19 replies to this topic

[zero_wlv]

Compression test - weak starter or leaky rings?

 

Hi all,

 

I've just done a compression test on a rebuilt engine that has done 650 miles so far.

(998cc + 0.060",  10.6:1 theoretical compression,  Stage 2 head,  MG Metro clone cam, Hif38).

 

I did the compression test because I was worried about apparently high oil consumption (0.5 litre / 150 miles or thereabouts) but I think I may have found the cause of that this morning  -  a slight leak from the back of the rocker cover gasket which was leaving a glistening film over part of the exhaust manifold.  Hopefully this is sufficient to explain the oil loss -  I'll know after my next 100 mile drive at the weekend :)

 

I am however,  still worried about the compression test results at 650 miles (warm engine,  full throttle etc.):

 

Results with my gauge (dry):     1:125  2:125  3:125  4:125

Results with my gauge (wet):    1:133  2:-    3:-    4:-

Results at local garage (dry):    1:120  2:122  3:128  4:124

 

As a comparison,  these are the results the engine builder got on the bench test at 0 miles:

 

Engine builder:                 1:190  2:185  3:190  4:185

 

The one difference is that the engine builder would have used their own Powerlite starter motor,  whereas I'm using the 30 year old factory original,  which does struggle to start the engine (maybe because of the 10.6:1 compression?) and needs quite a lot of cranking.

 

Could a weak starter be sufficient to explain low compression test results?   I've seen this suggested elsewhere but I really can't see how  -  thinking about it physically,  as long as the starter CAN crank the engine through complete turns,  surely the results depend on the engine and not on the speed or torque of the starter?

 

Obviously the alternative that I feared is that I might have bore-washed the engine on run-in,  BUT I've just had the local garage run an emissions test and the results are 0.58% CO and 268 HC.   

Does this indicate for certain that the engine is safely on the lean side  (as recommended by several threads on this forum for run-in)?    Also,  there is no smoke at all out the tail pipe  (just a hint of white which is presumably vapour)  -  surely if 0.5 litres of oil was blowing past the rings the exhaust would be very blue?   Would you also expect the HC to be much higher if oil was blowing past the rings?

 

Apologies for all the questions in one post,  but I'm just trying to convince myself that the engine is in a good enough condition to go for a rolling road setup in a couple of weeks time,   as it probably wouldn't be a good idea to thrash it on a RR if there's a fundamental problem to be sorted first.

Edited by zero_wlv, 22 January 2018 - 08:02 PM.

[nicklouse]

how did you do your compression test?

 

the best way is. plugs out. throttle open.

do test left to right and then repeat right to left.

 

this covers any battery issue. as they may be lower but the ratios will be the "same".

[zero_wlv]

All plugs out,   throttle wide open (as in pedal flat to the floor).

I did the test once (cylinder 1 to cylinder 4).

 

I'm think the battery's ok with a healthy voltage (12.55V at rest when I checked a few weeks back),  although as I said the car rarely starts easily,  even when warm.

Could a poor battery be the cause of low compression results?    I still can't visualise how this could be a problem as long as it's capable of cranking the engine.    If the engine's turning,  then the pistons are moving in their normal range of motion and the engine's completing its full cycle,   so I don't see how the compression results could be affected.   It would just take a little longer for the gauge to get up to its final reading.

 

Of course,  if it might be a battery problem that'd be fantastic,  as it's an easy fix :)

[zero_wlv]

Thanks for the suggestion Nicklouse,  I'll try doing another test later this week with the battery hooked up to my portable jump-starter.   Will be interesting to see if there's any difference.

[GraemeC]

With all the plugs out the starter is only having to overcome compression in one cylinder, so unlikely to cause it an issue

[mikal]

My engine is 11:1 compression ratio and the original Lucas M35 has no problems turning it over. Starting is very easy, normally starts 1st time. I did overhaul it a few years ago. Parts are easy to get and are cheap...

[ACDodd]

Cylinder leakage test will show up where the problem lies, my gut instinct says valves.

 

AC

[Spider]

Firstly, it takes between 2000 and 5000 miles before the rings are fully seated / run in. From around the 500 mile mark, you can drive it as you will as most of the running in is done.

 

The results between cylinders are similar and IMO look good.

 

In my books a compression test is really only good for comparison between cylinders but I wouldn't draw too much on the numbers produced.

[Homersimpson]

Can you borrow another gauge to check that yours is accurate?  As above the values are good across the cylinders so it seems that if there is a problem then its on all of them.

 

One thing, have you checked your valve clearances in case they are being held open now the engine is run in?  I doubt its this but worth checking.

[zero_wlv]

Hi Homer,

 

I was hoping my gauge might be inaccurate,  but I got the local garage to do a compression test as well and the results were the same as mine.

 

I did check the valve clearances at 400 miles (only 1 of the 8 needed adjusting) but I'll be checking them again very soon  (and the head torques at the same time).

 

As soon as the weekend comes I'll be trying again but with a jumpstarter pack hooked up to see if a fresher battery does make any difference.

Edited by zero_wlv, 24 January 2018 - 08:43 AM.

[Northernpower]

I'm with Moke Spider on this. I don't think you have enough miles on it to be fully run in. The fact all the cylinders are giving a similar reading would suggest there's not a problem, especially given your local garage had similar readings. I don't think it will be valve clearances given all the readings are similar, (unless you've set all the valves wrong to exactly the same amount!).

 

You don't mention how the performance is. If you'd said you did a compression test because you felt it was down on power then I would be more concerned, high oil consumption while not being welcome, and doesn't occur with every build, is not uncommon in the first 1,500/2,000 miles. A leaking gasket letting that much oil out would either leave a big mess on the block/floor, or, if it was being burnt off would make for a very uncomfortable smell in the cabin. Have you done a cylinder leakage test?

 

I think you may be trying to fix a problem that doesn't exist.

[ACDodd]

Either way a cylinder leakage test would show where, if any, problem lies.

Ac

[zero_wlv]

Thanks everyone,

 

I haven't tried a cylinder leakage test yet   (I assume this is the same as a "leakdown test",   where compressed air is used?)

I'll see if I can get a garage to do this for me as soon as I can.

 

It did occur to me that the rocker cover leak didn't look quite bad enough to be losing 0.5litres / 150miles,  but I'm still hoping that's the explanation   (haven't driven anywhere this week so I'm not sure yet).

 

As for Northernpower's question about performance,     the performance is fine up until 40mph on the flat or 30mph up hills.     At this point the engine starts to lose power quite suddenly and I can only keep going by 'feathering' on and off the throttle,  which makes it pick up again.     I haven't been worrying too much about this as I've assumed it's just a tuning issue that will be fixed by the rolling road.    I have on occasions tried pulling the choke halfway for a few seconds up steep hills and it does seem to help,   so I would think it's just running lean.       I've also considered the possibility of fuel starvation but I can't see any reason for this to happen.

 

It's also occurred to me that now I've covered 650 miles it's time I richened it up slightly so it's roughly stoichiometric (neither rich nor lean)  -   I've avoided doing this so far due to my general paranoia about borewash,  but after 650 miles maybe I should relax a bit about it.

 

 

p.s. when I say stoichiometric,  I don't have anything fancy like an AFR gauge,   I just mean adjusting the mixture as per Haynes to get not too rich or not too lean.

Edited by zero_wlv, 24 January 2018 - 01:18 PM.

[Spider]

 high oil consumption while not being welcome, and doesn't occur with every build, is not uncommon in the first 1,500/2,000 miles. A leaking gasket letting that much oil out would either leave a big mess on the block/floor, or, if it was being burnt off would make for a very uncomfortable smell in the cabin.

 

I think you may be trying to fix a problem that doesn't exist.

 

Yes, good point Northernpower. Early on you can expect some Oil consumption - especially if it has 2 piece oil rings - until it's gets some miles on it. It needs to be run in to seal.

 

In regards to a leak down test, sorry to disagree here, but IMO they are a waste of time. They are a static test at the bottom of the bore (unless you lock the crank at a particular point in the stroke). In higher milage engines, it's evident that most of any wear is at the very top of the bore and this is where the most sealing is needed (though should be the same the whole bore length). Also, being a static test, the rings and not going to behave as they would when running. Sorry, but I find it a very misleading test (and I don't rate a compression test much above this one).

 

A leak down test, though, can be a little helpful in pointing to the sealing of the Valves.

[Northernpower]

Just going back to the oil consumption, have checked it on the overrun?