The BSA A7-A10 Forum

Technical => A7 & A10 Engine => Topic started by: farmboyuk on 13.12. 2007 12:03

Title: Measuring compression
Post by: farmboyuk on 13.12. 2007 12:03
This may seem a really dumb question but I'm going to measure the cold compression of my 1952 A10 GF and I thought I would just need to have a idea of the psi/bar expected to read. All I can find in the instruction manual is the ratio of 7.25:1. Can someone please explain, what these numbers represent (is 7.25 the pressure itself?) and what should I do to make sure that the compression is correct. I've bought a compression gauge and just want to know what to expect to read.
Sorry for my ignorance, I've never done this before.  *conf*
Thanks for your time.
Title: Re: Measuring compression
Post by: groily on 13.12. 2007 14:04
hi, the 7.25 just means that the air/fuel mix is in a volume 7.25 times smaller when the plug fires it with the piston near enough at the top of its stroke, compared to the volume of the cylinder and combustion chamber when the piston is at the bottom of its stroke - ie the 'ratio' of the total swept volume to the volume the mixture is in when it goes bang, is 7.25 to 1. A compression test will give you some reading on a gizmo that looks like a beefy tyre pressure gauge screwed into the plug 'ole, and it'll be in pounds per square inch or in some unchristian metric thing. A decent reading would probably be 125-150 pounds per square inch plus . . . anything under 100 is probably not good. I'll bet someone here can give the exact answer. However, if the engine feels right, there's good 'compression' against the kickstart, and it runs right, it probably is right or right enough. If the compression is really bad, then it probably would smoke quite a bit from the exhaust, unless it was a valve giving you the trouble, in which case you might here the gases squeaking past a valve that wasn't closing or seating well when you turned the motor over slowly on the k/s . . .good luck - Groily
Title: Re: Measuring compression
Post by: farmboyuk on 13.12. 2007 15:27
Excellent, thats just the information I was after. Thanks very much for that.
Title: Re: Measuring compression
Post by: LJ. on 13.12. 2007 17:19

Not such a dumb question Ed... I have never been too certain of these measurements myself. I suppose its just a question of plucking up the courage to ask in front of the experts.

Any how... I remember reading about a sort of compression testing gadget on some BSA M20 guys website and he has made up some sort of clever device called a leakdown tester, I wonder if it could be used on twin cylinders? Take a look you'll find this interesting... (
Title: Re: Measuring compression
Post by: dpaddock on 13.12. 2007 20:33
The cylinder pressure reading will be the compression ratio times the prevailing atmospheric pressure, adjusted for incomplete cylinder filling. For example, 7.25 times 14.7 psia yields 104 psig times, say, 95 percent to allow for air filter restriction = 99 psi gage reading. In this case, any reading between 95 and 100 psig indicates good conditions.  90 to 95 psig would suggest an acceptably worn engine, whilst a reading below below 90 psig indicates too much leakage. But keep in mind it's not the pressure reading that's important - it's the variation between cylinders, which shouldn't be more than 10 percent. Obviously, the compression tester is of little use for a single.
The leakdown test, referenced in L J's reply, is a much better test. Testers of this sort are available in many automotive parts stores.
Title: Re: Measuring compression
Post by: groily on 13.12. 2007 21:17
knew someone would have an exact answer! and dead right about variation between cylinders too, that's something you don't want . . . groily
Title: Re: Measuring compression
Post by: RichardL on 15.12. 2007 16:20
OK, I am reposting this discussion. The first time I posted it there was some bad theory, which I will explain here. Regardless, in the mean time, dpaddock has provided a very good explanation with the obvious benefit of a lot of experience.

Standard atmospheric air pressure is 14.7 psi (appearing now, all over your dermis).  Compressing atmosheiric pressure by a ratio of 7.25:1 renders 106.6 psi (a minor correction to dpaddock's number). So this was all well and good and makes sense on the gauge. Regardless, I wanted to see if there was anything I was missing in the compression pressure equation. Looking in Wikipedia, I found that the equation for calculating pressure includes an exponent to be applied to the compression ratio. The exponent represents the amount of additional pressure due to adiabatic heating of air (as in the heat that is generated to make diesel engines fire).
Solving the equation, 14.7 x (7.25 exp 1.4) psi = 235.4 psi. The exponent 1.4 being needed to allow for the specific heat of air. (I know I learned about all of this in Thermodynamics, but a lot has been forgotten).  Wow, 235 psi! That was number I wasn't prepared for and which would never appear on a cyclinder pressure gauge. This was the point at which I removed the previous post. However, I believe I have figured out what the story is, but it is just my best guess. I think that the adiabatic heating process using the formula I've shown assumes that the change of pressure occurs instantaneously, or almost so. Kicking over the A10 is quite slow by comparison, so, it seems, there is not enough adiabatic heating to measurably change the pressure. So, the adiabatic heating does not mean much to us when measuring cyclinder pressure in a engine that is not running, but I think it is interesting because it probabaly does affect pressure in a running engine, particularly at high RPM. There would be quite a bit of power difference between 100 psi and 200 psi. I suppose this could be experimented with in a running car engine, but I haven't heard of it before.

I know this is tedious and long and seems like I am trying to teach physics, but I am just an audio engineer, as the name implies. Now, back to some other points from my original post. (dpaddock has explained this well from a different approach.) The 106.6 psi would be the gauge reading with absolutely no leakage. I would expect the reading to be slightly lower than this, even in a professionally built engine, as a little blow-by at the rings is inevitable. Also, the 7.25:1 ratio is probably not exact, as BSA wouldn't want to specify a compression ratio of something like 7.2342:1 (just a number I picked at random) if that was the actual, they would round it upwards. Another "also," air pressure where you are may not be, and probably isn't, exactly 14.7 psi, as it changes with weather and altitude. Yet another "also," 7.25:1 pistons would yield a different psi in over-bored barrels versus standard. You could go to a lot of trouble getting your actual atmospheric pressure and using that, but it probably is not worth it. 

Now, more tidbits:

I usually kick over four or five times just to be sure I have a good reading. Hopefully, your gauge is of the type that holds the peak pressure reading. It does not continue to get greater with each kick after having reached its realistic reading. Why more than one kick is needed, I am not quite sure, it could be that it needs to get an even smear of oil on the cylinder walls.

If you get a low reading, you won't quite know the exact cause without further testing: If it's the rings, a shot of oil in the barrel will seal them better and, likely, lead to a higher reading. If no higher reading, then its the gasket or the valves. From here, I haven't done the tests to know what diagnostic approach to use. Maybe someone else can chime in explaining further.

Title: Re: Measuring compression
Post by: groily on 15.12. 2007 18:29
blimey, hard to add to that one Richard .  . metaphorical hat metaphorically doffed! Dunno what made me think I'd be looking for north of 100 psi (higher comp motors maybe) but must admit, I've never tested compression in anything (tho I did buy my old dad a tool for measuring it once when i couldn't think of anything else to give him for Crimble). But seriously, it's usually been only too evident from the pile of broken bits/horrible noise/failure to run/smoke everywhere that summat's amiss and spanners are needed asap! Groily

PS - brass monkeys or what today! First 40 miles were OK (new gloves, bliss), but thereafter it was 'where are my feet, what happened to my fingers . . .  and what sex am I'? Nothing a few glasses of the old nectar can't cure, luckily, although it'll be her as is the judge I s'pose.
Title: Re: Measuring compression
Post by: RichardL on 15.12. 2007 19:17

125 psi seems like about the right number for 9:1 pistons, taking into account dpaddock's 95% factor.

Title: Re: Measuring compression
Post by: groily on 15.12. 2007 20:32
yep, you're right Richard - came to the same broad conclusion. Think the 125 + number was stuck in my head from years ago and some car workshop manual. Wonder what beast I ever owned with such oomph . . . Groily
Title: Re: Measuring compression
Post by: a10gf on 15.12. 2007 21:07
First 40 miles were OK (new gloves, bliss), but thereafter it was 'where are my feet, what happened to my fingers . . .  and what sex am I'?

And what a great thread about compression. Must do a test tonight!

Title: Re: Measuring compression
Post by: a10gf on 15.12. 2007 23:00
...the results: 650 iron head (7,5 to 1), maybe 3000 km since rebore and new pistons+rings and new valves\seats: approx 115 psi left, 120 right. 10 kicks or so.

And seems to be reasonably airtight, as there still was pressure in there when the release knob on the instrument was pushed after approx 30 sec.

Title: Re: Measuring compression
Post by: RichardL on 16.12. 2007 05:27

Looking in Bacon's Twins and Triples book, which seems to have a pretty good history of stock bikes, I don't see where 7.5:1 was an option. Perhaps an aftermarket source offered such pistons. I am curious about the profile of your pistons. If yours are, indeed 7.25 or 7.5:1, then I think there is a ceratin amount of adiabatic heating is taking place to reach the 115-120 psi marks. Otherwise, to truly torture this, perhaps you had a high barometer reading at the time. Maybe the reason that cyclinder pressure is not specified has to do with ambiguity in adiabatic heating and local atmosheric pressure.

To those who think has little to do with a fine working engine and enjoying the ride, I agree completely. It's just a science discussion to fill the time for want of good weather or a completed rebuild.

Title: Re: Measuring compression
Post by: a10gf on 16.12. 2007 11:49
For reference, I had a look in my xt350 book, 9:1, 350cc single:
min. 128psi, std: 156psi, max 171psi. Would think they provide these specs as an important tool to figure out the condition of the engine.

Using the 14.7 theory, x9 should then be 132.3, but that would be very close to min. Taking the standard 156 psi \ 9 comp = 17.33 psi to start with...  ;)

Interesting to see they quote 3 values: not good, good, too good ... maybe the acceptable spread is because of the possible varying conditions mentionned in the post above, combined with parts differences and wear ? ( btw, how can it ever be too good, with standard configuration and parts ? )

The 7.5 pistons are from SRM, is what they recommended as standard replacement.


Title: Re: Measuring compression
Post by: RichardL on 16.12. 2007 14:20

Yes, I am now officially obsessed with this topic, just to satisfy my engineering curiosity. Any time this gets boring, feel free ignore me.

If the Yamaha had lossless adiabatic heat, the pressure would be 318 psi. The following link gives a good description for those interested in a modicum of math:

We can be pretty certain atmospheric was not 17.33 to start, as the world record is 15.7. 

I expected to see a little more up-push on the top of your pistons. Yours look like they are flat and do not protrude above the top of the barrel at TDC. If that's true, are 6.5:1 depressed?

Another thing I want to understand better is the fact that the compression stroke is actually shorter than specified due to the intake valve closing 62 degrees ABDC (from '53 to '61). This amounts to a reduction in lenth of the compression cycle by 4.92 mm. I am very curious to find out if compression ratios are considered with this in mind. Clearly, named displacement does not consider this, or we (A10 riders) would all be riding "610"s instead of "650"s.

That's it for now, but don't be alarmed if I hit it again.


Title: Re: Measuring compression
Post by: a10gf on 16.12. 2007 15:03
The pistons I got are 7.5 according to srm, I do not have the old 6.5 to compare with, and you are right, they look like 6.5.

You got me hooked  *conf*

Keeping with the xt ref. numbers, the "std" = (14.7 x 9), + 15% to get to 156psi.
Applying this einsteinian theory, the A 6.5 "std" would be (14.7 x 6.5) +15% = aprrox 112psi. A 7.5 would be around 130psi. And does the cc, air humidity and intensity of kicking interfere in any way *doh*  *smile*

Back to practical reality, I am happy with the reading above 110 (supposing my gauge is showing passably correct values), and beeing quite even between cylinders. If anybody else could take a test on their engine, would be interesting to compare, and maybe come up with what may be some normal average. As the xt350 numbers suggests, there is quite a lot of room.
Title: Re: Measuring compression
Post by: bsa-bill on 16.12. 2007 16:30
I think I could keep up with this if I really really tried, however I do find it interesting and get the general grist of it.
So it gives me great pleasure    ;)   to be able to contribute a smidgin - 6.5 piston are indeed dished, I have a pair in the shed +60 if anybody is looking for some

All the best - Bill
Title: Re: Measuring compression
Post by: a10gf on 16.12. 2007 18:03
Couldn't let it go. Did a test on the xt350, around 145psi. Perfect according to specs would be 156psi. 9-1 ratio.

If the cc etc has nothing to say, this all makes sense. By adding 10 to 15% (the adiabatic heating factor ?) to the basic atm 14.7 x compression ratio, this should result in a good psi value, indicating a healthy engine.

And an atm x ratio with smaller or without any added %'s would indicate an engine on the decline  *ex* +  ????

Title: Re: Measuring compression
Post by: snowbeard on 12.09. 2008 05:10
I just picked up a compression tester at an estate sale, so I have to admit it could be off.  but the fellow was an mechanical engineer for NCAR here in Boulder, and had soldered the free spinning end of the hose even, so I assume he knew what he was shooting for.  still in the original package, tho well used.

I also don't know what my compression ratio is supposed to be, or what it actually is.  I have a 57 spitfire scrambler by the factory records for engine and frame numbers, but it could have had lord knows what done to it. I've thought a lot of things were in really bad shape, but they turned out to only need minor attention for the most part.  I think that someone who ran this bike really knew what they were doing.  (that wouldn't be me)

but I tested the old gal tonite, the altitude is 5300 or so, and it was raining pretty hard on a new storm just started.  what that might mean to the atmospheric pressure is beyond me?

but anyway, it tested at 120 psi.  took four to five good kicks to get it up there, and the engine was cold, but both sides were spot on at 120/119.

I have to say I was pleased after having been told of the smoke on my right cylinder recently.
Title: Re: Measuring compression
Post by: tombeau on 12.09. 2008 08:54
 I tested mine recently and got a reading of 170psi on both cylinders.
Initially id id it with the throttle shut getting a reading of 130 but was told I should have the throttle open whilst testing compression.
Title: Re: Measuring compression
Post by: snowbeard on 15.09. 2008 05:23
what are you running Tombeau?  any mods or stock A10?

btw, thanks for the posts helping me decide about using the lysette style seat, I am fond of it for now!
Title: Re: Measuring compression
Post by: tombeau on 15.09. 2008 07:42
Hi Snowbeard!
No problem for the help on the lycette seat.Have you got any pictures with it fitted?
My motor is a Road Rocket, its fitted with a Spitfire cam, but last rebore I couldn't get high compression pistons for it, so its running standard ones. It was infuriating at the time as I did notice a decrease in grunt, but considering that it used to pink a bit on the leaded fuel we got then, its probably for the best now. *sad2*
I must admit the difference in figures copared to everyone else here makes me wonder if the gauge I used was a little optimistic, at the time I was just worried that compression be the same on both cylinders.
Title: Re: Measuring compression
Post by: snowbeard on 15.09. 2008 18:39
ok, everybody heard him ask me for pics  ;)

this is the most recent, with the red plug wires from that fellow on BB that was selling sets for $25
att 1

this is a little earlier, but still the same seat mount
att 2

att 3

I made two plate brackets for the front since it didn't come with the standard yoke, they aren't attached to eachother, just bolted to the frame mount and each side of the seat frame, but they seem to be holding solid.  the rear is mounted with two bolts thru round pipe spacers to raise it a couple inches to level, since the springs I got with it were the 4" style.  with the swingarm shocks and the springs that are in the seat, it makes for a pretty decent ride.  I might try to set it back just a few inches someday, but it's fine for now as well.