Slow Monster rebuild

[350TSS]

Being very bored with mold filling/rubbing down I spent most of yesterday’s available garage time thinking how to fit my coils into the CF enclosure within an aluminium box which will act as my Faraday cage. Concerns are:
a) Containing the coils within the physical space envelope;
b) Maintainability – being able to get at the coils and associated input/output leads;
c) Restricting water ingress to the enclosure and preventing puddling if it does get in;
d) Space to permit air circulation around the coils for cooling;
e) Minimising the extent of holes required in the aluminium container for input and output leads so as not to compromise the effectiveness of the Faraday cage;
f) Necessity for resilient mounting to avoid vibration damage? And
g) A design that I can reasonably bend and is strong enough to bear the weight of the coils (347 grams bare without fasteners/ spacers)
CAD techniques were employed – cardboard aided design, about 10 templates were cut as I tried out various increasingly more complex possible designs and eventually it became obvious even to me that if I could not make it work in cardboard I was never in a million years going to be able to cut and fold it in 3mm aluminium successfully.
The thing with doing stuff like this is that it is very easy to become fixated with one element (in my case yesterday where the spark plug leads were going to exit/run) and you ignore the other elements of the problem. The end design is a compromise whatever you do and any solution that favours one element to the detriment of the others is a bad solution.
I had a cup of tea and another ponder and by thinking laterally came up with a very simple solution which I shall make next week.
This morning these arrived and I am now waiting for the garage to warm up so I can a) try out my new gauges, b) machine the first draft hub to completion (taking out the bore to an interference fit 40mm to take the bronze bushes which themselves will be bored from both sides to the depth of the bearing ( 7mm ) and bored to 37 mm interference fit for the ball bearings themselves, and c) make a completely new hub with a larger OD so that the countersinking for hanger plate attachment screws does not breach the outer ring.

[350TSS]

Being very bored with mold filling/rubbing down I spent most of yesterday’s available garage time thinking how to fit my coils into the CF enclosure within an aluminium box which will act as my Faraday cage. Concerns are:
a) Containing the coils within the physical space envelope;
b) Maintainability – being able to get at the coils and associated input/output leads;
c) Restricting water ingress to the enclosure and preventing puddling if it does get in;
d) Space to permit air circulation around the coils for cooling;
e) Minimising the extent of holes required in the aluminium container for input and output leads so as not to compromise the effectiveness of the Faraday cage;
f) Necessity for resilient mounting to avoid vibration damage? And
g) A design that I can reasonably bend and is strong enough to bear the weight of the coils (347 grams bare without fasteners/ spacers)
CAD techniques were employed – cardboard aided design, about 10 templates were cut as I tried out various increasingly more complex possible designs and eventually it became obvious even to me that if I could not make it work in cardboard I was never in a million years going to be able to cut and fold it in 3mm aluminium successfully.
The thing with doing stuff like this is that it is very easy to become fixated with one element (in my case yesterday where the spark plug leads were going to exit/run) and you ignore the other elements of the problem. The end design is a compromise whatever you do and any solution that favours one element to the detriment of the others is a bad solution.
I had a cup of tea and another ponder and by thinking laterally came up with a very simple solution which I shall make next week.
This morning these arrived and I am now waiting for the garage to warm up so I can a) try out my new gauges, b) machine the first draft hub to completion (taking out the bore to an interference fit 40mm to take the bronze bushes which themselves will be bored from both sides to the depth of the bearing ( 7mm ) and bored to 37 mm interference fit for the ball bearings themselves, and c) make a completely new hub with a larger OD so that the countersinking for hanger plate attachment screws does not breach the outer ring.

[Darren69]

Can you not use coil sticks and run a low voltage ignition? No HT leads and no bulky coils.

[350TSS]

Good suggestion but the blurb put out by the Fastbikegear (FBG) people suggests that coil sticks work fine with injection systems, they have a lower intensity spark (c20,000 volts) than “proper coils” (25 to 40k volts) and they work because the fueling is much more precise across the rev range. FBG did not say directly that coil sticks would not work on a carbie -just not part of their recommendations to get the optimum from their kit.
It is obviously a lot more complicated than that because even though the ignition trigger has said it is time to spark, the spark will only jump the gap when the air space between the electrodes reaches a certain ionization state (related I believe to the potential voltage threatening to jump the gap). The richness or otherwise of the fuel mixture and the load under which the engine is operating also affects this (worst case is with wide open throttle at low revs). With a fuel injection system the throttle position sensor in conjunction with the ECU which is also monitoring revs adjusts the timing of the spark to optimize the effectiveness of the burn (and in the certain knowledge that the coil stick can only deliver 20k volts).
With carburetor bikes the fueling is much less likely to be a precise / ideal air fuel ratio of 12.5 to 13.1 :1 and for the spark to actually jump outside this range and burn the mixture completely it needs and benefits from the higher voltage which is able to overcome the sub optimal sparking and burning conditions.
The ignition system I have bought includes an ignitech box which has a take off for a Throttle Position Sensor (TPS) and FBG state in their blurb that fitting a TPS to a carburetor M900 is a really worthwhile mod giving much snappier mid- range and take off from low revs.
I may consider this modification after I have got the thing running but from memory the space around the cable quadrant is heavily congested and I might take the view that it is just something else to go wrong. ????

[Luddite]

Originally Posted by 350TSS

... it became obvious even to me that if I could not make it work in cardboard I was never in a million years going to be able to cut and fold it in 3mm aluminium successfully.

Forgive me if I'm talking rubbish here but I'm basing this on 40 year-old O-level knowledge, which I'm not even sure I understood properly at the time!

Does the Faraday shield have to be solid? Would it not work just as well made from a tight mesh like this?



It would be easier to work into complex shapes and help with any cooling problems. You could perhaps have a solid base, strong enough to carry the coils with a mesh housing surrounding the coils themselves.

[350TSS]

Luddite:
I am no expert myself, but I do know mesh would work. In Victorian times there were circus acts with people enclosed in wire cages and a static generator was cranked up to maximum and discharged toward the poor unfortunate in the cage. The bolt of lighning then played around the cage in a spectacular fashion to the amazement of the paying onlookers and the guinea pig emerged from the cage completely unscathed.
Aluminium foil also works and is/ will be my solution. I will use a solid plate between the coils and the MUv2 control unit (the bit that should be 300mm away from the coils but is probably only 100mm away) and hang the coils in the CF enclosure which will be lined with aluminium foil- simples!!!!

[Luddite]

An ingenious solution! I look forward to seeing pictures of the finished article.

[utopia]

In that case (and having insufficient knowledge to comment on the Faraday cage idea itself) have you considered sticky aluminium foil tape ? My local pound shop sells it.
Or, how about the very thin expanded ally sheet that is sold for reinforcing patch repairs to car bodywork (Isopon ?). Perhaps a thin layer of that could be incorporated into the carbon fibre itself as an internal skin, covered with a thin layer of cloth and resin to secure and protect it ?
Or an even thinner metal gauze as used in sieves/screens .. though on second thoughts that might not lay flat very easily.
Or maybe these random thoughts will prompt further ideas.

[Darren69]

Won't the carbon fibre act as a faraday cage? It is conductive but not as much as metal.

[Mr Gazza]

Okay so the problem here is interference to the ECU from the coils, yes?

Do you perceive the interference to be a strong magnetic field or electromagnetic noise (RF or radio frequency noise).

If it is a magnetic field then I don't think a faraday cage will have any effect. In fact I'm not sure that it is possible to shield from magnetism?

If it is RF noise that is the problem, then there will be other sources of that from the bike in any case. So it is much more logical to shield the ECU (or whatever device is in question). Rather than the source of the noise.

A crisp packet will do this, but I daresay that you are after something a bit more sophisticated than that?
Aluminium foil will do the trick, provided that no inadvertent electrical connections are made with it.

A good way to do this is to first wrap the device with a rubbery type rigid foam, which will go some way to protect from vibration. Then wrap that with foil. Obviously leave the connection port(s) clear and preferably facing away from the coils... This also simplifies mounting the device, to a simple cradle or clip.
It is possible that some of the cables connected to the device might also pick up inference. Using coaxial cable will help with this. Or braided sheathing can be used. Again care must be taken to avoid an unwanted electrical connection, but the sheathing can sometimes be deliberately employed as an earth.

The mesh shield that you propose might not be effective against RF noise. There is a relationship between the RF frequency and the hole size in the mesh. and as the RF noise in question will be "White Noise" spanning a range of frequencies, it might well let through a fair bit of it. (or even all of it!)

Edit.... Carbon fibre will shield RF to some degree, but not totally.

[350TSS]

The joy of this forum is that you can get access to knowledge and assistance with ideas, thanks guys.
I do not know whether the MUV2 control box performance will be hindered by magnetic fields or RF interference, the instruction manual does not say. All it says is “have a 300mm gap from the coils”. If it is magnetic fields then I am flummoxed and as Mr. Gazza intimates there may be no way of protecting against them. Although I will research how naval minesweepers deal with magnetic mines - degaussing I think it is called – probably over thinking this but no harm in trying to find out.
Mr. Gazza, I think you are absolutely right the protection should be located around the vulnerable equipment not the source of the interference. This would be very problematic in the location I have selected for the unit as there are /will be around 20 wires connecting into and out of the box. It will be very congested there and the connections to the box are via small single grub screws into brass blocks so there is always the possibility of these becoming detached and enabling a short circuit/fire.
Also, the box has LED indicator lights which a) enable set up and b) indicate problems or circuits at risk. The control unit has to be mounted on a stout plate with a significant earth wire - 6mm square minimum. I will already have to remove the tank to see these LEDs. I could possibly cut up a mobile phone pouch (the ones designed to protect against unauthorized access to the phone) with a front opening Velcro fastening to enable viewing of the LEDs.
If the issue is RF interference I really like the idea of incorporating some mesh within the mold itself.
I think I will e mail the manufacturer of the MUV2 unit to ascertain whether it is RFI or magnetic fields that are problematic

[Mr Gazza]

Ah! Not such an easy nut to crack then?
The large metal mounting plate seems to imply some sort of heat sink, and so would in turn imply a disadvantage to enclosing the device.
You have the right idea to contact the supplier for the definitive answer to what kind of interference we are up against. They could well have a solution to the shielding regime too.

For what it's worth, I think it is less likely for the problem to come from RFI, as I think most of the sensors in question will work on electrical resistance and be fairly bombproof. Unless of course there is a microprocessor in there, in which case it could be vulnerable.

Thinking about magnetic fields; you have a Hall effect sensor inputting there, which works on a fluctuating weak magnetic field. If it is the case that coils also generate fields (which could pulse alternately?), then these fields will all have a relationship with engine RPM and therein would lay an interference problem.

All conjecture of course! It will be interesting to hear what the manufacturer has to say.

Out of interest, has anyone ever held a compass near a running coil. Or moved a compass around a switched on coil?

[Dukedesmo]

Originally Posted by Mr Gazza

Okay so the problem here is interference to the ECU from the coils, yes?

Mine has ignitech ignition/ECU and I was getting inconsistent ignition timing at idle (maybe elsewhere also but not so noticeable?) presumably due to the HT lead to the rear cylinder passing by the lead from ignition pickups to ignitech as they (must) cross to reach their objective.

I 'diverted' said ignition lead the other side of the frame rail and got it as far away as possible from the HT lead and the idle was noticeably smoother, as evidenced with a strobe on the flywheel now showing a (fairly) steady spark which, before was all over the place.

[Mr Gazza]

Originally Posted by Dukedesmo

Mine has ignitech ignition/ECU and I was getting inconsistent ignition timing at idle (maybe elsewhere also but not so noticeable?) presumably due to the HT lead to the rear cylinder passing by the lead from ignition pickups to ignitech as they (must) cross to reach their objective.

I 'diverted' said ignition lead the other side of the frame rail and got it as far away as possible from the HT lead and the idle was noticeably smoother, as evidenced with a strobe on the flywheel now showing a (fairly) steady spark which, before was all over the place.

So as well as sending a signal pulse from the crank sensor, the ignition cable was getting a (very slightly) delayed pulse via RF (or electro-magnetic radiation) from the HT lead.. Like an echo in fact.

[350TSS]

Typically, I go to bed between 10.00 and 11.00 pm and recently my rheumatoid arthritis has been waking me around 3.00 am (every bloody morning). I cannot go back to sleep and I have to get up to move and ease the stiffness and throbbing in my joints. Ibuprofen helps but concentrating on something I enjoy pushes the discomfort to the back of my consciousness. Hence the tedious and overlong (?) posts on here.
This morning I thought more about the situation I am in re the positioning of the electrics, worrying if the MUV2 electronic control box and /or the ignitech ignition box will be the victim of magnetic or electronic interference from the coils. My real concern was that the effects would be partial (at particular rev ranges) or intermittent. Reading on the internet about EMI/RFI and protecting for it there does not seem to be a “silver bullet” solution. In certain circumstances, if you screen cables and you do not do it correctly the screen itself can act as an antenna and make the problem worse.
The worst-case scenario would be putting it all together having stripped everything down and rebuilt it and it would not start or ran roughly. Is it the carbs, the ignition timing, the cam belt timing, the valve clearances, my wiring harness? – any one of a thousand things. Best not to go there.
My problem/source of worry is self-generated - a function of positioning of the equipment - so change the design and reposition the equipment. The sensitive stuff is the MUV2 and potentially the Ignitech box.
From the internet I glean that the potential sources of EMI/RFI are coils and plug leads, the rectifier/ regulator, starter motor and starter solenoid. The MUV2 manual only requires 300mm separation from the coils, no mention of any other equipment.
I need to move the coils from their current location but to where?
Options are:
1) To the rear of the battery tray (they will fit here but is underneath the carburetor float bowls so a potential fire hazard and not quite 300mm away from MUV2 unit)
2) To the under-seat tray (easiest option but too far away so long HT leads required and long power/trigger cables to coil)
3) Cut out a location on the underside of the tank above and behind the carburetors and attach to a plate bonded to the underside of the tank
(a lot of work to do this given current state of the mold and attaching coils to the tank is hardly ideal also it is likely to be a location that is hot being above and just in front of the vertical cylinder. Also concerns re the length of lead required to ensure that the tank can be fitted and removed ie lengths of trailing wire)
4) Cut out a location on the underside of the tank above and behind the carburetors and make a bracket(s) brazed to the frame and attach the coils to these, avoids the maintenance and lead length issues but it will still be hot here (no hotter than the original Monster location though).
Before I started writing this Option 1 was favoured but option 4 is probably the right answer, unfortunately it involves the most additional work.
The coil enclosure originally envisaged will now house the Ignitech box so it can be reduced significantly in size allowing better airflow to the rear cylinder so a bit more work as well but a better all-round solution.
My job list to conclusion (maintained on an excel spread sheet) is now 251 distinct jobs over an hour long totaling 449 hours.