DanW

WIth what you have, the relay operates wihen the system (bulbs) is OK, correct? Use a DPDT relay (instead of a SPDT) and use the 2nd set of contacts to drive a 5 second timer that turns a light on. If the system doesn't work, the relay doesn't work, not light. If the system works, the relay trips, the PLC gets its input, and the light lights. Dan

Nope, I don't know of a utility that can tell what's hogged the serial port. In addition to RSLinx, another culprit is Palm's Hotsynch manager (the app the synchronizes a Palm pilot's data with desktop data). Not as common now that USB is more prevalent than serial, but some serial units are still out there. Dan

I was thinking single phase when I wrote that last reply, even though you were explicit in stating 3 phase. That makes some difference in the decision between solid state relays (SSR) versus an SCR package. SSRs have long lives, if they are kept cool. Heat kills 'em. (So do SCRs if mounted where they can cool themsolves) So most people do some form of heat sinking, which can range from screwing them to the sub panel, or mounting them on commercial extruded aluminum heat sinks with heat transfer compound (white paste). Remember too that the output of the controller has to be able to drive 2 SSRs for 3 phase (the third leg is straight through). Some inexpensive controllers just can't handle the load of 2 SSRs in parallel. Both factors combine to close the gap on SCRs, because an SCR is engineered as a package, includes the heat sinking, and has paid attention to little things like bleeder resistors. For 3 phase, it's a smaller gap between SSRs and SCRs than it is for single phase. Dan

Cal Controls has a great little 1/32 DIN high limit temperature controller, but try to find it on their web site. Find your local distributor on the CALs web site and call them. You should absolutely, definitely, positively use a shunt trip circuit breaker to interlock the power to the SCR or solid state relays. I have a personal horror story I will bore you with. A manufacturer of high temp ceramic kilns refused to interlock his newly designed SCR control (the one I supplied) with any sort of switch, against my strong protests to the contrary. He called me on a Monday morning asking about device I had so strongly advised to use to interlock the SCR. I asked why he wanted to know, since it had been a dead issue. He told me he'd built his first kiln with SCR control the previous week and left it running over the weekend. He came back that morning to find the outside steel skin GLOWING RED HOT. The SCRs had failed, as SCRs do, in the 'on' state, pouring unmodulated heat into the kiln, on an uninterrupted basis all day Saturday and all day Sunday. He was fortunate that the kiln was left in the spot where it had been built, in a fairly open area of the production floor. If it had been located closer to a wall, the wall could have caught fire from radiated hea and burned his shop down. To save the price of a $400 shunt trip breaker and a $200 high limit, it cost him $30k in material and labor, a missed shipping date, and the possibility of burning down his shop. Consider the liability if he'd sold it and it burned someone else's shop down. So be warned about solid state output devices. They fail 'on'. As to the choice of an SCR or solid state relays: SCRs cost more than solid state relays. What justifies the cost? More precise control. Does your process warrant more precise control? Modern controllers (whether standalone or PLC) can be purchased with solid state outputs to drive solid state relays, without the wear and tear of yesteryear's electromechanical relays. So short, time-proportional cycle times (on the order of a second or two) are easy to do without wearing out the relay outputs. (this presupposes that the loop warrants short cycle times). It really depends on whether your ability to supply energy (KW) is on the low or high capacity side with relation to whatever's being accomplished in the chamber. If the KWs you supply is on the low capacity side, the tuning constants will approximate on-off control, ie. high gain (narrow proportional band) and a high rate term. Why pay for SCR control, if you have a loop that's essentially on-off? If, on the other hand, the ability to supply heat is high capacity (compared to the consumption of heat), then an SCR might pay for itself in slightly more precise control by doing what it's designed to so - modulate at fractional percentage outputs.. In other words, if an SCR runs in almost on-off mode, switching between 100% and 0, why bother? Just use solid state relays. CAL also has very nice ramp soak controllers. The folks on PLC lists always want to re-invent the wheel and do ramp soak in the PLC, to my utter amazement. But it's their time, not mine. CAL even has Windows PC monitor software and the ability to write tuning constants to the controller. They were on the OPC bandwagon early-on. I'm told that they even have a DIN rail mount process controller with PLC objects like timers and counters, but with thermal PID functions like ramp and soak. Again, call the distributor, unless you can wait through that silly intro screen on CALs web site. Dan

Text below is directly from a Honeywell hybrid controller with universal analog input (TC/RTD/4-20/mV) I would suspect that these numbers vary from input board to input board, depending on mfg. Dan RTD , PT100 3 wire, 40 ohms balanced maximum Thermocouples: 100 Ohms/Leg 100 (except Low), 500 & 1000 RTD: 100 Ohms/Leg 100 YIS: 100 Ohms/Leg 100-Low RTD & 10 ohm Cu: 10 Ohms/Leg

Ken, I'm really stumped why you'd get dropout on such a periodic basis. It's clearly a hazardous area, is it corrosive, too? You indicated you hadn't noticed anything in particular. Could copper wire to terminal screws really corrode over a 4 week period to the point you'd get drop-out? What about vibration? Are you running compressors close by to maintain the cryo? The drawback to changing out to RTDs is that the real problem is still an unknow, and RTDs aren't all that different than what you have: Type T T/Cs are copper alloy, not a heck of a lot different than copper wire from a transmitter. If you switch to RTDs you could have same problem - copper wire to terminal screws; all that cost and your boss mad at you. What else could cause this dropout? It's never the same circuit that goes out, right? Not like a broken component or bad solder joint in the barrier? Is the cold junction done by the AI card? cold junction would be common to all inputs. Could the cold junction be flakey? intermittent solder joint? By the time you wrestle the cover off, it fixes itself and you think you've done it by relanding all the copper connections? But if it's common to all inputs, all the inputs would flake out. What about an AI board swap for a couple weeks to see what happens? How hot does it get inside the enclosure? Can you monitor the cold junction to see? Could the temp be cycling somehow and affecting operation? One channel goes flakey when the temp goes up? Dan

It just occurred to me, maybe at cryogenic temps the self heating effect of an RTD is self-defeating? Pumping heat, albeit a small amount, into the process is not acceptable? Dan

Ken, Why not RTDs? I whipped open my Pyromation catalog to confirm that conventional Pt DIN style go down below -250° quote from catalog: LOW RANGE WIRE WOUND (-200 to 204) ºC [-328 to 400] ºF 1/8" OD 3/16" OD 1/4" OD 3/8" OD sheaths R1T185L ± 0.1% 100 Ω α = 0.003 85 ºC etc. etc. The thin film only goes to -40°F, wire wound goes lower. It must be your analog input card ? ? ? Usually 3 wire RTDs don't use the same IS barrier as T/Cs do, either, it'd be an expensive changeover. Alaric, are the connection pads on your mold the S type T/C elements or the copper extension wire? Dan

You should cover yourself, liability wise, and declare that your design in no way infers flame safety provision, because that's what flame rods are typically used for. Honeywell and Fireye make specialty products with US/FM approvals for flame safeguards. Failure modes are a big consideration in flame safeguard approval, and it doesn't seem to have been made clear to you what the end user is doing with this signal. Dan

If your situation is inside a factory plant, you'll have to encounter fairly high hard wiring costs to justify going wireless. Industrial I/O wireless costs are on the order of $700 - $1400 per point, at least an order of magnitude greater than PC 802.11x wireless (at less than $50/point). My experience shows that wireless wins when wiring involves trenching, getting across a road or pond, hazardous areas, really remote locations, like outside storage tanks, cooling ponds, water towers; or across a really large plant (steel mills and refineries that cover square miles). For a manufacturing plant inside a 20,000 square foot industrial building, it probably costs less to wire it, unless wiring encounters extreme obstacles, which could be physical or political, depending. When you mention "pulling network cable to each machine", I suspect you're not a wireless candidate. If you have to read a discrete switch, then you're talking industrial I/O wireless, rather than wireless data communications, say rack to rack, PLC to PLC or HMI to PLC. So that leaves out 802.11x stuff, which, by the way, does NOT have the noise immunity of industrial wireless in the same band. "Industrial wireless" is "industrial" for a reason. For an example of wireless flow switches, I worked on an installation of wireless transmitters at a refinery where standard mechanical flow switches monitored safety showers. Hardwiring was estimated to cost in excess of $100/foot, for the Class 1 Div 1 hazardous areas the safety showers are in. At that cost, given the widely spread apart locations of showers, wireless made a lot of sense (saving hundreds of thousands of $) The devices used are transmitters that take a dry contact, transmit the status of the contact to a central receiver. The receiver uses Modbus RTU to get data to the installed control system, by plant choice. The alternative could have been to recreate the switch outputs as DOs at the central receiver. The Honeywell transmitters are battery powered so they don't need power wiring, have IS (intrinsic safety) rating for hazardous area and are NEMA 4 for outdoor weather protection. The Honeywell model was close to ideal for that environment. However, each switch transmitter costs about $1300 (some with high gain antennas, even more) and the central receiver is about $2k. The layout was such that only one DI could be used per transmitter, due to the extreme costs of hard wiring a second DI to a transmitter. If it had been a safe area, not a hazardous area, where a couple switch contacts could be consolidated at one transmitter, transmitter costs could have dropped significantly. I've also used Phoenix Contact wireless, which is not battery powered (needs 24Vdc). It has standalone tranmitter-receiver sets that simply replace wire - put a discrete contact in at one end, get a discrete out at the other, at about the same cost: $1300 for a combo transmitter/receiver, which can handle 2 discretes. Other models can handle large number of discretes, but then one has to wire the field switches to a central point. Being a powered system, Phoenix Contact has much greater power and can handle longer distances. To see if wireless makes sense at all, get a number of your cost of wiring/foot (or meter, depending on where you are) from someone who's done a project recently, make a sketch of the area, pace off the distances, don't forget that indoor wiring typically goes up, across, and then down, so add in the ups and downs; total your distances and calculate where you come to for wiring costs. Dan

Great recommendation, Ken, for the Keyspan. I don't think I've ever heard such a comprehensive claim for reliable performance as you've made, but given the bad performance of the cheapies, a $40 converter is well worth the cost. The pdf manual for the Keyspan is impressive. The cheapie USB-232 converters have a quick start guide that amounts to "install drivers & start application" and that constitutes their support and/or knowledge base. I'm gonna order a Keyspan. Dan

Harbor Freight sells Chinese imported tools here in the USA, some stuff of dubious quality, but they have a great set of 100, 1/4" hex drive security bits, only 8 of which are duplicates (4 extra #2 phillips, 4 extra #2 pozidrives) all in a plastic case for $15 regular, $10 on sale. The only bit it doesn't have, to my knowledge is the tiny torx size used on cellular phones (my kids bought colored faceplates which require the removal of a tiny torx on Nokias) Set includes bits: SAE hex allen (1/16 - 1/4) SAE hex allen security (5/64 - 5/32) metric hex allen (1.5-8mm, no 7mm) metric hex allen security (2 - 6mm) torx (T8-T45) security torx (with the bore in the center) (T8-T40) straight blade screwdriver (metric designations from 3-8mm) spline drive (M5-M8) clutch drive (1, 2, 3) tapered square drive (0 - 3) 3 tine phillips* (1-4) 4 tine offset phillips* (6, 8,10) 2 pin spanner (4, 6, 8, 10) phillips (0-3) pozidrive (0-30) adapters: male 1/4" hex x magnetic female 1/4" hex 1/4" male hex (fits screwdriver) to 1/4" square male socket drive adapter, one short, one long Y thingie with 1/4" hex male (don't know what it is) Highly recommended because all the bits are in one place. You can find the set on www.harborfreight.com . Search for part number 91310. Dan *I don't know the technical name for these drives, only what it looks like

Advantech has a huge line of industrial PCs, some in small, enlcosed packages for embedding, others in 19" rack mount, others "on the back" of industrial LCD panels, usually a touch screen, others with touch panel buttons, like Nematron. Advantech appears to be all Asian manufactured. Support is sketchy, usually leave a messge, next day call back (when you're not at your phone). Documentation is translated and minimalist in nature. All industrial PCs in their line use processors (usually Intel) that receive an 'industrial' rating from Intel, which means that the processor was successful enough to have survived a year's market drumming, and consequently Intel intends to keep it in its catalog for 5 (?) years. This means that the "hottest" processor they have is last year's processor. Not necessarily a drawback. PCs are custom configured to what you spec. I have received devices with the wrong amount of memory and missing components. They've made good, but correction after the fact is always a hassle. In fact, I'm not sure that my distributor didn't actually buy the Compact Flash card that was missing in an Advantech web-box. He claimed he didn't, but I could tell something wasn't going too well in that case. One panel PC had its PS2 (the circular DIN connector for a mouse or keyboard) vertical, facing down in a difficult to access location (meaning buried). Periodically, vibration would loosen the keyboard connection and it was a contortion hassle getting it re-installed. In one case, purchasing an Advantech took a big hassle out of trying to retrofit an office PC with multiple serial ports. They offered it routinely, something for which there's not a demand in the consumer market. Like a lot of products, it comes down to how well the middle man, the distributor, works with you. Personally, I'd never buy Advantech on-line direct. Who's gonna take your side when the box comes in and isn't all there? I keep buying Advantech because my distributor backs it up. I can't answer for yours. Dan

I've used a Honeywell hybrid controller (HC-900) for some thermal process control apps with about the same level of I/O as you stated. Two big differences it has 1) its operator interface - it's a color LCD screen that comes with about 40 templates for screens that show trends, loops, digitals, variables, etc. As long as the end user is content with that screen format, the time spent on screen development was close to zero. Its operator interface is not a free form graphics screen, like the traditional HMI. The controller talks Modbus, so any HMI panel that talks Modbus could be used, but for those who come from the multiple, single loop controller world, Honeywell's display is ideal, because that's what they're used to. The Display trends up to 24 analogs, as is. I find screen programming tedious, so I like the concept. 2) function block programming only, no ladder. I found it very easy to configure PID loops - so easy, I'm loathe to call it "programming". It comes with an autotune that has worked on everything but a fast hydraulic loop. Digital programming is boolean, but I learned TTL logic back when, and boolean configuration looks just like laying out TTL, inputs from the left and top, outputs out the bottom and right. Batch processing with setpoint programming is a snap, given the function blocks available. One project was batch ramp/soaks with synchronized variables, temp + dewpoint, easy to implement. I've had to put HMI software on top of the controller for reporting functions. Honeywell has a special version of Specview (which is pretty low end) which auto-recognizes the controller and preconfigures some tags. Thus far, it has had sufficient power to do what people are looking for. But with an OPC server (or Modbus), it could could be used with any HMI software package. Dan

I'm not sure what the range of a residential garage door opener is, 50 feet maybe? (I don't have one, so I can't test it). Bluetooth is a short range RF, probably less than 100 foot. Dan

When I first used switchers, I was concerned about analog noise, but haven't ever had a complaint. Are small linears, 15W or 30W units, even available in DIN rail mount packages? I remember the inital appeal of switchers for small loads was the conveniece of packaging, like you mention - smaller, DIN rail screw terminal, not the open frame, solder terminal. And higher efficiency for less heat dissipation in small panels. Dan

I'm convinced GM will go into bankruptcy, primarily because I used be Mr. GM, until a year ago, when I bought a Toyota and advised my daughter to buy a Toyota. When GM loses long time customers like me, who resisted the lure of better designed, better built vehicles for decades, they're doomed. In 1979, the US congress provided support for Chrysler, which was nearing bankruptcy, in the form of government guaranteed loans. I opposed it then, as now, as being an unconstitutional form of fascism (violates the general welfare clause, being a form of special welfare). The 1979 Chrysler incident will be the model and precedent for the US goverment bailing out GM, in whatever form is required. When that happens, GM will be able to take actions under the cover of bankruptcy, that it can't take now. But even operation during bankruptcy, or actions like subsidized purchases of GM vehicles for government agencies, won't bring sales numbers up to the level of their current production capacity (at least, according to a Ford analyst I know), so production capacity HAS be shed, whether by closure or sale. Grim picture for long time employees expecting a pension, for contractors servicing GM, and for GM's suppliers. Dan

I didn't mention, it is NOT a safety system. It is a primarily a cascaded temperature control system on a batch that takes weeks to process with some batching functions. I agree, I/O is the most likely failure point, although power supplies run a close 2nd. I heard that the division that makes this hybrid controller can't have redundant I/O for marketing reasons because redundant I/O would compete with Honeywell's DCS systems. So the work around includes a pair of $5K specialty temperature transmitters (that they were already using) that analyzes the junction for drift, in addition to producing a temperature output. There are additional multiple dual element RTDs programmed for fall back if the specialty unit signals drift. The RTD elements are checked against one another for deviation and points are checked one another for deviation. They view the level of redundancy that they got as an insurance policy.

I just did a small redundancy project with a Honeywell hybrid controller. As a hybrid, it is designed for process, not sequence, so it isn't designed for speed and throughput like a conventional PLC. It programs in function block. There were assurances from Honeywell that redundancy would not affect processing time. Loops are processed on a 500mS basis, so that isn't surprising. The project was about 200 I/O points, fit into 2 racks. The redundancy included redundant controllers with a switchover module, redundant power supplies, but no redundant I/O. It could have had redundant ethernet to the HMI, but the customer was only concerned about actual processor redundancy. The redundancy is implemented in the hardware. Like Ken Moore mentioned about Foxboro IA, for this box there was no special programming, it just runs redundant when you turn it on. There is a system function block that has a taggable DO that indicates whether the controller has switched for failure reasons. The value of the processed material is in an given batch is in 7 figures, so an addtional $12K for redundancy was no big deal. Dan

There a technique known as a "loop back" test where data sent from a port loops back and is seen as received data on the same port. You could make an adapter for your serial port with solder-type D connector (from Radio Shack) and connect Tx to Rx. Then put the adapter on the COM port and use a terminal program like hyperterminal to see if what you send comes back. Dan

USB to serial converters are designed to run with a USB driver. Drivers for the USB-232 devices I"ve used are only available for Win 98SE and up. My brother tells me that he ran 386 DOS boxes on ethernet, the kind with the coax cable (not CAT5) way back when. You might find a PCMCIA DOS ethernet adapter with drivers from someone who never throws things away. I had an IBM 486 laptop running Windows 3.x for Workgroups, which supported ethernet comm through a PCMCIA ethernet adapter with a CAT5 dongle. It was half of my first home network. I'm not even sure that 1994 era stuff even shows up on ebay though. Dan

If you have a need to measure flow rate, a magmeter will do a great job on water/glycol. Then alarming from the flow signal is no big deal. But magmeters are expensive, compared to flow switches. For flow switches, there are mechanical types. I avoid the very thin stainless steel strip or vane type targets, (like the McDonnell Miller brand or Flotec brand), because the thin strip fatigues and breaks over time, creating problems downstream. The Kobold KAL series has similar mechanical paddle with much more robust target - a piece of metal that stands up very well over time. The switch is a tiny little reed switch though, OK for a low level DI input, but I wouldn't put 110Vac across one. There are piston and shuttle flow switches, but I've never used one. They seem to be more for machinery use than process use. Gems and Dwyer offer a bunch of either, in different materials and different flow ranges. An alternative to mechanical switches are thermal flow switches, which put heat up a small heater in the flow stream and then measure how rapidly the heat is removed from the heated element. It's called thermal dispersion. I've used both IFM Efector and Kobold and see that Dwyer has them now, found both Kobold and IFM suitable for low temperature fluids that don't change temperature rapidly. Both claim to be temperature compensated for changing media temperatures, but reality is that rapid step changes will fool either. The advantage of thermal dispersion is the intrusion is a stainless "knob" that sticks into the pipe, relatively little pressure loss, nothing to break off and float downstream. I've never had a heater element die, have had an occasional electronic controller module fail, but that doesn't involve messing with the plumbing.

Thanks for the update. Dan

A little off-topic, but there's a couple great video clips of some serious arcing at http://205.243.100.155/frames/longarc.htm see: - 230 kV (1.2 MB mpeg) and 115 kV image of three-phase air break switches opening "hot" - 500 kV switch creates a 100+ foot arc when one phase opens "hot" (1.5 MB mpeg) - Arcing fault causes a substation transformer to explode (5.1 MB mpeg) New - Another arcing fault in a substation - HV fuse fails to open properly (10 MB avi) Dan

Yes, I have used the 9300 RADES and interestingly enough, in the commissioning of it, it was discovered that the incoming dedicated line could not be dialed from a local exchange (only dial-ins from other area codes would go through), like we were attempting for testing purposes. Clearly a phone line problem. The phone company fixed that problem and the 9300 has been in service for 8 months now. So, yes they can work. Dan