Anatomy of a CX500 Cdi unit

[AshimotoK0]

Loaned my good mate RGP750 on here  my CX500 carbs and CDi to get his bike running properly. My carbs worked fine and member Woodside kindly donated Rich some spare carbs so he's sorted out now BUT I asked him to test my CDi ... and it was found to be duff. Hence my post on CX forum blah blah !

http://cx500forum.com/forum/technical-help-forum/89329-300-mile-only-bike-cdi-doesnt-work-another-bike.html#post824777

300 mile only bike CDi doesn't work on another bike
As posted on here when I started as a newbie earlier this year, I have a 305 mile only, 1982 UK registered, CX500B  which was left to rot in a damp shed presumably from 1982 because the DLVA records say it was last taxed in 1982. Everything about the bike, despite doubters on other forums, confirms the mileage (e.g. original fitment tyres with the moulding sprues still on the centre of the treads, oil colour like it came straight out of a new can etc, etc.). I am slowly gathering shiny parts for it with the aim to get it up and running but I have not attempted to run the engine yet. My friend in Kent recently had problems with a CX he just bought, so I sent him my carbs set and CDi to substitute, in order to identify the fault. My carbs were so clean inside that I didn't have the heart to put them in my ultrasonic cleaner. Anyway, he fitted my carbs and, without adjustment, his bike then ran perfectly with his own CDi. I therefore asked him to substitute and test my CDi but he said it didn't work, which surprised me a bit. Was there any changes to the later 1981-2 loom or CDi box which would explain this?

I appreciate that the later transistor assisted modules can be fitted but I would like to investigate why my CDi didn't work on his bike. I designed my own CDi for my Honda 400/4 in the late 70's as I am an Electronics engineer (ran OK for 20k miles until I sold it !) so I am well up on the technology, although picking out the encapsulation to investigate the cause of failure will be a bit tricky. However, if the unit is in fact duff then I have nothing to lose. Anyone got any tips or experience to share please? It is stamped TIA02-14 on the can with what looks like the Hitachi manufacturers logo.

Cheers for now .. Ashley

I did find this which is useful to some degree

http://www.cx500.50webs.com/



--------------------------------------------------------------------------------------------------------------------------------------------------------------------
Thanks for the replies. The CDi was tested on another bike but the same model as mine and that bike was fitted with the same CDi with the same part marking on the can. The 'test' bike ran fine with it's own CDi and ziltch with the one from mine.

I had a feeling of déjà vu here .. My mate bought a CB400N from a guy at work with less than 1k miles.... again left to rot in a damp shed. The carbs were horrid but after cleaning them, the bike ran OK for a week or so. Then the CDi failed.

Here is what I have done so far : Pics and write-up to follow.

I picked out quite a bit of the 'cheesy' type potting. I immediately noticed that the steel enclose was internally rusty on one side. I pick down as far as I could. Then I removed the black sleeving from the wiring and carefully pulled each individual wire through the grommet so that I could pick at the potting better around the fly-leads. I felt that by exerting force on the PCB this would possibly lead to grief as you have a large bonded area under the PCB. Drastic as it, is I slit the corners of the can with a 1mm cutting disc and folded the sides down, so that I could gradually peel the bottom ( normally top of the can) from the potting. This was interesting because there were several large voids in the potting between the PCB and the can. One was so bad that the bare underside of the PCB was exposed. When I designed potted 'gizmos' ( do you USA guys called them 'Hoo-dickeys'   ) I used to heat the module to around 45°C and pour a small amount of epoxy encapsulant into the potting box with its PCB in it . I then vacuumed it to remove entrapped air, filled up with potting and then re-vacced and cured. The Chinese sub-contract manufacturer we used, hated me for this (I went over there 5 times) but I was insistent that if we potted stuff it must be done properly. It got to the point that I made them mark the potting box with coloured marks to identify that the various stages of manufacture had be adhered to (no pun intended) . So shame on you Hitachi for such bad encapsulation. Can you image if the 'Price of Darkness' Lucas in the UK had done this ?

Anyway, I sat yapping to my mate Roger yesterday afternoon and at the same time picking away further at the encapsulant. Before I knew it I had the underside of the PCB fully exposed and most of the components revealed on the top side. Who is this guy ? I hear you all saying but I actually quite enjoyed it. Just the kind of thing you can do while the 'other half' is watching junky TV programmes. I was worried that some of metal-film resistors may have got damaged in the picking so I desoldered one end of each one so that I could make measurements, but so far so good they are all coming out as preferred 'E' series values and no duffers ...yet.This also enabled me to clean underneath them. I will replace every resistor as a matter of course with modern equivalents 'just in case'. There seem to be quite a few glass passivated , controlled avalanche diodes on there ....I already use the BYW56 diodes, which are similar and readily available. There are no electrolytic capacitors on there to 'dry'up with age. Most seem to be mylar/polyethelyne film. Again I will measure and probably replace these as they don't cost much. Nothing on that module isn't replaceable with modern components and everything is discrete, i.e. no IC's. Nowadays, something like this would have a custom programmed chip in it that, if found to be duff, you haven't a hope in hell of getting a new replacement for ( users of Ultrawave late edition and Sono-Swiss industrial ...mega expensive .. ultrasonic cleaners beware !)

So what now ? ... I will clean up the PCB, replace all of the passive devices and try to identify any 'duffers' in the process , that may have been the cause of failure and then make up some kind of 'off-the-bike' test setup for it, once I have convinced myself what makes it tick. Then I will rehouse it in a sealed box encapsulated in silicone oil, so that if it ever fails again I can just remove the PCB assembly, clean with a suitable solvent and fix without all of the fun of picking at potting ! The original PCB by the way is a good quality epoxy-glass material and has PTH (plated through holes) so no worries there.

Where I worked in the 1990's , we used to 'pot' electronics in epoxy/PU, when really a conformal coating would have sufficed. This was purely because the owner of the company was protective of anyone copying/repairing our designs even though they were patented and he insisted that everything was potted. He made us use the rock hard when cured, mineral filled, type of epoxy resin too, that was only removable by hot refluxing with THF or similar nasty aggressive solvents until it went 'cheesy' and could then be picked off. Invariably though, components would disintegrate beyond recognition in the process.

I will write all of this up and post with pics etc. Sorry if it's been covered already and thanks on the cracking advice so far. Any advice on an off the bike test setup gratefully received. First thing (after repair) will be to simulate the alternator and pulser outputs and hook up to a pair of Honda coils I have spare.

Cheers for now ... Ash

I did a similar write-up on fixing Honda ignition coils for another forum :- here

https://www.dropbox.com/s/x0a4fcpmrw...20fix.doc?dl=0

https://www.youtube.com/watch?feature=youtu.be&v=qxWtR_m80Fc

[Johnwebley]

 Ash,let me thank you for your posts,

 I am not clever enough to really understand them,but as long as you can strip,and repair the encapsulated coils and CDI units
 there is always hope to keep this lovely old bikes alive ,

 regards

  John

[AshimotoK0]

Damn Ash that's above my pay grade, I can understand some of it but not having any sort of electronic training most goes right over my head. Fair play for getting stuck in there.

Says the man who back in the early Honda 4 years could work out plain bearing codes when all his colleagues couldn't ! ... that was a black art then Ken, so everyone on here has their own area of expertise. I just can't stand it when someone says 'can't be fixed mate'. Hopefully, once I have finished, it may end up as a simple procedure. Along the lines of "Sit watching your favourite soap omnibus and pick out all of the potting. Chances are components X Y or Z may be suspect, here is where to get a suitable replacement part to solder in " .  Don't know how RGP750 got on but he bought a replacement used Cdi for about £100 and it was duff  even though the seller guaranteed it was a good 'un and worked perfectly

[RGP750]

Good luck Ash sorry to be the bearer of bad news .
I did get my money back from the CDI seller ,no argument at all so i recon he was taking a chance on it .
MOT on my CX this Friday so hopefully ride over the weekend.

[K2-K6]

Interesting to see what went wrong with this unit.

I'm ok working with electrical systems but internal knowledge of design and what the components do inside some of these parts is more of a mystery to me. I'll follow it in the hope of getting a more informed view of what happens inside and benefit from your skills.

I think I have to try and convert electronic capability into basic mechanical function to get my head round it. Something like the link fascinated me as his approach was to produce a mechanical component that mirrored an electronic design,  he views the two as doing the same thing.

http://www.racecar-engineering.com/articles/f1/understanding-the-j-damper/#

Thanks for your in depth analysis Ash

Nigel.

[mike the bike]

    . I always wanted to figure out how to make my own circuit boards, can't be hard. 

I've been making pcbs since I was a kid.  Let me know what you want done and I'll talk you through it.

[AshimotoK0]

Well I am a fish out of water on that CX Forum with an alive and kicking 'Bitsa' (R.I.P.) type old sage who keeps having a pop at me telling me I am wasting my time. But you all know me ..such comments just spur me on. plus I keep getting PM's saying they are enjoying it and 5 'likes' ..so not a waste of effort at all !

Here is the CDi stripped down to the bare PCB and a circuit diagram taken from a Dutch site.

The background on the picture  is for a member on there 'Shep' who is a much respected bloke on there from Hull ... but Orcadian on here, another 'ull lad,  will appreciate it too ! Now need to order modern bits and rebuild it.

[ Guests cannot view attachments ] [ Guests cannot view attachments ] [ Guests cannot view attachments ]

[Trigger]

As I have said before mate, if it can be built before, there is no reason why it can't be built again.
But, why do they go wrong in the first place? must be room for improvement 

[mike the bike]

Replacing components with higher spec ones usually solve the problem.   
I've lost count of the number of touch lamps that knacker the triac when the bulb blows and I've had to solder a higher rated triac in.

[Woodside]

im watching...very interested in the outcome..having one of these old armchairs the probability of a failiure is fairly high

[AshimotoK0]

As I have said before mate, if it can be built before, there is no reason why it can't be built again.
But, why do they go wrong in the first place? must be room for improvement 

Well mine was never potted properly at manufacture so water had crept in and got exposed to the bare PCB.  What I didn't want to do is buy one second hand like RGP750 did for over 100 quid  only to find it doesn't work or may die soon after installation because the components are nearly 40 years old. Now that is hassle and wasted effort in  my book.

[mike the bike]

The pcb is made from an insulating material such as fibre glass or synthetic resin bonded paper, with a thin film of copper bonded to one or both sides.  The pcb above is single sided.  You know the rest...

[Bryanj]

 Oddjob did a small amount of PCB work at college and you paint on a protector to keep the tracks you want and etch the rest away

[AshimotoK0]

The copper hasn't became detached, it's the epoxy resist resist that's come away and the copper 'foil' is still fully bonded on that PCB in most areas. Once I have finished with that PCB it will be nearly as good as new. The board I always use is FR4 epoxy-glass, 1.6 mm thick. Bonded paper PCB etc is used in cheapo appliances/audio etc and the last time I used it was when I was a kid. FR4 is pretty cheap these days. It comes pre-coated with UV sensitive emulsion so that you can print your design on laser film and use the 'contact' method (like photography) to make your own quality prototypes.
I can get 'Gerber' files done on CAD for that Hitachi CDi PCB  so that the board could be reproduced easily but no real need for me personal at this point. 'Gerber'  files are the manufacturing files you would give to a PCB manufacturer (a bit like the machining numerical  files you would give to a CNC company). I get PCB's made all of the time but I rarely make my own these days. Once the Chinese got into manufacture it became mega cheap with very low tooling costs even for multi layer 'plated through' boards. Funnily enough though I taught the young lad where I work to make them only last week, as I felt I needed to pass on the skills and my own kids aren't interested at all in any of my tech stuff.... I am a lovable anorak to them too. 

[mike the bike]

Best get to grips with single sided boards first.  Double sided boards are more involved in that to make contact between the two sides, you either need PTH (plated through holes), which are beyond the scope of most hobbyists,  or solder short pieces of wire between the layers.