More pneumatic control (tires filling this time)

[gleblanc 1]

Paul got me thinking about pneumatic control with his last post, and since I'm in the midst of an upgrade to our system here, I thought maybe I could solicit some feedback. We have an existing tire filling machine. It consists of incoming air regulated to 620 kPA (or 0.62 MPA, or 90 PSI). The filling equipment consists of a Mitsubishi Q00CPU with a few QX40 and QY50 digital input and output cards. This controls an SMC solenoid (VX2220-02-5G1) and gets it's input from an SMC pressure switch (ISG130-031). Currently all tires are filled to 85 PSI, at which point the machine shuts off the air supply. (Exact logic is still unclear, trying to get the tag names translated from Japanese to English). The end game state is being able to inflate tires to either 85 PSI or 105 PSI, based on parameters input by the operator (probably have them scan a barcode). I can think of two ways to solve this problem. 1) I could do what the original designers did, but add a second pressure switch, and select which input to use based on which tire pressure is selected. This sounds like an ideal case to pass something to a subroutine, assume that Mitsubishi PLCs can handle passing parameters. 2) I could replace the existing switch with some sort of analog output switch, and monitor the tire pressure in the PLC code. I'm sure this solution is a bit more expensive, but might allow more flexibility. I might also be able to add a regulator which is controlled by an analog output, in order to set the pressure. This might also act as an upper limit on the pressure, so that I couldn't have a solenoid get stuck, and end up putting 105 PSI into a tire destined for 85 PSI. Any thoughts, equipment recommendations, whatever? Thanks,

[BobLfoot 302]

Just some observations: 1. If your source air is 90 psi regulated you'll never reach 105 psi without a booster of some sort, but you've probably caught and planned for that. 2. I'd go the second switch and digital input rather than analog. my reasons are speed of response and ease of troubleshooting for Electrician Joe Relay. 3. If you go the digital input route sdon't take on the headache of a subroutine just use a simple branch or series condition with the existing 85 psi input.

[OkiePC 14]

I agree with Bob, with one exception. Next month they'll want you to add a third pressure option to the system...if you install an analog transducer you will have more flexibility. I have used MAC Valve PPC (proportional pressure controllers) to inflate tire assembly drums, but we only had to go to 60psi max. They were excellent devices that would nail the target pressure very quickly and accurately, and provided analog feedback as well as a digital signal to tell us when the target was reached. Our older machines had discrete valves with analog feedback and we could modulate the valve to attain the target pressure. That method was okay, but not quite as accurate or as quick. I think that the PPC would be overkill for your system. A simple transducer with good spike withstanding capacity should do it for you...if you want to go analog in anticipation of future needs...

[gleblanc 1]

I've put my responses inline below... Thanks for the advice. I'll shop around to see what sorts of transducers are available. Based on some observations of the process, I might need to have PLC control of the input pressure as well, so a transducer might not cut it. Continued from above: Tuesday afternoon I managed to get out onto the plant floor, and collect a few sets of data, and observe what the operators are doing. The operator interface consists of red and green illuminated pushbuttons. They push the green button to begin the process of airing the tires. The light begins to blink green at this point. Once the light switches to solid green, the valve has shut off the air supply, and (I would expect) the tire is filled. Out of the 18 tires I collected data on, it hit target pressure only 5 times. All other runs required hitting the green button again to restart the process. After a second run, the tire pressure varied anywhere from 93 to 100 PSI. I've since returned the inlet pressure to about 90 PSI, which seems to eliminate most of the variability in the final tire pressures, although it still requires two cycles to fill tires. I suspect that the existing control mechanism is failing completely, and the only regulation on the tire pressure is due to the input pressure, and some timers in the PLC. One thing that I didn't mention in my original post is the analog pressure gauges which are located adjacent to the pressure switches. These are just small 30 - 40 mm diameter gauges. The incoming air is piped to the solenoid, then to the pressure switch, then to the gauge, with about 80mm of 1/4" ID pipe between them. From the gauge, there is about 2 feet of rigid 1/4 ID pipe, then about 10 feet of 1/4" ID rubber air hose. Prior to the start of the filling operation, the pressure on the output side of the solenoid tends to leak down to significantly below input pressure. The operators hook the hose to the tire, hit start, and the gauge immediately spikes to the input pressure (based on the reading on the regulator), and stays their throughout the entire filling cycle. I would expect that the pressure gauge would read fairly low pressures initially, since there is no pressure in the tire. The only thing I can think of so far, is that there is a restriction in the flow due to the small hose and (relatively) long length. I'm going to try to get an inline gauge to put near the filling end of a hose today, to test my hypothesis.

[gleblanc 1]

Well, I got the gauge put on the end of the line, and it looks like the pressure switches are doing absolutely nothing for us. If I hook a completely empty tire up to the filling machine and hit "start", the pressure at the end of the filling hose is equal to the input pressure. It seems that all of the flow restriction is due to the valve stem itself, and none due to the hose. Seems like I'm going to have to regulate the source pressure, and just use timers to fill the tires. Or perhaps only measure the pressure at the tire with the solenoid turned off. Hmm, the later idea has promise, if I can insure that any excess pressure is allowed to leak off.

[J. B. Jennings 4]

Sounds like to ensure bleed off, you could add a second solenoid downstream of the first solenoid and open it to relieve the excess pressure.

[gleblanc 1]

Sounds pretty good. I've attached a PDF of my initial design summary. I haven't decided on any parts yet, except for the analog I/O cards from Mitsubishi (I'm guessing that adding new cards will be cheaper than switching platforms, even with me having to cut my teeth on GX developer for this project). Tire Filling Machine Upgrade Preliminary.pdf

[J. B. Jennings 4]

Unless you can get your analog cards for less than $139, you could get this Click PLC from Automation Direct, free software and a rather inexpensive programming cable. The biggest limit I see with using this system is that currently there are no add-on analog I/O and the only bases with analog I/O have DC inputs and outputs.

[gleblanc 1]

Only two inputs and two outputs. I need to fill 6 tires at once, which means 1 analog channel per tire. I'll probably post more of my thoughts on the cards for this project over in the Mitsubishi forum.