______________________________________________________

"CHANGING THE CHARGING SYSTEM AT FLORIDA STATE

UNIVERSITY" - BY VAN GRIFFIN

_____________________________________________________

McKAY: You were speaking about damage to the Glassman power supplies. I was wondering what people had done to protect against that, particularly the Oxford design spark suppression that Yale uses, whether that's been successful in eliminating that problem?

ASHENFELTER: We have not had a charging supply failure in six years, I would say, so the design is working very well obviously.

GRIFFIN: This would be Oxford?

McKAY: Dick Hyder designed those while he was at Oxford. They're fairly elaborate form of suppression but that's what I'm thinking of putting into our machine.

MUELLER: We installed the same type of suppression on our machine from the very beginning when we put the Pelletron in and haven't had a single failure.

______________________________________________________

"AN UPDATE ON THE AMMERAALL CHARGING BELT EXPERIENCES

AT ORNL" - NATHAN JONES

______________________________________________________

______________________________________________________

"EXPERIENCE WITH AN AMMMERAAL CHARGING BELT AT WESTERN

MICHIGAN UNIVERSITY" - STEVE FERGUSON

______________________________________________________

GRIFFIN: Is there any difference in the gases you use in your machines? I'm wondering why one is so yellow and the other one is so white?

JONES: Yeah, the difference in the gases actually would explain, I would think, the opposite. I run sixty-four percent (64%) nitrogen, sixteen percent (16%) CO2, twenty percent (20%) sulfur hexafluoride.

FERGUSON: We've added enough sulfur hex to bring the sulfur hex up to about two percent (2%), but it's still standard carbon dioxide and nitrogen gas. And we typically run around two hundred (200) psi.

JONES: I would think that would cause mine to be yellow for some reason, but then maybe the sulfur doesn't work that way.

GRIFFIN: You both use shim stock?

FERGUSON: Yes, we're both using shim stock.

JONES: I have the shim stock on the inside of the Oak Ridge EN, on the inside of the belt, just because I wondered why we were not trying to somehow control the charge on the inside of the belt. So I put one and attached to an inner belt guide just as the belt enters the column base across from the down charge screen, and I've really not seen anything that I could say is good from that, or really bad. It's just there and it will probably come out next time as part of my effort of simplification.

FERGUSON: I did try putting screen-shims on the inside of the belt and I couldn't see that the machine ran any better, but the currents were out of balance. I couldn't account for all of the charge I was putting on. I couldn't sum it up and figure out where it was all going, so I took the inner shims out and so we just have shims on the outside.

FERGUSON: And I should say that with the charging system the way it is now I can account for all of the charge. All of the charge going on the belt is coming off through the corona or the columns and there is no down charge.

HARPER: With the HVEC belts there were quite a few vulcanized joints and you could see distinct frequency components in the ripple. Could you see distinct frequencies with these belts also?

JONES: No. I see the fundamental and harmonics of, and when I'm in regulation mode I see some frequencies out in a range from twenty-five to forty (25-40) hertz as grass in that area. But the thirty-one point one two five (31.125) hertz, ten (10) times the belt frequency, that's the number of belt vulcanization joints times the belt frequency. Big peak is gone because this is continuously poured onto the belt on the inside -- they turn it inside out, they pour it on, and then they turn it right side out, and they pour it on, so there is no joints.

GRAY: Looking at your belt, it looks like to me that it would be interesting to scrape some of that stuff off and do an analysis, so I'm wondering if a little bit of sulfur hex is maybe not a good idea, a little bit, because you may still have discharges going on breaking the sulfur hex down and making deposits, if that's absolutely sulfur on that belt. I think it would be interesting to analyze some of that stuff on that belt and see what it is. Two percent (2%) might not be a good idea, five percent (5%), it might be some minimal cut-off level.

JONES: Could be, I'll take a hunk of it home.

LAMM: What would you estimate is the turnaround time on ordering a belt? If you call these guys up when can they put one in your hands?

JONES: If you do not order at the end of July when they go on vacation, you can usually get one within a month or less. My emergency replacement was received well within three weeks. Normal order I'd say would be a month or so. And you can tell them exactly what you want, and they build it, custom build it for you, but it's built in the Netherlands. Shipment is particularly important. Specify we won't pay for it if you don't ship right.

LAMM: What's the cost of the belt?

JONES: The cost for an EN belt is less than three thousand dollars ($3,000). I can't remember exactly. You could -- proportional probably would be whatever -- like I said, they have built them for the Vivitron and I can't -- twenty-eight hundred (2800), I think, something like that, for the white belt. It was just a little bit, three or four hundred dollars ($300-400) more for the gold belt. To compare I asked for a quote on one from Vivirad, and as I remember at the end it was around ten (10) grand. So, if it wasn't for the cost of changing it out, and I must get it on record, a comment, our craftsman at Oak Ridge, I often complained about being forced to use crafts. Our craftsmen at Oak Ridge can turn a belt around real fast for me, they do a good job, and they're very painstaking in doing it. And I have to give them credit, they're really good at doing that. The craftsmen that I generally use inside the tank are very good.

WESTERFELDT: I seem to recall last year you were talking about the weave of the carcass, and you were going to change the weave of it to eliminate stretch, reduce it.

JONES: I discussed with Jim Johnson the Ammeraal rep, a possibility of changing the weave. He was considering doing it. He spoke with the factory, they discussed it, and the recommendation they came back with was that we're seeing the amount of stretch they expected. And then it gives to a certain length and then stops, but they're not used to having applications where you can't get in and see both ends of it and everything. So, they didn't think there would be a problem. They decided that probably a weave change would not be prudent. And actually they recommended staying with the Ropanil on the inside which I'm agreeing with now, especially with the French experience, because he said, hey, if you find a place where it's working don't change it there, change the other side. And what I did was I installed a mirror at the terminal and a light bulb, and Rube Goldberg set up to turn the light on and off from the console, and a long TV camera lens, so I could see the belt position at the terminal and -- you know -- ENs and FNs, we don't have much space at the terminal to mistrack the belt. And although Chris made a good suggestion, I could have ordered a narrower belt, that would have helped. It allows me now to run the thing without dragging it off the top or the bottom of that alternator at the terminal, which was causing a lot of the fluff that I was seeing. And now I can continue to stretch it out, and I think that's helpful.

HILTBRAND: Maybe Steve would give a little explanation of the attempt to improve the charge pickup in the terminal and what resulted from that.

FERGUSON: I'd like to talk about that. It's in our lab write-up. Inside the terminal there is a holder for a charging shim, or in our case the charging shim just as developed just before the belt goes to the alternator. And we had two pieces of shim stock there. One is about a ninety (90) degree angle and the other at a forty-five (45) degree angle with respect to the belt. And that worked fine but I thought that maybe that one shim that was pretty tight might be wearing the belt more, and we're still trying to figure out what was wearing the belt. So I tried taking that shim out and then you get a lot of down charge so I didn't like that. And I thought, well, we'll put one shim in the original position and make a identical holder and put it on the opposite side so it is contacting the belt just as it comes off of the alternator and see what that does. We wound up with a lot of down charge. So, I was having to put more charge out on the up charge side, and I didn't like that very much either. So I went back to the original charging system, which is two shims close together just before the belt goes onto the alternator, and I'm happy with that now.

______________________________________________________

"AN IMPROVED CORONA PROBE FEEDTHROUGH AND NEEDLE

HOLDER ASSEMBLY" - E. PAUL CARTER

______________________________________________________

MUELLER: If you got rid of the d that's underneath the corona points, how do you adjust your corona point length?

CARTER: The initial assembly was precision machine, quite precision, so that the needles were at the right penetration, all right. We do realize that, and that's why I mention we've got another generation to go, that there needs to be some adjustment to make it easier. The reason for that adjustment is necessary because it turns out that the plastic ring that's glued inside the doorknob that you attach the assembly to, those all have different distances. So, if they were all the same distance then once you set your needles you would be set.

MUELLER: Well, aren't the corona points a little different from one

assembly to the next?

CARTER: Oh, no, no. I failed to mention that we've changed that assembly somewhat, also, from the original, and they're now pressed into this housing, and at a set depth, and the instrument shop does that, so all needles are set, and set in the holder. And the holder is part of the mechanism that you bolt down now, so there are no adjustments.

BERNERS: Have you always run with thirty (30) kilovolts on the needles?

CARTER: Oh, no, no. That's just something that occurs, and we actually have a spark gap right off that. And then we go from the feedthrough through about twenty foot (20') of high voltage cable to our probe controller.

BERNERS: So, that's just transient spiking of the voltage somehow?

CARTER: Well, yes, and depends upon who is running, and what the set up is, but as you -- as that -- as a charge card goes -- where the needles are, and how close they are to that electrostatic steel, they'll put a voltage up there.

BERNERS: Okay.

WESTERFELDT: Let me elaborate on that a little bit. The main problem when we went from the old high voltage stabilizer to the NEC TPS it operates quite differently. And it wants to operate the 6BK4 at a much higher plate voltage, and the old Ceramaseal standoff on the flange just wouldn't hold. It's spec'ed at fifteen (15) or twenty (20) kilivolts depending on which specs you look at. And we get problems with the high voltage stabilizer also for that matter. And putting TPS on it was just the last straw, we just couldn't make it operate satisfactory. So, the NEC TPS wants to operate over a wider dynamic range and you will up to thirty (30) kilovolts. I think the spark gap is around thirty-five (35).

BERNERS: Okay. I was surprised at the thirty (30), because we have the Rutgers-Bell Labs-Varian stabilizer system. That runs at eight kilovolts, and we never see it go over ten (10), and so we don't have to modify the feedthrough.

WESTERFELDT: At grid bias they recommend for the TPS somewhere, like ten (10) minus twelve (12) volts compared to about minus three volts, so the plate voltage is directly proportional to that. It's quite a bit higher.

BERNERS: OK.

CARTER: I have one little point to add. As you look through this material here, the first generation of high voltage engine that we did at the shop was to put needles in and have little pins, and they have soft points and we adjust them. We found, not only does this holder fit sloppily in the old system, that the pins sometimes were loose because of the soft point set screws. And also the needles were bigger. And we changed the size of the needles. It turns out, and what we really believe happened when this was modified, that the old needles were pulled out of the old high voltage engineering holder, and when that happened they had to redrill the hole, and they went to a next size needle. And what we've done is we've gone back down in a size needle that's somewhere around, maybe ten (10) to fifteen thousandths (15,000) smaller needle, has a slope that's somewhat longer. And so far, we've been pretty happy with it working okay.

GRIFFIN: We're making our own corona point assemblies also, and the material that I think that those are made of are some sort of thermal alloy or something like that that has a low melting temperature. So we pretty much have a mold for the base and a holder for the needles and allow the melted material to cool and solidify around the needles. That’s how we make our corona point assemblies.

CARLSON: Two questions: one, what is the lifetime that you get from point assemblies? And, two, does there seem to be an advantage of running the plate voltage on a 6BK4 or higher?

CARTER: Well, the answer to the first question, whether or not, or how long we get out of the needles, we put them in and we don't go in for replacing any needles specifically, so we've been up to months. So I don't have any real figures on the hours of corona curve being drawn by the needles. We did have some concern as to whether they would remain sharp enough to do very low energy operations and whether or not we would have to adjust them, but we feel that we go in often enough once a year or twice a year to change out, and we just change out the needles, and that's been good for us. And the second point, we like to operate the machine, or partially with somewhere near, I guess, a bias of about ten to fifteen (10-15) microamps of corona current. And that puts that 6BK4 plate voltage somewhere in the range of fifteen (15) kilovolts. We had a chart and I just didn't bring it. So, we like to stay somewhere in that area. Now, if someone operates with a higher voltage, they have to bring the needles in closer -- a higher corona current, the needles have to come in closer to -- and anytime they get closer to the dome, that's what drives up your 6BK4 plate voltage.

______________________________________________________

"A METHOD FOR DETERMINGING THE SF6 CONCENTRATION IN TANK GAS MIXTURES" - SCOTT WILBURN

______________________________________________________

LAMM: Could you discuss how the CO₂ concentration is measured?

WILBURN: Yeah, we measured that with a device, I think the trade name is Fyrite. And I think it's originally designed for measuring CO₂ concentrations and flu gases. It has a fluid in it, which I can't remember, that selectively absorbs carbon dioxide, and by measuring the change in volume as it absorbs the CO₂ you could determine the CO₂ concentration. That's independent of other gases, just an extra.

LAMM: And the same question. I apologize, somewhat unrelated to your topic, but in terms of tank gas mixture, does anyone like to discuss how they measure the dew point in the tank gas mixture?

GRIFFIN: We have a Shaw moisture meter that uses a capacitance measurement to indicate the dew point. Because a thin layer of hygroscopic material is used, absorbed water causes a change in the measured capacitance. They are commercially available.

McILWAIN: We do have moisture detecting equipment to detect this. There are commercial devices available.

WESTERFELDT: Responding to Larry about the CO₂ measurement, the Fhyrite is available from Granger. If you have a Granger distributor, it's in the new catalogue, if you can't find it, just call me and I'll get that information for you. It is a flu gas measuring device, and we periodically check it against these standard cylinders we purchase to help verify that it's accurate. We also purchased another model from Fyrite which is a oxygen sensor. We occasionally check our tank gas to make sure we don't have oxygen showing up.

WILBURN: Let me add just one thing to that. Because the velocity of sound in the CO₂ is much closer to the value for nitrogen, you're not as sensitive to the CO₂ concentration. You don't have to measure that as accurately to get the sulfur hex rate.

WESTERFELDT: Okay. My second comment about the dew point, we still have one working CEC which is an electrolytic cell where you have a pair of platinum wires and you pull gas through is associated with water vapor and get an ion current out. I think there is just one cell still alive there. I think you can still get them repaired, but it's not very cheap. And we purchased two Kahn Instruments, deypoint hydrometers which read out digitally, and you get them either in the Celsius or Fahrenheit or in PVMV, we purchased a PVMV, which would be more traditional for us. They work quite well. It's a capacitive type sensor, and we actually install it in our control room and run just a very small one sixteenth (1/16) I.D. tube from our accelerator and have a little needle valve, and we just flow gas through that and we can get a reading. You can also get a remote output from them.

LAMM: Where is the SF6 going, do you think, that's requiring you to add SF6 every time you make a transport?

WILBURN: Well, we've speculated on that. There are two properties of the SF6 that make you suspicious. One is the high density. You wonder if you pumped it, would you -- with as great an efficiency as you do the nitrogen carbon oxide. We're just using a single compressor. The second is that the SF6 liquifies at fairly low pressure, and so it's possible that in storage you're liquifying the SF6 there, and you're not getting all of it back out, but we're not sure.

KRAUSE: We use a Beckman trace moisture analyzer, and it uses an electrolytic cell, and I bought it in the late '70s, and Beckman still services it. And I really only sent it in once to the company to have it recalibrated and have a diaphragm replaced in it, but it's a very good unit.

McKAY: Sulfur hex, I think, there are a couple of things could happen there. I think it might be settling. It's so heavy that you can have that sitting in a beaker and you can see the stuff and see it pour. But I wonder if it's in the bottom of your storage tank?

WILBURN: We haven't checked the storage tanks directly I don't believe. We have seen, I think, some evidence for settling in the tandem tank.

WESTERFELDT: Yes.

WILBURN: That if you let the gas sit with the chains turned off for some period of time, and you take the sample at the bottom of the tank versus higher up you see a difference in concentration. It appears that running the chain stirs up the gas enough that that's not a problem.

McKAY: But it might be in the storage tanks?

WILBURN: That's a possibility. I think are we now actually pumping on the storage to recover the gas? Paul?

CARTER: Occasionally.

WILBURN: Okay.

McKAY: You can calculate --

CARTER: Yes, we've had some threshold control and other things so we've proved it could be done. Just to spend the money to do it, the commitment we've made, we haven't done it.

McKAY: You can calculate whether or not it's liquifying period fairly easily.

WILBURN: It certainly must be.

McKAY: I think it requires that two hundred and seventy pounds gauge --

WILBURN: We store at fifteen hundred (1500), yeah.

McKAY: We see liquification in our gas storage tanks when it's minus forty (-40), but you might not.

(Laughter.)

HOSEIN: Our storage tank is outside the building. And during wintertime we have to -- sometime we pump gas in it. But our tank is temperature controlled. We have band heaters on it, so it will remain at a fixed temperature all year.

WILBURN: Ahh!, that's a good idea

______________________________________________________

"TOURING THE LAB" - STEVE FERGUSON

______________________________________________________

McKAY: I'd like to build a little bit on what Steve has said about the importance of public relations. At ACL we see it a little bit differently. We -- this is a corporate public relations thing, but still the same thing, trying to sell science. And some of our people put together a liquid nitrogen show for the plowing match, of all things, a year ago. But what has happened with this show is that a lot of teachers have become interested in it. And through our public relations people we are now, they're going out a couple of times a month to give this show, and they're using things like copper coil and liquid nitrogen to show the increase in super conductivity to tie it into our super conductor cyclotron and things like that. But the idea of going out and showing kids that science is interesting, giving them a chance to ask questions is one of the best PR things we've done either for our lab or for the company. And it's an extension of the sort of thing you're saying. And I think very often we don't pay enough attention to it. I like the sort of things you're telling there, it's a very good way of doing it. It's really important in general terms of science, and hopefully science funding, to do this sort of thing whenever we get a chance.

JONES: A particularly timely subject since a week from today is community day at Oak Ridge National Laboratory. The only thing I would possibly add to what Steve said is one of the illustrations that I've used is explaining the physicist, is one who is inquisitive about nature, inquisitive about the atom. Take a small child with a small toy, say a truck; how do they find out how the truck is put together. I know how I found out, I took a hammer, and I studied the parts that flew off when I hit it with the hammer. Much the same way we provide the hammer to the physicists so that they can study the parts that fly off once we hit their little truck with our hammer. It's very simplistic but if some people who may not be at all interested in the physical sciences, this will explain to some extent a little bit better than the term atom smasher, what atom smasher really means.

GRAY: This point is particularly relevant given the Congressional attitude in our Country these days. And Tracy Tipping who many of you know, who is on our technical staff, one of his responsibilities is to coordinate tours and put things together for us. This even takes precedence over running. When we have our scheduling discussions and our Monday "nuts and bolts" meetings we have with the research group, where technical staff, students, and the research faculty, all sit together and discuss this, that and the other, that's always a point that's brought up, and they always reside in the schedule first because they understand the importance of maintaining public relations with the people we ask to help support us. They're talking with the people who provide funding for us. And we do a lot of tours for schools. We've had family week not too long ago, and it's a big deal. It's a campus-wide thing, and we participate in those things very actively. But it takes a specific coordinating person, the people outside need to know who to contact, the teachers know that, for example, and we do a lot of stuff like that. We think it's been important for our laboratory. And certainly we'll see long-term what that means. But physicists, as a rule, generally have a tendency to sit around and say, well, the stuff I do is so beautiful I shouldn't have to tell about it anyway, you should have read my papers and see how good I am. That's an ego speaking. We don't have to deal with those people, right? Right. But the point being is that we need to all be more informed, I think, about the need. I'm glad that Steve has brought this to our attention so well, because it's part of our job, maybe we'd rather be down fixing something, or designing something, or building something, but it's becoming more and more prominent in the things we do.

BERNERS: On our tours, I usually show a set of transparencies that try to explain how a scattering experiment gives you information about what the particles are being scattered from. And transparencies are maybe a little bit like what Greg showed on the tanker truck, but not really exactly like that, but it's the same kind of thing, I guess. But I give the "tourees" a problem of being border guards at a place where there is a lot of traffic going back and forth across the border, many wagons loaded with hay and other things, and suspicion is that these things have weaponry embedded in them. The border guards, of course, have machine guns, and their problem is to use the machine guns to find out what's embedded in the hay loads. So the transparencies show a beam of collimated machine gun bullets following parallel paths coming into the bale of hay, and you put different shapes in there. You put in, say something that's wedge-shaped like this, like a rocket launching ramp. And if you make it at forty-five (45) degrees and send the bullets in, some of the bullets go straight up when they're reflected, some of them go straight down when they're reflected. Some of them go straight through if they don't hit anything, and so you can deduce from -- without showing the shape I just show the kids, usually it's kids, the path of the scattered particles and ask them what's in there. And it doesn't take them very long to decide that there is a sloping surface at forty-five (45) degrees. If you put a spear in a cannonball, and assume the laws of reflection, the angle of incidence is equal to an angle of reflection, you can get a pattern of reflective bullets coming off of the sphere, and it doesn't take too long for the kids to deduce that that's a sphere. Then you get the sphere spinning. Spinning, you tell them it's part of a gyroscope of missile guidance system, and so you really want to detect this. And the spin, of course, changes the angular distribution of the scattered bullets, and it doesn't really take them very long to see that the thing that the bullets are scattering from is spinning. And they seem to think that this gives them an idea of how a scattering experiment can develop information about something that you can't see at all. And this has worked out fairly well.

CARTER: I'd just like to touch on one thing, and that's -- when we visited the Museum of Science I had to kick a couple of kids off just to get to play one or two of those things (laughter). And we know kids are interested in getting involved in doing things. And by the same token, adults are, too. So, I don't think we should forget that.

______________________________________________________

"NEW IDEAS FOR TERMINAL POTENTIAL STABILIZATION" - ROBERT DARLING

______________________________________________________

DARLING: Well, the first thing I'd like to do is apologize. I did not come here intending to give a talk. I just commented, when I called up to make arrangements to come here, that I wanted to talk to some people about some ideas that I had on terminal stabilizers. I was somewhat surprised when I got here and discovered that it was my turn But it's a subject that I've become very interested in, and unfortunately I've not been able to pursue the last two years because I've been a programmer at NIS2 at Los Alamos Labs. That's no longer true so I can get back to thinking about things that I enjoy. A number of years at Los Alamos they sent in an automation system for their accelerator. It consisted of, if you can believe it, a mode comp computer, essentially a room full of computers, and a bunch of stepping motors that were turning pots. And their comment about this was that it cost them more in manpower to keep the system running, to keep it repaired, than it was to throw it out and replace it by a bunch of bin turning pots. So as a consequence every time I suggested any kind of computerization or automation of the actual operation of the tandem, it can't be done! It was universal, it can't be done! Well, it can be done now. And there are so many things going on out there in the electronics industry that we can make changes, and we can do an awful lot with computers that just couldn't be done. This is a very rough diagram. We have at the top, the tandem, and we have at the bottom, the brown box. Now, the orange lines there represent the first attempts at that. Mainly we use tubes, we use simple components, and use that to stabilize the machines. As a matter of fact, at Rutgers University we had a two MV machine that used vacuum tubes throughout the stabilization system. I see some nods. This was later replaced by an electronic circuit using transistors then using ICs, using bricks, and then using real linear ICs. But we now have the capability of going to the third mode which is we can replace everything but the outer fringes of this box with a computer. Programs like National Instruments Lab View which I'm sure most of you seen running next door, lab instruments assists and a number of other programs that do essentially the same thing can be used not only for data acquisition, can be used not only for control, you turn a pot, you crank a knob, you touch a screen and it changes the magnet. It can be used to do this to essentially replace everything that we used to do with a lot of these op amps. And it has so many advantages. We can design a circuit, we take a couple of log amps for slits, we take a few linear amps for measuring beam current, for measuring voltage off a generating volt meter, off a capacity takeoff unit. We put the last three into intergrators, we feed the into the computer, let the computer do the rest of the stuff. If we don't like the way the computer is doing it, we re-program it. Now, if we were to try to re-program, say, the Variun stabilizer, that's a thousand dollar ($1,000) project, hundred thousand dollars ($100,000), maybe, if you come right down to it. And if something changes we may have to consider doing that again. For a little bit of time we can do the same thing with a computer, by gosh, if we don't like it we can reload the old program, and we're back to the way we were. This is the way, I think, that we need to go with things like terminal stabilizers. If we can replace the Varian, some of the custom units that we have built, and now we have a source for a consistent supply of something that somebody understands, namely NEC. We have essentially standard ways that we can do this, but with the computer we have the capability of not just doing that but making the system a lot more flexible. We can do things, for instance, Lab Windows has a program called PID, Lab View has the same thing, that's Proportional Intergral Derivative. What that means is if you've got an error, you feed back, because the signal is a part of that error; that's proportional. When that signal is changing, you want to start turning that change off before it gets to the point where it belongs. That's called derivative, that says, okay, if I'm at six, and I'm ramping up one per second, I turn it off now it will ramp into seven. Okay. That's derivative. And then integral is used, because you have a limit on the gain that you can get. But if you have an integral circuit that can actually integrate the error, end up with a delta that compensates for that error that you get, if you have a signal, your proportional signal, you turn up the gain, you cut your error down. But you turn it up too high when you have a delay which we have when we're running coroner current, or anything else to stabilize the machine, you have a delay. At some point you can't turn gain up without the system going to oscillation. But with the integral you now have a delta that goes over the top of that, so it essentially zeros your error. This is a PID loop. We can now do that with a computer. And if we want to change the parameters of it, it's very simple. There's a program out there called Expertive, and I'm sure it's not the only one, but it allows you to essentially plug into a computer an operating system. It looks at it and says, ah-ha!, these are the time constants. So it develops a time constant for a PID loop. If you had something like that as part of your package, you could possibly have your computer dynamically adjust the machine as you change the voltage, and other things change, so that it constantly is correcting from these errors that arise. And that is a few comments have come up while I've been talking to people around here for the last few days, these are safety systems; you've got a hundred (100) microamps going into the machine, and it's happy. The next thing you know, it's gone. Okay. The computer is watching it, and knows it, it shuts the machine down before you have damage. And then you go out there and say, well, what's going on, let's find out before we damage the machine, because you people are the people that have to fix the machines, and you don't like them broken. In any case a computer can be used to do quite a bit beyond just what we used to think of this job, mainly data acquisition. You can actually replace almost anything that you can imagine in the analogue world and a lot you can't. Non-linearities -- well, you can do that with a computer, you can't do that very easily with an analogue world. So there are a number of things that can be done with this. All right. In addition that that I'd like to throw in a little circuit. There have been a lot of comments recently about problems with machines not knowing what the -- I hope you can read this okay. (Putting up a diagram) -- what's going on the with the various equipotentials on tubes. You have resistors separating the equipotential planes on your columns. This is a circuit which can be used -- the max nine thirty-one (931) is a rather interesting circuit. It combines a two-and-a-half volt precision reference with comparator. It will operate down to approximately two point seven (2.7) volts and up to, I think, possibly fifteen (15). Operating current is in the tens of microamps between five and twenty (20) microamps, for everything. With something like this, a lithium cell, coin cell, a hundred twenty (120) milliamp hours. You could actually have this thing operating inside your tank for approximately a year without it running out of power. Or if you run it quite often, that's going to be less. But the point is, you take a large resistance, ten thousand (10,000) megohms, four microfarad nylar capacitor, and this cirucit -- this is a -- I guess the value on that's wrong (indicating entry on overhead slide). Anyway this is a hysteresis, this is the sense, this is the LED. What you do is put this onto the slug that you use to short the planes on your machine. And what it actually does is it measures voltage between those two points, it's about a ten percent (10%) error. But the point is that you can run it up and down and you can see the potentials while the machine is actually in operation. And this means that if something is acting up this could show it, whereas meggering the machine when it's at atmospheric, you might never see it in any case. If anyone is interested, I'd be glad to give you any more information I can. Thank you.

McILWAIN: I gather that last circuit provided a pulse train that was proportional to the voltage --.

DARLING: Yeah, basically with four microfarad capacitor and ten thousand megs here and a two-and-a-half voltage here, that gives you approximately one pulse per microamp here, which would mean therefore one pulse So you can feed that into a period counter, and then actually it should measure your potential to ten percent (10%). But certainly you would catch some of the things you wouldn't catch otherwise.

______________________________________________

CLOSING DISCUSSION

JONES: There was something that recently came out in the last couple of weeks at Oak Ridge that I think everybody should be alerted to. If you use Fluke digital volt meters of the '70 series; that's like '70, '74, '77, get your serial number off of it, write it down, call the Fluke Company. They have found a software error, that if you're measuring minus one kilovolt, which these are designed to do, they read zero, that can hurt you. So, these kind of safety things, I think, if anybody else has anything like this that you can throw out to protect each other.

LAMM: We've got a similar thing on a Keithley unit that we've got that has a data logger on it where you can record values over a certain range. When you initially go into that mode, it gives you an enormously low or high reading, I can't remember now. We were watching on our low-energy column current with this meter, trying to see if it was stable or not. According to the meter it never was, so we called Keithley and talked to them. They offered to replace ours with a different variety, and I'm trying to remember now, I'll talk to some other electronics technicians when I get back and put it out on the net, but I believe it's the one that has the RS-232 link in the back of the meter, and there is some problem with that that causes them to feel they can't correct that problem. But I'll get that information out. Thanks for reminding me.

WESTERFELDT: I believe we have some information from Fluke I can put on the net, it gives the range of serial numbers of the affected units from Fluke. I'll see if I can pick that out and put that out here the first of the week, especially for those people that didn't make the meeting.

MCKAY: Well, if it is really the last thing, I would like to repeat what I said in the business meeting. Thank you very much for running a very fine meeting. It's been most enjoyable. And I think the pacing has been very good, and we've had lots of time for discussions. And it's been very pleasant surroundings, pleasant hospitality, so thank you very much.

WESTERFELDT: Well, thank you. We really enjoyed having you. We're delighted that this has turned out as well as it did. Our staff is rather limited.

GRIFFIN: Van Griffin, Florida State. I was just wondering what are the possibilities of adding some of highlights from the electronic mail posting into the SNEAPS? I know it makes more work for you.

WESTERFELDT: Who determines the highlights? Are you voluntering to edit?

GRIFFIN: I don't know, it's just an idea.

BERNERS: Ed Berners, Notre Damm. Do you mean of the proceedings?

GRIFFIN: Enter the proceedings on the net --

BERNERS: Oh, the editors determine the highlights then.

WESTERFELDT: Okay. We certainly will look at that. I have them all safely stowed away.

JONES: There had been some question about the hundred and ten pound (110) test Kevlar fishing line that we use to hold things in the terminal. I've got a spool of it in my bag, and if you want to get the name and number off it, or like a fifteen (15) or twenty foot (20') sample of it, go ahead and get some.

WESTEFELDT: Okay. Well, then we'll close this session and this meeting of SNEAP, and we'll see you all next year at Woods Hole. Thank you very much.