I am finishing up my 68 xr7 and although no longer a stickler for day one correctness (electric headlight doors, color change, carbon fiber wrapped dash, deleted side markers), I do like for all the features to work and in this car it’s mainly the convenience lights in the upper console.
Door, brake and belts are easy but after two new thermistors I thought there had to be a better way. I knew there was an electronic version that was pricey but reviews were split so I looked into adapting an aftermarket low fuel warning to the existing hardware. They were almost always available in Europe (eBay.eu) but they all had a “comparator module”. Enter Amazon and a KNARCO comparator module. 12 v 2 channel…$16 shipped. This device, when on, will compare the voltage to the fuel gauge to your selected value and turn off or on a relay. The voltage limiter on the dash has to be electronic to provide a constant 5 volts to the sender but you can search building your own on google and the parts are less than 2 dollars (technically you have to buy 10 of each of the 2 components to get that price so if you are an accountant add $11 which blows the $20 mark but you only need 1 of one and 2 of the other…you get the point. I played with the concept on the bench for a while and it worked. I wired it up and with an 1/8 th tank I adjusted the circuit to just come on and ground the green wire to the low fuel relay thus turning on the light. After adding fuel the module ungrounded the low fuel relay…light off. I placed the module and wiring in a small black case (not included in the $20 but $8 from Amazon) and connected in to the harness in the trunk using the passenger drop off as it’s home (much like the sequential unit on the other side) and got power from the rear defroster feed to the trunk (unplugging the connection near the glove box and running switched power from an accessory plug). Now there may be some on off issues when the threshold is reached but I hope the sluggishness of the sender will reduce that.
Sorry to be so long and in the Cougarverse the low fuel light may not be that big of an issue but after 2 thermistors I tried something else and it appears to work…thanks
There are versions available that work with any fuel sender. I think I paid $40 about 6-7 yrs ago. Very simple to install.
That is very creative. Let us know how it works. I suspect that you are going to see a lot of flashing from the light if you can’t put some sort of hysteresis into the circuit.
I don’t know if you were aware but I rebuild these senders including replacing the thermistor and calibrating them to show correct fuel level. It’s a lot more than $20 but it fixes a number of other problems, namely being able to actually tell how much gas is in the tank.
What happens is that over time the gauge will always read low, progressively getting worse. So even with a full tank you get about a 3/4 reading and it shows empty when you still have 1/4 tank of gas. The after market senders do the same thing but it only takes a few months for them to start reading low. I think that you can work around this to get the low fuel light to come on at the right level.
The factory system uses an instrument voltage regulator that pulses 12 volts creating an average of 5 volts over time. You know this already, that is why you had to build the solid state regulator to get constant 5 volt power. However the way the gauges work is that they have bimetallic strips wrapped with a heating wire. What is important in making heat is power, so the pulses of power are turned into to more or less constant heat. Here is the important part. The thing that keeps the fuel gauge needle from swinging rapidly back and forth is the amount of time it takes to heat that bimetallic strip causing it to bend and then drive the needle. The sender is not sluggish in any way it is constantly moving, constantly changing its resistance value. All of the damping comes from the two bimetallic devices.
The thermistors that are being offered as replacements are not the correct value for the Cougar and unless you have a really strong charging voltage and extremely good electrical connections in the circuit to cut down voltage loss, you will probably never get one to light up the low fuel light. They sent samples to me for testing and I have advised them on a correct part. I believe they will have one on the market soon that does work, or alt least as soon as they sell through what they have in stock. The one they have does work in the 64 through 67 Thunderbird. They incorrectly thought that they were all the same.
I actually disassemble the stock thermistor housing, clean everything up and then install an improved thermistor that can handle modern solid state voltage regulators with out burning up. Another very common problem with the aftermarket thermistors is that you have to solder the housing to the sender bracket. This takes a lot of heat and the thermistor is easy to over heat and destroy. The individual component is even easier to kill. If you don’t solder it in, the resistance between the housing and the bracket makes it even less likely that it will ever light up.
Thanks xr7g428…i see your point about the constant 5v but do you think the longevity of the gauge will be an issue. I rebuilt this car from a shell and a ton of parts and an original sender wasn’t one of them. The sender I used was backup repop from a 72 cougar project from 7 years ago…other than the angle of the exit tube it appeared the same and I adapted the fuel line to it, I read of your service and would have preferred that option however without a suitable core I knew I was trapped in thermistor wasteland. I had assumed the old style voltage limiters were all they had at the time and hadn’t considered the ramifications of a constant 5v.
Now I’ve got to look up “hysteresis”…thanks for your time and explanation
Michael
Very interesting but Bill’s got a good point on the sender fluctuation. I guess I’m still waiting for a proper thermistor. Unless a thermal (bimetallic) switch was used on the relay output is used to damp it. Or use an RC circuit with a voltage divider from the first relay feeding the second voltage comparison circuit on the board.
Thanks
If ACP doesn’t get around to selling a working thermistor I will do that. I will have to tool up the little brass housing and the insulator at the end. I still have to reconcile making it possible for people to attach it without having to solder it. I have a very old very powerful soldering gun that I use and I have ground the tip to the right shape to efficiently make that connection and I still check every single one to see if it got overheated.
Can you solder a lead or conductor trace to the case before you put the thermistor into it? Then we could solder the pickup tube or bracket to the conductor instead of soldering to the case.
On the first part, yes there would be lead coming out to avoid consumer soldering directly to the thermistor. It is the second part that I am thinking through. The brass housing that contains the thermistor has to be soldered to the bracket that holds it. That bracket and the housing have quite a bit of mass and you have to heat that up to get the solder to melt. The key is to get the immediate area very hot before the heat can travel very far. It also benefits from the use of flux to clean both surfaces. So my thinking right now is make a clamp that would attach the thermistor housing to the main tube of the sender. Basically it would provide a very solid mechanical connection. I think this can work but I have not tested the design.
Update on low fuel light fix…xr7g428 …thanks for the hysteresis heads up and as you predicted…it would occur during a slosh and when the low level was being reached I would get the on off on off…so an additional of a Akozon delay relay which would reset each time the comparator module closed the ground has seemed to solve that issue. Basically the first module grounds the relay which doesn’t close for 10 seconds…if during this time the relay loses power (on off on off) the timer resets. The delay interval is variable up to 60 seconds but 10 seems to work fine. So now the $20 fix is up to $35 but the challenge was priceless.
I’ve been gathering parts to try an RC (resistor/capacitor) circuit on second voltage switch on the voltage switch board. Charge the capacitor slowly and switch the second relay and the low fuel light on the voltage increase the time delay. The board you’ve used may be simpler than messing around with all that. I has not yet measured the voltage switch board input impedance, if too low it could make charging the cap problematic. Just ordered the second board.
Thanks for sharing.
Note that these modules are not designed for use in a 1960s automobile. They do not have any transient voltage spike protection.
Our cats generate voltage spikes that will randomly (seemingly) destroy unprotected electronic devices.
Take a look at the SAE J1113 or ISO 7637-2 load dump specifications for example. Your circuit must be protected against up to 90V and -50V spikes that are generated by the alternator and other relays in the car. Modern alternators have some built-in protection but not our alternators.
I built a high voltage “Zapper” to test my products. It dumps a 90V 1/4 second spike into the part to test survivability.
That said, you can use a 5KW 20V TVS across the power source to snub the high voltage spikes and insert a 1N4003 diode in series with the power source and any other input to block the negative spikes for protection.
Thaks Vic…but you lost me pretty quickly…so a simple question…would stepping the voltage to these modules down to 5 volts help avoid the spikes you speak of?
The solid state voltage regulator that you are using to drop the 12V down to 5V will help smooth the positive high voltage spikes.
However, the upper voltage limit ranges from 20V to 36V for the devices used in these regulators - well below the potential high voltage spikes.
It also will not handle the negative voltage spikes.
Now, these spikes do not happen that often but when they do, they can kill your circuit.
An intermittent battery cable connection, bad alternator voltage regulator , a headlight relay (without a suppression diode), or an original 1967-68 K9/K10 relay can generate these spikes. There is also a cumulative effect from these spikes.
Don’t let these potential issues dissuade you from developing your circuit design,
just be aware and build it so critical parts can be replaced or add the TVS and blocking diodes to protect it.