Differential Pressure Gauge Installed

[Jim Allen]

Originally Posted By: AITG
Originally Posted By: Jim Allen
Originally Posted By: AITG
For the sake of discussion: how hard is it to swap the gauge units? I would want to verify that some of the 6 psi difference isn't just variation between the two displays.


It's a 30 second job. Good idea. That's something I hadn't thought to do, but I will. However these are very accurate stepper motor type gauges so I predict there will be no difference.


No problem. I'll send you a bill ;-)

I've seen hundreds of hours spent chasing 'problems' indicated by gauges that turn out to be a gauge problem. We have a tendency to believe any instrument - human nature, I guess. Think Three Mile Island, for example.


UPDATE: Good idea but when I swapped the gauges, I realized it wouldn't work because with the Isspro Performax system, the gauges are daisy chained on one wire (from a signal processor) and coded to read only the signal it was programmed to read.

Tried my first datalog a short time ago. Can only get the laptop to accept about four minutes of data for some reason (working on that) but on the cold start with oil temp at only about 44F, initial DP was 6.9 psi, even when the engine flared at start. When it quit recording, DP was 5.9 psi and oil temp was about 90 degrees.

 

[dnewton3]

Thanks Jim - was unaware of that setup.

Did the gages come with any kind of R&R statement or accuracy statement, either in the packaging or in the literature?

 

[ZeeOSix]

Originally Posted By: Jim Allen
Originally Posted By: AITG
Originally Posted By: Jim Allen
Originally Posted By: AITG
For the sake of discussion: how hard is it to swap the gauge units? I would want to verify that some of the 6 psi difference isn't just variation between the two displays.


It's a 30 second job. Good idea. That's something I hadn't thought to do, but I will. However these are very accurate stepper motor type gauges so I predict there will be no difference.


No problem. I'll send you a bill ;-)

I've seen hundreds of hours spent chasing 'problems' indicated by gauges that turn out to be a gauge problem. We have a tendency to believe any instrument - human nature, I guess. Think Three Mile Island, for example.


UPDATE: Good idea but when I swapped the gauges, I realized it wouldn't work because with the Isspro Performax system, the gauges are daisy chained on one wire (from a signal processor) and coded to read only the signal it was programmed to read.


You would have to physically swap out the two gauges' location and keep the signal setup the same on the data acquisition system. Doubt you want to go through that trouble.

 

[ZeeOSix]

Originally Posted By: Jim Allen
Tried my first datalog a short time ago. Can only get the laptop to accept about four minutes of data for some reason (working on that) but on the cold start with oil temp at only about 44F, initial DP was 6.9 psi, even when the engine flared at start. When it quit recording, DP was 5.9 psi and oil temp was about 90 degrees.


Was that all at idle speed, except for the brief engine flare in RPM at start-up? I'd be interested so see PSID with some cold oil and a few RPMs.

 

[Jim Allen]

Originally Posted By: dnewton3
Thanks Jim - was unaware of that setup.

Did the gages come with any kind of R&R statement or accuracy statement, either in the packaging or in the literature?


Microprocessor-controlled stepper motor gauges (google it) and my contact at Isspro told me they are accurate to .5-1%.

 

[ZeeOSix]

Originally Posted By: Jim Allen
Originally Posted By: dnewton3
Thanks Jim - was unaware of that setup.

Did the gages come with any kind of R&R statement or accuracy statement, either in the packaging or in the literature?


Microprocessor-controlled stepper motor gauges (google it) and my contact at Isspro told me they are accurate to .5-1%.


That accuracy is better than what you can read with your eyeball on a small analog gauge with 5 PSI hash marks.

I'd guess one needle width is about 1 PSI on those gauges.

 

[Jim Allen]

Originally Posted By: ZeeOSix

You would have to physically swap out the two gauges' location and keep the signal setup the same on the data acquisition system. Doubt you want to go through that trouble.


And you would be correct... especially given the gauges are accurate to .5-1%. Even I am not that anal.

 

[Jim Allen]

Incidentally, the datalog reads in 10ths.

 

[ZeeOSix]

Originally Posted By: Jim Allen
Incidentally, the datalog reads in 10ths.

... nice.

 

[jegs]

Originally Posted By: Jim Allen
Originally Posted By: jegs
How does port 1 get pressure from the clean side of the filter? Isn't the threaded steel part that the filter screws onto (clean side) a solid tube?


There is a gallery around the center bolt with a passage leading to port one. The center bolt is o-ringed and also has a port in it. When used as designed in a parallel flow bypass filter situation, my port 1 is the return from the filter.

Thanks for the reply. I tried to find some pictures online of the center section but didn't find anything. A lot of these "sandwich" adapters only give access to the dirty side of the filter and have a solid center tube. They are advertised for oil temperature/pressure gauges,turbo oil feed ports etc. I have never seen one like the one you are using. Very interesting.

Thank you for sharing your findings with this test. Knowing when,how long and under what conditions a filter goes into bypass has pretty much been a mystery. This will provide some great information on that. Thanks again.

 

[dnewton3]

Given it's a garage experiement, and not a lab study, I'd say the accuracty of the instruments is moderately sufficient, if their statements are true (1% max).

I am currious if that is 1% total deviation, or +/- 1%? I have looked over their website and not seen any specific claims for accuracy; I have emailed them for details.

It does matter, to some degree, even in your garage though ...
Consider it if was +/- 1%; that would be a range of 2%. And don't forget that you have two instruments for each system (an oil sending unit and a gage unit). Now consider that you actually have two gage sets (two sending units with two gages). If system #1 was at the total low end (at -2%) and system two was at the high end (+2%), then you would have a total spread of 4%! At 60psi, that would equate to 2.4 psi. Simply put, while "1%" error seems like a nice low number, it compounds with ferver as you stack instruments into systems. See how things pile up so quickly? In fact, if you wanted to REALLY get picky, we'd have to include the resistance variation of the wires themselves.

Typically, to have faith in a gage system for lab and statistical purposes, you want the R&R to be no greater than perhaps 20% of the least measurable magnitude. If the system reads in tenth's (0.x psi), then we should expect the R&R to be process capable at 20% of that one-tenth (or in effect, that R&R would need to be process capable down to .02psi!). See how ugly that "1%" error statment seems now! What is possible, is that their public statment would be that they are accurate to 1 psi, and they meausure to the tenth as their R&R factor assurance. If that were the case, then we really should not read any display number in tenths at all, and only look at whole numbers. We should truncate the data one magnitude above their R&R level; if they read in tenths, we should only take data at the whole. And if there is no markings on the gage at that 1 psi level, then it makes it VERY hard to delineate data.

I'm not picking on you; please understand that. In fact, I'm excited to see the results! But let's just keep a clear head about this before we jump to any conclusions. And let's stay pragmatic with the results; we should not profess success or failure until we understand all the limitations of the inputs and the equipment. Let's get the DOE right and assure how much credibility the data system has!



Like I said; I've emailed them. I'll hope to get better details on the gages and sending units soon.

 

[Jim Allen]

I know and can directly call the engineer at Isspro who designed the system to get whatever accuracy information they will give me. He's the one who sent me all the equipment to set this up and that's where I got the .5-1% figure in the first place, though I only wrote down the numbers and didn't get any other details. Whatever accuracy the gauges are... they are. Unless you or someone else is willing to cough up the cash for true lab tests, this is what we're going to get, +/- whatever.

In this genre, stepper gauges are as accurate as you can get. Ordinary gauges are known to be in the range of no better than 5% and often not even 10% accurate according to what I have read. I see some stepper gauge manufacturers advertise "2% accuracy," others ".5% without any other context. Don't know if Isspro is at the high or low end of of "industry standard" for the stepper types.

In any case, this was never going to be anything more than a "garage test," mainly to determine how often the filter reaches the DP to potentially go into bypass. Secondarily, it's interesting to me to observe the effect of time and miles on DP and later to test a few filters and get a snapshot of their starting DP in relation to each other, as measured on the same equipment. I'll share what I get with my friends at BITOG, who are free to discount it as they choose.

 

[ZeeOSix]

Originally Posted By: dnewton3
In fact, if you wanted to REALLY get picky, we'd have to include the resistance variation of the wires themselves.


It's probably a very high impedance system, so the length of the signal wires is a non-issue.

On a side note. Only way to know if both gauges are reading the same output with the same input would be to swap them and see if there is variance between them. Not looking for an absolute readout, but a consistent readout between them since the difference is what is the value everyone is interested in.

 

[Jim Allen]

Originally Posted By: ZeeOSix
Originally Posted By: dnewton3
In fact, if you wanted to REALLY get picky, we'd have to include the resistance variation of the wires themselves.


It's probably a very high impedance system, so the length of the signal wires is a non-issue.

On a side note. Only way to know if both gauges are reading the same output with the same input would be to swap them and see if there is variance between them. Not looking for an absolute readout, but a consistent readout between them since the difference is what is the value everyone is interested in.


I did swap the gauges but quickly realized the futility. The gauges each have a microprocessor coded to read only the input from a certain pin on the ESP (the ECU/signal processor). Isspro could reprogram the two gauges to read from different pins but I'd have to send the gauges back. I have a call in to my engineering contact there so I'll have a feel for accuracy fairly soon, as well as an answer to my datalogging issue.

I think I could swap the connectors at the senders, though, to see if the gauges reverse exactly their readings. I have enough wire to do that. Seems to me that would answer the questions, right?

 

[ZeeOSix]

Originally Posted By: Jim Allen
I think I could swap the connectors at the senders, though, to see if the gauges reverse exactly their readings. I have enough wire to do that. Seems to me that would answer the questions, right?


That sounds like it would work as long as there isn't something special about each sensor (ie, more unique "coding" for each sensor, etc). Something to ask the guru at Isspro.

 

[Coprolite]

What is the approximate MSRP for this kind of setup?

 

[Jim Allen]

Originally Posted By: Coprolite
What is the approximate MSRP for this kind of setup?


About $1500, considering the Isspro Performax gauge system and filter base. If you found individual stepper gauges, you could make it a lot more cheaply.

 

[Jim Allen]

Dave, Zee: I got the error info on the senders and gauges. Dave, I asked my contact to look for your email (he found it) and answer it personally but basically the senders are accurate to 1.5% @ 100psi =/- and the gauges are the same. Total stackup error for the sender and the gauge is tested at a maximum of 2% +/-.

This afternoon, I started the truck and recorded the pressures in a specific situation, then a switched the sender plugs and the values switched almost exactly (like within .2 to .3) so I'm not too worried about them now.

I also got the answer to my datalogging problem... I had version 1.0.0.0 of the software ( : < ).

 

[ZeeOSix]

Originally Posted By: Jim Allen
This afternoon, I started the truck and recorded the pressures in a specific situation, then a switched the sender plugs and the values switched almost exactly (like within .2 to .3) so I'm not too worried about them now.


Sounds good for pressure difference measurements, which is what you're after the most it sounds like (ie, PSID across the filter). Let the testing begin!

 

[dnewton3]

Originally Posted By: Jim Allen
Dave, Zee: I got the error info on the senders and gauges. Dave, I asked my contact to look for your email (he found it) and answer it personally but basically the senders are accurate to 1.5% @ 100psi =/- and the gauges are the same. Total stackup error for the sender and the gauge is tested at a maximum of 2% +/-.

This afternoon, I started the truck and recorded the pressures in a specific situation, then a switched the sender plugs and the values switched almost exactly (like within .2 to .3) so I'm not too worried about them now.

I also got the answer to my datalogging problem... I had version 1.0.0.0 of the software ( : < ).



Given that little check, I'd say your GTG for the garage experiment! Going to be close enough for what you're after. Very excited to see the results of your experiment!


I'll look for your contact's email response; still interested in the details, though they likely won't effect your results.