The vehicle is a 1996 Rover 100, 1.1, K Series engine.

On its latest MoT test, it had extreme difficulty passing the emissions checks. Only great patience and a little magic from the MoT inspector got it to pass.

Listening closely to the exhaust note, I detected the occasional, faint, missed beat, and it was possible to smell hydrocarbons in the exhaust gas.

Having encountered these symptoms 8 years ago, I had a fair idea where the problem lay.

I set up my 4425 PicoScope to monitor the inlet manifold vacuum (using a Fluke pressure sensor), the No 1 HT voltage (using a capacitative pickup), and the exhaust pressure pulses (using a home-made sensor based on a fuel-tank pressure sensor).

Six years ago, and 2 years after refurbishing the cylinder head, the same test was carried out and the waveform obtained is shown below:
The exhaust pulse is the red trace, with positive pressure in the downwards direction; the inlet manifold vacuum is in green; and No 1 HT voltage is in blue. It’s quite clear that the exhaust pulses are uniform, with one exhaust pressure pulse for each inlet-manifold vacuum pulse, or 4 exhaust pressure pulses for 720 degrees of rotation. That ideal waveform should be contrasted with the most recent one, also taken at idle:

There is nothing uniform about the exhaust pressure pulses, and instead of seeing one exhaust pulse per inlet-manifold pulse (or 4 pulses per 2 engine rotations), we can see 8 pulses. Furthermore, in the second half of the waveform there are 2 events, in the same part of the cycle relative to No 1 cylinder firing, where a much higher positive pressure pulse is detected in the exhaust.

Similar high-pressure exhaust-pulse events were recorded on the following waveform, which was captured at high RPM, whilst the vehicle was stationary:

At these higher RPM, the exhaust pressure pulses are looking a little bit more uniform, and, in this more dynamic state, there are now only 4 pulses for every 2 revolutions of the engine, as indeed there should be.

So, it appears there are 2 faults: leakage of pressure past the exhaust valve seats, especially at low rpm, and the occasional sticking exhaust valve.

The disappointing part is that, since the last overhaul in 2016, as little as 18,000 miles ago, only premium 4-star fuel has been used, with Esso Supreme only being used in the last 4 years (because seemingly it contains no ethanol). However, my driving style - too many short journeys and not enough long ones - may well have contributed to these faults appearing so soon after the cylinder head was last refurbished, though I’ve no idea of the precise mechanism.


As for the home-made pressure sensor, this is based on a fuel-tank pressure sensor 16238399 for GM, Buick, Chevrolet, Cadillac, bought for £9 on ebay. I can power it either from a USB supply or a 9V PP3 battery.
It took a little experimentation, using the Venturi effect, to arrange the plastic tubing such that the detection of pressure (and vacuum) was at its greatest.

Hi Martin, I hope you are well and thank you for taking the time to post this study

I do remember with the years I spent at Rover, just how fickle the K, M and T Series engine could be with regards to carbonising valves

It was often attributed to short journeys and low mileage vehicles and it was not unusual to see a cylinder head removed twice in 18 months where the vehicle covered little in terms of mileage (Oil consumption was also a contributing factor on older vehicles)

I remember a "fix" being issued by Rover entitled "Carbon busting valves" which were to be installed on such engines. Whilst it sounded dramatic, it turned out these valves were machined with a thinner stem on approach to the seat face.

This would simply allow for a larger growth of carbon deposits before ultimately the same symptoms appeared again (sticking valve) I am not sure if there was more to these "carbon busting valves" but from the outside it did feel like a hack!

It does look like you were onto cylinder 1 here topic23536.html and whilst it is nice to obtain more conclusive evidence, it does leave you with another conundrum (to fix or not fix)

Thank you again for posting, take care......Steve

What kind of pressure sensor is this? For the ones I have founds on any car(intake, exhaust, differential, low/high pressure fuel, turbo, brake vacuum, ..) the output signal voltage follows the pressure variation. If measured pressure rises, the voltage rises, and vice versa. For most of them, at atmospheric pressure, bias voltage is 0.5V, but can be 0.8V or 1. Have you tried to calibrate a pressure/voltage scale for this sensor? Did you use the same sensor and Venturi tube on both captures? What sensor did you use for INT pressure? Here you are a capture from 1.5DCI K9K engine. RED-injector 1, BLUE-exh, GREEN- int. Dead misfire on cyl 3 as injector faulty. EXH pressure drops as cyl 3 builds up vacuum due to lack of combustion.

Steve Smith wrote: ↑Fri Aug 16, 2024 11:17 am Hi Martin, I hope you are well and thank you for taking the time to post this study

I do remember with the years I spent at Rover, just how fickle the K, M and T Series engine could be with regards to carbonising valves

It was often attributed to short journeys and low mileage vehicles and it was not unusual to see a cylinder head removed twice in 18 months where the vehicle covered little in terms of mileage (Oil consumption was also a contributing factor on older vehicles)

I remember a "fix" being issued by Rover entitled "Carbon busting valves" which were to be installed on such engines. Whilst it sounded dramatic, it turned out these valves were machined with a thinner stem on approach to the seat face.

This would simply allow for a larger growth of carbon deposits before ultimately the same symptoms appeared again (sticking valve) I am not sure if there was more to these "carbon busting valves" but from the outside it did feel like a hack!

It does look like you were onto cylinder 1 here topic23536.html and whilst it is nice to obtain more conclusive evidence, it does leave you with another conundrum (to fix or not fix)

Thank you again for posting, take care......Steve

Hi Steve,

(Sorry for the slow reply: despite having every notification/email box ticked in my preferences, I didn’t receive notification of reply)

Thank you very much for your comments, which, are always greatly appreciated.

No real conundrum: whilst, economically, it makes not a shred of sense, for personal satisfaction and enquiry, I’m going to replace the cylinder head with my spare head (which is the original head and which was removed at 104k miles, with exactly the same symptoms). (Because of an ageing brain, as well as nostalgia, I use RAF maintenance paperwork to document every step of the removal and refit, so nothing gets overlooked, and I have future references as to what I did, including part numbers etc; so that’s an added reason for looking forwards to the job. Plus, I will ensure I frequently monitor the state of the valves using the Pico to capture the pressure pulses.)

The current head, which has only covered only 18k mikes, I bought on ebay to refurbish whilst keeping the car on the road. And I did indeed fit “carbon break” exhaust valves on the rebuild; so my experience fully backs your theory that it is indeed a useless hack. Slightly disappointing in that I used only premium fuel since the rebuild, and only Esso Supreme during the last 4 years.

But I’m certain you are bang on: short journeys is the cause.


Now here is a photo of the face of an exhaust valve from the original head:

When I come to remove the current head, I expect I’ll see exactly the same. I see impacted carbon flakes on the valve face,

and I also see “dimples”

I presume the dimples are caused by flakes that get hammered in and eventually burn away.

Whilst searching for replacement new valve guides, I discovered a system whereby the existing valve guide is drilled to take a bronze liner - a far better idea not least because the guide and seat remain aligned.

http://www.britishv8.org/Articles/Rover ... Guides.htm

And I know an excellent engineering firm who will fit this K-Line system during the re-skim.

https://www.jandeengineering.co.uk/valveguides.html

wiki38andy wrote: ↑Sun Aug 18, 2024 5:36 pm What kind of pressure sensor is this? For the ones I have founds on any car(intake, exhaust, differential, low/high pressure fuel, turbo, brake vacuum, ..) the output signal voltage follows the pressure variation. If measured pressure rises, the voltage rises, and vice versa. For most of them, at atmospheric pressure, bias voltage is 0.5V, but can be 0.8V or 1. Have you tried to calibrate a pressure/voltage scale for this sensor? Did you use the same sensor and Venturi tube on both captures? What sensor did you use for INT pressure? Here you are a capture from 1.5DCI K9K engine. RED-injector 1, BLUE-exh, GREEN- int. Dead misfire on cyl 3 as injector faulty. EXH pressure drops as cyl 3 builds up vacuum due to lack of combustion.

Thanks for the reply. The sensor I used for vacuum pulses in the inlet manifold was a MAP sensor from a Renault Laguna.

https://www.ebay.co.uk/itm/291878331529 ... media=COPY


And the exhaust-pipe pressure sensor is 16238399 Fuel Pump Tank Vent Pressure Sensor EVAP For Chevrolet Buick Cadillac

https://www.ebay.co.uk/itm/133063347306 ... media=COPY

Now, I did have a slight problem with this sensor in that I had to be careful when building it to ensure it didn’t max out on either its pressure side or vacuum side, I can’t remember which one it was. But once I got the plastic pipes exactly in place to stop the maxing out, I locked it all in place with self-amalgamating tape. I think I probably picked up these ideas on this Pico forum or certainly elsewhere on the Internet. I do have a “proper” pressure sensor “Fluke”, and I wasn’t at all bothered about calibration of these sensors, I was more interested in looking for anomalies, although I fully accept the logic behind your question.

As to a bias voltage, it looks like I mistakenly had it DC coupled on the original capture, so a bias voltage of a little over 3V shows. In the later captures, I had it correctly set to AC coupled. Not that it matters: I can see where the neutral point is (around 3.5V on the DC coupled capture), and I’m only interested in the variations, the shape of the curves, the number of cycles etc.

Yes, I did use exactly the same pressure sensor in the exhaust (and the intake) on both occasions.

wiki38andy wrote: ↑Sun Aug 18, 2024 5:36 pm What kind of pressure sensor is this? For the ones I have founds on any car(intake, exhaust, differential, low/high pressure fuel, turbo, brake vacuum, ..) the output signal voltage follows the pressure variation. If measured pressure rises, the voltage rises, and vice versa. For most of them, at atmospheric pressure, bias voltage is 0.5V, but can be 0.8V or 1. Have you tried to calibrate a pressure/voltage scale for this sensor? Did you use the same sensor and Venturi tube on both captures? What sensor did you use for INT pressure? Here you are a capture from 1.5DCI K9K engine. RED-injector 1, BLUE-exh, GREEN- int. Dead misfire on cyl 3 as injector faulty. EXH pressure drops as cyl 3 builds up vacuum due to lack of combustion.

I think I found the original source for my choice of sensor for the exhaust pressure pulses:

https://www.autonerdz.com/cgi/yabb2/YaB ... 1373552797

Steve Smith wrote: ↑Fri Aug 16, 2024 11:17 am
I do remember with the years I spent at Rover, just how fickle the K, M and T Series engine could be with regards to carbonising valves

It was often attributed to short journeys and low mileage vehicles and it was not unusual to see a cylinder head removed twice in 18 months where the vehicle covered little in terms of mileage (Oil consumption was also a contributing factor on older vehicles)

I remember a "fix" being issued by Rover entitled "Carbon busting valves" which were to be installed on such engines. Whilst it sounded dramatic, it turned out these valves were machined with a thinner stem on approach to the seat face.

This would simply allow for a larger growth of carbon deposits before ultimately the same symptoms appeared again (sticking valve) I am not sure if there was more to these "carbon busting valves" but from the outside it did feel like a hack!

Thank you again for posting, take care......Steve

Going back in time to the 1980s I used to rebuild engines when they failed in service like many others in the trade, with what appeared to be an exception to the norm, and that being I like knowing why things failed and what caused it! The major problem I've always found in the trade being that people can ask questions, but the person answering the question must be very clever! I found many that got their backs up because as engineers etc, I looked up to them, but it seemed they didn't have the correct training or experience, thus getting myself a bad name because I think differently than the average trader!

Putting that now to one side, if I may, I'd like to comment on the pictures of the valves Martin has posted and maybe a little science understanding about valves! I too have seen many exhaust valve contact faces damaged for various reasons, and asking those wrong people got me a bad name.

Recently I've rebuilt a cylinder head from a Peugeot 307 because the valves became damaged due to auxiliary fanbelt failure. Looking at Martin's examples of burned valve seats and the pitting he refers to, this condition can also be seen on my Peugeot valve seats (exhaust only)

You mentioned that Rover came up with a "Carbon busting valve" to work around the problem! You mentioned they used valves with "thinner" valve stems that they thought would maybe reduce or remove the problem of valve seat problems?

As said above, going back to the 1980s I've seen a lot of these types of problems and has years have gone by much has been improved, the reliability and long life of engines without the need for cylinder head removal every six months or so as the case used to be back then.

Martin mentioned that he thought the valve seat pitting was possible caused by the carbon deposits being crushed between the rubbing mating faces as the valves closed, which is probably a very good reason and exxplanation for the pitting being there, as this phenomenon only seems to occur on the hotter exhaust valve seats.

The manufacturers/designers must be well aware of this condition given that much has been improved since the 1980s. In my younger days seeing this problem week in and week out and not being able to ask about the condition and why it occurs sent me down another road in life! Reading, yes taking in the designers/manufacturers research material and learning and understanding from them reasons why components deteriorate and fail in service. With regards to the engine valves and engine breathing my research suggested that manufacturers worked along the lines of these ideas to try and reduce the damage effects to the valve seats...

The exhaust valve gas flow is the reverse of the inlet valve flow with the exception that the inlet gas flow in uncontaminated. An exhaust valve that operates at low lift when opened only causes the gas flow to seperate ever so slightly at the corners, which then re attaches itself on both adjacent seats so that it fills the gap before it discharges into the relatively large port passage as a convergent conical jet stream. This is the point for understanding I think...

Consequently, the sudden expansion of the exhaust gas dissipates the kinetic energy of the gas with only the minimum being converted to pressure energy downstream of the valve port seat, hence slow moving gas particles becoming trapped between the seat rubbing face.

Now, if we increase the valve lift, the gas flow seperates from the outer seat corners and remains free from both sides so that it establishes a jet with the centre of the valve stem as its axis. The jet being being of sufficient size relative to the port cross sectional area to convert some of the gasses kinetic energy into pressure energy, that is a pressure rise down stream of the valve port seat.

From this point forwards I could do some math to justify the research above, but its probably easier to say that there is a ratio between the valve lift, the duration its open and the piston working travel to push out all the spent gasses, and let the overlap period be sufficient to pass clean gas through the exhaust valves to reduce and or prevent the exhaust valves seats from pitting and guttering due to the contamination and seat grinding effects occurring.

Some information for thought provoking ideas maybe?

Could you elaborate these ideas, please?
"the sudden expansion of the exhaust gas dissipates the kinetic energy of the gas"
"gas particles becoming trapped between the seat rubbing face"

This will be an ongoing case study, possibly with a couple of offshoots that will probably be better as separate case studies.

In the meantime, I took my original cylinder head, removed 8 years ago because of a sticking exhaust valve, to a specialist machining workshop, J and E Engineering, in Rossendale, in preparation for swapping over with the misbehaving cylinder head now fitted to the engine.

Back in 2016, an analogue compression-gauge test during cranking on that original head, was filmed on a smart-phone, which captured the valve-sticking incident as a zero-compression event during the compression stroke on No 2 cylinder. On removing that exhaust valve, it was seen to be covered in oil, far more so than the other 3 exhaust valves.
Time will tell if this is also the cause of No 1 exhaust valve sticking on the cylinder head now in place. But J and E Engineering use the original and genuine K-Line valve-guide insert system, which is a very elegant solution to the renewal of worn valve guides. It negates the need for valve-guide removal, the phosphor bronze liner wears far less than steel or cast-iron valve guides, and the internal helix grooves in the liner act as oil reservoirs but are also designed to send any oil that gets past the seal, back up the guide.

This case study is yielding some very interesting PicoScope traces, as well as helping me get more familiar with PS7 and some of its new features, such as the Annotations in the Early Access version.

To be continued …….