Suzuki TS185/TS250 CDI Box Replication [Update](Again)

[jeffheath]

I know you guys aren’t “bikers” but I feel as though posting here is my best bet at making this info
“visible” on the net for those who need it.
I feel this is important because for these models of bikes, no replacements are currently made, and
owners are depending on 40 year old brain boxes to keep their bikes running.
(I tried 4 different readily available 5 pin AC CDI boxes from ebay, and all were way off timing)

Basically, I have successfully “replicated” a CDI box for Suzuki TS185/TS250 motorcycles. My main
goal was crack one open, figure a circuit diagram, find modern replacements for all components, and
arrange them according to the circuit.
I wanted all components to be easily available from Mouser (although I’m sure they would be just as
easy to find anywhere else)
In the schematic below, you will find manufacturer part numbers for all components to avoid confusion.
For the discharge capacitor, I used a Panasonic metallized polyester, the part number for which is ECQ-E6225JF.
(C4 on schematic)

Thyristor, (Q2), Part number VS-30TPS12LHM3
SCR, (Q1), Part number BT169D,116
Zener, (D3), Part number TZX10C-TR
Diodes are part number 1N5623GP-E3/54

My particular bike is a 74, but this should work on most TS185/250s. The engine seems to be doing fine so far,
but I will continue to test it and see how it goes.

[Update] Thought I would add the gerber and schematic files, the gerber files should be in the zip file, and should be everything needed to get a pcb made. I've also included the cad files for the case, for anyone who needs to make this cdi box themselves or make it better!
[Another Update] Added a parts list since someone mentioned it, includes part numbers for every component.

[m3vuv]

well done,i had same issues on a yam xs-750 a few years ago.

[jeffheath]

Thanks, it's just that I've seen people restore these bikes to the hilt, but still use a 40 year old cdi box. I just don't understand how they're okay with that.

[m3vuv]

they will be ok with it untill the day they have to push it home with no spark! lol,73 m3vuv.

[joeqsmith]

I know you guys aren’t “bikers” ...

https://youtu.be/6PpAJeAurr4

https://youtu.be/bAJ8AY-cOUQ

[jeffheath]

Wow, a simulated hayabusa   Correction: Most of you guys aren't "bikers" 

[joeqsmith]

The real one can be seen here:
https://youtu.be/h9pixcy3GOU?list=PLZSS2ajxhiQA9yWh-hvVtv2ib-3ObOHGV&t=477

[NiHaoMike]

Is the kill switch just a short to ground when activated? It looks like that if you're unlucky, it would fire a spark right in the middle of the intake stroke and make a rather exciting backfire through the intake manifold. I would assume that in the absence of any way to shut off the fuel by electrical means, the engine would be shut down by disabling the trigger pulses so the ignition doesn't fire.

[jeffheath]

The kill

Is the kill switch just a short to ground when activated? It looks like that if you're unlucky, it would fire a spark right in the middle of the intake stroke and make a rather exciting backfire through the intake manifold. I would assume that in the absence of any way to shut off the fuel by electrical means, the engine would be shut down by disabling the trigger pulses so the ignition doesn't fire.

The kill switch is a short to ground. How would hitting the kill switch cause a backfire?

[mikerj]

Is the kill switch just a short to ground when activated? It looks like that if you're unlucky, it would fire a spark right in the middle of the intake stroke and make a rather exciting backfire through the intake manifold. I would assume that in the absence of any way to shut off the fuel by electrical means, the engine would be shut down by disabling the trigger pulses so the ignition doesn't fire.

It's a two stroke engine so the intake stroke is the same as the compression stroke, but you could still potentially get an advanced spark. 

The kill switch is a short to ground. How would hitting the kill switch cause a backfire?

Because the kill switch will ground out the non-coil side of the capacitor, discharging it into the coil. This arrangement is unusual, in most CDI designs one diode connects the igniter coil to the capacitor, and another diode connects the igniter coil to the kill switch, which prevents the kill switch triggering a spark e.g.

[jeffheath]

Update- I've made a board up in kicad, and will try to order one, populate it with parts and see how it goes.

[mikerj]

Looking at your schematic again, are you sure the kill switch wasn't originally connected in the middle of the two diodes?  That would also prevent it producing a spark.

[jeffheath]

Looking at your schematic again, are you sure the kill switch wasn't originally connected in the middle of the two diodes?  That would also prevent it producing a spark.

Attached is a picture of the original cdi board, and a snippet of the service manual, showing 2 megs resistance between kill(b/y) and excitor(b/r) wires. In the picture, two 1meg resistors are shown in red(they are on top of the board). The manual snippet mentions 2megs resistance between kill and excitor wires.

[floobydust]

Similar circuit and discussion:
https://www.eevblog.com/forum/beginners/diode-selection-102979/msg1412985/#msg1412985
 but with added CD4001 timer to prevent false triggering of the SCR, on Kawasaki KX80. Everyone else has a thermistor to back off gate-drive when the module heats up.
Check polarity on the 3.3uF cap, although I think you are using a film cap.

When I looked into it, there was a sea of patents back then so manufacturers could not come up with a simple decent circuit, they all got cute. The important part is where the ignition timing advance is being done. Usually in the magnetics.

[jeffheath]

Check polarity on the 3.3uF cap, although I think you are using a film cap.

3.3uF? As stated in the first post, the part number for the discharge capacitor is ECQ-E6225J (2.2 uF, 630v), I didn't put those values on the schematic because there are tons of different 2.2uF caps on mouser, and I wanted to make it simple for those who need to build this.

The important part is where the ignition timing advance is being done. Usually in the magnetics.

All I know about timing is that the service manual states that the timing is 16 degrees btdc at 1000 rpm, and 24 degrees btdc at 6000 rpm. The plan is to print a strip numbered 0 to 180, tape it to the flywheel, and check the timing at those rpms. But for now, with my test board, I haven't had any backfires when starting (it's kickstart) and it seems to rev up just fine.

[SeanB]

Your capacitor for C4 might not have a long life in there, the requirements for the capacitor in this CDI application are severe, and you might have lifetime issues with the capacitor.

https://industrial.panasonic.com/cdbs/www-data/pdf/RDI0000/ABD0000COS20.pdf

Generally the manufacturers use a foil construction to give high pulse handling ability, and the metallised film might not have a good enough endurance in this application. However it will work there, provided it does not get too hot from both self heating and from ambient temperatures.

[jeffheath]

Your capacitor for C4 might not have a long life in there, the requirements for the capacitor in this CDI application are severe, and you might have lifetime issues with the capacitor.

https://industrial.panasonic.com/cdbs/www-data/pdf/RDI0000/ABD0000COS20.pdf

Generally the manufacturers use a foil construction to give high pulse handling ability, and the metallised film might not have a good enough endurance in this application. However it will work there, provided it does not get too hot from both self heating and from ambient temperatures.

Is there ANY chance in hell of being able to determine the type of the original? (shown below)

[floobydust]

3.3uF? As stated in the first post, the part number for the discharge capacitor is ECQ-E6225J (2.2 uF, 630v), I didn't put those values on the schematic because there are tons of different 2.2uF caps on mouser, and I wanted to make it simple for those who need to build this.

I was referring to the chinese copy, the 3.3uF gate-cap is backwards and D4 not across the SCR 


You're doing great work, it's very difficult to unpot these and get the part values, let alone guess the Japanese components.

I was trying to figure out the SCR triggering, and stumbled onto the little SCR seeing some very high peak currents because it basically shorts the pulser-coil, once triggered. edit: updated pickup coil resistance to 87 ohms and it's OK 300mA guessing based on pickup coil voltage.

Is R1 really 130 ohms and in the right spot? You need over 12V there which is almost 1W impulses. It just seems a bit low.

The circuit is like a double capacitive discharge. C1 gets charged and the little SCR dumps that into the big SCR to trigger it, which dumps C4 into the ignition coil. Sims gives 37A pk for 0.7R ign coil primary.

[Mr. Scram]

Nice job so far. Following with interest.

[jeffheath]

You're doing great work, it's very difficult to unpot these and get the part values, let alone guess the Japanese components.

Thanks.

updated pickup coil resistance to 87 ohms

Service manual states pulser coil resistance of 75 ohms, so I guess the amperage on the little SCR should still be OK at that resistance.

Is R1 really 130 ohms and in the right spot? You need over 12V there which is almost 1W impulses. It just seems a bit low.

Resistor colors are brown orange brown, disconnected from circuit it measures about 128 ohms. I have double checked it's place on the board and circuit, but what's funny is that it's just tacked on the bottom of the board, like an afterthought.

[jeffheath]

Nice job so far. Following with interest.

Thanks ;  I have ordered the boards from jlc, they cost all of 2 dollars! (but 20 for shipping) Will post back once they arrive.

[SeanB]

Your capacitor for C4 might not have a long life in there, the requirements for the capacitor in this CDI application are severe, and you might have lifetime issues with the capacitor.

https://industrial.panasonic.com/cdbs/www-data/pdf/RDI0000/ABD0000COS20.pdf

Generally the manufacturers use a foil construction to give high pulse handling ability, and the metallised film might not have a good enough endurance in this application. However it will work there, provided it does not get too hot from both self heating and from ambient temperatures.

Is there ANY chance in hell of being able to determine the type of the original? (shown below)

Well, it is a polyester capacitor, and the Panasonic one will work there, it at least is rated to 100A pulse current, though it will run warm at that. As the coil is the current limiter, it will most likely work for a few years there. Just keep in mind it will, like the original one, fail with time, going lower in value. Make it easy to replace, probably by either potting in a flexible compound or just using a conformal coating.

[floobydust]

I thought R1's value and location look weird, ahhh "it doesn't work!" kind of fix but at the expense of kick- start harder faster, lol. Big difference in easier kick-start (triggering) with it at 1k but I could not find the consequence.

To dampen ringing from the pick-up coil due to the reverse-recovery time of diode D4, or limit the big SCR's gate pulse current, I'm not sure what the low value is trying to do and can only guess at the magnetics.
I was using TS250 coil values, the TS185 a bit different. Maybe R1 is for one model.

[jeffheath]

I was using TS250 coil values, the TS185 a bit different. Maybe R1 is for one model.

Finally got a hold of a ts-250 service manual, and it says "87 ohms on TM250 and 75 ohms on all other models" so that explains that. Also according to partzilla, the ts185 and 250 cdi units have the same part number, 31900-30020

[jeffheath]

Just keep in mind it will, like the original one, fail with time, going lower in value. 

The thing is, the original capacitor still measures just fine at 2.19 uF with my multimeter. I don't have the equipment to do leakage tests, but maybe it would lower in capacitance under load? (This is actually part of how I determined capacitance of this part, as I could only very faintly read 2.2 on the capacitor itself.