IGN Module

Having problems with a Seagull? - ask an expert here

Moderators: John@sos, charlesp, Charles uk, RickUK, Petergalileo

Daryl
Posts: 261
Joined: Tue Jun 02, 2009 12:56 pm
Location: South Australia

Re: IGN Module

Post by Daryl »

The offer of a free module has been taken up by a UK member.
Daryl
Posts: 261
Joined: Tue Jun 02, 2009 12:56 pm
Location: South Australia

Re: IGN Module

Post by Daryl »

Well!! there apears to be a reliabillity problem just destroyed 3 modules on a 40+ seagull with Wipac coil. On dismantling a module find the switching transistor is faulty. The next problem is to find a suitable replacement transistor with a higher voltage rating (that isnt too expensive) Oh well will keep looking. Daryl
electrosys
Posts: 130
Joined: Wed Sep 14, 2011 3:08 pm
Location: Boston, UK

Re: IGN Module

Post by electrosys »

Daryl wrote:Well!! there apears to be a reliabillity problem just destroyed 3 modules on a 40+ seagull with Wipac coil. On dismantling a module find the switching transistor is faulty. The next problem is to find a suitable replacement transistor with a higher voltage rating (that isnt too expensive) Oh well will keep looking. Daryl
Hi Daryl
It seems that I've taken an interest in this thread very late in the day ... and to be honest I'm totally muddled by much of what has gone before - which hasn't been helped by half of the photographs showing as blanks on my browser. Dunno if this last point can be rectified ?

One of the prima facie problems I had was understanding exactly WHAT this device was intended to do ... but then I spotted the Patent number you gave (... thanks ...) which explains much.

However, there's still something of a muddle here - for much of the Patent is concerned with coils being wound specifically for the purpose of 'breakerless ignition' (Table 1 near the end giving a summary ...).
So - can I first check with you - does the commerical module you have work OK with both the Wipac and Villiers coils ? (sorry if you've already covered this, but I'm playing catch-up here ...)

Now - assuming the answer is 'yes', which of the numerous diagrams in the Patent (another source of 'muddle' !) are you working to ?
Now I'll have to wait for your reply on that one, but I've taken Fig.20 of the patent as being representative of this type of circuit, and I've highlighted the flyback diode which protects the semiconductor driver from destructive 'back-emf' spikes from the collapsing magnetic field of an inductor. I note that your DIY module doesn't contain any such diodes. Perhaps this is why your switching transistors are dying ?

One other question - have you conclusively identified the switching (driver) semi-conductor in the commercial module ? I'm wondering whether it has an exceptionally high reverse-voltage rating, or whether it may have an inbuilt flyback diode ? Most modern power-FETs have built-in protection diodes, but they're not commonly seen incorporated into bipolar transistors.

Colin
Attachments
DIODE.gif
DIODE.gif (5.76 KiB) Viewed 618 times
Daryl
Posts: 261
Joined: Tue Jun 02, 2009 12:56 pm
Location: South Australia

Re: IGN Module

Post by Daryl »

Hi Colin. Used fig 3 & the transistor i used was a BD651 which is a darlington transistor with a built in diode betwean the collector & emmiter. & the transistor is rated 120Vce. Pity you cant see the pics as there is one of the orig Atom ign module split a part. Its switching transistor just has A2 on it, must have been custom made. The orig does work on a villiers coil & on a Tohatsu & a Tanaka outboard.
ATOM IGN.JPG
electrosys
Posts: 130
Joined: Wed Sep 14, 2011 3:08 pm
Location: Boston, UK

Re: IGN Module

Post by electrosys »

Hi Daryl - thanks for the info and diagram. Makes a lot more sense now - it was difficult trying to follow the thread without pictures - a bit like listening to only one side of a telephone conversation !

Well - you seem to have had everything covered - I was rather hoping that the reliability problem was simply due to an oversight.

I'll take a couple of hours off tonight and wade through that Patent - there's a heck of a lot of stuff there - and it ain't exactly light reading ...
Colin
electrosys
Posts: 130
Joined: Wed Sep 14, 2011 3:08 pm
Location: Boston, UK

Re: IGN Module

Post by electrosys »

Following a couple of days online research at the Patents Office into this whole breakerless ignition stuff - the following might be useful to anyone else interested in this subject.

For an easy read re: the principles, with an easy to understand schematic, I'd recommend reading through the US Patents of Robert Bosch: #4175509, #4188929 and #4329950, which are each concerned with developments of the same basic circuit.

For a not-so-easy read, try #5058543 - which, although my searches invariably turned-up the names of big players such as Bosch, Briggs and Stratton, Yamaha and so forth, this patent was awarded to "Sten's Lawnmower Parts" - which I found rather intriguing ...
Even more interesting is if you do a Google for "Sten's Megafire" (or Megafire II), you'll find that these modules are currently on sale for around $20 US.

Another patent which has intrigued me is #5111798 which, if you turn to Fig.5 you'll find an almost exact copy of the 'Atom' circuit (with just one unimportant diode added) - for which a patent is being applied for. Naughty.
Daryl
Posts: 261
Joined: Tue Jun 02, 2009 12:56 pm
Location: South Australia

Re: IGN Module

Post by Daryl »

Hi electrosys. Thanks for the extra patents, have printed them & checking them out. Modifyed one of my modules by fitting 2x 56V 5Watt Zenner diodes in series to limit the voltage to the switching transistor. After running for half an hour on a 40+ with a Wipac ignition coil. The Zenner diodes got up to 50 deg C & got hard to start. Tried the module on a Villiers coil for the same time & the diodes only got to 30 deg C. It seems this Transistor & Zenners cant cope with a Wipac coil. The real problem is to find a suitable transistor. I can find higher voltage Darlington transistors that have suitable current ratings but not hfe ( DC current gain) needs to be at least 750 to work in this circuit. Will keep looking.
ATOM IGN.2.JPG
ATOM IGN.1.JPG
Seagull 40+.1.JPG
Seagull 40+.2.JPG
electrosys
Posts: 130
Joined: Wed Sep 14, 2011 3:08 pm
Location: Boston, UK

Re: IGN Module

Post by electrosys »

Hi Daryl - yes, these circuits are somewhat more demanding than they first appear ...

Although several circuits use a 4 volt trigger level voltage from the primary coil, that coil is then reported to give a back-emf pulse of some 100-150 volts. One author even reports 300 volts !

So much depends on the coil in question, and it's magnets. If you run some soft-iron (or a 6" nail) around the outside of a Seagull flywheel, the Villiers hardly causes a stir, whereas the Wipac magnets have a fairly strong attraction. Which very much fits-in with your experimental observations.

I've got some BU806's on order, which are good for 200v, and plan to protect them with some L8C150C bi-directional zeners which I've still got in my spares box from the mid-80's (!). I know they don't need to be bi-directional - but that's what I've already got. Unfortunately, the hFE of the BU806 is only around a 100, so it'll need to be really well driven.

There is one device which may be of interest - the BU931 - which is good for 400v, with an hFE of around 300. That may or may not be enough gain. However, as these cost six times the price of the BU806, I'm giving them a miss for now.

'best, Colin
User avatar
Top Cat
Posts: 28
Joined: Wed Nov 02, 2011 2:27 pm
Location: Aylesbury

Re: IGN Module

Post by Top Cat »

The villers flywheel incorperates a circular soft Iron keeper whilst the Wipac does not.
electrosys
Posts: 130
Joined: Wed Sep 14, 2011 3:08 pm
Location: Boston, UK

Re: IGN Module

Post by electrosys »

Daryl - one solution to your low hFE problem with Power-Darlingtons is to 'triple' the Darlington. I've never done this myself (never needed to), nor even seen it done in practice - but here are two examples I've found.

The diagram with the speed-up diodes is from a heavy current driver module where an hFE of 3000 was required.
speed-up diodes.gif
speed-up diodes.gif (4.37 KiB) Viewed 552 times
The diagram showing a complementary driver set-up is from a HiFi system, where apparently triple Darlingtons have some kind of cult following (no, I don't understand that either). I've taken the liberty of highlighting one of the triples in that diagram.
fig1.gif
fig1.gif (5.26 KiB) Viewed 552 times
electrosys
Posts: 130
Joined: Wed Sep 14, 2011 3:08 pm
Location: Boston, UK

Re: IGN Module

Post by electrosys »

Bit of an update on the 'Atom' Ignition Module ...

It appears that the widely-held view that just one module was intended for 'universal' application is something of a myth. During their production days (pre-2005), AtomIndustries(.au) made several different types depending upon their end use, and colour-coded them accordingly.
So - for instance - the Red module was for twin-cylinder outboards (one module per cylinder) which have variable ignition timing.
The Purple module was for single-cylinder outboards with fixed ignition timing.
Yellow was for chainsaws, Brown was for something similar, and so on ....

So - that the same circuit is behaving differently with dissimilar magnet/flywheel set-ups is nothing unusual, and actually, it is to be expected.

BTW - if anyone's interested, there is still some residual stock for sale:

Red Atom
http://www.bantasaw.com/catalog/viewpro ... 199&p=2205

Purple Atom
http://www.bantasaw.com/catalog/viewpro ... 199&p=2206

Prices are typically around $20 Canadian (c. £12).
User avatar
Top Cat
Posts: 28
Joined: Wed Nov 02, 2011 2:27 pm
Location: Aylesbury

Re: IGN Module

Post by Top Cat »

Considering this thing was designed in the 1970s it is perhaps in need of an upgrade....rated 600v 315 watts Try blowing this one...
http://uk.farnell.com/fairchild-semicon ... dp/1971834
or this is a bit cheaper
http://uk.rs-online.com/web/p/products/ ... 4E4F4E4526

As its your baby Daryl I am sure you would like to build it but if you need any help with the design let me know.
electrosys
Posts: 130
Joined: Wed Sep 14, 2011 3:08 pm
Location: Boston, UK

Re: IGN Module

Post by electrosys »

Top Cat - regret to say you can't use a power Mosfet in a magneto ignition. Gate voltage limitations. Which is a shame ...
Daryl
Posts: 261
Joined: Tue Jun 02, 2009 12:56 pm
Location: South Australia

Re: IGN Module

Post by Daryl »

Hi Top cat. Its open to who ever can crack the puzzle of finding a suitable transistor or combination of transistors with suitable gain/voltage & current ratings. I suspect the original may have been a custom made transistor, it just had A2 on it. Colin will be interesting how you go with the BU806 transistor as they are avalable here.
User avatar
Collector Inspector
Posts: 4181
Joined: Sun Jun 29, 2008 4:32 am
Location: Perth Western Australia
Contact:

Re: IGN Module

Post by Collector Inspector »

Hi Guys, just a thought here with WIPac coils (Seagull) having a built in condenser?

You still have the condenser in circuit just removing the points etc.

Redraw one (or more) of the schematics above and include the condenser across?

What happens then?

Regards

B
A chicken is one egg's way of becoming others
Post Reply