Gauge Magic – Fuel, Temperature & Pressure Gauge Matching / Calibration / Linearisation

[Benton]

Thanks for the update, take all the time you need to get it sorted and above all enjoy your break! 

[revilla]

Design finished (I think!):

Switched to a MOSFET output driver for better saturation characteristics. The coil has a maximum DC resistance of 2.5 Ohms. The MOSFET has a maximum ON state resistance of 5.3 Ohms. So the output can in theory simulate a sender resistance as low as 7.8 Ohms (my fuel gauge sender only goes down to 21.2 Ohms so it will easily drive the gauge hard over).

It all fits on the original size board (after extensive rearrangement). The switch to a MOSFET saved a couple of bias resistors. The specified inductor should fit neatly between the boards in the position shown. Note that the layout of parts on the board has to fit around chips on the UNDERSIDE of the Arduino board, so placement is critical. That specified inductor is 1mH inductance with only 2.5R of DC resistance and cam take 280mA DC. I wanted a 100uF capacitor but that really meant going to a tantalum component and they were just too tight to fit. Finding large multilayer ceramic capacitors that would fit and are economical isn't easy, but the ones shown are 47uF each at 25V so I used two in parallel to get 94uF.

The new firmware for the device is finished (the board rearrangement meant some of the Arduino pin assignments were swapped around and all of the frequency configuration has been stripped out, as it runs permanently at 15.625 kHz now to maximise the filter effectiveness).

The new Windows application is also finished. The only real difference there is the removal of the frequency controls.

And here's a video of the prototype actually driving a microprocessor controlled electronic gauge that misbehaved badly on the old PWM output, which had it hunting around apparently at random. Now it is responding predictably. It is also easily driven over the full range.

The prototype has the new filter (off-board) and a 2N7000 MOSFET switch just to prove the concept.

And here's the wiring diagram for use with the newer Caterham gauges:

I'll be able to supply it ready-assembled with the connectors as shown above, ready to plug and play between the loom and the gauge. There's one additional wire as the later gauges have a lighting feed on the same connector where the earlier gauges had a separate bulb holder.

[revilla]

OK back from holiday and the first Mk II unit is built to the new design:

Everything fits on the board just as planned. The prototype has the what I believe to be the correct connectors for the Caterham gauges. Mk II units are colour coded with blue heatshrink, Mk I units were black.

It's working for me, with both "old" (plain moving coil) and "new" (electronic microprocessors controlled) gauges.

The hardest part was working out how to successfully crimp the tiny Micro Fit 3 terminals. In the end I had to use a good Japanese Hozan two-step crimper with 1.4mm jaws for the conductor, 1.7mm jaws for the insulator, and it's critical to strip exactly 1.5mm of insulation - any less and the crimp doesn't hold, any more and he terminals won't lock into the housings. Very fiddly! But once I'd got the hang of it and got my eye in, all doable.

So ... Before I start mailing them out in bulk, would anyone who bought a Mk I from me and who has the "new" gauges be prepared to volunteer to test one for me first quickly?

To be honest it doesn't need all of the time consuming emptying the fuel tank and calibrating the sender doing, as none of that side has changed. Testing basically means:

• Unplugging the cable from your gauge.
• Plugging this unit into the loom and gauge and making sure I've got the correct connectors.
• Connecting to a laptop, installing the configuration utility and checking that it connects.
• Dragging the gauge slider up and down to make sure that you can drive the gauge smoothly to both ends of the scale, that it gives a steady and predictable reading etc.
• Making sure the gauge illuminates with the sidelights as normal.

If all of that works, they're good to go.

 

[AlastairA]

Hi Andrew,

I have the MkI design and would be very happy to test the Mk2 version on my fuel gauge (new version). Apologies again that I can't currently remove the gauge and send it to you for testing, but very happy to do whatever tests are necessary for you.

Alastair

[revilla]

Thanks Alastair, I'll get the prototype device and some instructions off in thepost to you!

[DerekH]

My gauge magic board and LED is now fitted calibrated and tested.

My SV tank takes 38 usable litres having emptied and filled it during the calibration process.

Tank got down to halfway on Saturday and took 18 litres to refill.  Gauge at half full is set at 20 litres. Pleased with that.

Started on Sunday with a half tank and ran it down to 12 litres, to flashing orange warning. LED

Got to 10 litres and flashing red LED, then 6 litre warning with the solid red LED. 

Filled with 28 litres, so maybe some more margin but knowing what the gauge is reading is much more helpful than just a vague indication.

Maybe I could have got more fuel in but covering your shoes in petrol is never nice.

Well done again Andrew.

DerekH

[BigCol]

long time listener, first time caller on this thread - am in absolute admiration of not only the solution for a problem most of us have but also the attention to detail  

Chapeau Sir!  (and that was my opinion before reading the VVC thread!  Awesome work!)

re #53...  revilla said it's critical to strip exactly 1.5mm of insulation - any less and the crimp doesn't hold, any more and he terminals won't lock into the housings. Very fiddly!

I have been contemplating buying the DIY version (as am confident with a soldering iron!)  Would the very fiddly bit mentioned be relevant?    

If so, I'll spank a few more quid on the pre-assembled version!

 

[revilla]

The very fiddly crimping is relevant to anyone who has the new gauges, as it's the terminals for the connectors used by the new gauges in place of spade connectors which are difficult. For those you will need the MkII device at least. Because if the need for the right tools for crimping I wasn't planning to offer that as a kit. The other issue I found with the kits was it took me more time to carefully count out all the parts, put them in labelled plastic bags, double-check everything etc. than it did for me to just solder them onto the board! 

AlastairA has kindly done some testing of the MkII module for me. To be perfectly honest, he thought it was OK but there were a couple of things I'm not 100% happy with myself! I will write it up on here as soon as I get the time, my Dad's not been well and I've been a bit tied up with hospital transport over the last few days.

I'm still on it!

[Benton]

revilla - Could you please respond to the messages I've sent via Blatchat and eBay regarding my order.

Thank you.

[revilla]

I've replied to your latest BM. Included my email address. I'll keep an eye out for your reply. 

[Benton]

As suggested I have replied to you via your email address. Cheers

[Gridgway]

As I am fitting a new tank with a new fuel level sender, I thought it would be good to get it accurate.  Are you producing any more kits please Andrew?  Thanks!

[BigCol]

@revilla, is the Gauge Magic available again (link to eBay page still draws a blank)…?

Would love a unit for this style if fuel gauge: