Ducati.ms - The Ultimate Ducati Forum banner

1 - 20 of 21 Posts

·
Registered
Joined
·
75 Posts
Discussion Starter #1
Thanks to the current economic environment, I find myself with some unexpected time on my hands. The financial manager has decreed spending must be curtailed so I thought I had nothing to lose by trying to fix the tachometer on the ST4S.

People have said that they can be opened and fixed but I couldn’t find any pictures etc after a search so thought I would rectify that situation. It was fiddly but well worth the time spent as (for now at least) it works well.

Symptoms were that the needle was sticking at 7k and not responding at tickover. The first tear down to check it over followed by tightening up the brass screws at the back made it worse as tickover was then showing as 4000 rpm so it was at that point that I thought I had better work out how it works….

First check is that the feed to the tach via the 3-pin connector is working properly. There are two bullet connector feeds for the immobiliser led but this is completely separate to the rev-counter function and only shares the common housing. The grey and black wire on the main connector for the tach is the earth whilst the light blue wire is +12V. These are obviously easy to check. The green wire feed is from the ECU and feeds a square wave signal to the tach electronics and so is harder to check. I managed to borrow an ancient oscilloscope to check the signal and it was a nice clean square wave with a frequency of about a third of the engine rpm (ie about 333 hz at tickover). As expected, the frequency increases proportionally as the engine is revved. Measuring the voltage at the centre pin with a multimeter set to DC gave me about 1 volt less than battery voltage with the ignition on and about 2.5V with the engine running (not changing with rpm) so that may help diagnose for those without oscilloscope.

Now confident the wiring was working, I set to with the Dremel and a cut-off disk and split the housing about 1” below the bezel. There is nothing to damage at that level. Removing the 3 brass sleeve nuts at the back then allow the circuit board to be separated from the needle movement. It’s a fairly basic circuit and a bit of research into the IC (CS289) reveals that it is an industry standard air-core tachometer drive circuit. A simple internet search revealed a data-sheet which allowed me to understand the basics. The circuit takes the signal from the incoming square wave and converts it to three DC feeds (corresponding to the 3 connections mediated via the brass sleeve nuts to the meter movement). There is a reference voltage feed Vz along with Vsine and Vcosine feeds. Basically, the meter movement has two sets of windings at 90 degrees apart connected together at the Vz feed. Varying the magnitude of the Vsine and Vcosine voltages forces the needle to a certain degree of rotation. Each meter winding should be about 200 ohms. In my case, a quick check showed only one working winding. Thus, I popped off the needle and unscrewed the dial to allow me to remove the meter movement. It then became clear that when I had tightened one of the brass screws, the whole mounting post had twisted and broke the very fine wire feed from the coil (hopefully clear from the picture). I managed to re-solder the connection and then used Plastech to reinforce the post to (hopefully) prevent it twisting on reassembly. I then reassembled and checked that all is well at the brass connections at the rear (200 ohms across two of the connectors and 400 ohms across the other). The difficulty is now putting the needle back in the right place! My first guess resulted in a working tacho but it was showing a tickover at 2,500 rpm. I suspect they apply a calibrating signal when they build them and then pop the needle on at the appropriate place. I resorted to letting the bike tickover at what was previously a reliable 1000 rpm and then popped the needle back so it read what it did before. Its probably not as accurate as the factory method but will be close enough for normal riding. Now happy it was working again, I used Plastech to ‘glue’ the housing back together again and finished it off with Duck tape. Once mounted back in the dash the repair is invisible. Not sure how long it will last but it kept me quiet for a few hours :)

The main message should be that be careful not to tighten up those brass screws too tight or you risk breaking a connection in the meter in the same way I did. They should only be nipped up. If tightening them doesn’t work then get a multimeter and check you have 200 ohms across two of the brass connectors and 400 ohms across the other. If that is OK then you have no need to dismantle the meter movement so check the circuit board for dry joints or fracture. It’s a pretty basic circuit so any electronics repair guru should be able to trouble shoot and fix it.

Hope that helps,

Keith
 

Attachments

·
Premium Member
Joined
·
5,642 Posts
Nice work! On the board I see what looks like a variable resistor. I wonder if that's a calibration? Many, myself included, have working but very inaccurate tachs. Maybe a signal generator with a fixed square wave at a frequency representing, say 5K RPM, then adjust for the proper indication on the tach. Did your research find a schematic for the actual tach? Now I’m interested… I don’t have a sig gen, but this could also be done on a running bike with something like VDST reading RPM off the ECU as a reference… OK. Getting ahead here! Don’t know if that VR is even a calibration point!
 

·
Registered
Joined
·
75 Posts
Discussion Starter #3
Hi Dan,

I didn't play with the variable resistor though assumed it was for calibration. If I had a signal generator I would have spent a bit of time feeding a square wave in and seeing how the needle responded. I might then have been brave enough to try altering the calibration and seeing what happened. I have not got a specific schematic for the circuit and didn't trace it in detail but the datasheet for the CS289 gives generic wiring information;

http://www.datasheetcatalog.org/datasheet/CherrySemiconductor/mXrqzqs.pdf

It should be possible to trace out the circuit that from the pics I took (which is mainly why I got the camera out) Figure 2 in the data-sheet clearly shows a potentiator for tweaking output/calibration. As my fault was in the meter and my electronics is very rusty (A-level physics was a very long time ago) I didn't spend the time to better understand the circuit.

Keith
 

·
Member
Joined
·
7,733 Posts
For what year(s) ST4s/ST's is this repair relevant? All pre 2004's?
 

·
Registered
Joined
·
75 Posts
Discussion Starter #5
I'm not sure what changes were made to the tachometer after the face-lift. Certainly, tachometers on the ST4S are all the same pre-facelift (and are shared with the 998). However, I would put money on the view that the basic circuitry and design will be pretty much the same as the majority of Ducatis of that era and before. I suspect the later ones are a much different design as (if I remember right) they go through a start up sequence so are probably stepper motors?

Kei
 

·
Member
Joined
·
7,733 Posts
I'm not sure what changes were made to the tachometer after the face-lift. Certainly, tachometers on the ST4S are all the same pre-facelift (and are shared with the 998). However, I would put money on the view that the basic circuitry and design will be pretty much the same as the majority of Ducatis of that era and before. I suspect the later ones are a much different design as (if I remember right) they go through a start up sequence so are probably stepper motors?

Kei
So, is it safe to say your post covers all ST tachs from 1998 to 2003?
 

·
Registered
Joined
·
75 Posts
Discussion Starter #7
I'm fairly certain. The earlier tachs don't have an immobilser light but, as said above, that is completely independent of the tach.

Keith
 

·
Registered
Joined
·
75 Posts
Discussion Starter #8
I've been playing with DucatiDiag 24 today while setting my TPS and doing my TB balance. I noticed this has a rev counter test facility which could be very handy for checking calibration. When its activated my tach indicates 3,900 rpm. I didn't run it before so have no comparison. Does anyone know what the test is supposed to give - I'm gessing 4,000 rpm?

Keith
 

·
Premium Member
Joined
·
416 Posts
An interesting read. I went through something similar last year for the tach on my 996.

The variable resistor is indeed for calibration of the needle, but it sets a "gain factor". Simple to recalibrate based on a know input frequency, but only if you've not moved the needle.

If you have removed the needle calibration is more tricky. The way I calibrated mine was to use a 12 volt reference sqaure wave input. I don't have fancy equipment, so I made my own signal generator using a 555 timer circuit, two resistors to make it 50/50 and a vr to set the frequency.

The input is actually 33.33hz / 1000 for the two cylinders. So I set to input to 33.33 and placed the needle on the dial at 1000. Then set the imput to 333hz and checked where the needle was driven to, if it wasn't 10,000 I trimmed the VR to make it 10,000 and then changed the input back to 33.33hz. If the needle then wasn't at 1000rpm, I carefully removed it and placed it back on at 1000, and repeated the whole process a number of times to dial it in.

I then used the VDSTS software to check the tach operation. I emailed their tech support and asked them specifics. To be exact, their "test" drives the tach to 3750RPM for 2 seconds. When I performed this test, the needle was pretty much right on, about 3800rpm. So I was happy with that.

Electronics is a hobby of mine, the reason I began "playing" with the tach was because of the "Bounce" we sometimes experience, mostly when system voltage drops when indicators or lights are activated. I added additional capacitors to the circiuit to smooth out the voltage drops. It helped but didnt completely resolve.

The CS289 chip has it's own shunt voltage regulator to provide the reference voltage. I have a feeling that the chip has "aged" and the on board regulator is not as good as it once was.

I'm looking for a new project, I would like to replace this circuit with a micro controller and stepper motor drive. I would like to use a "non working" tach though... Hmmm, maybe one day.
 

·
Premium Member
Joined
·
1,877 Posts
Does anyone have a south Georgia (USA) translation for this thread????
 

·
Registered
Joined
·
75 Posts
Discussion Starter #12
He he.

Its good to know that VDSTS feeds a 3750 rpm test signal as I suspect DucatDiag will be the same. Thus, it looks like I am only about 100 rpm out despite taking my needle off.

Keith
 

·
Premium Member
Joined
·
1,877 Posts
If it moves and shouldn't - Duct Tape
If it doesn't move and should - WD40

Tony, I save the hammer for those repairs that show two wrench's....
 

·
Premium Member
Joined
·
106 Posts
They are just trying to keep the blue smoke inside the wires.
 

·
Mr Leakered
Joined
·
8,820 Posts
They are just trying to keep the blue smoke inside the wires.
Geez, I didn't know that that tach was made by Lucas, the Prince of Darkness!

Getting red smoke means a new Pope, though. Right?

Have a good one.
 

·
Premium Member
Joined
·
106 Posts
Geez, I didn't know that that tach was made by Lucas, the Prince of Darkness!

Getting red smoke means a new Pope, though. Right?

Have a good one.
I can tell you have seen the smoke leaking out before!

Lucas produced vacuums for a time....they were the only thing they ever made that didn't suck!
 

·
Mr Leakered
Joined
·
8,820 Posts
Yeah. I'm a Lucas minion. But then, I think that hammers can fix electrical problems.

I do actually like warm beer. . . sometimes.

Have a good one.
 
1 - 20 of 21 Posts
Top