Andymark 1.1 Compressor Data
 

My team just did some tests to compare the new 1.1 pump to the VIair pump and the larger Thomas pump and I thought it would be helpful to share it with the community. TL;DR The new 1.1 pump is the fastest at getting 3 574mL tanks from 80-120 psi by a significant margin(on average 7-11 seconds). Also Thermal Overload will be tested at a later date.

We set up our three compressors with fittings so that they could be hooked up to the same set of tubes and tanks by taking out only 1 tube end. We then would charge the air tanks to about 80 psi with the current compressor we are using, then stop to adjust the amount before starting the test. We would then start the compressor and time how long it took to reach 120 psi. We did this 5 times with each compressor, and each compressor had its own fully charged battery to use. We ran each trial one after another, meaning that the battery was only on full charge for the first trial of each experiment.
Now on to the data(please excuse the poor formatting)
Andymark 1.1
Trial 1 - 30.68 sec
Trial 2 - 26.02 sec
Trial 3 - 27.38 sec
Trial 4 - 27.83 sec
Trial 5 - 28.02 sec
Average - 27.99 sec
VIair 90C
Trial 1 - 42.35 sec
Trial 2 - 37.47 sec
Trial 3 - 38.82 sec
Trial 4 - 38.25 sec
Trial 5 - 38.90 sec
Average - 39.17 sec
Thomas
Trial 1 - 37.42 sec
Trial 2 - 34.82 sec
Trial 3 - 34.08 sec
Trial 4 - 35.82 sec
Trial 5 - 34.97 sec
Average - 35.42 sec

From this data, it is clear that the new Andymark 1.1 pump is the fastest to fill the tanks by a sizable margin. Even comparing the slowest fill from the Andymark pump to the fastest of the VIair and Thomas, it still comes out the fastest by about 2.5 seconds, and on average comes out at around 7-11 seconds faster than the other two.
Something rather interesting to note is that for every compressor, the first trial is significantly slower than the rest, which is odd. Another interesting trend is that the second trial seems to be the fastest, while subsequent trials tend to get slower as you continue. Maybe there is an optimum operating temperature for the compressors, but further testing would be required that is out of the scope that we are capable of doing.
Possible causes of error include using different batteries for each compressor, which may be at different stages of being able to hold a charge, and not having a fresh battery for every trial. I do believe that the significant difference in charging times makes the change that this would have caused minimal.
All in all, it seems clear that the new pump is superior to both older styles of compressors. Though it may weigh about a pound more than the VIair, it seems to make up for it in its output. The only thing that worries my team is the thermal overload tripping in the middle of the match. We will be doing another experiment to see how that works out and how it compares to the other two compressors in the near future.

Re: Andymark 1.1 Compressor Data
 

Doing a quick bit of statistical analysis: Each of the three options is significantly different from the other two, with p<.01. (Using ANOVA)

Re: Andymark 1.1 Compressor Data
 

Which model of Thomas compressor are you referring to when saying this?

Re: Andymark 1.1 Compressor Data
 

Good job on the data collection.
One minor upgrade to your testing setup would be to only switch valves and not air tubes and have a separate compressor/battery charge the system to 80 psi.

I mention this as minor as the data seems fairly consistent.

My next question is the power needed worth the speedier air fill?
This might be hard to answer for an unknown situation.
So, how about the efficiency of each unit?
Another useful measure might be amp-hrs needed to fill a tank.

The AM1.1 has a max draw of 16 amps
The Viair 90c has a max draw of 11 amps
(from the AM site)

Dave

Re: Andymark 1.1 Compressor Data
 

Just curious...why 80-120?

Re: Andymark 1.1 Compressor Data
 

My guess is, that is the approximate range the pressure sensor switch swings the pressure during a normal match.
Again, that's my guess. Let's see what the OP says was their reasoning.

Re: Andymark 1.1 Compressor Data
 

I'm referring to the pump that came in the KOP years ago.

I will also try to get data on amp hours at the 80-120 psi mark and filling up tanks from 0psi.

Re: Andymark 1.1 Compressor Data
 

Thomas 405ADC38

Also, for reference: http://findrobotparts.com/compressors/

Re: Andymark 1.1 Compressor Data
 

It's interesting that the Thomas compressor is pushing the cost limit for a legal COTS item.

Re: Andymark 1.1 Compressor Data
 

80-120 was to simulate the pressure that would be mainly relevant during a match

Re: Andymark 1.1 Compressor Data
 

I know for 254's robot last year we mainly sat in the 70-90 range during matches. I also saw pressures as low as 40 and we almost never exceeded 100, so Im not sure if the 100-120 section is even relevant in your test.

IMO it would be better to test the 60-100 range to see how long it takes for your robot to recover between actions.

I would hazard to guess the reason why slower in first trials may have to do with bearings, and the cylinder being cool.

Re: Andymark 1.1 Compressor Data
 

My team isn't too terribly pneumatic-heavy, unlike the cheesy poofs(see their 2014 robot). We also store air before the match, and try to avoid running the compressor constantly during a match. A way we get around not running it constantly is having more air tanks, giving us more air to work with between the compressor turning on.
Out of curiosity what compressor did you use last year, and how often were you actuating cylinders?

Re: Andymark 1.1 Compressor Data
 

What your average pressure is would depend on how much tank capacity you have compared to your usage. It sounds (not surprisingly) like 254 is optimizing weight a bit more than we are. For us, 100 psi is about average while the compressor is on - we usually set the switch to about 115-117, and our low pressure under moderately heavy use is about 85-90 psi.

It is possible that the test being centered on 100psi is based on my post in the other AM1.1 thread.

Does anyone have any good gouge on the duty cycle of the Thompson compressor referenced above? None of the web pages or data sheets listed it. Unless by "continuous" in some labels they mean 100%? If so, the Thompson would outperform the 1.1 in a demo/practice mode, and in many cases in competition mode. That would also explain the much higher price.

Re: Andymark 1.1 Compressor Data
 

Don't mean to sidetrack the discussion too far here, but...

What do you mean "usually set the switch"? Are we allowed to use adjustable pressure switches again? I remember being part of a lively discussion a couple of years ago where the Chief Robot Inspector took us down a rule interpretation that I didn't anticipate. The rules got more specific after that point, including R78c from last year.