975 replies to this topic

[mjr]

DBrown67 - I think the answers you're looking for are all in the TV ON Switch chapter.  If you don't want to build the power sensing circuit, read the section about Using Windows commands - that explains how to use a batch script to pulse the TV switch relay.  You do still need to build the TV relay circuit, of course, since otherwise there won't be any relay for the software to control.  To find that circuit, scroll down to Standalone wiring (no expansion boards).

[ronaldvg]

I am running into a problem with the Pinscape Config Tool. Could not find the problem here so here it goes: If i go to the setup and change any of the Feedback device Outputs, I choose a pin from Board 1 (for instance JP5-3) and after I selected it, it shows in the list as Board 2

 

On top of the setup page where it says "Number of MOSFET power boards" I entered "1" 

 

I also noticed that when I choose any of the other boards, the pin that the mouse is on is indicated by a green circle on the picture of the board, except on the powerboard picture where it indicates multiple green circles but not the one I am pointing to.

 

Firmware of the Pinscape is   2020-04-21-2212 and also using the latest Pinscape Config Tool downloaded today.

Edited by ronaldvg, 14 July 2020 - 01:28 PM.

[DBrown67]

Trying to configure the nudge/tilt with Pinscape controller (KL25Z) in VPX. It seems to work in that I can get the ball to wobble about on the plunger and flipper. But if I nudge a bit harder the ball will suddenly jump forward as if I've pressed space bar for a forward nudge. Pressing Space bar does look almost exactly the same but in my key config I have all those keyboard nudge options set to "NONE". Yet it still works from keyboard. Can the keyboard nudges be turned off to use analogue only? Is this even my issue?

 

I managed to get the tilt screen shake removed by setting that value to zero. But this ball jumping is really unrealistic and I don't know what's causing it.  I have the X and Y Max both set to 80 (default) with gains at 100 each. Deadzone 2%. Altering these doesn't get rid of the jumping ball.

 

EDIT:  Feel like I'm getting somewhere but still not great. I renamed the NudgePlugin (mrj one) but it was no good for me. I don't have a real tilt bob installed so using that plugin just caused a tilt as soon as any nudge was detected. So I've skipped that now. I did find the command in the script for the centre nudge key and set that to zero, but it still nudges the ball forward a bit when I hit space. How can I get rid of those key nudges altogether? Is there a global setting or do I have to manually edit every table script? The nudging does seem to work with the KL25Z.

Edited by DBrown67, 19 July 2020 - 11:32 PM.

[xtrouble]

I'm trying to disable the pinscape with usbdeview from nirsoft, but however I set the pinscape software, it always tries to reconnect every 10 seconds. I disabled the "Reset on disconnect" setting in config tool first, and also tried changing the default reconnect time to 60 seconds, but it still tries to reconnect every 10 seconds. Am I missing something obvious ? (Answer - probably)

[PIN667]

 

Sorry, should have specified- Sharp GP2Y0A51SK0F) It has an analog output and would be a good linear pot replacement if the output were linear and a bit higher amplitude.    

 

Rick - got it.  That could definitely be worth a try.  I haven't tried any of the triangulation sensors, only the older reflectivity sensors, which were terrible at being distance sensors.  Hopefully this will produce somewhat better results.  

 

Hi mjr,

have you ever tried to get the Sharp GP2Y0A51SK0F sensor going with Pinscape? It looks like a clean, simple and cheap solution for a plunger. Someone already made a working plunger with it. See here: https://www.tested.c...inball-machine/ sadly he doesn’t use the KL25Z with Pinscape. 
I know Pinscape is open source, but I’m not smart enough for making it work.

[mjr]

have you ever tried to get the Sharp GP2Y0A51SK0F sensor going with Pinscape? 

 

No, I'm afraid I'm not too fond of any of the available distance sensors, since they're really slow and noisy.  It would certainly be possible to add support for this one - I'm not planning on it myself, but if someone wants to get it going and send me a pull request, I'd be happy to merge it into the mainline.  In the meantime, if you really want to try a distance sensor, the current state of the art (as far as I've found) is the VL6180X, which is already supported.  That's a time-of-flight sensor, which measures distance by timing the round-trip light time for reflected laser pings.  The Sharp and Vishay IR sensors all work by measuring the analog reflected intensity instead, which is much less precise.  Several of the robotics vendors sell ready-to-use breakout board for the VL6180X  (Pololu, Sparkfun, Adafruit - I think Pololu's is the cheapest), so it's quite easy to try; wiring instructions are in the build guide.  I think you'll find that much superior to the Sharp sensors, but much inferior to a potentiometer.

[rickh]

I recall playing with the Sharp GP2Y0A51SK0F and the biggest problem I had was that the output was not linear to the distance.  The cost for one sensor was over $9 a couple years ago.   Although it can be used as is, it would need some type of polynomial curve fitting to make it perform as good as a simple linear pot.  As Mike mentioned you still have some noise and offset to deal with.

 

Regards,

 

Rick

 

have you ever tried to get the Sharp GP2Y0A51SK0F sensor going with Pinscape? 

 

[mjr]

I recall playing with the Sharp GP2Y0A51SK0F and the biggest problem I had was that the output was not linear to the distance.

 

Definitely - all of the IR prox sensors report an analog reading of the intensity of the reflected light.  At the physics level, that'll follow a power law - reflected intensity varies with the inverse of distance raised to some power.  It would be nice if this were just 1/R^2, but in practice it's not.  If you look at the data sheets for the sensors, they'll usually show a log-log graph of the response curve, and that'll look vaguely linear over some distance region.  A sensor's suitability depends on that graph looking as close to linear as possible over a range from about 10mm to 100mm.  In general, if you have a nice line plot on a log-log graph, you can read the slope of that line to get the power law exponent, so you can get the approximate exponent by looking at the data sheet graphs.  But you probably won't want to use that, since the data sheet graphs are just meant to give you a ballpark figure, and since your real-world readings will probably be quite different anyway because your actual plunger setup won't exactly duplicate the test conditions they used to prepare the data sheet.  So you'll probably want to calculate a best fit based on live calibration readings.  Even then, the result won't be perfectly linear - even under the ideal test conditions they use for the data sheet, the graphs don't come out exactly linear on a log-log graph.  But maybe you can get it close enough to look decent for the on-screen plunger animation.

[qwerty71]

Hi all,

I just complete soldering the PCB and experience some issues with testing.

Hope someone more intelligence then me is able to help.

 

What is the situation:

Currently focusing on the main board only. I have detached the power and chime board and removed it from the configtool.

On the mainboard, the input buttons are working properly (validated with the "test button" option in the configtool).

The first simple goal is to control the button lights as feedback devices. For that I like to use the outputports on Jumper8.

The problem is that I dont see the output changing on the jumper ports when using the output tester in the configtool.

I attached a volt-meter to the port LED1R and the +5v. When toggling the on/off nothing changes on the output. It consistently stay at 1.8v.

Tried other pins in the same JP8. The are all steady 1.8v and dont change when try on the output testing in the configtool.

Perhaps I completely miss something basic? Still new to this project...

 

Not sure where to start the troubleshoot? I checked the power to the TLC5940NT chip: I got 3.3v on pin 21-22, which was correct.

I swapped the TLC5940 chips (they are in sockets, so it easy to remove). Didn't make any difference.

For R5 resistor I use a 2.2k.

 

Not sure if the JP8 ports are some kind of special. Happy to use other ports if that is better.

At this stage, I am just tying to get some control over this wonderful project.

 

I did a proper reading of the build-guide, but couldn't find any useful tips and trick or how to get this going..

Hope MJR or some other expert is online to give me some hints.

 

Tx,

John

Edited by qwerty71, 22 June 2021 - 09:02 AM.

[mjr]

qwerty71 -

 

> I attached a volt-meter to the port LED1R and the +5v. When toggling the

> on/off nothing changes on the output. It consistently stay at 1.8v.

 

The way that the TLC5940 chip works is kind of like a switch connected to Ground (GND, or 0V, zero volts).  When one of the ports on JP8 is switched on, it conducts to ground.  When a port is switched off, it doesn't conduct anywhere at all.  So connecting a voltmeter between one of these ports and 5V isn't really the right test.  What you want to do is see if the port is conducting to ground or not, so if you want to do a voltmeter test, what you could do is attach a resistor (pretty much any one you have handy, say 1K or 10K) between 5V and one of the port pins.  Then you can connect voltmeter (+) (red) to the port, and voltmeter (-) (black) to one of the power supply Ground/0V pins.   When the port is off, you should read 5V, and when the port is on at 100% brightness, you should read a small voltage, maybe 0.5 to 1.0 V (it won't quite hit zero, because it's a transistor switch, so there's always a little bit of a voltage there even when fully on).  The other caveat is that the PWM chips can sometimes show incorrect voltmeter readings because of the way they rapidly switch on and off, even when a port is fully on..  That can sometimes make it hard for a voltmeter to get a lock on a DC voltage, since the voltage is constantly changing.  But usually when a port is on at 100% brightness, the switching is fast enough that the voltmeter will read it as an average and get a stable reading.

 

With the TLC5940 chips, if none of the ports are working, it's usually an indication that you have a bad connection (like a solder joint that's not quite solid) to either a power/ground pin or one of the control pins.  All of the control pins obviously have to be working perfectly for the chip to work at all, so if even one of those has a bad connection, the chip will just sit there and act like a brick, like you're seeing.  You might also want to double-check the configuration to make sure that you've got the right number of TLC5940 chips selected and that there aren't any red circle error icons anywhere for GPIO pin assignments.  The red error icons indicate conflicts, like when the same GPIO is used for multiple functions - click on any error icons you see to get the details of the error.

[qwerty71]

qwerty71 -

 

> I attached a volt-meter to the port LED1R and the +5v. When toggling the

> on/off nothing changes on the output. It consistently stay at 1.8v.

 

The way that the TLC5940 chip works is kind of like a switch connected to Ground (GND, or 0V, zero volts).  When one of the ports on JP8 is switched on, it conducts to ground.  When a port is switched off, it doesn't conduct anywhere at all.  So connecting a voltmeter between one of these ports and 5V isn't really the right test.  What you want to do is see if the port is conducting to ground or not, so if you want to do a voltmeter test, what you could do is attach a resistor (pretty much any one you have handy, say 1K or 10K) between 5V and one of the port pins.  Then you can connect voltmeter (+) (red) to the port, and voltmeter (-) (black) to one of the power supply Ground/0V pins.   When the port is off, you should read 5V, and when the port is on at 100% brightness, you should read a small voltage, maybe 0.5 to 1.0 V (it won't quite hit zero, because it's a transistor switch, so there's always a little bit of a voltage there even when fully on).  The other caveat is that the PWM chips can sometimes show incorrect voltmeter readings because of the way they rapidly switch on and off, even when a port is fully on..  That can sometimes make it hard for a voltmeter to get a lock on a DC voltage, since the voltage is constantly changing.  But usually when a port is on at 100% brightness, the switching is fast enough that the voltmeter will read it as an average and get a stable reading.

 

With the TLC5940 chips, if none of the ports are working, it's usually an indication that you have a bad connection (like a solder joint that's not quite solid) to either a power/ground pin or one of the control pins.  All of the control pins obviously have to be working perfectly for the chip to work at all, so if even one of those has a bad connection, the chip will just sit there and act like a brick, like you're seeing.  You might also want to double-check the configuration to make sure that you've got the right number of TLC5940 chips selected and that there aren't any red circle error icons anywhere for GPIO pin assignments.  The red error icons indicate conflicts, like when the same GPIO is used for multiple functions - click on any error icons you see to get the details of the error.

 

Thanks for your thoughts on this.

 

I check the wiring connectivity between the KL25Z and TLC5940 as describe in the troubleshooting section of your guide. I also measure more power levels: I found one port on JP8 (port 4G) which was connected to ground (it had 5v) where others were no connected (measured 0v) and the remaining had as wrote above 1.8v or 3v depending from where you measure (5v or ground). What was common is that all ports were not changing no matter you changed the status in the software.

 

I continue my search and connected the powerboard and did similar test. All failed in not responding to the on-off toggle from the computer. But on 2 ports, port 11 on both jumpers (one on both sides) I measured port was on. Now i am starting the suspect the TLC5940 chips really...

As extra test, I removed the opto-couplers (they were in sockets). The ports which was on went off without the opto's. So I concluded that the ports were on because of a signal from the TLC chip.

Since I bought 10 TLC chips in bulk, I replaced them with new once, but not luck.

 

I noticed that the Chime boards were not using the TLC5940 so I connected them. I was hoping this to work as it is not using the TLC5940 and indeed (YES) all ports were working fine: in the software I could switch them on-off and i saw my LED go on-off as well (even the timer function worked), so I get my chime board working.

 

So I think I narrowed down (in my limited view...) to a suspicious operation of the TLC5940.  

Although the chips looks original, I did buy them from Aliexpress, so you never know what you really get. 

I am now ordering new once from a local, more expensive supplier. Hopefully this will resolve my problem. 

Stay tuned for the results.

[mjr]

qwerty71 - yeah, it sounds like bad TLC5490 chips might be the problem.  I've heard from a few other people who had that happen - I don't think anyone has run into a whole batch being dead like you seem to have, but these chips have been out of production so long that I can believe it.  The Aliexpress seller could be selling counterfeits, or possibly just old stock that wasn't handled properly in the years since they left the factory.  Hopefully you can find a good batch and get this working.  It's definitely a good sign that the chime boards are working for you.

[qwerty71]

 

qwerty71 -

 

> I attached a volt-meter to the port LED1R and the +5v. When toggling the

> on/off nothing changes on the output. It consistently stay at 1.8v.

 

The way that the TLC5940 chip works is kind of like a switch connected to Ground (GND, or 0V, zero volts).  When one of the ports on JP8 is switched on, it conducts to ground.  When a port is switched off, it doesn't conduct anywhere at all.  So connecting a voltmeter between one of these ports and 5V isn't really the right test.  What you want to do is see if the port is conducting to ground or not, so if you want to do a voltmeter test, what you could do is attach a resistor (pretty much any one you have handy, say 1K or 10K) between 5V and one of the port pins.  Then you can connect voltmeter (+) (red) to the port, and voltmeter (-) (black) to one of the power supply Ground/0V pins.   When the port is off, you should read 5V, and when the port is on at 100% brightness, you should read a small voltage, maybe 0.5 to 1.0 V (it won't quite hit zero, because it's a transistor switch, so there's always a little bit of a voltage there even when fully on).  The other caveat is that the PWM chips can sometimes show incorrect voltmeter readings because of the way they rapidly switch on and off, even when a port is fully on..  That can sometimes make it hard for a voltmeter to get a lock on a DC voltage, since the voltage is constantly changing.  But usually when a port is on at 100% brightness, the switching is fast enough that the voltmeter will read it as an average and get a stable reading.

 

With the TLC5940 chips, if none of the ports are working, it's usually an indication that you have a bad connection (like a solder joint that's not quite solid) to either a power/ground pin or one of the control pins.  All of the control pins obviously have to be working perfectly for the chip to work at all, so if even one of those has a bad connection, the chip will just sit there and act like a brick, like you're seeing.  You might also want to double-check the configuration to make sure that you've got the right number of TLC5940 chips selected and that there aren't any red circle error icons anywhere for GPIO pin assignments.  The red error icons indicate conflicts, like when the same GPIO is used for multiple functions - click on any error icons you see to get the details of the error.

 

Thanks for your thoughts on this.

 

I check the wiring connectivity between the KL25Z and TLC5940 as describe in the troubleshooting section of your guide. I also measure more power levels: I found one port on JP8 (port 4G) which was connected to ground (it had 5v) where others were no connected (measured 0v) and the remaining had as wrote above 1.8v or 3v depending from where you measure (5v or ground). What was common is that all ports were not changing no matter you changed the status in the software.

 

I continue my search and connected the powerboard and did similar test. All failed in not responding to the on-off toggle from the computer. But on 2 ports, port 11 on both jumpers (one on both sides) I measured port was on. Now i am starting the suspect the TLC5940 chips really...

As extra test, I removed the opto-couplers (they were in sockets). The ports which was on went off without the opto's. So I concluded that the ports were on because of a signal from the TLC chip.

Since I bought 10 TLC chips in bulk, I replaced them with new once, but not luck.

 

I noticed that the Chime boards were not using the TLC5940 so I connected them. I was hoping this to work as it is not using the TLC5940 and indeed (YES) all ports were working fine: in the software I could switch them on-off and i saw my LED go on-off as well (even the timer function worked), so I get my chime board working.

 

So I think I narrowed down (in my limited view...) to a suspicious operation of the TLC5940.  

Although the chips looks original, I did buy them from Aliexpress, so you never know what you really get. 

I am now ordering new once from a local, more expensive supplier. Hopefully this will resolve my problem. 

Stay tuned for the results.

 

I can confirm my new TLC5940 chips everything works like a charm...

[mjr]

I can confirm my new TLC5940 chips everything works like a charm...

That's great!  

[LynnInDenver]

Yeah, buying chips off Aliexpress is a crapshoot, honestly. Especially during a shortage of chips of various kinds. You could get a legitimate working chip, or could get a chip that was actually a non-functional reject from the same fabrication line, or a sanded/painted and relabeled chip that may or may not be compatible electrically with what you ordered.

[deltaford]

I'm having a weird problem with enabling night mode. If I connect the KL25Z and then enable night mode, before it receives any other command, it enables all outputs(!).

Anyone seen this?

The issue is not triggered if it has already received output commands, like if I have been playing for a bit or used the config tool.

I'm running latest firmware 2021-06-02 with outputs on three TLC59116 and a 595.

 

UPDATE: This issue is resolved in latest Pinscape firmware (build July 16, 2021)

Edited by deltaford, 16 July 2021 - 10:13 AM.

[redferatu]

Hi, I am struggling to get my pinscape recognised by windows 10. I updated through windows 7 and it works on that, but still not windows 10. Should I just keep trying? Anything odd that I can try to get it to work? Thanks.

[mjr]

Hi, I am struggling to get my pinscape recognised by windows 10. I updated through windows 7 and it works on that, but still not windows 10. Should I just keep trying? Anything odd that I can try to get it to work? Thanks.

 

It should just work.

 

It might help if you could post the full step-by-step of exactly what you've tried, and exactly what you're seeing at each step.  Extreme detail might be helpful - it can come down to seemingly small things like plugging the cable into the wrong port on the KL25Z.  It's also helpful if you can define your terms when they're potentially vague - for example, when you say "I updated through Windows 7", I don't know if you're talking about the Pinscape firmware or the PE Micro boot loader.  The distinction is important because the PE Micro boot loader is the part that can cause problems on Win 10, if you have an old version of it.

[redferatu]

I followed your write up on the website, update the micro, but yeah I can do it again with detail and photos tomorrow for me.

 

Thanks for getting back to me @mjr. Hopefully it was just a bad update to the bootloader and that will fix the issue.

[redferatu]

Hi mjr, I tried another two times following the guide. I get one yellow flash every three seconds in the win10 machine. There are pictures if you need them. One of the bootloader and the sda, another of FDRM, one of pinscape in win7 recognised and one in win10 not recognised. If i need to set it up for the first time after the flash, that step didn't happen for it to be recognised in pinscape. So that might be where I am going wrong.

 

Also in the case of the detail I would be just copying your guide on the website. I understand I may not be following it correctly, but as far as I can tell I am following it to the letter without any indication otherwise.

 

Cheers mate.

 

https://ibb.co/gM3d8C6
https://ibb.co/Pc7Z11J
https://ibb.co/4KXhbQx
https://ibb.co/qBcfmmq

Edited by redferatu, 26 July 2021 - 01:02 AM.