hap
Junior Member
Posts: 89
Registered: 30-1-07
Location: happeninghap@gmail.com
Member Is Offline
|
![[*]](./images/corporate/default_icon.gif) posted on 12-7-08 at 01:33 AM |
|
|
testing 123
I played with the Leakalyzer this morning in my kitchen. I ran some 5 gal bucket
and eye dropper tests after reading the manual. One of the points the manual
discuss' is air, water, and cell temperature. I live in Phoenix so that got my attention
right away. It was inside all night so the cell temp was stable relative
to my indoor temp. I have an infrared thermometer and checked the temp of
the tap water. Everything looks okay within one degree of each other. So I
setup the device on my bucket of water and just let it sit there for a 5 minute
test after I set the evaporation rate to zero. The first tests showed the bucket
filling. It took about 45 minutes of testing before I finally got a flat line on the graph.
Next I used my eyedropper to remove 27 drops of water, wait a minute, and add it back.
The leakalyzer showed an amazing 33 "leakalyzer units" of loss and then put 25 units back in.
Later I loaded up my new toy/tool and headed over to a pool that is half full and set it
up on the second step. I ran several 5 minute tests on a 350 square foot parameter
with .25" evaporation under a cloudy sky all showing 150 to 250 gallons loss per day.
The water was at the bottom of the light conduit but the Leakalizer showed some pretty
fast loss still. That left main drain, 7 floor heads or structure leak. I was able to flat
line the Leakalyzer in a few minutes after plugging the deep end heads on some 10 minute
tests.
My first experience at testing and using the Leakalyzer is that it's very very sensitive.
It's graphing and units of loss is unstable while it's cell temp is adjusting and stabilizing.
This morning in my kitchen the units of water gain were small, 7 to 10 units. On the job
the clouds started to move out during the latter time I was there and you could see the
graph line move due to the heat gain or loss depending on cloud movement. If the pool
being tested is loosing water fast the early unstable tests won't matter because the loss
will over take the unstable graphing jitters anyway. On smaller leaks I suspect I'll need
to wait out the graph jitters until the cell stabilizes.
The customer showed up during the latter part of my visit to his pool. He was eager to
see 5 of the saved tests I did on his pool. He was very impressed and I looked like a
well equipped pro.
Lance, I suggest you add 5 more to the save list in future software versions.
Hap
|
|
|
Lance
Super Administrator
Posts: 48
Registered: 16-3-06
Member Is Offline
|
| posted on 12-7-08 at 09:55 AM |
|
|
Hap
I strongly suspect that the "instability" you found when doing your test on a 5 gallon bucket had to do with the instability of either the bucket or
the surface on which you placed the Leakalyzer. It probably has less to do with temperature changes. In our testing we found that most any plastic
material (wether it be the bucket or the surface on which the Leakalyzer is placed) relaxes over time when put under load. This relaxation causes
measurable changes in the relation of the Leakalyzer base to the water surface.
The pool deck and the pool do not usually experience this type of movement.
Regarding the changes as clouds move in and out. Especially for you in the Phoenix sun, you should have best results if you are able to keep the unit
shaded throughout the test.
|
|
|
ALD
Junior Member
Posts: 38
Registered: 11-12-05
Member Is Offline
|
| posted on 13-7-08 at 03:47 PM |
|
|
Wow, the leakalyzer sounds awesome. Can't wait to get one.
|
|
|
hap
Junior Member
Posts: 89
Registered: 30-1-07
Location: happeninghap@gmail.com
Member Is Offline
|
| posted on 15-7-08 at 09:49 PM |
|
|
I revisited my kitchen bucket test and made sure every thing was very stable and got very stable results. I could even see the evaporation differences
with a room fan on and then off.
Today I was able to test the Leakalyzer for several hours on a non leaking pool with the auto fill off and under a large pool shade. I left everything
as it was being sure not to change any variables on the pool or equipment. Pump on.
I tested this pool many times today and never moved the test cell only the perforated water hose. The results were frustrating because the graphs were
not stable and I could not get the instrument to repeat the results. Later I taped all the holes in the hose and left the end plug hole open. Then I
drilled a small hole in a 5 gal bucket close to the bottom, added some lead weights and let it equalize on the top step.
Next I filled the hose with water again and carefully attached the cell end and held my finger over the hole and put it in the bucket for a smoothing
buffer. Wow what a difference!! I've uploaded two images of the graph screen so you can see for yourself.
To read the graph left to right top row in the first image
55 is the amount of loss in Leakalyzer units. Whatever that is. Positive numbers are loss.
750SF is the square footage of this pool.
20:06 is how long the test ran. You can end it early and save it.
The next two numbers left of the graph are the graph resolution. In this image it's 0 to 60. That's why 55 is almost to the bottom.
Okay, the next row is the test length, choose 5:00, 10:00, or 20:00 minutes. I did 20 on this one.
The bottom row is how much I think the pool is evaporating per day in inches.
The graph shows the evaporation line as a doted straight line sloping down at different angles depending on the resolution.
At any time during a test you can hit the detail button and read water loss and evaporation loss data in gallons/inches per day or hour.
You can save 5 graph/data screens. I wish this was much more. Like 25 or something.
http://img26.picoodle.com/img/img26/4/7/15/f_PICT1915m_cd3e059.jpg
http://img29.picoodle.com/img/img29/4/7/15/f_PICT1916m_f31e53c.jpg
This is how this version of the software is working on my unit. I spoke to Lance today, he's got new ideas already.
Todays Leakalyzer thought for the day is..... if you're thinking of purchasing this unique instrument, wait a while until....I don't know, just
wait. I love to tinker and I'm sure it shows. This thing will be around for many moons to come. If you like to tinker buy now and email me often.
happeninghap@gmail.com
hap
|
|
|
AquaLeak
Junior Member
Posts: 2
Registered: 9-12-05
Member Is Offline
|
| posted on 16-7-08 at 06:30 AM |
|
|
Thats what I need, can you send me a picture of the bucket setup. That line is very stable. After the modifications were the rest of the tests
consistent (how many tests)? I will have some pics at the end of the day today to share.
-rob
|
|
|
hap
Junior Member
Posts: 89
Registered: 30-1-07
Location: happeninghap@gmail.com
Member Is Offline
|
| posted on 16-7-08 at 09:32 AM |
|
|
Rob it's just a 5 gal bucket with a 1/16" hole close to the bottom and some lead weights in it to keep it from moving on the top step. After sealing
the perforated hose with black tape I held my finger over the open end and submerged it in the bucket.
I don't remember how often I tested the Leakalyzer yesterday. A bunch. About 5 or 6 with the bucket. I got some mixed results with the bucket but a
huge improvement over all. One of the problems is waiting for the cell and bucket to settle down. I over filled the bucket about a quarter inch and it
took over a half hour for the graph to stop it's dive down. The last 1/64" is what takes the most time. There's several ways to resolve this.
The drama continues......good luck testing and let us all know what you learn.
hap
| Quote: | Originally posted by AquaLeak
Thats what I need, can you send me a picture of the bucket setup. That line is very stable. After the modifications were the rest of the tests
consistent (how many tests)? I will have some pics at the end of the day today to share.
-rob |
|
|
|
Lance
Super Administrator
Posts: 48
Registered: 16-3-06
Member Is Offline
|
| posted on 16-7-08 at 01:57 PM |
|
|
Hap:
I knew you’d have stuff to say if we got one of these in your hands.
Thanks for posting the pictures. . . . As I look at both of the graphs I’m actually pretty impressed with the first one as well as the second. You
should have seen the jags on the graphs from some of our earliest testing!
The function of the Leakalyzer is to take actual data (water surface measurements at different times) and present it in a form that allows the user to
make decisions about long term trends. Specifically, we need to get a sense of what the constant slope of the line is. As you have noted, this is
easiest to do when we eliminate the wild gyrations or variations that the unit actually detects. There are three ways of minimizing the “noise”
of actual data in order to identify a useful trend line:
1.Mental– The users interpretation of the observable data (which includes outside conditions in addition to that presented by the Leakalyzer)
2.Mechanical – by controlling where the water is sampled and the flow of water from the pool to the sensor.
3.Mathematical – by using software to modify the collected data into the displayed data.
Our product development challenge has been to figure out how to balance Mechanical and Mathematical noise reduction while maintaining responsiveness
that offers the opportunity for the user’s Mental interpretation. The more we mechanically and mathematically “muffle” the presented data the
less the user is seeing the “actual” data.
We can mechanically reduce jags in two ways; either by sampling the water in different locations to get wave “average,” or by reducing the flow of
water to the sensor.
By using your bucket with the small hole you are minimizing the flow of water to the sensor. This does a great job of smoothing the graph but also
means that your results are delayed. For example if you pulled a gallon of water out of the pool the drop on the chart will be delayed and
“softened.” This is all fine and good, but not exactly representative of what is happening in the pool. I am going to send you a new Baffle Tube
that has an adjustable valve just before the connection to the sensor at the top. This will allow you to reduce the flow if desired, but it will
still allow you to sample water from various parts of the pool which is an important means of minimizing wave motion without reducing responsiveness.
Regarding Mathematical muffling of the data, keep in mind that the Leakalyzer allows for 5, 10 or 20 minute tests. The difference between these tests
(in addition to the obvious) is that the longer tests average data over a longer period for each point marked on the graph. Most simply stated, each
point on the 20 minute graph is an average of the previous 100 seconds. Each point on the 5 minute graph is an average of the previous 25 seconds.
When you want a smooth line do a 20 minute test (and check it every 5 minutes). Keep in mind however that what is displayed is a delayed response to
what has actually happened, (removing a bucket of water will show up 100 second later on the graph). When you want a more responsiveness do a 5
minute test.
You will also find that the line plotted on the graph will look much smoother when viewed at the largest (0-120 unit) scale. We have found that this
scale is usually the easiest one to see trends on when conditions are flaky.
As you first start using the unit I can appreciate your interest in trading responsiveness for ease of interpretation. However I think that as you
become more familiar with the unit and develop your interpretive abilities you will also appreciate and eventually demand the responsiveness it
provides.
While I’m sure there will be a few tweaks to the unit, I think it’s pretty much ready to go. I’d like to keep hearing your feedback as you
continue to gain experience.
Lance
|
|
|
hap
Junior Member
Posts: 89
Registered: 30-1-07
Location: happeninghap@gmail.com
Member Is Offline
|
| posted on 16-7-08 at 05:39 PM |
|
|
Lance
I did do one test yesterday with the bucket and at 10 minutes into a 20 minute test I removed a gallon of water from the pool. The the graph almost
instantly started moving down and kept going down as expected. It took about 2 minutes to bottom out. That's fast response.
For me I think buffering the sawtooth appearance of the graph line will make my transition(and my customers) to the Leakalyzer easier. I also think
that it doesn't matter how the buffering is done..... math, software, valve, or bucket will work. It's pretty apparent that the Leakalyzer buffering
software is set for response not smooth lines. I suggest doubling the initializing period.
One other test I did yesterday that I forgot to mention was with the baffle hose holes tape closed and no tape on the holes. I cold not tell any
difference in these two modes of use.
Also worth mention is yesterdays tests were in very still wind or no wind. Very calm surface water.
hap
| Quote: | Originally posted by Lance
Hap:
I knew you’d have stuff to say if we got one of these in your hands.
Thanks for posting the pictures. . . . As I look at both of the graphs I’m actually pretty impressed with the first one as well as the second. You
should have seen the jags on the graphs from some of our earliest testing!
The function of the Leakalyzer is to take actual data (water surface measurements at different times) and present it in a form that allows the user to
make decisions about long term trends. Specifically, we need to get a sense of what the constant slope of the line is. As you have noted, this is
easiest to do when we eliminate the wild gyrations or variations that the unit actually detects. There are three ways of minimizing the “noise”
of actual data in order to identify a useful trend line:
1.Mental– The users interpretation of the observable data (which includes outside conditions in addition to that presented by the Leakalyzer)
2.Mechanical – by controlling where the water is sampled and the flow of water from the pool to the sensor.
3.Mathematical – by using software to modify the collected data into the displayed data.
Our product development challenge has been to figure out how to balance Mechanical and Mathematical noise reduction while maintaining responsiveness
that offers the opportunity for the user’s Mental interpretation. The more we mechanically and mathematically “muffle” the presented data the
less the user is seeing the “actual” data.
We can mechanically reduce jags in two ways; either by sampling the water in different locations to get wave “average,” or by reducing the flow of
water to the sensor.
By using your bucket with the small hole you are minimizing the flow of water to the sensor. This does a great job of smoothing the graph but also
means that your results are delayed. For example if you pulled a gallon of water out of the pool the drop on the chart will be delayed and
“softened.” This is all fine and good, but not exactly representative of what is happening in the pool. I am going to send you a new Baffle Tube
that has an adjustable valve just before the connection to the sensor at the top. This will allow you to reduce the flow if desired, but it will
still allow you to sample water from various parts of the pool which is an important means of minimizing wave motion without reducing responsiveness.
Regarding Mathematical muffling of the data, keep in mind that the Leakalyzer allows for 5, 10 or 20 minute tests. The difference between these tests
(in addition to the obvious) is that the longer tests average data over a longer period for each point marked on the graph. Most simply stated, each
point on the 20 minute graph is an average of the previous 100 seconds. Each point on the 5 minute graph is an average of the previous 25 seconds.
When you want a smooth line do a 20 minute test (and check it every 5 minutes). Keep in mind however that what is displayed is a delayed response to
what has actually happened, (removing a bucket of water will show up 100 second later on the graph). When you want a more responsiveness do a 5
minute test.
You will also find that the line plotted on the graph will look much smoother when viewed at the largest (0-120 unit) scale. We have found that this
scale is usually the easiest one to see trends on when conditions are flaky.
As you first start using the unit I can appreciate your interest in trading responsiveness for ease of interpretation. However I think that as you
become more familiar with the unit and develop your interpretive abilities you will also appreciate and eventually demand the responsiveness it
provides.
While I’m sure there will be a few tweaks to the unit, I think it’s pretty much ready to go. I’d like to keep hearing your feedback as you
continue to gain experience.
Lance |
|
|
|
hap
Junior Member
Posts: 89
Registered: 30-1-07
Location: happeninghap@gmail.com
Member Is Offline
|
| posted on 17-7-08 at 08:03 PM |
|
|
More of the same today with the Leakalyzer. I had a pool that the customer says is leaking 5-6 inches per day. The Leakalyzer says nope no way with
the EQ off. I turn on the pump, still no way loosing that much. I turn on the Polaris booster pump and bingo.
I hung out and played with the Leakalyzer for several more testing runs just to get used to the thing. Talked to Lance again yesterday. He's sending
me stuff he wanted me to have and I asked for two of his very cool baffle hose connectors so I can hook up to my buffering bucket quick. I'll try
some more stuff in the weeks to come.
I must admit that was very cool this morning when I determined which pipe was leaking without getting my hands wet in less than 20 minutes. And did it
while the customer was yak, yak, yaking about his tenet's poor pool skills.
I'm still lobbying for more virtual buffering in the software. I really think that slower leaks are going to need it. Especially in small square
footage bodies of water.
hap
|
|
|
