RE / Impossible to find datasheets

[gijim]

Okay, so I have this Mitubishi/Dodge tachometer with a Denso/Nippondenso SE094 IC, which is an air coil driver in a "similar" style as LM1819. However, the components are different, and I can't find anything close to the charge pump circuit that is used there.

I need to find a datasheet for this chip, or find a similar chip/app note, so I can convert the tach to read a V8.

I've looked at:
SA5775
CS4192
LM1819
IL8190

and none have any similar circuitry. I'm expecting to find a combination of resistor(s)/capacitor that adjusts the deflection vs frequency in.

I'm just about ready to ditch the stock board and make a new one.

 

[geekEE]

You put a V8 in a Mitsubishi? Nice!

 

[gijim]

No, my car's just a hodgepodge of parts.

Pontiac Fiero
4.6L DOHC V8 from Cadillac
11.25" rotors from Chrysler Lebaron
Calipers from a Chevy S-10
Mitsubishi 3000GT Dash w/ gauges
Tail lights made from the reverse lights on a Corvette

 

[TurboXS]

Perhaps you build a divide by 2 circuit for the input. Assuming the original circuit was for a 4 cylinder.

 

[VEBill]

Simply 'recalibrate' the resistor that is in series with the meter. If the original calibration was for a 4-banger and you've now got a V8, then you'd ROUGHLY double the resistor (actually total resistance including the coil) to drop the needle by half (compensating for twice the spark rate). Trivial.

If you can't find a series resistor, then try other resistors until you find one that works. It's mostly an analog circuit and you're just trying to move the needle. Nothing more than Ohms law really.

Does your 'Cadillac-ak-ak-ak' engine do the 4/6/8 cylinder disabling? Any impact on the tach?

 

[Comcokid]

I used to design this kind of product. I doubt I still have a document, but I did find a datasheet online. Check out the Cherry Semiconductor CS289. Funny - I didn't realize that OnSemi acquired them. Air-core (bi-torque, or sin/cos) meter movements are the general meter movement type used by US vehicle manufacturers.

It's just a matter of reselecting a resistor. It also possible a simple R/C input filter may need to be opened up slightly to keep from removing the high-end of the tach range. You could also find a Marine tach - they have a selector switch on the back for the #cylinders.

http://www.onsemi.com/pub/Collateral/CS289-D.PDF

 

[gijim]

The wierd thing is that resistor doesn't exist here! On all of the air core driver datasheets I have seen, Rt is always over 100k, typically 300+k. In the past, I have found that resistor and changed it with a pot, and was able to recalibrate it for anything. Worst case scenario, I was stuck having to replace the Ct...

In this case, if you replace what appears to be the correct cap, it makes the gauge movement non-linear.

So best I can figure, this is a REALLY wierd driver.

 

[VEBill]

They (ND) wouldn't have designed a "4-banger specific" tachometer chip. There has to be a single, cheap passive component to calibrate. If you can't find the resistor, then look again for smaller, hidden resistors.

It shouldn't be the capacitor.

If really no resistor, then simply add one (equal value to the meter coil, in series with the meter coil). This should knock the meter back by half.

Depending on the circuit, maybe in parallel with the coil.

 

[Comcokid]

Hummmm - its made by Denso/Nippondenso. Over ten years ago when I was last involved with this kind of thing, there were almost no Far East companies using the air-core/charge pump driver approach. There were some air-core, but driven PWM by microprocessors or being driven by mysterious multi-component circuits. It hard to guess what's up.

Now, does the movement automatically move back to zero RPM with no power? If so, it may be magnetically biased to return-to-zero. This creates the need for a non-linear drive method, or a non-linear dial scale.

If your dial is printed linear, maybe you can get the movement and circuit board from another tach, and move the pointer and dial over to it. Of course this will be difficult because the dial mounting holes and shaft diamenter would be different. It'll take a bit of work.

 

[gijim]

It does move to 0, but it has a spring return, so that is probably not a good indicator!

So I've accepted the fact that I will likely have to build a new circuit, this time with a CS4121, which is similar in operation to the other air core drivers. So now my question is, how do I figure out which is the Sine and Cosine coils, and which end is + and -? I assume applying power to one will make it flip to 0 degrees, and the other 90 degrees. I'd make an educated guess the Sine coil is the one that will cause it to point to 90 degrees/vertical.

If the cosine is the 0 degrees, which polarity would that be?

 

[VEBill]

Sorry, I didn't catch the importance of the 'air core' meter movement. All the aftermarket tachs that I've seen used plain old two-lead meter movements and my comments were directed at that type, not the 'air-core' type.

 

[Comcokid]

OK. Spring return will most likely change linearity, so the circuit probably is 'reversed' in it's nonlinearity. Just a guess.

Building a new circuit with the same spring-return air core may not work. You could end up with a dial that needs new markings. That's why I suggested transplanting the dial (if linear in degrees) and the pointer to a new movement and circuit board.

But first, you may want to experiment more with trying to rescale the circuit you have.

Connecting sin/cos coils is no problem - one winding goes to the sin output, one to the cos output If the tach moves backwards, swap the connections on a coil. Power the tach up with the engine off, and install the pointer at zero.

Of course, another method of using what you have would be to build a frequency-to-voltage converter feeding a voltage-to-frequency, and use this circuit to rescale your tach. You could take the all-analog approach, or use a microprocessor. Again, a bit of a task. I bet if you do a search, you might find a after-market module that accomplishes this function. Industrial engines, transmissions, and axles are frequently changed around, and some manufacturers make various adapters. Someone has to make an adapter for this purpose.

VE1Bll - Don't be embarrsed. I recognize your handle from the many RF, antenna, and communications forums. This is a pretty small nich area and there are a lot of D’Arsonval type gauges out there. The air core type is simple, inexpensive, requires almost no precision manufacturing to work, is very accurate, can be driven through 360+ degrees with the right circuit, and extremely rugged.

 

[MacGyverS2000]

For all of the trouble, I'm thinking this is a 6-pin PIC job

Dan - Owner

http://www.Hi-TecDesigns.com

 

[gijim]

There are two coils, 4 connections. Sin +/-, Cos +/-... When building a board, it's difficult to swap traces after the fact