Compression test results - Advice?

Anonymous · January 16, 2011, 20:23

Argument over. If Rogue said it, thats it. s:lol: :lol: s:lol:

Anonymous · January 16, 2011, 20:34

I'm glad this thread has come up though,healthy debate is no bad thing and a few good bits of info has come from this. s:)

s:)

mrzwei · January 16, 2011, 20:47

Quote from: "filcee"
Quote from: "muffdan"Is it possible to over-read through bad procedure?

Warning: I am not an engineer

Maybe not, but I bet you're a mathematicion s:lol: :lol: s:lol:

K T M Rider · January 16, 2011, 21:27

Quote from: "filcee"Warning: I am not an engineer, so make of this what you will ...

I suspect that getting the process wrong will cause readings to be 'different' to those that might be expected. However, Toyota state 184psi +- 15psi. Someone else has quoted a range of 169 - 199 psi as being 'within tolerance'. Now, if the kit itself is not calibrated, it might be reasonably be expected to have some error factor 'built in'. Let' say 10%, just to keep the numbers easy. This gives a range of approx. 153 - 219 psi - and now we're starting to see numbers at the "extremes" of those quoted on here. If a "bad procedure" is introduced, this might make further variation to these numbers - and we'll get even more extreme figures at the boundaries.

Would we be better, with uncalibrated equipment/amateur approach (apologies to those who are more knowledgeable and experienced than me) to look at the percentage variation between cylinders as read by the current compression test device: 184 psi +- 15psi is about an 8% variation, anything outside of this should be viewed as 'suspect'. Any truly wacky numbers - lets say outside of the +-15psi range with an added (or subtracted) 10% in view of using uncalibrated equipment should be a strong indicator of the need to get things checked by someone you trust using equipment that can be considered of good quality and well calibrated.

I am IMHO no longer an engineer (though I do have a couple engineering qualifications s:)

s:) )

Your approach sounds perfectly logical. It should be noted however that the link:

m http://madstyle1972.com/Repair/14/200f0c04/i740001.pdf m

Indicates a desired pressure of 189psi / a minimum pressure of 145psi & a desired difference of no more than 15psi. So (assuming a device in perfect calibration) 4 cylinders all reading 145psi would be in spec, whereas 3 cylinders reading 189psi and one reading 173psi would not, even though it would probably indicate a healthier engine s:? :? s:?

So to take a hypothetical example of an uncalibrated device that under - reads by 10%.

A reading from this device of 131psi would still be in spec.

Assuming uncalibrated equipment, then as Mrzwei has already stated, the variation between cylinders would generally be more significant than the actual readings & the best way to compensate for using a cheap gauge is to take 3 sets of readings and average them. If this approach doesn't give readings (and in - cylinder variation) in the right ballpark, that is when it could be prudent to hand it over to a Pro.

K T M Rider · January 17, 2011, 00:34

Quotethe best way to compensate for using a cheap gauge is to take 3 sets of readings and average them.

Thinking about it a little more - an approach sometimes used when gathering engineering data (particularly if the procedure or equipment used are slightly suspect) is to 'drop the highest and lowest and average the rest'.

The idea behind this method is to eliminate any 'outliers' that are due to measurement error. So, FIVE pressure readings for each cylinder would be taken but only the middle 3 readings averaged, the highest and lowest reading are ignored.

Can't imagine anyone having the patience to do this on a V12 though s:lol: :lol: s:lol:

aaronjb · January 17, 2011, 10:17

Quote from: "KTM_RIDER"The idea behind this method is to eliminate any 'outliers' that are due to measurement error. So, FIVE pressure readings for each cylinder would be taken but only the middle 3 readings averaged, the highest and lowest reading are ignored.

That's a good idea in theory, but I bet after two sets of readings you'd start to see them drop significantly as the battery got tired - unless you had an external jumper pack or another car there to keep the battery charged s:)

s:)

K T M Rider · January 17, 2011, 13:05

Quote from: "aaronjb"
Quote from: "KTM_RIDER"The idea behind this method is to eliminate any 'outliers' that are due to measurement error. So, FIVE pressure readings for each cylinder would be taken but only the middle 3 readings averaged, the highest and lowest reading are ignored.

That's a good idea in theory, but I bet after two sets of readings you'd start to see them drop significantly as the battery got tired - unless you had an external jumper pack or another car there to keep the battery charged s:) s:)

don't know, 'cos I don't own a compression tester s:)

s:)

the following thread appears to be written by someone much more knowledgeable than me in the art and advises taking 3 or 4 readings per cylinder (with a warm engine and a fully charged battery)

m http://www.vauxhallownersnetwork.co.uk/ ... rol-engine m

Anonymous · January 17, 2011, 16:06

wow this thread is mad lol. surely as long as all cyclinders are around the same then its all good. and there is also to many variables to take into account to say you should have this figure. every engine is different rings wear differently depending on how its driven. but hey thats just my 2p's worth.

richard

Cap · January 18, 2011, 03:01

Just got done doing a 'High Compression' Model T engine ( 1926 ) .. Mostly Stock.. Except for the Hi Compression Head.. This raises the Hp from 20Hp to 22Hp.. on a 'Splash and Pray' oil System..

Not knowing what the Compression Pressure should be, but knowing some Model T have recorded 45 to 50 PSI, all we have to go on is the Pressure Variance.. we ranged from 87 to 92 PSI.. So all is good on this build.. Now we will record the pressures from time to time and get a history for this engine..

K T M Rider · January 19, 2011, 16:31

Quote from: "rbuckingham"surely as long as all cyclinders are around the same then its all good.

richard

I agree.

But 'around' isn't much of a spec.

and what about if they aren't?

Anonymous · January 20, 2011, 11:14

i see your point you dont want stupid low figures on all cyclinders. if they are out by more than 15 - 20 psi then the low one has a compression prob and that could be down to loads of things.

Anonymous · January 24, 2011, 23:14

Quote from: "life of bryan"I bought a cheap one like this m http://cgi.ebay.co.uk/AUTOMOTIVE-COMPRE ... 3ef937829b m but I havnt tried it yet,I will let you know what I think when I do.

This type of tester in the red case is completely pants,I couldnt use anything with the kit to fit it to the the spark plug holes. s:evil: :evil: s:evil: Might try one of these Draper ones. m http://cgi.ebay.co.uk/Draper-Compressio ... _500wt_922 m

m http://cgi.ebay.co.uk/ws/eBayISAPI.dll? ... 2632wt_905 m

nathanMR2 · January 24, 2011, 23:16

Let us know how you get on with this new one mate s:? :? s:?

mrzwei · January 24, 2011, 23:22

I think you're right there. I've got an old 'rubber bung' type guage which was great before ohc engines when plugs stuck out of the head. That one looks to have the reach and I think there are only two types of thread on modern plugs and that seems to cover them and will have the reach.

michaelasaunders · February 15, 2011, 23:16

So I have now done 600 miles after my first compression check

My first results were.
180 150 180 180

My new check came up with
180 160 180 180

The difference in the car has been running with a new oil (Mobil Helix 10w40) and using Shell V power fuel.

Could these two changes make a difference or this is just down to test variation?

I would of expected greater compression on all cylinders with V power.

Edit - My local garage did both checks, not me.

markiii · February 15, 2011, 23:29

Why woul you expect vpower to affect your compression?

ChrisGB · February 15, 2011, 23:57

Quote from: "markiii"Why woul you expect vpower to affect your compression?

Actually it can! If your cylinder head and piston crown have substantial deposits on them, running good super unleaded like VPower can take off the deposits leaving you with a slightly lower compression measurement.

I still think the 180s are ok, but the bore running 150 ish is suspect.

Chris

mrzwei · February 16, 2011, 00:08

My take on that would be that there may be a slight (very) issue with No.2. If the oil used was thicker when cold than the original then that could (I say could again) account for the better compression. Your readings are well within factory tolerance so don't start ripping the engine apart. If you are using a lot of oil then that basically means you have to check it more often and buy more.
If you're not using any oil, then don't worry.

michaelasaunders · February 16, 2011, 12:57

Quote from: "markiii"Why woul you expect vpower to affect your compression?

Maybe my fundamentals are a bit iffy here.

My assumption was that a higher octane fuel gives more power, so more compression.

I guess not, as according to wikipedia the higher octane fuel is more useful in a high compression car as the fuel
will detonate at a higher compression.
[urlhttp://en.wikipedia.org/wiki/Octane_rating][/url]

So my thinking is that the 180`s are fine and the 160 is just out of spec as it was +- 15 for normal variation.

My question was, I was wondering if the change from 150 to 160 (Gain of 10psi) could be the fuel or oil or test variation?

It seems like the fuel is not the one, as Chris says the Vpower cleaning abilities will lower the reading, which is not what has happened. So maybe the oil, or its just test variation.

Wabbitkilla · February 16, 2011, 13:21

It's possible that the oil grade and warmth in the engine have brought about the change in results.
The V-Power doesn't make a difference, but it won't do any harm with its extra cleaning ingredients possibly cleaning up the rings and valves therefore allowing them to seal better.