Page 2 of 2

Re: Ohm reading for fuel sender?

PostPosted: Wed Sep 12, 2018 10:00 am
by AZD
The nonlinearity is definitely a problem when mixing and matching parts. Neither the sender nor gauge are linear. The way the stock sender with the arm works, there is very little arm movement from FULL to maybe 3/4 or 2/3, then at the lower levels there is more arm movement (degrees of change) relative to the change in level near EMPTY. In other words, the sender degrees of movement are not linear relative to fuel level. It’s a trigonometry thing. I experimented and found this to be the case. A few degrees movement at the FULL end have much more effect than a larger movement at the EMPTY end. My suspicion is that the gauge itself is made to compensate for the nonlinearity of the original sender, or vice versa. A different sender, even one with the “correct” resistance, may read incorrectly in the middle values, yet be ok at full and empty, if it does not mimic the same output curve, if you want to call it that.

Now this bit is somewhat speculative, but I suspect the principle is sound: The gauge itself is a heater which bends as the current (and heat) increase, which causes a deflection of the needle. In electronics your resistive losses, or copper losses, are current squared times resistance. Run the formula around and it would seem that the losses (heat, and therefore needle movement, ignoring the potential nonlinearity of the bimetallic strip itself) are proportional to the square of the current, i.e., nonlinear, so the sender and gauge kinda balance out. The fun part is that in the real world of 1950s electromechanical gizmos, we’re probably getting close but not exact, which is good enough for not running out of gas, which is the whole point to begin with. When you think about it, it’s amazing this stuff EVER worked to any degree of accuracy.

I will kindly disagree about using a non-pulsed IC voltage regulator, though. A few months back I converted to 12V and wanted to keep all my gauges. Figuring that the 6V system charges at about 7.1V, I set up an LM317 regulator for that value. It feeds the supply bus to the gauges via the ignition switch. Works perfectly, or at least as well as it ever did. The gauges still “see” the same voltage as before, so that’s to be expected. Results may vary depending on your application, of course.

Re: Ohm reading for fuel sender?

PostPosted: Wed Sep 12, 2018 10:53 am
by cornbinder89
In your case you were supplying the 7.1 volts that the gauge is designed for. The early K/S senders (at least oil pressure and temp) have a pulsed strip in them, not a flat resistance type sender. The later K/S used a pulsed "regulator" ahead of the gauge and this is the one I was saying can not be replaced by the IC unit.

The change in gauge systems took place sometime in the 50's, near to when 12 volt electrics came about.
In both cases, the gauge relies on a pulsed current, the early system had the pulse mechanism in the sender, the later placed it before the gauges and used a conventional resistance card type sender.
Look up the description of the gauge system in the K service manual or an old Motors manual.
I don't fully understand why the pulse current is necessary, but in both cases they went thru some trouble to make sure it is there, so there must be a reason, my guess is the pulse width varies with the sender to make the thermal strip in the gauge head more responsive to small changes at the sender.
Ford, Chry and IHC all used K/S gauges into the 70's and I know from experience, that there is no fixed regulator setting that will make the gauges function properly.
The old system with the pulse units in the senders, will respond well to a fixed input regulator set to running voltage of a "6 volt" system. The fixed regulator will hold the voltage constant regardless of current draw ( up to the cap of the reg) just as a 6 volt battery would in the truck.
A search of this site should turn up a lot of info on the K/S system. Both Monsonmotors and Kevin have done a lot of work researching the system.
Edit:
To clarify, the early K/S used a "fixed input" with a pulsed sender, the later K/S used a pulsed input with a simple resistance card type sender.
This is why a "modern" resistance card type temp or oil pressure sender will not work with the K series gauges. Repop senders for 40's Fords can work, but the one s for the V8 use a different "card" because two are used, but are close.
The later system used the common senders with resistance card and can be had easily.

Re: Ohm reading for fuel sender?

PostPosted: Wed Sep 12, 2018 11:59 am
by worthy
Brianz's post....
""Hi Worthy,
I did get it to work ok, not great. If I am above half a tank, the gauge reads about 1/4 higher than it is. From about half to empty, it's pretty close. I am unsure why it acts like this given my previous testing. The tank is a new 15 gallon tank I mounted on the frame like a saddle tank, so I am dealing with a pretty small tank. I used a sender from ISSPRO. I was happy with the product and the support. isspro.com
They do 6 volt, 12 volt, custom built units, etc.""

This sounds like a solution, thanks Brianz. I have a custom built tank that fits where the old one did. Mine is rectangular with square corners so I assume is just slightly larger that the original. If this works well when the tank gets low that's good enough for me. I just don't want to run out of gas. Thanks! I will order one soon.

Re: Ohm reading for fuel sender?

PostPosted: Thu Sep 13, 2018 8:19 am
by Seymour
cornbinder89 wrote:I think most of the problems people have are the gauges are thermal and non linier. It is hard for the layman to understand. I don't fully understand them but it is the reason for the pulse unit in the senders (older K/S) or the pulsed supply (newer K/S). Fitting a flat resistance sender without a pulse doesn't work well. The gauges fail to respond accurately. Nor does fitting a IC "5 volt" regulator work to take the place of the pulsed OEM unit, which didn't supply a steady current.


For what it's worth, I put a 5 volt regulator in mine. It reads correctly at empty and full. How linear it is I'm not sure, but seems to go down with some correlation with fuel use.

John

Re: Ohm reading for fuel sender?

PostPosted: Thu Sep 13, 2018 8:43 am
by cornbinder89
Seymour wrote:For what it's worth, I put a 5 volt regulator in mine. It reads correctly at empty and full. How linear it is I'm not sure, but seems to go down with some correlation with fuel use.

John

Early K/S system or late? if early and you used it to save the gauges from 12 volt, than yes it should work. A better setting would be 7.1 volts like the other poster used.
If a later K/S system that had a pulse unit ahead of the gauges, the accuracy will be off by a lot. Temp and oil pressure will not read correctly. I never said the gauges will not move or function at all, but the whole reason for a gauge over an idiot light is to let you know with some accuracy what is happing.
Back in the early 70's when these gauges were still in use in Ford, Chry and IHC and the pulsed regulators were failing, I did a lot of work experimenting to see if there was a constant voltage setting that would work with some accuracy. I found none, and the response time and accuracy was appalling. If you read the description of the way the unit functions, no where does it say it supplies 5 volts, rather that it pulses between 12 and about 3 volts that if read with an analog voltmeter, it will hover somewhere close to 5 volts. Put a test light across it and you will see if flash from bright to dim. Put a digital voltmeter on it and you will not get a constant reading, it will always be in flux.
I know the early K/S system work well as designed, having compared the readings to mechanical gauges in the same engine at the same time. I was surprised how well the two gauges corresponded, and the electrical oil pressure was faster to react than the mechanical with 1/8" tubing.
Link to MM post about them:
viewtopic.php?f=5&t=7927&p=60400&hilit=king+seeley#p60400