Building a defeat EXUP circuit or Dump that Servo

[jewilson]

The resistors trick

One R1 list/forum stated that you could get around the problem using some resistors in the loom plug to "trick" the CDI into believing the exup servomotor is plugged-in. This is rubbish. If you try it you will see that the fault code is maintained. If you actually think about what the CDI is doing you can see why. The CDI is trying to see if the servomotor or its built-in potentiometer is broken. If you put a resistor across a couple of terminals you are telling the CDI that the EXUP is stuck i.e. it is not functional. It is thus understandable that the CDI issues a fault-code.

R1 EXUP Replacement Circuit
The schmatic will not srink down to good view size.


Having decided to develop this I decided to measure the various signals on the EXUP-servomotor loom. The design of the servomotor is pretty simple and is in fact a classic servomotor with feedback control to the CDI

The B/Y and B/R wires look like they power the motor so these were the first wires I looked at. What you find is that these are wired straight across a dc motor. They are energized in turn, with the opposing wire being earthed. Thus the motor is rotated first in one direction, and then in the other. I saw that the B/R wire gets driven to +12V almost instantly power is applied to the CDI. As it turns back to 0V (after 0.5sec or so) the other wire, B/Y is driven to +12V for around 2secs. After this time both are powered down and the CDI concludes no EXUP is present and starts its fault-codes.

So we now know that these wires power the motor. The others are obviously for the potentiometer feedback. A quick check with a multimeter will tell you which wires are used here.

You find:
Blue - +5V output from CDI to power the Pot.
Black/Blue - 0V (never driven and buzzes out to earth)
White/Red - floats (will not drive current) at 4.5V - the feedback input

Having seen the pulsing of the motor wires at the start we know that the CDI is trying to exercise the servomotor across its full extents. We know the time it takes to do this and we know that this will result in a varying resistance as it happens. This gives us the design brief for the gizmo as we now know signals, times and function required.
There are a dozen ways to solve the problem but the most elegant is to use a digital-potentiometer chip.

This can be setup to move the resistance up and down with time. I used a Xicor digital-pot for my design. The circuit is shown below:
Circuit Diagram

The circuit operates by monitoring the two signals that would normally drive the dc motor. When one is ON you want the resistance to go UP. When the other signal is ON you want it the resistance to go DOWN.

If you look at the datasheet for the Xicor part you can see that you can use the device as simple up/down resistor very easily. If you pick out the logic control lines you get the following truth table:

CS U/D INC Action
0 1 1 -> 0 Resistance UP
0 0 1 -> 0 Resistance DOWN

If we now put in our control signals that would normally drive the dc motor we get a full table.

States
B/R B/Y CS U/D INC Action
0 0 X X 1 -> 0 No R change
0 1 0 0 1 -> 0 Resistance DOWN
1 0 0 1 1 -> 0 Resistance UP
1 1 X X 1 -> 0 Don't care

Some conditions are shown as dot-care (X) as these shouldn't occur in our system. These allow the logic to then be determined. The U_/D can be driven directly from the Blk/Red signal direct (once the signal is "potted-down" to prevent damage to the Xicor part.

The /CS must be driven from a combination of the two lines. The function is actually a NOR function. This is implemented using diode/transistor logic.

The /INC line is provided by a free-running clock. We only need to count up 100 steps (pot spec) in 2seconds. This yields a clock frequency of 50Hz or so. This was provided by a simple 555timer IC1 a CMOS 4047 can be used also.

The circuit was found to work 1st time! The CDI really believes the exup servomotor is plugged in! In normal operation it will issue commands to the gizmo which will faithfully modify the resistance as required.

If you have some electronics skills you can make one of these units for your bike or car for just a about $10. Parts can be otained from Mouser Electronics or DigiKey. If you build the circuit you will need to ground your self because static kill semiconductors. A 20 to 40 watt solder iron should be use a high powered iron will burn up the circuit when soldering.

Use the circuit diagram and the prototype picture as guides.
Just to let you know I found this circuit on the internet and Ihave not build it yet but I will.

http://biketransplant.tripod.com/exup_replacement.htm You can also view the circuit at this site. The circuit is free for all to copy and use Thanks to Marc Bell.

[tgitchel]

Jim,

Did you already check the color-coding of the EXUP wires on the FZ1 as compared to the R1 that the circuit was designed for? I have all the parts to build the circuit and just haven't verified the pinout for the connector yet.

TG

[eskort]

The color code is the same as he describes for the GEN 1--don't know about Gen 2.

[jewilson]

The connection is very simple we have power, ground, the 2 control signals and the ouput signal. My FZ1 is a first gen1 bike. Poor me.

the 9713 http://www.intersil.com/data/fn/fn8183.pdf
is a Digitally Controlled Potentiometer. I have attached the block diagram if you don't have it.

The signal at Q1 pull the transistor low, it's state is normally high. Also, this transistor can be a 2N2222 or a 2n4401 it just a generic NPN transistor. This is basic diode transistor logic.

These are both control signals. If you have a scope and the color code is off you will be able to tell that.
Black/Yellow
Black/RED

You should also check the power signal at the connector. If it's not 5VDC you will have to put a regulator in series with the curcuit, something like 7805 three pin regulator will work.
All the analog ground have a start connection back to the controller-ECU. So you have a node all ground wire one from each circuit go to that node then it connects back to the ECU.
The Ground wire should be a 20 AWG wire minimum.
I would also put a 10uf cap on power and parallel it with a .01uf ceramic cap for noise.

Of course you need to put this thing in a small box.

If you have any problems let me know.

[Marboy]

Hi guys

From the german Fazer Forum, another way to eliminate the EXUP (works for both the Gen1 & Gen2 FZ1, from here):

There's a guy named "metulszki" who sells them, for EUR 20 (about 30 USD). They work like a charm!

[jewilson]

Yep that nice and simple for sure. The one thing I don't like is he has no resistors in series with the input of the 2mosfets. Some times a mosfet will oscillate when it does not have any real input resistance. No wires that good.

I like the way he used the connect and the pins on the little circuit boards.

Thanks Marboy

[tgitchel]

Parts for the simplified circuit are less than $3 (not counting the board) compared to about $12 for the other one. Plus the smaller circuit doesn't require a project box. I'm glad to hear the first circuit may be superior since I have already started on that one.

[Dek]

Hi Marboy
I've emailed 'metulszki asking for purchase details.
Thanks for the link.
Dek

[jewilson]

I went looking through the German site and found two circuits that are really easy.

[Bruce Bohannon]

so..... is somebody local gonna build a pile of these things and have em ready to ship to us for some reasonable price?

I'm in for the right price.

[ssparky77]

Originally Posted by Bruce Bohannon

so..... is somebody local gonna build a pile of these things and have em ready to ship to us for some reasonable price?

I'm in for the right price.

What is the right price?

[patricjft]

Originally Posted by Marboy

Hi guys

From the german Fazer Forum

Looks like I will have to get my mother to translate that page and order the item for me. She is native German. I, on the other hand, had no clue what was being said on that page.

[Bill A]

Originally Posted by Bruce Bohannon

so..... is somebody local gonna build a pile of these things and have em ready to ship to us for some reasonable price?

I'm in for the right price.

I'd sure appreciate that, too.
I am unable to build one, but I'd like to buy one if someone is willing to build multiples

[Northstreet]

Several workable designs - pretty neat! Some pretty ingenious and varying implementations.
I like the isolation diodes and input resistors on these last two German versions. This would certainly provide some addedc protection for the motor drive circuitry on the FZ1 side of the connector. I'd also lean slightly toward the cap and resitstor version over the zener version - just slight personal preference. Perhaps this design with a simple small plug-in packaging simialr to the second design that had the FETs and chip?

Also any thoughts on reliabilty of the transitor based solution versus the FET and chip version?

Any thoughts on overall robustness in terms of long term circuit performance (dirfting values, component tolerance changes over their lifetime, etc)?

[Northstreet]

Originally Posted by patricjft

Looks like I will have to get my mother to translate that page and order the item for me. She is native German. I, on the other hand, had no clue what was being said on that page.

That's one of the great things about being a geek engineer - language doesn't have much to do with electrica schematics. For the most oart a circuit schematics are pretty international. Well - with the possible exception of some of the Russian schematics I come across now and then in my job. When you add their cyrillic alphabet and abreviations, I sometimes wish my grandmother could interpret!

[jewilson]

Yes the site is Greek to me; however the circuits are quite simple. Since my FZ1 has header, I’m tired off dragging this worthless servo and I’m not going to pay $100 for the circuit.

The two circuits on post number 9 are the same topology, however the top circuit uses an opamp and the bottom uses a single transistor to amplify the signal. The circuit is easy 2 hours maximum if. However, if the servo does not bug you leave well enough alone.

The transistor used in the bottom circuit can be 2n2222 or 2n4301. The capacitor in the circuit is the trick in integrate the signal to DC level.

Of course these would be really easy to buid with circuit boards. Having said that, most companies that build PCB want you to build a couple of hundred.

I would use Vishay Resistors for all the values, radial non polarized 220uf cap at 16VDC a cheap Panasonic should do fine and. As for longevity the circuit should last a long time.
The transistors are not as sensitive to ESD as the MOSFETs and integrated circuit and like I said before the mosfet can oscillate if a resistor not in series with the gate, which is the input.

[eskort]

The transistor used in the bottom circuit can be 2n2222 or 2n4301. The capacitor in the circuit is the trick in integrate the signal to DC level.

Please note that the BC 517 is a darlington transistor--this is necessary only for the second transistor to minimize dropoff of the capacitor voltage as the darlington "reads" it. You could use 2n2222 but you should use two of them piggybacked in darlington configuration.

radial non polarized 220uf cap at 16VDC

A polarized electrolytic will be fine also and much smaller. The rating could be dropped to 12v as it will only ever have 5v on it.

[jewilson]

Bob,

I did not believe it's necessary to use that Darlington transistor, a 30000 HFE part., However, driving that 220uf it going to need that current quick.

I selected 16VDC cap because the input voltage is 12 and some one could get care less and pop the 12VDC cap or a failure could blow it. Of course you could put a 5.6 volt zener across then you have some safety for a low voltage part. Also, a bleeder resistor say 1 meg could not hurt either. It not critical to use the non polarized parts it just adds another degree of linearity to the circuit, which is not that important.

Jim

[eskort]

I selected 16VDC cap because the input voltage is 12 and some one could get care less and pop the 12VDC cap or a failure could blow it.

Jim you are 100% correct. I failed to notice that in this circuit, the capacitor voltage can run up to 12v. Rather than use a zener, though, I would just use a cheap clamping diode from the capacitor positive to the 5v supply. This would keep the capacitor voltage no higher than 5.6 volts.

I did not believe it's necessary to use that Darlington transistor, a 30000 HFE part., However, driving that 220uf it going to need that current quick.

I'm assuming that you mean the first transistor, and agree it will not need to be a Darlington. The current charging rate will be limited by the 10K series resistor--R1 and the discharge rate by teh collector resistance R4--4.7K so the cgain of the first transistor is not critical at all.


I would not put a bleeder resistor on the capacitor because the capacitor is the analog "non-volatile memory". You want R1 and R4 to be the only charge/discharge paths, that's why a darlington is used for the second device, to minimize discharge of the capacitor.

[ancientone80]

And someone said the Russian Crylic was hard to understand !! ..

Keep up the good work though guys.