| ||
| AgTalk Home | ||
| ||
 Reliability/Accuracy of yield monitors in S-series JD combinesJump to page : 1 Now viewing page 1 [50 messages per page] | View previous thread :: View next thread |
| Forums List -> Precision Talk | Message format |
dpilot83 |
| ||
In this thread I've had some dialogue with a few people in regards to how to post-process yield data. Before that conversation I was under the impression that you could get some useful data even if calibration wasn't perfect. Now I'm leaning towards the calibration process being extremely important, especially if multiple combines are involved. We hire all of our harvesting done. The crew we hire runs six S670's with 2630's. I just talked to them and throughout the conversation I sensed the same disenchantment with technology that I have frequently felt when upgrading equipment with high expectations for good results. He explained to me that when they were down south cutting irrigated corn (before our corn was ready) they spent a lot of time carefully calibrating two of the machines. When they were done with that field the machines indicated that the field had yielded within 1 bpa of what the other machine said. At the end of the next field the first combine said that the field had averaged 20 bpa higher than the second combine said. At the end of the third field the first combine said that the field yielded 20 bpa lower than the second combine said. After their initial and thorough yield calibration process, they did not change anything, they were just harvesting. The operators of the two machines were brothers and partial owners of the crew so they knew what they were doing. This description is in stark contrast to what I frequently read on NAT and hear from others. Many times I hear that a properly calibrated yield monitor will read within 1% to 2% of what the scale tickets say. I know there is variability in fields but these combines were just following each other back and forth staying close together on fields 1 mile long so that the grain cart could be more efficient. There is no way that they could have experienced yields that were 20 bpa apart from one another. So what's the problem? Are John Deere yield monitors not as accurate as Ag Leader monitors still even after they have Ag Leader sensors and an Ag Leader style calibration process? Is our harvest crew doing something wrong? Is it something else? Any thoughts would be much appreciated. Thanks. | |||
| |||
| Wheat77 |
| ||
| Yield monitors in general are not very accurate in my opinion. I have calibrated to a known weight through an elevator scale or grain cart scale, and even when doing it again on the next field, being off 3-5% is good, often can be much more than that. The hardest thing for a YM is uneven yielding crops such as soybeans that may be from 15 - 50 bus. in the same field. | |||
| |||
| premiertech |
| ||
Alden and Osage, IA | dpilot83 - 2/20/2013 15:05 In this thread I've had some dialogue with a few people in regards to how to post-process yield data. Before that conversation I was under the impression that you could get some useful data even if calibration wasn't perfect. Now I'm leaning towards the calibration process being extremely important, especially if multiple combines are involved. We hire all of our harvesting done. The crew we hire runs six S670's with 2630's. I just talked to them and throughout the conversation I sensed the same disenchantment with technology that I have frequently felt when upgrading equipment with high expectations for good results. He explained to me that when they were down south cutting irrigated corn (before our corn was ready) they spent a lot of time carefully calibrating two of the machines. When they were done with that field the machines indicated that the field had yielded within 1 bpa of what the other machine said. At the end of the next field the first combine said that the field had averaged 20 bpa higher than the second combine said. At the end of the third field the first combine said that the field yielded 20 bpa lower than the second combine said. After their initial and thorough yield calibration process, they did not change anything, they were just harvesting. The operators of the two machines were brothers and partial owners of the crew so they knew what they were doing. This description is in stark contrast to what I frequently read on NAT and hear from others. Many times I hear that a properly calibrated yield monitor will read within 1% to 2% of what the scale tickets say. I know there is variability in fields but these combines were just following each other back and forth staying close together on fields 1 mile long so that the grain cart could be more efficient. There is no way that they could have experienced yields that were 20 bpa apart from one another. So what's the problem? Are John Deere yield monitors not as accurate as Ag Leader monitors still even after they have Ag Leader sensors and an Ag Leader style calibration process? Is our harvest crew doing something wrong? Is it something else? Any thoughts would be much appreciated. Thanks. The newer yield monitor system on S series combines 'should be' every bit as accurate as the Ag Leader system. They are using the same calibration procedure and method that Ag Leader has used for years. It sounds like your operators did a good job calibrating the yield monitor system. That being said, I know there were some issues (published or not) with S series combines having variances in elevator to deflector clearance at the top of the clean grain elevator. As far as we could tell last fall, there wasn't much rhyme or reason to what combines had proper clearance and which ones did not. These issues were present from the factory. Maybe it has been traced down by now, I haven't heard. This clearance irregularity could definitely be a reason that your yield monitors were right, then off one way, then off the other way. Those particular systems acted as if it just couldn't hold a calibration. | ||
| |||
| CaseFarmer |
| ||
Flora IL | Should be. Had better be. Our data is always very accurate..... neighbors would like to be as accurate as ours. Thst and not have to replace moisture sensors yearly... | ||
| |||
| KMech |
| ||
Missouri | Some variability is bound to happen, but what you described sounds excessive. When I run a combine, I know how many bushels my tank holds when the tank full buzzer sounds. I periodically compare the number of bushels counted in a binful. It's not precise, but I can tell if it gets off by more than 5%. Heavily loaded swaths have a tendency to under indicate. Low flow (that long last two row swath with six row head) usually tends to over indicate. Intermittent flow (lots of short rows in a terrace point or corner of field) usually under indicate, but in some situations I've seen them over indicate. [excuse this paragraph. It's a mess, but I've written a long reply, and have lost interest in revising] A yield monitor is not, and was never intended to be, a weigh scale replacement. It is convenient to look at the monitor total tab, and with a good calibration, that should be within a few percent of scale tickets, but if the total bushels is what you want to know, it is not a substitute for a scale. They don't weigh grain, they measure grain flow, and from that grain flow calculate an *estimated* yield volume. A yield monitor does not weigh or measure grain. It measures the force applied to the impact plate from the grain thrown at it as the paddles cross over the top of the clean grain elevator. From that force, it calculates a grain flow. From that grain flow, it further calculates an estimated yield volume. When you consider how all of this is done, it is really a wonder that they are as accurate as they are. If I can send you an email, I have two small, 6mb videos that illustrate this very well, and show some things you can look for when troubleshooting an issue like this. I have been calibrating yield monitors for 10 years, and I learn something new every year. Achieving the perfect calibration is really an art form. Conditions will change, machine parts will wear, and eventually it will be necessary to perform a recalibration to restore accuracy to the changed conditions. The machine is always the most accurate in the field in which it was calibrated, and in the hybrid in which it was calibrated. So many things can affect it. Kernel/seed size, test weight, moisture, all affect the way the grain hits the impact plate, which affects the force measured by it, which in turn affects the grain flow calculated from that force, which in turn affects the yield volume calculated from that grain flow. I've seen a corn combine get set to thresh too tight, and get too much cob in the bin. Excess cob cushions the grain flow at the sensor, lowering the indicated grain flow. Readjusting the separator brings it back. The same thing can happen in beans with lots of pods in the grain flow. Mechanical issues can also have effects. As the elevator chain wears and stretches, that affects the way it throws the grain at the top of the elevator too. That's not the only issue. We had a straw chopper that was so badly out of balance you could feel the vibration in the cab. The year in which we repaired that issue, the yield monitor had half as much error as the year before. Even the way the machine vibrates affects the way the flow meter measures grain flow. The real value in a yield monitor is not the total bushes on the summary screen. It is the map that shows you, proportionately, how many bushels came from where. Having an accurate calibration is important, but if it is off a few %, that does not substantially reduce the value of that data. Say you were wanting to divide your field into three zones, to identify the high, average, and low yielding thirds of the field. If the calibration is off a few percent, the numbers listed on the breakpoints for the legend of that map will be off, but the overall map will be almost unchanged. At least, that would be the case with a single machine harvesting a field. Having multiple combines in the same field does somewhat complicate that, and having as accurate a calibration as possible becomes more important in this case. dpilot83 - 2/20/2013 15:05 So what's the problem? Are John Deere yield monitors not as accurate as Ag Leader monitors still even after they have Ag Leader sensors and an Ag Leader style calibration process? Is our harvest crew doing something wrong? Is it something else? Any thoughts would be much appreciated. Thanks. I don't think the issue breaks down along brand lines. I have seen that kind of difference from one Ag Leader equipped JD combine to another. My Dad's combine has a two year old yield system on a 60 series Deere. With older systems, we got into a habit of calibrating annually, in each crop. In the second year, I calibrated his corn, and using the previous years calibration, it was only off about 2.5%. He never called me to calibrate beans. I asked him about it after the season, and he said he had compared the monitor to scale tickets, and it was too close to recalibrate. Another customer, with another 60 series deere, had an issue we never figured out. On the day we calibrated, it was amazingly accurate. Day by day, as the season went on, it was under indicating more every day. He sold that combine before I ever got the chance to find the problem. I uninstalled his yield kit, and reinstalled it on a 70 series deere, and he never had a complaint on that combine. I'll never know what caused that issue. Most of them work really well. There's always the exception that performs flawlessly, and there's always the exception that performs unsatisfactorily. The hard part is getting a customer with an unsatisfactory system to let go of his combine during harvest long enough for you to work on it. I don't know how the deere machines manage their calibration, but with an Ag Leader system, you can always add in additional calibration loads to further dial in a calibration, at any time. It's a good idea to cut a test load anytime an operator has reason to suspect the accuracy is beginning to diminish. | ||
| |||
| Scott (OSU) |
| ||
Dalton, OH | We've been running Ag Leader yield monitors for a few year, but decided to use the 2630 when we switched to out new s660 this past season. We mostly decided to switch since we knew JD was using Ag Leader parts. One thing we found out is that while there software is similar it is not the same. One big thing that was missing, maybe they changed in a recent update, was the ability to make the calibrations retroactive. That is such an important feature in my mind.One issue that you point out and we saw first-hand is our combine's calibration not holding. We would calibrate it and it would work good all day long. Next day we would start out and it would be low 20% the whole day, next day it would be high 20%, and then one day it would be right back in calibration. We tried multiple things and never did have any luck with it. On our larger fields that we ran in multiple days it kind of makes the yield maps useless. I'm going to look into the elevator clearance that you mentioned. Thanks | ||
| |||
| Wheat77 |
| ||
Scott (OSU) - 2/21/2013 19:07 We've been running Ag Leader yield monitors for a few year, but decided to use the 2630 when we switched to out new s660 this past season. We mostly decided to switch since we knew JD was using Ag Leader parts. One thing we found out is that while there software is similar it is not the same. One big thing that was missing, maybe they changed in a recent update, was the ability to make the calibrations retroactive. That is such an important feature in my mind.One issue that you point out and we saw first-hand is our combine's calibration not holding. We would calibrate it and it would work good all day long. Next day we would start out and it would be low 20% the whole day, next day it would be high 20%, and then one day it would be right back in calibration. We tried multiple things and never did have any luck with it. On our larger fields that we ran in multiple days it kind of makes the yield maps useless. I'm going to look into the elevator clearance that you mentioned. Thanks And did you know, the Deere Fed. Crop Ins. arm, wants to use combine yield data as proof for production. What a joke. | |||
| |||
| berta |
| ||
| Nothing has changed on the s series accurasy for yield monitoring quite frustrating when you go thru process of putting data in and yield data all over the map and inaccurate .thought 2016 Jd s series would be better then our older s series models but not data basically useless | |||
| |||
| Jump to page : 1 Now viewing page 1 [50 messages per page] |
| Search this forum Printer friendly version E-mail a link to this thread |
| (Delete cookies) | |
