Top Dead Centre
Discussion
Without it accurately calculated you are sheding loads of power, but it is a bugger to accurately get.
Three weekends spent now with the cam covers off. Unfortunately the double rotation of the cam means that any error in crank position is magnified at the cam and you have an awful lot of Cerbs not running on cam.
The gold standard for a positioning system then is to measure the bucket heights on two consecutive attempts to find TDC and see the error. So here are my findings.
Spark plug dead stop. error 36%
Wire through spark plug hole error 23%
Borescope error 22%
compressor error 11%
Obviously I ended up using the compressor. Meanwhile contemplating the various co-incidences in life I noticed that the nut release tool on a simple angle grinder is a perfect fit in the head of the allen key bolts to turn the cam relative to the crank.
Simple co-incidence. I think not
Now going to spend next weekend getting a life.
Three weekends spent now with the cam covers off. Unfortunately the double rotation of the cam means that any error in crank position is magnified at the cam and you have an awful lot of Cerbs not running on cam.
The gold standard for a positioning system then is to measure the bucket heights on two consecutive attempts to find TDC and see the error. So here are my findings.
Spark plug dead stop. error 36%
Wire through spark plug hole error 23%
Borescope error 22%
compressor error 11%
Obviously I ended up using the compressor. Meanwhile contemplating the various co-incidences in life I noticed that the nut release tool on a simple angle grinder is a perfect fit in the head of the allen key bolts to turn the cam relative to the crank.
Simple co-incidence. I think not
Now going to spend next weekend getting a life.
Yep that was the borescope. You can pass the tube into the spark plug hole and look at the top of the piston. A horrible sight. The problem with the borescope is the same as that with the wire. The piston spends quite a few degrees of crank time hovering motionless at the top of its travel. During this motionless time the cam lobes are moving like an express train still 
I'll dig it out tonight, but its in his very interesting book on engine managment - worth a bedtime read just to get an understanding of ECUs.
But basically, you use a dial gauge (preferably two...) to get the same ammount of overlap on the inlet and exhaust valves. I'm sure it avoided the need to know the exact TDC, as you were measuring either side of it.
But basically, you use a dial gauge (preferably two...) to get the same ammount of overlap on the inlet and exhaust valves. I'm sure it avoided the need to know the exact TDC, as you were measuring either side of it.
Don't bother trying to guesstimate where TDC is.
Work it out from the crank angle at which the piston moves up and down.
Note the angle at which the pistong stops moving up, then take it past TDC and when turning the engine backwards note the angle again when the pistong stops moving. half way between those angles is TDC.
Work it out from the crank angle at which the piston moves up and down.
Note the angle at which the pistong stops moving up, then take it past TDC and when turning the engine backwards note the angle again when the pistong stops moving. half way between those angles is TDC.
The cams run at half engine speed. Using cam position as a reference to TDC is folly.
Although I realise that an engine strip is probably out of the question, the only way to reference TDC is to:
Locate a DTI on the axis of the bore and zero at nil (top) piston travel, fit a degree disk to the crank pulley and pointer to the timing cover.
Turn the crank anti clockwise and then clockwise until the piston stops and note the degree disc figure. Go over TDC (minimum 90 degrees) and turn the crank anti clock until the piston stops; note the degree disk figure. True TDC is half the difference between the recorded angles. Lastly, turn the crank anti clock at least 90 degrees, then turn clockwise to the deduced degree disc angular point. With the engine in this position, you may now make a suitable mark for all time in perpetuity.
Apologies if this is what you didn't need to know, but I am bored
.
Although I realise that an engine strip is probably out of the question, the only way to reference TDC is to:
Locate a DTI on the axis of the bore and zero at nil (top) piston travel, fit a degree disk to the crank pulley and pointer to the timing cover.
Turn the crank anti clockwise and then clockwise until the piston stops and note the degree disc figure. Go over TDC (minimum 90 degrees) and turn the crank anti clock until the piston stops; note the degree disk figure. True TDC is half the difference between the recorded angles. Lastly, turn the crank anti clock at least 90 degrees, then turn clockwise to the deduced degree disc angular point. With the engine in this position, you may now make a suitable mark for all time in perpetuity.
Apologies if this is what you didn't need to know, but I am bored
. Edited by sjwb on Monday 14th May 15:32
Well....
Both you guys are dupicating a sparkplug dead stop. Doing it with a dial gauge is mistaking accuracy for precision.
The big flaw in dead stop measuring seems to be all the losses measured when rotating the engine backward. (I actually had a dial gauge sellotaped to my bit of wire system but it didn't improve accuracy)
Secondly determining piston TDC by looking at camshaft pushing on the valves then using that to reset the camshaft is rather like standing in a bucket and trying to lift yourself up. It didn't make any sense to me as its the camshaft you are actually trying to reset.
Strangely my measurements (above shown) the losses incured using a deadstop were actually greater than trying to estimate using a length of wire on the top of the piston, so not sure I'm very impressed in using the deadstop method.
My gold standard for attaining the TDC was reproduceability. Anyone thinks that they can do better than my compressor. The test is that you have to set the crankshaft to TDC on cylinder 1 three or four times. Then measure the two cam followers on number 8 cylinder and attain an error in the measurement of the two cam followers each time of less than 11% amongst your three or four readings.
If anyone thinks they can do better than my score, you better try first cos I pretty much think 11% is the best anyone will get. Either that or I will travel a long way to learn from you.
Both you guys are dupicating a sparkplug dead stop. Doing it with a dial gauge is mistaking accuracy for precision.
The big flaw in dead stop measuring seems to be all the losses measured when rotating the engine backward. (I actually had a dial gauge sellotaped to my bit of wire system but it didn't improve accuracy)
Secondly determining piston TDC by looking at camshaft pushing on the valves then using that to reset the camshaft is rather like standing in a bucket and trying to lift yourself up. It didn't make any sense to me as its the camshaft you are actually trying to reset.
Strangely my measurements (above shown) the losses incured using a deadstop were actually greater than trying to estimate using a length of wire on the top of the piston, so not sure I'm very impressed in using the deadstop method.
My gold standard for attaining the TDC was reproduceability. Anyone thinks that they can do better than my compressor. The test is that you have to set the crankshaft to TDC on cylinder 1 three or four times. Then measure the two cam followers on number 8 cylinder and attain an error in the measurement of the two cam followers each time of less than 11% amongst your three or four readings.
If anyone thinks they can do better than my score, you better try first cos I pretty much think 11% is the best anyone will get. Either that or I will travel a long way to learn from you.
is it not possible to use the crank sensor? I mean, its easy to determine when piston 1 is at/near the top by using the wire/coat hanger down the spark plug hole, and then when both valves are closed (cam lobes on the heel) then you should be very close to TDC. from there can you not determine what the crank sensor needs to be looking at? and from there determine TDC?
failing that, the degree wheel is usually the best solution, but then that is old skool american engine stuff
failing that, the degree wheel is usually the best solution, but then that is old skool american engine stuff
I've been using DW's lift-on-overlap method for years. In fact I'm not even sure he'd lay claim to it being "his" technique but I'm happy to credit him along with one other person (another Dave) for introducing it to me.
The secret is not to rotate backwards when setting the cam timing. You can find TDC with a dial gauge forwards or backwards but cam timing should be only set forwards. You take up slack when running backwards and slew the results.
Of course to use this method, you need to know what lift figures you are trying to achieve. Most often cam timing is expressed in degrees, for which you need protractor measurements.
Also the Cerb V8 has a fixed relationship between inlet and exhaust timing (one cam per bank) which means you can set only one (probably inlet - though I haven't done it yet on the Cerb). This all assumes your tappet adjustment is accurate too.
I've done it before on K series and Vx 4-cylinder engines with great (and repeatable) accuracy once I'd worked out the best techniques to use. It was a long-time coming though, and I bent a set of inlet valves getting it wrong early on in my learning curve. Get it wrong with exhaust valves and they'll punch a hole through your piston (they're generally much harder than inlets to account for the increased temperatures they operate under).
For a 16-valve 4-cyl engine you need 3 dial gauges to do this work easily. I found it easiest with a couple of digital gauges on the two cam buckets and an analogue dial gauge on the TDC measurement through the plug hole. If you only do one bank at a time, you could get away with only one digital gauge but it's so much easier to do them both at the same time, and the cost of the digital dial gauges nowadays...
The secret is not to rotate backwards when setting the cam timing. You can find TDC with a dial gauge forwards or backwards but cam timing should be only set forwards. You take up slack when running backwards and slew the results.
Of course to use this method, you need to know what lift figures you are trying to achieve. Most often cam timing is expressed in degrees, for which you need protractor measurements.
Also the Cerb V8 has a fixed relationship between inlet and exhaust timing (one cam per bank) which means you can set only one (probably inlet - though I haven't done it yet on the Cerb). This all assumes your tappet adjustment is accurate too.
I've done it before on K series and Vx 4-cylinder engines with great (and repeatable) accuracy once I'd worked out the best techniques to use. It was a long-time coming though, and I bent a set of inlet valves getting it wrong early on in my learning curve. Get it wrong with exhaust valves and they'll punch a hole through your piston (they're generally much harder than inlets to account for the increased temperatures they operate under).
For a 16-valve 4-cyl engine you need 3 dial gauges to do this work easily. I found it easiest with a couple of digital gauges on the two cam buckets and an analogue dial gauge on the TDC measurement through the plug hole. If you only do one bank at a time, you could get away with only one digital gauge but it's so much easier to do them both at the same time, and the cost of the digital dial gauges nowadays...
Julian64 said:
Well....
Both you guys are dupicating a sparkplug dead stop. Doing it with a dial gauge is mistaking accuracy for precision.
The big flaw in dead stop measuring seems to be all the losses measured when rotating the engine backward. (I actually had a dial gauge sellotaped to my bit of wire system but it didn't improve accuracy)
Secondly determining piston TDC by looking at camshaft pushing on the valves then using that to reset the camshaft is rather like standing in a bucket and trying to lift yourself up. It didn't make any sense to me as its the camshaft you are actually trying to reset.
Strangely my measurements (above shown) the losses incured using a deadstop were actually greater than trying to estimate using a length of wire on the top of the piston, so not sure I'm very impressed in using the deadstop method.
My gold standard for attaining the TDC was reproduceability. Anyone thinks that they can do better than my compressor. The test is that you have to set the crankshaft to TDC on cylinder 1 three or four times. Then measure the two cam followers on number 8 cylinder and attain an error in the measurement of the two cam followers each time of less than 11% amongst your three or four readings.
If anyone thinks they can do better than my score, you better try first cos I pretty much think 11% is the best anyone will get. Either that or I will travel a long way to learn from you.
Both you guys are dupicating a sparkplug dead stop. Doing it with a dial gauge is mistaking accuracy for precision.
The big flaw in dead stop measuring seems to be all the losses measured when rotating the engine backward. (I actually had a dial gauge sellotaped to my bit of wire system but it didn't improve accuracy)
Secondly determining piston TDC by looking at camshaft pushing on the valves then using that to reset the camshaft is rather like standing in a bucket and trying to lift yourself up. It didn't make any sense to me as its the camshaft you are actually trying to reset.
Strangely my measurements (above shown) the losses incured using a deadstop were actually greater than trying to estimate using a length of wire on the top of the piston, so not sure I'm very impressed in using the deadstop method.
My gold standard for attaining the TDC was reproduceability. Anyone thinks that they can do better than my compressor. The test is that you have to set the crankshaft to TDC on cylinder 1 three or four times. Then measure the two cam followers on number 8 cylinder and attain an error in the measurement of the two cam followers each time of less than 11% amongst your three or four readings.
If anyone thinks they can do better than my score, you better try first cos I pretty much think 11% is the best anyone will get. Either that or I will travel a long way to learn from you.
Well the DTi method is used by all our engine builders where I work and produces results 100% acurate everytime. Once you have set TDC and noted it's position then you use this to set the cam timing. I assume you are using Maximum opening point on Cyl 1 as your set point.
As Mars has said, for cam timing you only ever set rotating forwards due to the slop in timing chanis, bearings etc. Always wind past where you need to be then rotate up to the point in the direction of engine rotation.
ridds said:
Julian64 said:
Well....
Both you guys are dupicating a sparkplug dead stop. Doing it with a dial gauge is mistaking accuracy for precision.
The big flaw in dead stop measuring seems to be all the losses measured when rotating the engine backward. (I actually had a dial gauge sellotaped to my bit of wire system but it didn't improve accuracy)
Secondly determining piston TDC by looking at camshaft pushing on the valves then using that to reset the camshaft is rather like standing in a bucket and trying to lift yourself up. It didn't make any sense to me as its the camshaft you are actually trying to reset.
Strangely my measurements (above shown) the losses incured using a deadstop were actually greater than trying to estimate using a length of wire on the top of the piston, so not sure I'm very impressed in using the deadstop method.
My gold standard for attaining the TDC was reproduceability. Anyone thinks that they can do better than my compressor. The test is that you have to set the crankshaft to TDC on cylinder 1 three or four times. Then measure the two cam followers on number 8 cylinder and attain an error in the measurement of the two cam followers each time of less than 11% amongst your three or four readings.
If anyone thinks they can do better than my score, you better try first cos I pretty much think 11% is the best anyone will get. Either that or I will travel a long way to learn from you.
Both you guys are dupicating a sparkplug dead stop. Doing it with a dial gauge is mistaking accuracy for precision.
The big flaw in dead stop measuring seems to be all the losses measured when rotating the engine backward. (I actually had a dial gauge sellotaped to my bit of wire system but it didn't improve accuracy)
Secondly determining piston TDC by looking at camshaft pushing on the valves then using that to reset the camshaft is rather like standing in a bucket and trying to lift yourself up. It didn't make any sense to me as its the camshaft you are actually trying to reset.
Strangely my measurements (above shown) the losses incured using a deadstop were actually greater than trying to estimate using a length of wire on the top of the piston, so not sure I'm very impressed in using the deadstop method.
My gold standard for attaining the TDC was reproduceability. Anyone thinks that they can do better than my compressor. The test is that you have to set the crankshaft to TDC on cylinder 1 three or four times. Then measure the two cam followers on number 8 cylinder and attain an error in the measurement of the two cam followers each time of less than 11% amongst your three or four readings.
If anyone thinks they can do better than my score, you better try first cos I pretty much think 11% is the best anyone will get. Either that or I will travel a long way to learn from you.
Well the DTi method is used by all our engine builders where I work and produces results 100% acurate everytime. Once you have set TDC and noted it's position then you use this to set the cam timing. I assume you are using Maximum opening point on Cyl 1 as your set point.
As Mars has said, for cam timing you only ever set rotating forwards due to the slop in timing chanis, bearings etc. Always wind past where you need to be then rotate up to the point in the direction of engine rotation.
Do you actually mean 100% accurate? No offence but has anyone ever checked their ability to TDC. Its easy to do. Just ask them to do it twice without looking at the cam then measure their results at the cam. I pretty much guarantee they couldn't do it twice within an accuracy of 10% at the cam.
The only going clockwise is a given, errors are massive if you ignore this which is why the deadstop doesn't work well.
The one I wanted to use was mentioned earlier on, in that the crankshaft sensor fires for each revolution. The problem is that is doesn't fire at TDC, and reports vary a little on how many degrees its meant to fire at. If someone knew this info for a cast iron fact it could be used to put a permanent TDC on the flywheel although getting a protractor accurately on the flywheel would be a pita.
Julian64 said:
Do you actually mean 100% accurate? No offence but has anyone ever checked their ability to TDC. Its easy to do. Just ask them to do it twice without looking at the cam then measure their results at the cam. I pretty much guarantee they couldn't do it twice within an accuracy of 10% at the cam.
You're confusing me here. You don't measure TDC at a cam. I'm sure you know this already but I can't work out what you meant above.
As for repeatability, I can set cam timing using the lift on overlap method, then run the engine for 5 mins to 5 days and recheck. The settings will be consistent. The area of "no movement" of the piston at TDC is very small but can be accounted for by averaging the entry and exit points.
Julian64 said:
The one I wanted to use was mentioned earlier on, in that the crankshaft sensor fires foreach revolution. The problem is that is doesn't fire at TDC, and reports vary a little on how many degrees its meant to fire at. If someone knew this info for a cast iron fact it could be used to put a permanent TDC on the flywheel although getting a protractor accurately on the flywheel would be a pita.
Crank sensor, which is actually a flywheel sensor, detects movement every few degrees. There are several flywheel triggers around its circumference but it shouldn't be used as a TDC indicator anyway. It's only relevance is mapping ignition and fuel timing [#1], not cam timing.
Remember also that your crank TCD indicator (the protractor) lines up with a mark on the front pulley which could be *out* by the odd few degrees here and there depending upon the line-up during assembly. Before relying on it, a quick dial gauge check on TDC would at least ensure you understand where your reference is for the future.
#1: most engines don't care **exactly** when their injectors fire as they fire well before they're required anyway. Where it becomes slightly more important is sequential injection (again, not terribly important) and direct injection (diesels most notably) when it's of paramount importance.
Julian64 said:
The big flaw in dead stop measuring seems to be all the losses measured when rotating the engine backward.
Hi Julian,
Where do you think the errors in this method are coming from? As the piston moves up the bore it doesn't know which way the engine is being rotated. If your degree wheel (the larger the better) is attached to the crankshaft and you rotate the engine the right way to end up at your calculated TDC point the only error is your repeatability at getting the crankshaft to line up with your calculated degree mark.
I tend to think any errors you are seeing at the cams are due to repeatability in the cam chain and cam/tappet mechanics. You've successfully found TDC.
Rich.
Julian64 said:
ridds said:
Julian64 said:
Well....
Both you guys are dupicating a sparkplug dead stop. Doing it with a dial gauge is mistaking accuracy for precision.
The big flaw in dead stop measuring seems to be all the losses measured when rotating the engine backward. (I actually had a dial gauge sellotaped to my bit of wire system but it didn't improve accuracy)
Secondly determining piston TDC by looking at camshaft pushing on the valves then using that to reset the camshaft is rather like standing in a bucket and trying to lift yourself up. It didn't make any sense to me as its the camshaft you are actually trying to reset.
Strangely my measurements (above shown) the losses incured using a deadstop were actually greater than trying to estimate using a length of wire on the top of the piston, so not sure I'm very impressed in using the deadstop method.
My gold standard for attaining the TDC was reproduceability. Anyone thinks that they can do better than my compressor. The test is that you have to set the crankshaft to TDC on cylinder 1 three or four times. Then measure the two cam followers on number 8 cylinder and attain an error in the measurement of the two cam followers each time of less than 11% amongst your three or four readings.
If anyone thinks they can do better than my score, you better try first cos I pretty much think 11% is the best anyone will get. Either that or I will travel a long way to learn from you.
Both you guys are dupicating a sparkplug dead stop. Doing it with a dial gauge is mistaking accuracy for precision.
The big flaw in dead stop measuring seems to be all the losses measured when rotating the engine backward. (I actually had a dial gauge sellotaped to my bit of wire system but it didn't improve accuracy)
Secondly determining piston TDC by looking at camshaft pushing on the valves then using that to reset the camshaft is rather like standing in a bucket and trying to lift yourself up. It didn't make any sense to me as its the camshaft you are actually trying to reset.
Strangely my measurements (above shown) the losses incured using a deadstop were actually greater than trying to estimate using a length of wire on the top of the piston, so not sure I'm very impressed in using the deadstop method.
My gold standard for attaining the TDC was reproduceability. Anyone thinks that they can do better than my compressor. The test is that you have to set the crankshaft to TDC on cylinder 1 three or four times. Then measure the two cam followers on number 8 cylinder and attain an error in the measurement of the two cam followers each time of less than 11% amongst your three or four readings.
If anyone thinks they can do better than my score, you better try first cos I pretty much think 11% is the best anyone will get. Either that or I will travel a long way to learn from you.
Well the DTi method is used by all our engine builders where I work and produces results 100% acurate everytime. Once you have set TDC and noted it's position then you use this to set the cam timing. I assume you are using Maximum opening point on Cyl 1 as your set point.
As Mars has said, for cam timing you only ever set rotating forwards due to the slop in timing chanis, bearings etc. Always wind past where you need to be then rotate up to the point in the direction of engine rotation.
Do you actually mean 100% accurate? No offence but has anyone ever checked their ability to TDC. Its easy to do. Just ask them to do it twice without looking at the cam then measure their results at the cam. I pretty much guarantee they couldn't do it twice within an accuracy of 10% at the cam.
The only going clockwise is a given, errors are massive if you ignore this which is why the deadstop doesn't work well.
The one I wanted to use was mentioned earlier on, in that the crankshaft sensor fires for each revolution. The problem is that is doesn't fire at TDC, and reports vary a little on how many degrees its meant to fire at. If someone knew this info for a cast iron fact it could be used to put a permanent TDC on the flywheel although getting a protractor accurately on the flywheel would be a pita.
Well to get MoP cam timing within 1° of specification every time then yes we can do it. It's accurate enough that we can set it run the engine and then measure it all again and come out with the same number. And that's with 4 cams 2 of which are phased off of the other 2.
diycerb said:
Julian64 said:
The big flaw in dead stop measuring seems to be all the losses measured when rotating the engine backward.
Hi Julian,
Where do you think the errors in this method are coming from? As the piston moves up the bore it doesn't know which way the engine is being rotated. If your degree wheel (the larger the better) is attached to the crankshaft and you rotate the engine the right way to end up at your calculated TDC point the only error is your repeatability at getting the crankshaft to line up with your calculated degree mark.
I tend to think any errors you are seeing at the cams are due to repeatability in the cam chain and cam/tappet mechanics. You've successfully found TDC.
Rich.
Thats a good question I don't have a good answer to. The obvious answer would be the piston conrod connection and the conrod crankshaft connection. However there really shouldn't be any play there at all. Which makes me think its my technique thats at fault.
Could someone explain valve overlap technique for finding TDC that people keep mentioning. If its what I think it is, it makes no sense mentioning it on this thread?
Hi Julian,
I just had a quick look and its about three pages worth in the book, but I did refresh my understanding of the process.
It does require knowing where TDC is (even roughly) and then setting the cams relative to that, using the lift on overlap method. What the book does say is that final cam timing can only really be determined by running the engine on the dyno and making small changes. This does make sense as you can set things up to theoretical values whilst static, but of course the engine is a dynamic thing and its whats happens when its running that really matters.
Of course, I know this is no help to you unless you have a dyno tucked away somewhere.
So knowing where TDC is will be required, but unless you measure the results with a dyno, you might set up everything perfectly according to TDC etc., but it might not be optimal for the engine when its running.
Is it worth contacting anyone who used to work on the Tuscan racer engines to see if they have any ideas?
Simon
I just had a quick look and its about three pages worth in the book, but I did refresh my understanding of the process.
It does require knowing where TDC is (even roughly) and then setting the cams relative to that, using the lift on overlap method. What the book does say is that final cam timing can only really be determined by running the engine on the dyno and making small changes. This does make sense as you can set things up to theoretical values whilst static, but of course the engine is a dynamic thing and its whats happens when its running that really matters.
Of course, I know this is no help to you unless you have a dyno tucked away somewhere.
So knowing where TDC is will be required, but unless you measure the results with a dyno, you might set up everything perfectly according to TDC etc., but it might not be optimal for the engine when its running.
Is it worth contacting anyone who used to work on the Tuscan racer engines to see if they have any ideas?
Simon
I was tapping the rick vein of opinion/info here in Cerb land. I pretty much rekon we can sort anything out here with others knowledge and my willingness to experiment/make mistakes on my car.
Now trying to work out why there should be losses with the dead stop method, and why the inaccuracy compared with the compressor.
Out of interest as well as the spanner on an angle grinder being a dead fit for the cam sprocket. I also found a conventional set of digital calipers is the perfect tool to measure the difference in cam follower heights. These have now replaced my two DTI's as first choice. Eventually I'll take a photo of the actually quite simple tools required for the job.
Now trying to work out why there should be losses with the dead stop method, and why the inaccuracy compared with the compressor.
Out of interest as well as the spanner on an angle grinder being a dead fit for the cam sprocket. I also found a conventional set of digital calipers is the perfect tool to measure the difference in cam follower heights. These have now replaced my two DTI's as first choice. Eventually I'll take a photo of the actually quite simple tools required for the job.
Julian64 said:
I also found a conventional set of digital calipers is the perfect tool to measure the difference in cam follower heights. These have now replaced my two DTI's as first choice.
Hi Julian,
Thats interesting as I found I couldn't balance the digital calipers well enough over the two buckets, the depth gauge just didn't quiet make it onto the second bucket well enough. I thought about making up a tool with the right depth of step which would sit nicely on both buckets when the correct lift was found.
On the dead stop, I wonder if its worth using a differnt length dead stop and repeating the test, mines about an inch and a half long from the end of a spark plug.
Rich.
Gassing Station | Cerbera | Top of Page | What's New | My Stuff