This presentation ties into and is built on last year's presentation which was published in the first proceedings. It is suggested that a careful study of that first more basic paper precede a study of this paper. In this paper the key word is systems. A system is a group of interacting components. That is what I will emphasize here; the interacting or the contribution of each component to the system. We have basically three major components in our system. The instrument, the diaphragm type compensator, and the sailplane.
Figure 1 is a plot of sensitivity change with altitude. The conditions are standard atmospheric and the instrument is at the same temperature as the atmosphere outside the sailplane, and this is pretty close to the case in the sailplane. These data were obtained using an altitude chamber in my basement lab.
The important thing here is that the leaky capsule type retains essentially a constant sensivity with altitude change, and for these conditions the slope of the plot is only 3 percent change for about 30,000 feet of altitude. The next important thing borne out by this figure is that the thermal type variometers are essentially mass rate of flow instruments and their sensitivity drops about 10 percent for each 3000 feet of altitude increase.
Of the three system components mentioned above, the instrument itself gives the least trouble. If it is at fault, it is easiest to identify and the easiest to fix. The second component, the diaphragm type compensator, is connected in the system from the pitot line and its function is to provide a signal that exactly cancels the unwanted variometer signal that would be seen on the variometer when a change in altitude occurs due to a speed change, We will discuss the response of the diaphragm and also the time constants which will include the restrictions and the capacities in the system.
Figure 2 is a plot of response needed if the variometer system had a reference capacity of 525 cc and were being tuned for 4000 feet msl. The initial slope is 1.49 cc per inch of H20 pitot pressure change. Notice that the solid line departs from the initial slope especially in the high speed ranges. It is the nature of this type of plot that this important little difference doesn't stand out. This information is replotted. in Figure 3 as slope (diaphragm response) vs. pitot pressure.
Figure 3 is the most significant information in this presentation and should be studied very carefully. It contains a lot of information. The required total energy response, and I have selected 4000 feet msl, is equal to C2 effective divided by the pressure Altitude. C2 effective includes the volume in the thermos flask plus the volume in the flask side of the variometer line plus the volume under the diaphragm, C 3 Figure. 5.
These small extra capacities were really not too important when we were flying over very limited speed ranges. However, as the speed range increases -we are actually considering speeds up to 200 mph as far as this design work goes -- it becomes important. In Figure 3 the straight, solid top line actually represents the solid line on the previous figure -- the required diaphragm response. This is what happens. Our initial slope, as we saw in Figure 2, is 1.49 cc per inch of H20 pressure change on the diaphragm. This is the left, or starting point, in Figure 3. As we dive and pick up speed, we move to the right in Figure 3 and the diaphragm is displaced toward the bottle. We find that as we move to the right side of our plots we have displaced our diaphragm because of the 20 inches of pitot pressure. We have actually moved the diaphragm and decreased C2 effective by 28 cc. At the same time, we are not where we started. We have dropped; because at altitude you can trade each inch of pitot pressure for one inch of pressure altitude; and we have actually descended 20 inches of pressure altitude, so the pressure altitude of 352 inches of H20 is not increased to 372. So, if we work it out again, the response required on the right (or high speed) end is now about 1.35 cc per inch of pitot pressure. That says that we really don't want a perfectly linear device, we want a device that is almost linear but falls off on the high end.
The test that we perform with the dynamic bench calibrator involves the total energy system that is set up according to these rules. We start at high air speeds and climb at a given rate, either 1000 fpm or 2000 fpm. This is what the sailplane would actually be doing. While we are doing this maneuver we will read the variometer in the total energy compensated system. Any non-zero reading is an error in the system. If we are climbing at 2000 fpm and we show 200 feet on the compensated variometer, this is a 10 percent error. We find that we are able to adjust the compensator so the error will be less than 10 percent above 100 mph and less than 5 percent between 100 mph and stall. Figure 3 shows the measured response of a typical unit adjusted for 4000 feet msl.
To check out the diaphragm response, I want to offer another system that is less complicated than the bench calibrator. The set-up in Figure 4 will do the job. What we are checking here is the response, that is, the change in volume divided by the change in pressure such as we started with in Figure 2. Figure 4 shows two simple U-tube manometers. Syringe 1 and syringe 2 are simple throw away 10 cc capacity types. They are good for over 100 mph. We set up the system just as in Figure 2. A trick in running this system is that there is some interaction between syringe and syringe 2. To start with, you would have syringe 1 extended (pulled out) and syringe 2 run in to zero (pushed in). Then start applying the pressure with No. 1. While you are applying the pressure with No. 1 with your left hand, just back syringe No. 2 out with your right hand and maintain the null. You can perform this test quite easily and you and limited only by how much you want to refine the test. You could use electric manometers in place of the U-tubes and measure pressures to 100ths inches of water. You can also refine the volume measurement, but generally speaking, these throw-away syringes will resolve 2/10th of a cc, and I think that is close enough for this type of test. This concludes the discussion of the diaphragm.
The third part of the system is the sailplane itself. This is what we face. We take our system out and put it in the real world. In Figure 5, the variometer system is on the right and the sailplane is shown on the left. RP and RS are the pneumatic resistance of the pitot and the static lines. C 4 is the biggest culprit. It represents the collected volume of all the other instruments in the sailplane that require a static connection. The manner in which C 4 trouble manifests itself in the sailplane is as follows. You get the vario system on the right in Figure 5 all tuned up and working correctly on the bench at the altitudes that you have selected and then take it out and put it in the sailplane and go to altitude (your compensated altitude). Then perform a very rapid pitch maneuver. This is a pretty abrupt maneuver which has about a 5-second cycle, about 2-1/2 seconds down and 2-1/2 seconds up. You rotate just about as fast as the ship can rotate in pitch. With this situation, and with tail static ports and some extra capacity on the panel (C 4), your first indication (with back stick) is a big down. And it's a big down! And the correction for this erroneous reading is to add restriction at R 1 to slow the pitot until the signal from the pitot and the static lines are in phase. You can do more to correct your total energy system by changing R 1 than any other one thing you can do. We'll make a little proviso. The diaphragm response has to be pretty good. Figure 6 describes the bench calibrator. It has an air speed indicator, rate of climb, and the variometer. If you were making one of these units, your variometer would be the one shown here as the indicator 1. One of the questions frequently asked in letters I get relates to R4 and R 5' I think the thing I failed to do last year was to get it across that these really had to be instrument quality fluidic elements. They must support 40 inches of pressure from pitot (Figure 6) to atmosphere; the reason being that we are operating to 20 inches at the virtual static point which is only half of the total pressure that R4 and R5 have to support. These restrictions must operate under linear conditions. The idea is to make R4 and R5 operate at high flow rates so that we can ignore the small flow required by the static side of the variometer.
Another point is that you should have a filter on the intake to the pump. In my case, the pump is the power unit from a tank type vacuum sweeper. The filter that comes with the sweeper is sufficient to take all air pollution out so that you won't be blowing small airborne particles into the restrictions and through your instrument system.
Figure 7 shows construction detail for R4 and R5' These are constructed by taking 8 pieces of 1/32 inch I.D. brass tubing and placing them inside a thin walled brass tube that is 1/4 inch O.D. The space between them is filled with epoxy so that the only air path is through the bore of the small brass tubes. A convenient way to do that is to buy these small tubes in 12-inch lengths. You can get them at the local hobby shop. Seal the ends of the small tubes by soldering the ends shut; fit them inside the big tube; suck the epoxy up through the space between tubes with a syringe; then let the epoxy cure. Saw the ends off and cut them to length and you have the two restrictions. Don't try a short cut and cut your little tubes and mix them so that they don't run continuously through R 4 and R 5' This whole exercise is really a beautiful way to measure the I.D. of small bore tubes in that the pneumatic resistance of these small capillaries is dependent upon the bore of the tubes to the 4th power of the inside dimension. If you will do as I say and use continuous lengths and then cut them in two, you can come out with units that are fairly well matched. If you want to tune a variometer at something other than ground level, you must adjust the ratio of R4 to R5'. For example, at 3000 feet above the level of the bench calibrator, R4 should be 5 inches and R5 should be 5.6 inches. This is just the pressure ratio of the two altitudes.
Figure 8 details the compensator that we showed last year and it included all the restrictions and the precautions and the provisions that we thought necessary to make a satisfactory total energy unit to fly over a wide speed range. There are some things that we found out. The first thing was that the adjustable spring was very complicated and it required such a large volume of cutout area above the R 2 restriction that we had, effectively, a second capacity (C3 Figure 5) that was not working through a restrictor. This caused enough trouble that we redesigned the unit and made it much simpler to build -and probably much easier to copy -- if you are interested in building one.
Figure 9 is the unit that we make now. The figure is somewhat misleading because it is not to scale. Try to keep the volume under the diaphragm as small as possible.
We recommend for a starter that you use 2 inches of 0.025 inch capillary for R1 and R2' This will give you about a 2-second time constant. Then you fill the reference capacity with three Chore Girl pot cleaners. These are little metal scouring pads that have no soap or tramp metal, and they are very clean. You just place three of them in a pint bottle and it makes a nice loose pack and does a very neat job of making the process very close to isothermal. The change in scale factor is insignificant.
I have some numbers here that you might like to note. The spring is one inch long, 0.475 inch O.D., is made of 0.050 inch steel wire, and has 6 free turns. The course adjustment is made by adjusting the number of free turns. If you determine that you have too much compensation, just increase the height of the epoxy button and short out, so to speak, a half of turn or a turn until you get up into the range that can be tuned easily just by changing the tension on the latex diaphragm. I would say that, typically, you have about 2/3rds of your spring constant in the spring and about 1/3 in the latex. The only way you can change your response is to change the spring constant. You don't get anything by changing the tension. You must change the constant and the constant here is the parallel combination of the spring rate of the wire spring and the spring rate of the latex diaphragm. We can get enough adjustment in the latex diaphragm that we can adjust the compensator response quite readily. This concludes my comments on the diaphragm compensated total energy system. Are there any questions?
Moore: What this new device does is to provide a couple of fingers that extend through an O-ring seal and actually turn the little washer inside the PZL from the pilot's side of the panel. This will change your spring tension but it will not change the spring constant. You must change the spring constant to effect the response. It won't do the job that I'm talking about. All it could do would be to allow the pilot to move the spring back to where it should have been in the first place. In other words, he can find the factory range where it is against the diaphragm with no preset in the spring. You can probably turn it enough to get in a lot of trouble if you screw the tension up too much. Then you will run out of compensator entirely when you slow to thermaling speed and then your instrument will be entirely uncompensated right at the range in which you want to fly. If the spring bottoms out at 60 mph and you want to fly at 50, you won't have any total energy at all.
Question: Gene, if that were a variable wound coil spring it would work wouldn't it? Different spacing?
Moore: No. However, if you put a little shaft inside the coil spring with a bridge that would wind down through the coils so you could change the length of that spring, it would work, The basic problem is that the PZL simply doesn't have enough capacity to get much of an altitude adjustment anyway. The way you must fix this is to change the volume. If you have a PZL system that is working now at your lower altitudes and you're going to go to Marfa, the safest thing you can do is take some volume out of your bottle. Put a little section of closed cell polystyrene or something in there. About 40 cc would give you 3000 feet altitude increase. It will change the calibration of your instrument, it should be noted, but that's a minor penalty to pay for a good working total energy system.
Question: Over how much speed range will the normal PZL compensator operate?
Moore: Well, of course, that depends upon the altitude. We had some plots in the Symposium Proceedings last year and I believe they showed some typical responses. Generally from about stall to over 100 mph. Where you really eat up the response is when you get to the high speed range. A little bit of air speed change here is a lot of pressure.
Question: (Squillario) I have the eternal question, Gene and Wil Schuemann. When are you going to make some of these instruments?
Moore: I think we answered that last year. We're tinkerers, we're not manufacturers.
Question: Surely you don't think that the 10 percent drop in sensitivity with altitude is critical?
Moore: Well, yes, it is. I think that you should have your system set up for what you think is going to be your mean altitude throughout a contest period. Over the past year, in working with these things, I en. more convinced that we have more of a problem in the time constant than we have in some of the responses not being right. We can set the response on the bench calibrator but we can't set the time constant. The sailplane is involved in that problem too strongly and you should have a flight test. That is really why Wil and I wouldn't be interested in doing any consulting on variometers unless the guys could come to Cumberland and have a flight test and get it checked out. It won't work out without a flight test. You'd be unhappy about it.
The only instrument that you can buy right now that has all the elements to do this job is the Ball, and it does it very nicely.
Question: Why did you decide not to make the leaky capsule type like the Ball, where you don't have to worry about altitude?
Moore; I was looking for something that I could do in the basement. It is more difficult for me to work with his technology with a soldering gun and a pair of pliers. The thermistor is a much simpler unit to build starting from scratch.
Question: Is it faster than the Ball, for example,
Moore: Yes. You can make a thermistor unit as fast as you want, but we don't need all that speed.
Question: What, if any, is the objection to the BSW?
Moore: Well, the BSW leaves the pilot with the problem -- what is he going to do for total energy. You see, that is the important thing. They need to furnish a working total energy device with their variometer. It's otherwise a beautiful system. A very nice instrument. It is well made and has a beautiful tone, if you're a musician. It's quite pleasing. There are a lot of them sold but it still leaves the pilot stuck with solving the total energy problem.
Question: In other words, the ideal total energy compensator -with the BSW system would be as good as any system?
Moore: Yes. However, it would fall off in sensitivity with altitude. If that doesn't bother you (it wouldn't bother me) then it's O.K. It will decrease in sensitivity as you increase altitude.
Question: Gene, what do you recommend as a compensator to be used with the BSW?
Moore: I would recommend one that was made using the ideas discussed earlier. Whether you or someone else makes it. Incidentally, you can make one with a couple of Skippy peanut butter jar lids and a spring and a diaphragm. It doesn't have to be cast like Figure 8. If you want to do the job, there is enough information here to do it.
Question: If the Ball variometer has all these advantages, why isn't it more popular?
Moore: I think that pilots really haven't found out how good it is. That's my personal opinion. It's got an altitude adjustment that will go from sea level to about 5000 feet msl. Incidentally, I found that it requires about five turns per 1000 feet. You might want to make a note of that if you have a Ball.
Question: Have you found that the Ball variable total energy compensator works well?
Moore: Oh, yes it works. If it isn't working, you may need to take out or add a little restriction to R1. Ball has an R 1 in the pitot approach so you may have to adjust this restriction to compensate for C4 in Figure 5. Question: For a quick fix in your pitot, could you just change the length of line? Maybe by putting some extra line in and then tuning it? Moore: That would be a lot of line because one inch of 0.025 inch restrictor is equivalent to about 20 or 30 feet of line. A needle valve would be the best.
Question: Is there a needle valve type restrictor available?
Moore: I think many instrument companies have a suitable type of needle valve. Anyone selling scientific equipment would have it in the gas chromatography field or industrial instrumentation. For example, Imperial Eastman Company has a one degree taper needle metering valve in their Poly-Flo fittings that is fairly inexpensive. About $6 to $7.
Question: But this would slow your air speed, wouldn't it?
Moore: Don't put it where it bothers anything but the variometer. In other words, put it after you tee off the pitot. No. Don't get into your other instrument system.
Copyright Soaring Symposia All rights reserved. Permission to copy this article is granted for non-commercial use, in its entirety, and with this copyright notice attached.