Igniter with two diodes in series

[aquamon]

Hi,

I'm currently completing my first reverse engineering - on a Suzuki GS450's igniter (CDI or powerpak).
1982 model so that you may have an idea about electronics era.

I figured out most but I'm having doubts about two diodes in series placed across a Darlington **broken link removed** (sub 2sd1409a) from its base to collector.
Emitter is grounded as with all switches (transistors) on this board (other than 1 jfet).

In addition to this duet-series-diode, the collector (driving/interrupting the ignition coil) has a 0.15uF on it to ground.

I read somewhere that these diodes connected in series would give 2(0.7V)=1.4V drop - acting as a voltage regulator (which makes sense since automotive applications have a varying 12V supply).

If it's a general diode, I assume that this configuration would regulate the base to a voltage ALWAYS 1.4V less than the collector, is nonsense as the diode is pointed towards the collector???

Silkscreen suggests that it's a Zener which I'm a bit reluctant to check as I have done only up to 12V with no indication...

The markings on the "2-pac zener" diode are (each diode): T 1Z 175 1.l
Yes, that is a 1.l and NOT 1.1

I searched for a 1W 175V zener (1z175) but it seems extremely rare if in existence.

Is this really a 175V zener diode or a general series-doubled-diode?
If these are really Zeners, will this give protection, value, regulation, reference or what?

Zeners in Series: **broken link removed**

General diode series calculation by Dean Hustler: 175Vx3=525V

Note that this drives the primary winding on the ignition coil which gets a constant 12V supply from battery.

It operates similarly to the points in the older stuff (sparks when points open).

 

[aquamon]

Here's the schema:

 

[aquamon]

Of the Board:

Picture of the board:

 

[kchriste]

It seems like an odd place to put it, but it could be a MOV.
Can you post a picture of the actual part?

 

[Boncuk]

To me it looks like a fast recovery diode.

 

[aquamon]

Pls. note that I made those pictures/diagrams.
They are not original.

Regarding posting actual picture of the part, it simply looks like two black Axial-10 diodes crimped & soldered in the middle.

The "wife" stole both my digi-cameras and destroyed them.
After she bought a Sony thingy, see my other posts (hahaha!).

I really do believe that **broken link removed** is it (bottom L/H of page).

Also, remember that the white wire "W" goes to the ignition coil (primary winding) which is fed battery power 12DCV-14DCV on the other end of this primary winding.

The secondary winding goes to spark gap (spark plug) which is grounded through the cylinder head (engine).

The primary winding is series connected to the secondary winding.
This is where the "W" or white wire is tapped in.

I assume this diode set is to protect the Darlington (Using a transistor as a switch).
"Protection diode
If the load is a motor, relay or solenoid (or any other device with a coil) a diode must be connected across the load to protect the transistor from the brief high voltage produced when the load is switched off."

 

[kchriste]

Yes, it looks like the 2 zeners are designed to turn on the 2SD799 when the voltage across it rises to apx 350V, thus clamping the maximum voltage across the transistor. The transistor has a max CE voltage rating of 400V.

 

[tonigau]

Modern Ignition IGBT's have this zener built in, its part of what makes them robust for ignition coil switching.

Tonig.au

 

[aquamon]

Igbt

You guys RULE!!!


If I use IGBT, what would it replace and what p/n would you suggest as it's now practically impossible to obtain those zeners?

Absolutely no experience with IGBT's!

Looks like it would be more feasible to adapt IGBT though...

How would you do that?

R18 is 300 Ohms 1/4 watt carbon.

 

[tonigau]

You can always use more than 2 Zeners to get the total breakdown Volts.
75V zeners should be readily available.

The original XV ignitor used 2SD1071 darlingtons, these I assume have internal zener EB clamp.

Can only see part of your circuit so is a bit hard to advise with detail.

However it should work as a "dop in" replacement, (please check pins, I have'nt had time to check now),
Check the gate threshold volts of the IGBT & the voltage swing of your existing base drive circuit.
Do a bit of research on IGBT & darlington transistors should help.

I have attached schematic of my ignition board it has IGBT part numbers, I have issues with this board (software & noise on reset IC), Please don't ask for software or board layout, I may have it sorted next year if I get the time.
It works well on the bench though.
Soft turn off is still a WorkInProgress.

 

[tonigau]

Found the datasheet for the 2sd1071, can see the built in zener just like IGN IGBT.

------

In my circuit T5 main function is to hold IGBT's off while 2051 is in reset.

I will release this TCI as a full project once I have sorted the problems.

Tonig.au

 

[aquamon]

Thank you Toni,

I removed the complete schemas due to probably infringement protection.

You know how that goes.

The company that made it is still in business.

If anyone needs it to repair their own Suzuki, pm...pls.

Awesome how you found that datasheet!!!

 

[tonigau]

Awesome how you found that datasheet!!!

Some transistors & IC's have no datasheet available as the parts are often made just for the automotive companies & they don't usually like to share.

I tried to get data on a M59201P made by mitsubishi but nothing found.

The hardest part sometimes when reverse engineering an ignition is working out the triggering scheme.

On the SR500 yam they have CDI with a trigger pulse from a small pickup coil, the advance scheme uses a sine wave at RPM frequency, the leading edge of this wave increases in amplitude with rpm increase & relative to the trigger pulse earlier in time.

XS1100 uses the other (ch2) pickup to initiate dwell on ch1 & V.V.

XV1k is independent ch1 & ch2. but it uses both peaks (+&-) of the pickup pulse. First peak is full advance, second peak is idle timing & the advance curve is in between.

Smart cookies those autoelectronic engineers of days gone by.

My ignition uses just one negative peak but it makes the software more complex to calculate timing & dwell especially during RPM increase but at the end of the day it is more versatile & easy to adapt to different machines.

 

[aquamon]

http://member.ic37.com/sell/search.asp?keyword=M59201P#

I couldn't find it either but I'm sure if you go **broken link removed** and send a request for the data sheet to all the email addresses on the page, one will send it to you?

Especially if you tell them that you would be buying 9999999 million of them and need it for development of your circuit.

I think the noise you are getting could be from the reset pin?
Just a thought as I saw a post where someone got rid of this by putting power to it???
But on a counter circuit.
I personally would increase voltage ratings on C14/C15 to at least 630V (1KV)

Have you checked out the L484 (pg. 8)?
Looks really simple and check out BU930Z (NTE98?).

 

[aquamon]

Yes, it looks like the 2 zeners are designed to turn on the 2SD799 when the voltage across it rises to apx 350V, thus clamping the maximum voltage across the transistor. The transistor has a max CE voltage rating of 400V.

About the 50 Volt protection=Exactly 100% correct!
As I found out in this on page 3-4 :

 

[chrisjpitt]

XV1k is independent ch1 & ch2. but it uses both peaks (+&-) of the pickup pulse. First peak is full advance, second peak is idle timing & the advance curve is in between.

This sounds very much like the analog BB105 type Kokusan Denki ignition units that Ducati used on a lot of their carburetted 2V belt engines (my M900 has these). The flywheel/ign rotor has two independent raised timing "bumps" (offset by 90 deg,) that give the advanced and idle ign timings, by virtue of their length - as you point out.

Have you any idea what is sensed to advance between the two timings? I suspect it is frequency based, but it could also be done with pulse amplitude?

 

[tonigau]

any idea what is sensed to advance between the two timings? I suspect it is frequency based, but it could also be done with pulse amplitude?

I would guess frequency based, The pulse waveform peaks are very short & contains no data for advance curve that I can see.
The XV ign uses M59201P IC's, absolutely no data available for these chips.
The old SR500 uses sine amplitude from a secondary pickup to apply advance.

In my IGBT ignition I use the first peak only of the pickup pulse & delay to spark timimg based on frequency using uC code.

Attached is anatomy of the pickup pulse of the XV1k

Tonig

 

[chrisjpitt]

I would guess frequency based, The pulse waveform peaks are very short & contains no data for advance curve that I can see.
The XV ign uses M59201P IC's, absolutely no data available for these chips.
The old SR500 uses sine amplitude from a secondary pickup to apply advance.

In my IGBT ignition I use the first peak only of the pickup pulse & delay to spark timimg based on frequency using uC code.

Attached is anatomy of the pickup pulse of the XV1k

Tonig

Yes, I suspect you are correct – the timing of the pulses gives the rpm far more accurately than attempting to do this from the frequency/amplitude response of the pulse coil.

I have not had one of the BB105 type Kokusan Denki units apart, so I don’t know if they use that same chip - but it’s likely they do, if the principle of operation is the same. There are several versions (p/n BB1xx) of this type of ign unit around that I know of, but I don’t know what the differences are – possibly rpm/advance curve or rev limit? Kokusan Denki also state these units are “analog” and I guess that just means they don’t contain a uController.

The ‘scope trace looks very familiar – except, on the M900, there is a noticeable underlying higher frequency sine wave (from the multipole flying magnet alternator?).

 

[aquamon]

This sounds very much like the analog BB105 type Kokusan Denki ignition units that Ducati used on a lot of their carburetted 2V belt engines (my M900 has these). The flywheel/ign rotor has two independent raised timing "bumps" (offset by 90 deg,) that give the advanced and idle ign timings, by virtue of their length - as you point out.

Have you any idea what is sensed to advance between the two timings? I suspect it is frequency based, but it could also be done with pulse amplitude?

Yes, it is made by Kokusan Denki for Suzuki in my case (still in business-google).
It uses the crude and reliable spring mechanism/centrifugal weights to advance the timing at higher rpm's with an induced voltage around 1.5V

Very analogue indeed!

 

[chrisjpitt]

Yes, it is made by Kokusan Denki for Suzuki in my case (still in business-google).
It uses the crude and reliable spring mechanism/centrifugal weights to advance the timing at higher rpm's with an induced voltage around 1.5V

Very analogue indeed!

Yes, I am sort of familiar with that type of Suzuki ignition, it was on the GSX250 I had in the early 80’s – I can’t exactly remember, but I think it was twin pickups with a single ign unit and twin 12V (3-4 ohm) coils – is that what yours is like?

I remember the ign unit itself being pretty reliable (unlike the generator/regulator), but if the engine suddenly dies (and when you try to restart it, all you get a nasty rattling sound from the bottom end), check to see that the long M6 bolt that holds the trigger magnet/advance unit on the end of the crankshaft has not come loose, or has broken off? Certainly do this before you whip the engine apart looking for something more serious (don’t ask how I know this).