the lights (daisy-chained) from PSU2

-0v from the lights to the power board outputs (connected individually)

Then set up the DOF tool to correspond to the correct button/port

Create new .ini files and replace the current ones.

 

also in that post, you mentioned that the lights might need to be wired a specific way? Or did you mean the terminals for lights/button should all be grouped together? I was unclear on what you meant by this.

 

The thing that makes the button lights a little hard to think about (it was for me at first, anyway) is that the it seems like the "input" side of the buttons should somehow be coordinated with the "output" side (the lighting).  It's easier if you just think about the lights as output devices in their own right (which they are!) and forget that they have anything to do with the buttons (which they don't!, apart from being physically co-located).

 

Edited by Fusionwerks, 05 October 2017 - 09:10 PM.

[mjr]

 

well besides having a temporary under-powered PC, i managed to hack my way through the software loading and DOF configuration and all my stuff works like a champ...

But... here i am with another stupid question:

I was reading previous posts on this thread regarding the cabinet button lights. I didnt wire them up on my "test cabinet", but i am ready to now. I have the virtuapin button kit that i believe has the 6.3v lamps. Im fine with them being a little dimmer by hooking them to 5v. And i would like them to react off DOF instead of always lit. So here it is...

 

Correct me if im wrong:

+5v to the lights (daisy-chained) from PSU2

-0v from the lights to the power board outputs (connected individually)

Then set up the DOF tool to correspond to the correct button/port

Create new .ini files and replace the current ones.

 

All correct - you just wire them like any other output device and treat them in DOF like any other output device.  The DOF config tool has entries for Start, Exit, Launch Ball, Extra Ball, and maybe a few others.

 

If you ever did want to increase the voltage to 6.3V, an easy way to do that is to buy a "DC-to-DC step-down voltage converter" from eBay, give it 12V input from your PSU, and adjust the output voltage to 6.3V.  Converters are available that have little adjustment screws for setting the output voltage to anything you want (anything lower than the input voltage).  The #555 bulbs work okay at 5V, so it's really not necessary, but they're a lot brighter at their rated voltage.

 

 

 

also in that post, you mentioned that the lights might need to be wired a specific way? Or did you mean the terminals for lights/button should all be grouped together? I was unclear on what you meant by this.

 

The thing that makes the button lights a little hard to think about (it was for me at first, anyway) is that the it seems like the "input" side of the buttons should somehow be coordinated with the "output" side (the lighting).  It's easier if you just think about the lights as output devices in their own right (which they are!) and forget that they have anything to do with the buttons (which they don't!, apart from being physically co-located).

 

I was actually trying to say the opposite of "...need to be wired in a specific way".  I was just trying to say that you should think of the button SWITCHES as one thing and the button LIGHTS as a whole separate thing, even though they're all mashed together in one little plastic doo-dad.  For the switch part, just wire them exactly like any other switches.  For the light bulb part, just wire them exactly like any other feedback device.

 

And no, there's no need for grouping or anything else special.

Edited by mjr, 05 October 2017 - 09:45 PM.

[hlr53]

My button kit from virtuapin contained 161 wedge bulbs which handle 14V. I'm powering with a 12V supply. First time I applied 6.3V they were really dim, so I popped one out to see what type it was.

[mjr]

My button kit from virtuapin contained 161 wedge bulbs which handle 14V. I'm powering with a 12V supply. First time I applied 6.3V they were really dim, so I popped one out to see what type it was.

 

Yeah, the big rectangular buttons use those 14V bulbs.  They'll look noticeably brighter on 14V than 12V, by the way - you can get a DC-DC step-up converter on eBay to accomplish that if you want to be completist.

[Fusionwerks]

Well, so far so good, I'm about 3/4 way done with the woodworking part of my cabinet, and i was putting in my 5 LED flasher panel this weekend and it led me to questions about it.

 

These are the specs of the LED stars i ordered, based on gtxjoes tutorial on RGB flashers: http://www.vpforums....als&article=153

 

LED Emitter: 3W RGB (Red,Green,Blue)

   

    DC Forward Voltage (VF) : RED:2.3-2.6V,GREEN:3.4-3.8V,BLUE:3.4-3.8V

    DC Forward Currect (IF) : 350-1000mA

 

He suggests 27 Ohm 5w for the Green and Blue and 33 Ohm 5w for the red

 

I have put this into a resistor calculator and it only matches up when i enter 12v as power source.

 

Doesn't the JP11 header on the expansion board provide 5v for the flashers?

 

If it does in fact provide the 5v+, i guess i have to daisy chain the power to the LED's and then put the resistors on the "-" side before going back to the controller correct?

 

But at 5v these resistors will not be the same. What confuses me is the forward current number... 350-1000mA? What number do i actually use to get the correct resistor value?

 

Or better yet... maybe you can just tell me 

Edited by Fusionwerks, 16 October 2017 - 04:36 PM.

[mjr]

 

Well, so far so good, I'm about 3/4 way done with the woodworking part of my cabinet, and i was putting in my 5 LED flasher panel this weekend and it led me to questions about it.

 

Good progress!

 

 

 

These are the specs of the LED stars i ordered, based on gtxjoes tutorial on RGB flashers: http://www.vpforums....als&article=153

 

LED Emitter: 3W RGB (Red,Green,Blue)

   

    DC Forward Voltage (VF) : RED:2.3-2.6V,GREEN:3.4-3.8V,BLUE:3.4-3.8V

    DC Forward Currect (IF) : 350-1000mA

 

He suggests 27 Ohm 5w for the Green and Blue and 33 Ohm 5w for the red

 

I have put this into a resistor calculator and it only matches up when i enter 12v as power source.

 

Doesn't the JP11 header on the expansion board provide 5v for the flashers?

 

Correct about the 5V.  

 

For LEDs in general, you can use any voltage that's higher than the "Vf" (forward voltage) quoted for the LED.  HOWEVER, I always like to use the lowest convenient voltage, because the higher the voltage, the more energy gets wasted in making the resistors get hot.  5V is always conveniently available with a PC power supply, and it's above every LED Vf I've seen, so I like to use 5V.  So you CAN use 12V, but I always prefer 5V because it's less wasteful of energy and heat.  And that means you don't need those giant brick-sized resistors pictured in the tutorial - those are so huge because of all the wasted energy and heat they have to shed.

 

And you're absolutely right that you shouldn't use the resistor values based on what anyone else is using.  You have to figure the right values for your particular setup, and you did the right thing by using an LED calculator to figure it for your combination of parts and voltages.

 

 

What confuses me is the forward current number... 350-1000mA? What number do i actually use to get the correct resistor value?

 

That's extremely confusing, all right.  I can see three possibilities:

 

1.  The seller doesn't know what they're talking about and made a wild guess or quoted wrong information

2.  The seller meant "each color is 350mA, and all together when all three R+G+B are running, it's 1000mA"

3.  "These parts will work at anything from 350mA to 1000mA, so you won't burn them out at 1000mA"

 

My guess is that these it's #2, because as far as I've seen, everyone selling these parts on eBay uses the same basic LED components, and those run all run at 350mA.  So I'd go with 350mA and ignore the "up to 1000" part.  Even if I'm wrong and they actually meant #3, 350mA is STILL the right answer, because LEDs don't really get much brighter when you increase the current - they just get hotter.  Better to run them at the lower end of the range, if the seller actually did intend to give you a range.

 

So, with all of that in mind, we go to the LED calculator and plug in the values and get the resistors you need:

 

Values from http://led.linear1.org/1led.wiz

 

Red:  

   Power supply 5V

   Vf 2.3V

   If 350mA

   ==> 8.2 ohms, 2W

 

Green/Blue:

   Power supply 5V

   Vf 3.4V

   If 350mA

   ==> 4.7 ohms, 1W

 

Note that the ranges quoted for the Vf numbers ARE valid (e.g., they said Red was 2.3-2.6V).  They quote those are ranges because there's some variability from one part to the next.  If you plug each end of the range into the calculator, you might get slightly different resistor values.  I like to use the LARGER resistor value, which you get by plugging in the SMALLER end of the voltage range, because that will give you better protection if the part you end up with happens to be at the low end of the voltage range, and it won't noticeably reduce the brightness if the part you get is at the high end of the range.  It's that same thing I said earlier about how LED brightness doesn't vary much by current.  That's why I plugged in 3.4V for Green/Blue instead of 3.8V.

 

 

If it does in fact provide the 5v+, i guess i have to daisy chain the power to the LED's and then put the resistors on the "-" side before going back to the controller correct?

 

That'll work fine.  You can put the resistors on either side, actually, as long as the resistor is in series with the LED:

 

 

+5V  ------- RESISTOR -------- +LED- --------- Pinscape port

 

+5V --------- +LED- -------- RESISTOR -------- Pinscape port

 

They're exactly the same electrically and either one will work equally well.

[Fusionwerks]

Thank you Mike, and oh i forgot to ask.. should they be fused? i see a lot of discussion about that, but im not sure if i need to do 1 larger fuse on the 5v + side or 15 smaller fuses on the port side...

[mjr]

Thank you Mike, and oh i forgot to ask.. should they be fused? i see a lot of discussion about that, but im not sure if i need to do 1 larger fuse on the 5v + side or 15 smaller fuses on the port side...

 

I like to use fuses for everything, but I'd say there's only a small chance they'd ever be needed for LEDs. 

 

In general, the entire point of using fuses in feedback device circuits is to protect the controller board (the Pinscape board in this case).  They're not to protect the device, as the device has essentially no exposure to any harm from overloads.  The fuses are really there to protect against two possibilities:

 

1.  The feedback device malfunctions, and causes an overload that fries the output port on the Pinscape board.  

 

2.  You drop a screwdriver into your cabinet with the power on, shorting something out and causing an overload that fries an output port on the Pinscape board.  (Or a wire comes loose on its own and falls onto an exposed power connector, or various other things like that.)

 

In a case like a motor or solenoid, possibility #1 is a real risk.  Those devices are mechanical, so they can get stuck, the coils can overheat, and all sorts of things can happen that could cause an overload.  So I do think it's a good idea to use fuses for the mechanical parts - motors and solenoids and so on.

 

In the case of LEDs, though, they're about as stable and predictable as the fuses themselves, so the chances of #1 happening are pretty close to zero.  That leaves risk #2. That's definitely a risk, but if you're careful about securing all your wires and you're careful to turn the power off when doing any work inside the cabinet, you can make it a pretty small risk.

 

So it really comes down to the trade-off between the cost, work, and added complexity of adding the fuses, vs the risk that you'll accidentally fry something if you leave them out.  I think the risk for LEDs is pretty small, so depending on how cost-conscious you are and how much trouble you think it is to install the fuses, it would be reasonable to decide to leave them out just for the LEDs.  But I would recommend them for most other devices.

[Fusionwerks]

Thanks again. Im going to build a small baord that has 5 headers for the lights to plug in to, the resistors, and a 16 pin header so I can connect it to the controller board with a ribbon cable. I was just considering the fuses on it, but i agree it is probably low risk given the location of them and i will just eliminate them. I do have a 12 position fuseblock for the 10 contactors, shaker, and gear motor

[mjr]

Thanks again. Im going to build a small baord that has 5 headers for the lights to plug in to, the resistors, and a 16 pin header so I can connect it to the controller board with a ribbon cable. I was just considering the fuses on it, but i agree it is probably low risk given the location of them and i will just eliminate them. I do have a 12 position fuseblock for the 10 contactors, shaker, and gear motor

 

Sounds perfect.  With the ribbon cable especially, the chances of anything going wrong with the LED wiring once you have it set up are pretty minimal.  

[Fusionwerks]

ok, so i did a prototype board that has a matching header for JP11, and i got to thinking about it... do you think that the ribbon cable will be sufficient to provide the 5v for the 5 LED's? It's going to be about a 6-7 foot run by the time i route the ribbon cable. Its a really small wire in there... or should i just provide 5v from another source with a larger gauge wire?

[mjr]

ok, so i did a prototype board that has a matching header for JP11, and i got to thinking about it... do you think that the ribbon cable will be sufficient to provide the 5v for the 5 LED's? It's going to be about a 6-7 foot run by the time i route the ribbon cable. Its a really small wire in there... or should i just provide 5v from another source with a larger gauge wire?

 

The ribbon cable should be good.  I have a long run of it myself supplying two sets of flashers, and it works fine.  

 

In practice, the flashers never fire for very long at a time, so the supply wire isn't carrying significant current for any length of time.  If you want to be sure it's okay, go into the output tester and turn on ALL of the flashers and just leave them on.  Monitor the wire by hand to see if it's getting hot to the touch.  If it's not getting hot after a few minutes, it's never going to be a problem.