Fox T-Bird/Cougar Forums

The stock digital tach is apparently very inaccurate. I recently bought an Autometer tach/shift light so I can shift my car manually at the track (it has an epoxy mod AOD). I hooked it up and set the Pulse Per Revoluation setting to 4 (what Autometer says a single coil V8 should be set to). I knew from previously using a tach/dwell meter on my car that the stock tach is around 200-300rpm off at idle. Now you can see it in the car.

(http://i198.photobucket.com/albums/aa33/thunderjet302/Thunderbird%20web/20150307_1650481_zpsapg3jrjz.jpg) (http://s198.photobucket.com/user/thunderjet302/media/Thunderbird%20web/20150307_1650481_zpsapg3jrjz.jpg.html)

It gets worse at higher RPM. According to the stock tach the car's cruise rpm at 70mph is 3000rpm. It's actually 2300-2400rpm. The yellow shift light is on as I set the shift light to come on at 2250rpm for testing.

(http://i198.photobucket.com/albums/aa33/thunderjet302/Thunderbird%20web/20150307_1623041_zps0sszaoex.jpg) (http://s198.photobucket.com/user/thunderjet302/media/Thunderbird%20web/20150307_1623041_zps0sszaoex.jpg.html)

It gets worse. At 4000rpm on the stock digital tach the Autometer tach is displaying 3400rpm. Above idle I observed the stock digital tach reading 500-700rpm high. The Autometer tach is being fed a signal from the ignition coil wire under the hood so it's getting a good signal. So the moral of the story is the stock digital tach is way off. Don't use it for any kind of performance driving.

I'm not really suprised, stock gauges basically only give you a general idea of where it's at. although, my speedo is dead-on with my phone's speedo app.

so what I have learned here is .........
when I go and buy any new duplicate part to compare to my car, if the new thing say something different, my car must be wrong.

are you running the stock size tire???? 
wheels?

this isn't any way to troubleshoot, you gotta get more than just one sample.
maybe its time for a dealership calibration inspection,,

Well the speedometer has nothing to do with tachometer calibration. I just revved the engine to 3000rpm (what the stock tach shows when the car is traveling 70mph) and the autometer tach reads 2300rpm. I'm running stock size tires and the speedometer driven gear matches the rear end gears so it's fairly accurate. A rpm calculator shows 2400rpm at 70mph with my setup. The Autometer tach is pretty dead nuts on with the calculator, the stock tach not so much. I've run two different tach/dwell meters on my car to check base idle. Both have said 700rpm in park, just like the Autometer tach. The stock tach has always read 1000rpm at that same idle speed. Checking the stock tach against the Autometer unit the stock tach reads around 25% faster than actual engine speed. I'm pretty sure that it's just py calibration by Ford. Stock fox body Mustang tachometers have been know to be up to 25+% off. I'm betting our cars are probably the same.

Also note my car is not moving during this test.

I think the inaccuracy engine rpm displayed has nothing to do with tire size or rear gears.    I always figured the digital tach wasnt totally spot on, but it sure looks cool......

I'm wondering if Ford didn't calibrate the digital tachometer for the V6/V8 but instead just used something that worked "close enough" for both engines. That would save them the trouble of calibrating the tachometer for each different engine. The factory tachometer in my car is between 20-25% off at different rpm. I wonder how far the tachometer is off on V6 cars with the full digital dash....

Remember what I've always said: everything concerning our era digital gauges are representative only. Nothing was accurate. buttstuffog wasn't much better. But it was closer than digital.

One bar on the tach represents...what? 250 rpms, maybe? But that is wrong, as proven above. It makes for an interesting light show but it is not a true picture of the actual rpms.

The closest we have to an accurate digital gauge is the speedo. Everything else, take it with a grain of salt. ;)

I had a 90 LS, which had the full digital, mine would be around 18-1900 rpm @ 65mph.

True. Each bar represents 200rpm. So when the bars are lit up to 3000rpm it could be any engine speed between 2800-3000rpm. I wish it was slightly more accurate. Then I wouldn't have to run the aftermarket tach/shift light.

as much as I like these cars more than the MN-12, I will say this, the ONE thing that I feel was better was the layout of the electronic cluster in them, oh, and the HO 5.0 :-D the 85-88 just has a kooky layout that I've never really liked.

Also, I forgot about the one that could be had in the Aerostar, looks a lot like the Fox one, doesn't it?

The Aerostar tach is close to a Fox tach. I don't think it has a multi gauge function though.

I contacted Autometer about tach accuracy just out of curiosity. The tach I'm running has an accuracy of +/- 50rpm. Apparently Ford's factor tolerance is +500rpm or so.

As a note I checked the AutoMeter tach against the factory tach across several rpm points. The factory tach consistently reads about 500rpm higher than the AutoMeter tach. However if I switch the pulse per revolution setting on the AutoMeter tach to 3 (the correct setting for a V8 is 4, which is what I have my tach set to) it perfectly matches the factory tach. 3 pulses per revolution is the correct setting for a V6. Now I have receipts for repair work on my car dating back to when it was new so I'm doubtful the instrument cluster has been swapped to a V6 one at some point (according to the receipts the instrument cluster has never been out of the car). I've come up with three options as to why the factory tach in my car is so far off:

1. The tach in my car has gone out of calibration over time.
2. Ford didn't calibrate the tach well at all. It's the same one used in both V6 and V8 cars so they split the difference and called it good.
3. My car accidentally had a V6 tach installed at the factory.

My bet is on your number 2 thought.

That's my guess. The guy who designed it probably didn't think some "idiot" was going to be drag racing the car almost 30 years later. You know when you need an accurate tachometer for shifting :burnout:.

I must have a freak then.  Mine's always been pretty spot on.  My speedo gear is correct for my rear gear ratio/tire size combo and it shows what it should.  I've verified the accuracy with a timing light as well.  Most anyone I know who knows these cars that I've shown how accurate it is are shocked.

The left panel/tach looks spot on save for some color variances.  I wonder if the center would work as well or if the pinouts on the ribbon in back are not the same....

You're pretty lucky then. I've had reports from so many people about the inaccuracy of the digital gauges that it's pretty much a given that they're wrong most of the time LOL.

Lou, you said your DASH tach is telling you 3000 rpm at 70 mph?
Whether I'm running 225-60s or 205-70s (have had both on the rear while on the highway at some point in time), I see 2600-2800 rpm at 80 mph, 3.73 gears, 3000 rpm non-lockup converter.

The math tells me this is what I should be getting too. Interesting stuff. Maybe you do have a V6 tach!

We actually around 2800rpm :hick:. Yeah the stock tach is around 500-600rpm off above idle. With 225/55/16 tires and a non-lockup converter the engine is actually spinning around 2300rpm at 70mph, which the Autometer tach confirms. At idle (700rpm in park) the stock tach shows between 800-1000 rpm. The same thing is displayed in drive at idle (real reading 650rpm). I knew the stock tach was off at idle as I confirmed it with two different tach/dwell meters. The Automater tach I installed has an accuracy of +/-50rpm. I couldn't wrap my brain around how off the stock tach was until I installed the Autometer tach. 2000rpm stock tach, 1500 on the Autometer. 3000rpm stock tach, 2400 on the Autometer. 4000rpm stock tach, 3400 on the Autometer.

The stock tach has been costing me at the track. I have an epoxy mod valve body in my Thunderbird so when I drag race I shift it manually 1-2-3. I used to upshift at what I thought was 5500rpm. Turns out I was up shifting around 4800rpm, which is defiantly costing the car some trap speed at et.

Yeah if you're looking to get everything you can at the track, good instrumentation is a must.

Man oh man - I saw this thread and decided to dig into this. I used the electrical diagram book to find the I-O's for power and the tach signal and hooked up the module in my lab which I had removed from the spare electronic cluster I have. I used a signal generator to create a square wave and used my oscilloscope to measure the frequency of the signal into the tach. Apparently Ford thought keeping it simple was better than accurate? The thing is very precise, but it's (deliberately) off by 25%!  :nono: It's all very simple - 50Hz = 1000RPM on tach display, 100Hz = 2000 RPM 150Hz = 3000 RPM and so on. The math is simple:

50Hz * 60 (convert to cycles/min)/4 (power strokes/revolution) = 750 actual RPM - which agrees with the original post.

100Hz (2k RPM on tach) = 100*15 or 1500 RPM

150Hz (3k) = 150*15 or 2250 RPM

I will be poking around this spare tach (electronics shown below) to see if I can find a way to trim this thing so it's more accurate.


(https://i.imgur.com/1KTM4Y7.jpg)

I completely forgot about this thread.

If you come up with something let us know. I'd love to mod the stock tach so it worked accurately, and ditch the Auto Meter one.

It's the V6 tach used also in Aerostar. They don't reconfigure anything when they decided to use it in our cars. That's why everyone finds it off 25% use one of these to correct it. https://www.dakotadigital.com/index.cfm/page/ptype=product/product_id=127/category_id=694/mode=prod/prd127.htm

But if it can be fixed for a couple bucks of parts soldered to the stock tach I'd go that route. I can solder to circuit boards easily enough, and I'm cheap  :mullet: .

Hmmm, I may have a solution if the microprocessor will cooperate. The silver can in the above photo is a crystal oscillator. Sets up the timing for everything, including the sampling of the pulses from the negative of the coil (or whatever tach input you're using.) It's marked 4 MHz. I think it counts the number of clock cycles between spark plug fires, the less pulses counted (higher RPM as plugs are firing more frequently) the more bars you get on the display. I needed to think of it the other way, the more clock cycles counted between spark events, the less bars you'll see. So if we have more clock cycles for the same engine speed, it should mean a higher clock speed should show less bars at the same RPM, which is what we want. So, my guess was that if we increase the frequency of the clock to 5.333 MHz (it reads 250RPM high @750RPM = 250/750 or 33% error), it would then read accurate. I wanted to verify this before ordering a crystal. I have some crystals from the "crystal upgrade" thing that I was going to do to my A9L, so I tried sticking one in there. Went crazy, 18MHz must be beyond what it can sync to. I then realized that my function generator could run to above 2 MHz. I maxed it out at 2.16 MHz and then tapped it into the crystal location. I then realized I had no way to simulate the pulses from the ignition coil  :frown:  Wait! I have this old tube audio frequency generator from like the 1960's! Now I'm definitely looking like a mad scientist.



I was hoping the input for the tach was robust enough to handle a 50 Hz AC sine wave. I tested it with the original 4 MHz crystal and it worked. Sweet. I then removed the OEM crystal and injected my 2.16 MHz signal from my function generator. It powered up fine, and then when I connected the 50 Hz signal, I saw 9 bars on the tach!  :banana:  I sat down at my laptop and did the math - it should have read 9.2 bars, perfect. So, assuming the micro can sync up to the higher 5.355 MHz (closest frequency available) crystal that I'm ordering, it should read pretty much on the money.

The last experiment I did was to confirm that it reads continuously up to 5000 RPM (3rd yellow bar) with anything 5200 RPM or higher being the red bars all lit. In our cars presently, the red bars all illuminate at anything higher than about 4000 RPM.  :grinno:

I'll post back with results using the new oscillator crystal probably within a week since Mouser has both crystals that I am going to order (5 Mhz & 5.355 Mhz) in stock.

Wow!!!  That is some impressive work right there.

Nice job.  If it works Ill send you my tach.

Let us know how it works out when the crystal arrives. You've really gone above and beyond here.

Parts will be here Thursday. Unfortunately I won't due to a death in the family. I should be back by the weekend, however.

My condolences. No rush at all. My Thunderbird is put away for the season. Got 2" of snow here on Halloween and 3" yesterday. Car isn't moving out of the garage for awhile. 

Well, I finally got home Monday night and my first chance to try to new oscillator was last night.

It worked as expected. The only issue I saw is now that it's sampling up to a higher frequency (5k RPM is now 333Hz vs 250Hz), it had some issues at higher RPM when the (positive) duty cycle was less than 20%. I couldn't really measure the actual duty cycle at the coil (engine running) with the cheap portable oscilloscopes I have, I'll just have to try it with the dash hooked up to the signal from the neg of the coil and see if it lights smoothly to redline. Will update assuming the weather cooperates enough to let me try this out in the next few days.

Cool. Keep us posted.

 :smile:

Yay. It worked like a charm, see the attached pic. You'll have to zoom in but it's tracking perfectly with the Holley EFI dash. I free-revved it to the yellow and it was nice & smooth. Can't drive the car tonight because it's wet out there and I'd kill myself, but I was more than pleased to know my tach is calibrated now.  :headbang:  :roxor:  :headbang:

That's awesome! It's accurate all the way to 5000 RPM correct (then the tach just turns red)?

Can you list the parts you used and a write up so I can duplicate this with my tach?

Sure, I bought this 5.355 MHz crystal from Mouser (https://www.mouser.com/ProductDetail/449-LFXTAL011100BULK) and replaced the one on the PCB (silver can shown in the picture below.) That's it.  :grinno:

Disassembly.

I'm guessing some know how to remove the tach module from the dash, but for those that don't:

Remove the clear plastic face from the dash (6 screws.) Flip the dash over and remove the 3 screws which hold the tach module to the dash, one is in the middle center (behind the tach) and two are at the top on either side of the illumination bubs for the tach. Next remove the screw holding the ribbon cable to the dash, then carefully separate the ribbon cable from the dash which is held in place by two white friction-fit posts. Now flip the dash again. Remove the Lamp Out and Check Engine light sockets (twist type) from the lower left face of the black bezel, then pull the ribbon cable off the two black friction posts. Lift the top of the bezel off the 3 white posts and note that there's a rubber pad at the bottom that fits over the steering column. This pad ties the top black bezel to the white dash body and can be rather annoying to separate from the bezel as just when you think it's going to go, you note there's a tongue on the bezel that interlocks with a slot in the dash body and this dumb rubber pad is stuck to that tongue. I probably should have just used a razor to make clean cuts to separate the two, but one way or the other you need to un-stick this very sticky rubber pad that likes to tear apart under stress.  :beatyoass: Once you can fully lift the bezel, you'll note that there are two multi-wire cables that connect to either side of the dash from the bezel. You'll need to disconnect the 5 pin one (only 4 wires) that connects to the tach (left side.) It's taped so you'll need to cut or remove the tape first. Once that cable is unplugged, fold the bezel back and grab the tach module and lift it. Note the ribbon cable will require you to 'slide right' to remove the tach module cleanly. Fold the bezel back onto the dash and put it somewhere safe.

Disassembling tach module and replacing crystal:

Firstly the back part of the plastic clam shell must be removed from the tach module.  This part stressed me the **** out and I'm assuming it'll be the same for others. There are two tabs that hold the (probably yellowing) back of the tach to the rest of it. Both tabs must be pushed in to then rotate the back piece out of the way. I used a screwdriver to push one side in (easy), then try not to pull it out to far while working on the other side. It takes patience and you may wonder if you're going to break the plastic tab off, but it'll even eventually pop off. Once the back is off you'll need to carefully slide the blue connectors on the ends of the kapton (brown/orange) ribbon cables from the right angle connectors. There's two per side. Try to slide them off evenly so the pins don't get bent. Once all 4 are removed, use your thumb and forefinger to hold the ribbon cables out of the way while you lift and remove the stacked printed circuit board assembly from the tach. Turn it over and you see this:

(https://i.imgur.com/1KTM4Y7.jpg)

The final step before desoldering and replacing the crystal (oval shaped shiny metal can) is to separate the two printed circuit boards (PCBs) that are held together with little plastic stand-offs. Carefully wedge something between the PCBs where no components are and pop the boards apart. Start at the end away from the little white ribbon cable that connects the two boards together. The boards can then be "unfolded" being careful not to damage the ribbon cable connecting them.  Now you have access to both sides of the PCB so you can desolder and remove the old crystal, then solder the new one back in.

Reassembly is much easier, just be careful when finagling the ribbon cables back into place and don't forget to plug the 5-pin cable back together also. Might be a good time to clean up the sockets and replace some of the finicky dash bulbs. I recommend getting a can of CRC QD Electronic Cleaner and using it on all bare copper/tinned connections.

Oh, and try doing a post like this on short attention span Facebook  :giggle:

I did buy 10 crystals just in case I'm tasked with helping out someone who isn't comfortable doing this (for a fee, it's a lot of work to do as a favor.)

Wow, this is so helpful! Definitely deserves a write-up on coolcats.net. Just for the sake of dospoogeentation. I have a spare (semi-broken) cluster that I may need to try this out before I go messing with the one in the car.

So, I made a little video of the modified tach in action. I even free-revved it to >5k RPM so y'all could see. Seeing as it reads accurate now, I had to really stab it hard to get it to kiss 5k RPM.

https://youtu.be/PWwLk0xy6g8

Thank you for going through the trouble of writing all this up and testing everything.  :bowdown: You've gone above and beyond.

I'm going to do this modification in the spring when the car comes back out. It's cheap and pretty easy to boot. It's too bad that the tach still just goes solid red above 5000 rpm. If that could be fixed it would be perfect. I couldn't even begin figuring out how to change that.

Wow. Great work, above & beyond for sure.
It always amazes me the amount of knowledge on this forum. Thank you sir.