PDA

View Full Version : temp gage



mdelapp
07-22-2008, 12:45 PM
I just replaced my temp sending unit. The gage seems to check out in that it will peg if grounded.
After connecting all components and running the engine. as it warms the temp gage will start a normal increase. However with only the hose temp about 105 (temp gun check) the gage continues to climb and pegs itself. When I shut off the engine, the ohm meter reads about 100. Turning the key back on the ohm meter pegs at 0. I have a second gage and get the same results.
Anyone have any ideas on further troubleshooting? I have checked the wire from the sending unti to the gage for continuity and it checks ok.
The gage has a voltage reducer on it to take the supplied 12volt to 6v

http://i81.photobucket.com/albums/j231/druefer/Stude4.jpg

41 Frank
07-22-2008, 12:52 PM
Question: Are you using a 6 volt sender because if you are not it will never function properly. Just trying to establish some basics.

mdelapp
07-22-2008, 02:39 PM
Yes,it is a 6 volt sender. It acts as if it is a thermo switch. When it gets about 105 the ohm reading goes to zero. I am confident in my parts source this is the right part. I have just tried putting a 200ohm resistor in line from the sending unit but it must be too much as the gage doesnt move. re-clipping the aligator clips direct it starts to zoom. Maybe I should try some various resistors to get it to move about center.
Does anyone know what the resistance should be a various temps?

http://i81.photobucket.com/albums/j231/druefer/Stude4.jpg

StudeRich
07-22-2008, 05:18 PM
Mike; you know what all this points to, is a lot simpler than what you are trying to do. :)

Do you remember that fuel gauge post (topic) where Gord explained that there is a coil pulling the needle on each side? One requires power, the other ground resistance, sounds like one is open causing the needle to peg after a certain amount of resistance is applied. So although the gauge WORKS, it has failed! The Temp. works the same. Because of this design Gord said the gauge has to have a ground as well as the "controlled" ground (OHM's Resistance) at the sender! [:0]

StudeRich
Studebakers Northwest
Ferndale, WA

Jeff_H
07-22-2008, 06:32 PM
Here is some material on the '53 gauges I posted on the old alt.autos.studebaker group a few years ago:

The instruments are not like a meter movement. There are 2 coils inside
them. One is connected between the IGN terminal and the chassis ground.
The other goes between IGN and the sender terminal. The sender is a
resistor from the sender terminal to chassis ground. I found that the
polarity of the applied power does not matter. You need to use from 6-7V on
the gauge to test it. I found that 5V for example affects the "accuracy".
To test the gauge, apply power to the IGN terminal and ground to the gauge
case or cluster chassis. Now, you can connect resistors from the sender
terminal to gnd to see how it works. The guages are slow reacting. I think
they are somewhat thermal as well as magnetic. The two coils are set up so
that one pulls the needle up and the other down so the needle position
depends on the "balance" between the 2. The gas gauge response (needle
position vs sender resistance) is fairly linear whereas the heat gauge is
more log response due to the nonlinear resistance vs temp of the thermistor
used in the heat sender.


For reference, here is what I got for results on 2 sets of used gauges and 1
set of NOS:


Gas Gauge:


IGN to sender R: 75.5, 75.1, 75.5
IGN to Chassis Gnd R: 97.8, 98.4, 96.8


On 6.5V power the needle position vs sender R is
<E >90 ohm
E 90 ohm
1/4 55 ohm
1/2 29 ohm
3/4 15 ohm
F GND


Heat Gauge:


IGN to sender R: 60.7, 62.1, 61.1
IGN to Chassis Gnd R: 96.3, 97.8, 94.5


On 6.5V power the needle position vs sender R is
<C >130 ohm
C 129 ohm
Center 51 ohm
H 21 ohm
pegged Gnd


The Ammeter is direct reading. I did not have access to a power source
capable
of checking the full scale rating but as best I could tell its about 45A.
30A will make the needle go a bit over half way.


I hope this helps. I thought these gauges were like a meter movement until
I got puzzling results when I first toyed with them. I found some good ref
material someplace on how they work and that let me get them hooked up
correctly.

I am not sure what sort of voltage reducer you have. First thing I would do is measure the voltage at the IGN terminal of the gauge with respect to the ground with key on. Should be somewhere between 6 and 7 V. If that checks out. Disconnect both the IGN and Sendor wires from the gauge and with your ohmmeter measure from ground (gauge frame) to the IGN terminal and then from the Sender terminal to the IGN terminal. From my measurements you should see around 60 ohms from the Sender terminal to the IGN terminal and around 90-100 ohms from the IGN terminal to the gauge frame/housing. If both those check out, we can "assume" that those 2 coils inside the gauge are OK. You can use some fixed resistors of the values I showed above and it should cause the needle to move to those approximate locations. Maybe find potentiometer instead and set it with your ohmmeter. The symtoms you describe sound similar to what I was seeing when I toyed with them on the bench and did not have a ground connected to the gauge frame/housing. Any reason to think your gauge cluster is not grounded well? One way to tell may be to see if your voltmeter shows any voltage between the battery ground or the engine block and the gauge housing or dash. Should be none. As for the sender resistance vs temperature, I would guestimate from the measurements I got that it should be 130 ohm or greater at room temp (higher at freezing!), and 20 ohm or less at boiling. You could drop the sensor in a pot of water on the stove with a thermometer in the water and hook up your ohmmeter and check it out.

Jeff in ND
http://i156.photobucket.com/albums/t5/ee-engineer/53byalaketiny.jpg
'53 Champion Hardtop

mdelapp
07-22-2008, 07:05 PM
Thanks to all for the detailed response. I will pick up some varying resistors at radio shack and do some testing. Even though the wire has continuity from the sender to the gage I have discovers it has been spliced. The splice appears good with a solder connection. Not sure whether this would have any effect on the resistance enough to cause the gage to malfunciton. In any case I look forward to trying some more testing. It was soooo nice to get the gas gage working

http://i81.photobucket.com/albums/j231/druefer/Stude4.jpg

Jeff_H
07-22-2008, 08:31 PM
To rule out the wire, run a separate one for testing. Disconnect the sender wire on each end (gauge and sender) and run a new pc. Don't even bother to fish it through the firewall grommets, just run it out the hood side and tape it to the fender then go into the car through the front vent window. See if it works better. Sometimes bad connections get worse when heat and vibration come in and then seem to "heal" themselves when cooled off. Another thought is that something (sender even) shorts itself out to ground when it gets hot. A possible way to determine that is to disconnect the sender wire at the gauge and connect your ohmmeter from that wire to ground at the dash and then run the engine up to temp, maybe even drive around while eyeballing your meter. The resistance should decrease from the 130 or so up to something around maybe 50ish +/-10 ?? when the engine reaches stable temps.

When troubleshooting any electrical or electronic system, I always start by making sure the power supply to it is present and stable. Then, either trace the "signal" through the system or separate the component parts if possible and test them separately. If they all check out then its on to the wiring. This all assumes it was a system that was well designed and it worked originally. I usually work on new designs and R&D stuff so there is no assurance it EVER worked! Finding the problem(s) is more "interesting" then ;)

Jeff in ND
http://i156.photobucket.com/albums/t5/ee-engineer/53byalaketiny.jpg
'53 Champion Hardtop

mdelapp
07-23-2008, 03:53 PM
Ive run through some of the tests. I will list the readings in a later reply. Currently my best luck has been with two 10 ohm resistors in series. The temp gauge will rise to the left edge of the middle bar. With one 10 ohm it will still works its way over and then peg. I got a potentiameter and hooked it up No gauge on it and when all wire the ohmeter will peg. So have to set it about 15 ohms with a direct reading and then hook it up. Playwing with the pot will move the needle but it is slow and eratic. this and the pegging, (which requires me to smack the edge to get it back) makes me think that the coils are not too good. I get about the same results with either gage I use.
I'm going to put the resistors in line and see about the needle location for the resistors.
More data to follow.
(BTW, has anyone ever rebuilt the coils? Is there anyway to improve thier response and steadines?)

http://i81.photobucket.com/albums/j231/druefer/Stude4.jpg

StudeRich
07-23-2008, 04:52 PM
Jeff H in ND; Mike brings up a good point on his 6V setup, what can we buy for commercial use or at Radio Shack in the way of a voltage limiting diode I think it is called, like the one Chrysler & Ford use to stabilize & speed up their 12V Temp. and Fuel gauges? [?]

It also would protect them from the deadly, all to common "sticking Voltage reg. over-voltage syndrome"! [:0]

Hopefully they would be easier to find and cheaper than OEM Chrysler/Ford. :)

Of course to be used on the more popular '56-'66 12V Stude. models.

StudeRich -Studebakers Northwest Ferndale, WA

N8N
07-23-2008, 05:01 PM
I'm using one of the runtz voltage dropper units - not sure what they are internally, but I believe they do offer some limited voltage regulation - in my '55 with no probs. fuel and temp gauges both work correctly although I might throw a small value resistor inline with the temp gauge so that 180F is the middle of the scale (right now, I think that it's set for 160F which was the stock thermostat temperature for a '55 engine; 180F is about 3/4 through the white bar.)

I think I also attempted to do this the hard way a long time ago when I was working on my '49 Chevy; a calibrated resistor in line and a 6.something volt zener diode between the gauge and ground works, albeit it's inelegant and the Runtz are cheap enough I didn't bother once I found out they were available.

nate

--
55 Commander Starlight
http://members.cox.net/njnagel

StudeRich
07-23-2008, 06:14 PM
You are close to correct Nate, all Studebakers shipped to the lower 48, had 170 degree stats standard.

StudeRich
Studebakers Northwest
Ferndale, WA

N8N
07-23-2008, 06:44 PM
Rich,

'55 shop manual says 160, parts book says 170, not sure which to believe. My part book was printed in '58; shop manual is a reprint. I had an original 170 stat in the first engine that was in the car and the gauge still read high of center leading me to believe that original was in fact 160 but I can't 100% confirm that. When I went back to the original thermostat housing to use the molded hose I had to get the bigger Chrysler style thermostat and had to choose between 160 and 180 (nobody makes 170 stats anymore) I chose 180 for better efficiency, but if I end up having detonation or cooling problems I will swap to a 160.

nate

--
55 Commander Starlight
http://members.cox.net/njnagel

Jeff_H
07-23-2008, 07:24 PM
I'll apologize now if I get too technical or rambly with this reply, I've been working on a very technical report on some failure analysis investigation I have been neck deep in for the past month ( 20 pages so far [xx(]) ......

I built a 6.5V regulator out of stuff I had around for my car. I think the runtz units are probably very similar. I have everything on neg ground (gauges don't care) so the homemade unit I did works on that. Bad things may happen if I reversed the polarity to it. I don't recall if I put in a clamp diode and fuse or not, but I probably did (or should have!).

For reverse polarity to the regulator, one approach is a diode connected backwards across the power input so that if its reversed, the diode tries to short out the power and clamp it to about 1v. The fuse must be sized with the diode so that the fuse blows and not the diode! I remember having a fun time testing this sort of thing in the lab with a battery, fuse, diode, and a knife switch. I had a oscilloscope and current shunt to look at the current in the circuit. We were using some pretty slow blowing 40A fuses spec'd by our customer and it was tough finding a diode that could withstand 400-800A for a few 100ths of a second it took to blow that fuse.

I have both the gas and temp gauges running off the one regulator. Protecting the reducer output from a overvoltage can be brute force or complicated. Brute force approach is similar to the reverse polarity protection: put a big zener (clamping) diode on the output and then a fuse of the "correct" blow current at the regulator input. The hard part is figuring out the combination of zener and diode that does what you want but does not sacrifice the diode too. I've not seen much at radio shack these days that would help here :(

The electro-mechanical "regulator" used in at least 80s vintage fords are "crap" IMO. Basically they are like a Christmas tree lights flasher, only they "blink" about once a second or so. The shop manuals will say the average output is supposed to be 5V. What is really happening is they turn on and put out 12v battery, then shut off (0V), then turn on again, etc. If you hook a voltmeter on the gauge power from one of these it will jump up and down. Cheap digital meters often won't show much at all since they update the display too slowly. EE's will call this operation PWM (pulse width modulation). The gauges are slow enough responding that they sort of "smooth" this blinky power. It works well I guess until the "regulator" gets old and then the blink rate and the on and off times get a little off. The apparent average voltage then fluctuates and you get "dancing" gauges were the needles start to move around and it may look like your gas dropped a 1/4 tank at the same time as the temp rose from OK to Hot in the span of a few seconds. There are articles on ford sites on how to build your own solid-state electronic ones. My winter beater has this problem but I am too lazy to rip the dash apart to fix it.

Back to the problem at hand tho....

This is still sounding to me like a gauge ground issue. The gauge needle position depends on the relative amount of current flow between the two coils inside (IGN to Ground and IGN to S). The IGN to ground current is inside and controlled by that coils resistance. The S current depends on the sender to ground resistance. If the IGN to ground coil were open, perhaps that is the trouble. When I was testing my gauges before I knew what was in them, I did not ground the gauge so there was only current in the IGN to S coil and I got simlar behavior to what you are seeing. No ground on the gauge is basically the same effect as a open coil internally as far as the gauge needle goes.

Another thought.... Mike, I hope you are not leaving your ohmmeter in series with the gauge sender wire when the gauge is powered are you? The ohmmeter is attempting to measure the resistance by putting out a voltage of its own and measuring the current. So, if the gauge is putting out voltage on the S terminal to t

mdelapp
07-24-2008, 12:48 PM
I am not leaving ohmeter attached during operational tests. I found that it did indeed screw it up. When I test the extra gage outside the car with 6V from a 6v battery charger I get a 47 ohm resistor to bring the needle about half way. I have to insure that I run a separate ground wire from the meter plate (simulating the dash) to the ground post of the charger as well. This results in a good test. Also when I use 20 ohms it brings it to hot but doesnt peg it so that I have to move it back by hand. If I remove the ground from the gage plate I get the needle to peg regardless of what resistor is in the line. This would seem to validate your comment about the necessity for a proper ground on the dash. I tried this same set- up on the in-car gage and put an extra ground on the plate but it still pegged. I plan to remove the voltage reduced on the in-car gage, put on the out-of-car gage and simiulate it through the battery to see what the results are.
I have not run a separate wire from the sending unit yet to check that. I may have mentioned that I received a good result with everything hooked up and a 15 ohm resistor in the line from the sending unit. I am just not comfortable that in an over-heat problem that it would move the needle appropriately. I may also try another voltage reducer. The one I have has a small lead coming from the reducer that is grounded to the gage plate(This is the instruction from the supplier) This same type reducer is on the gas gage and it works pretty well. Not totally accurate on the empty side but about half a tank when it is at 1/4. At least it will go to full and gives me a comfort level.
I will report more results later today.
thanks again to all for particpating and providing advice and a sounding board
Mike

http://i81.photobucket.com/albums/j231/druefer/Stude4.jpg

Jeff_H
07-24-2008, 01:25 PM
Mike,

Sounds like the spare gauge tests produce the results I would expect. The resistance vs needle and open ground test are the same I got. That ground wire on your reducer leads me to believe its a electronic regulator and not a simple resistor. Make sure its connected. You should get 6V or so on your IGN terminal relative to the gauge ground if its working correctly. The homemade regulator I built has a similar ground wire requirement. There is a transistor circuit inside it that needs a ground to operate. 12V comes into it from the key switch and 6v should come out for it to go to the IGN terminal on the gauge.

Since you have a spare gauge, I would be tempted to connect it up on the car and do a driveway test. You can use your 6v charger to power it and connect ground of the charger to the car. Run a separate wire to the sender and idle the engine while watching it. If it acts strange, then I think maybe the sender shorts out internally somehow when it gets hot. On the other hand, it it works OK, then its likely the car wiring or the gauge in the dash.

Jeff in ND
http://i156.photobucket.com/albums/t5/ee-engineer/53byalaketiny.jpg
'53 Champion Hardtop

mdelapp
07-24-2008, 04:58 PM
I hooked up the spare gage and used the 6v battery charger for power and ran a separate line from sending unit to spare gage. Gage worked well but a little hot. Put a 10 0hm resistor on it and it held steady just below middle. Also did this with battery power and it checked OK. I assumed it might be the sending unit wire. I went back under the dash with the in-dash gage and the separate wire from sending unit and it moved up and shot to peg.
attached spare gage under dash with spare wire from sending unit and same results as outside car. Used existing thru-firewall wire, dash power, grounded to dash and voltage reducer. Held steady with 10 ohm reducer, used 15 ohm and it held steady just below center-line. I assume from your many comments that these can be sensitive enough, with their age and design that any variation in voltage or resistance can have a visable impact. I think I will touch-up the face of the spare unit, intall it using the existing fire-wall sending unit wire and the 15 ohm resistor. Would you think this is about as close as it can get to "stock" (excluding 6v to 12V) as I can get.
I took readings on the sending unit firewall wire and spare wire after shutting off and they both read about the 50 ohms with the water temp about 182. (as close as 70 year old eyes can do on an analog meter)
(I must remind myself to clean the temp gages or my next roast will have an anti-freeze taste)

http://i81.photobucket.com/albums/j231/druefer/Stude4.jpg

55 prez
07-25-2008, 12:02 AM
Mike you have really been through the mill switching from 6v to 12v and I hope the following will be of some help. I have a '55 President that I converted from 6 to 12v and had the good fortune of a chapter member telling me that I could just switch gauges and sending unit to 12v and forget about all the voltage drops and resistors. He told me that '63 gauges were a perfect replacement for my 55's fuel and temp gauges. I bought a new 12v temp sender from SI and everything works great. I did have some problems getting the fuel sender and gauge adjusted but finally succeeded. I would hope that you might be able to find a match for your '53 gauges that function on 12 volts. I also put a closed water system with overflow bottle on the radiator since I have installed air conditioning. It does not overheat with the 160 degree thermostat even on 95/100 degree days with electric pusher fan in front of the radiator.

mdelapp
07-25-2008, 03:33 PM
Well I guess I can say Mission Accomplished. Everything is back in place. The final result is I utilized the existing sending unit wire, the sending unit was replaced. I went back and forth between a 15 ohm resistor or 20 ohms. I decided upon the 15 ohm and the needle rest pretty much dead center. The operating temperature is 182 degrees with a thermometer stuck into the coolent at the radiator cap. Not sure if under pressure it will rise or be lower. I seem to think that pressurized the temp may bet higher. Maybe I should install a cooler thermostat.
i have not found any gages that will fit into the dash of the 53 that would be 12 volts.
After a couple of car shows this weekend I am going to undertake making a decal for the flat panel behind the instruments. Right now it is painted black with gold stripes I added with a ruling pen. Another thread gave great information on decal making.
thanks to all for the help
Mike

http://i81.photobucket.com/albums/j231/druefer/Stude4.jpg

Jeff_H
07-25-2008, 05:03 PM
Mike, Did you end up replacing the gauge in the car with the spare one too or was it the sending unit that was bad?



Jeff in ND
http://i156.photobucket.com/albums/t5/ee-engineer/53byalaketiny.jpg
'53 Champion Hardtop

mdelapp
07-26-2008, 11:45 AM
I used the one in the dash. I was prepared to switch and had done a little "restoration" on the face of the extra one, but both were giving the same results. I am not sure in the final analysis is the original sending unit was bad. Maybe it just needed a little resistor in the line. When I first tore into the thing there was a 200 ohm resistor in the line from the sending unit. Of course that would let the needle move at all. I am not disappointed in replacing the sending unit. $45 is not a bad cost for so much fun. I make now take some of your gas gage readings and see if I can tweak the gas gauge to be a little more accurate. It now registers full with a littl help on the float arm but drops off quickly to a a 1/4 or slightly less when only about 10 gallons down. Not sure whether adding some resistance to the line will alter that but with a bunch of alligator clips I am finding a home under the dash.
Another problem I have to sort out is I have a slight voltage draing when the car is sitting. When connecting a charge there is a little spark, which others tell me is a sign of this problem. Didnt seem to have that before. It was before I even started on the gages. Not sure where to start. Instruments and switches all seem to be off but not sure if something in the altenator or other place could be it. Fortunately I have a cut-off onn the battery so not too big an issue of disconnecting when parking it. I would like to find that as I want to get the clock hooked up. It works but of course can't keep it connected. It was fine last year.
Sorry to have added so many other issues but you have been great to help me.
Mike

http://i81.photobucket.com/albums/j231/druefer/Stude4.jpg

55 prez
07-27-2008, 12:52 PM
Mike on your voltage drain check on the wire to the glove compartment light under the dash and the dome light wires that go to the door switches. They both cause me problems on my '55.

ROADRACELARK
07-27-2008, 02:44 PM
Just to add to 55Prez.....if your car is equipped with a clock, cigarette lighter, an under-hood light, or trunk light, check or disconnect, then retest. Hope this helps.:)

Dan Miller
Auburn, GA

[img=left]http://static.flickr.com/57/228744729_7aff5f0118_m.jpg[/img=left]
Road Racers turn left AND right.

mdelapp
07-27-2008, 05:10 PM
Thanks
I will make that test tomorrow and report. It sure provides a source of power that is active when the swithc is off. I have a cigarette lighter and two door swithces. will isolate each one and check.
I am advised that an easy check is connecting a battery charger should not spark unless there is a drain. Is that a valid observation?

http://i81.photobucket.com/albums/j231/druefer/Stude4.jpg

BobGlasscock
07-27-2008, 07:24 PM
Valid, yes, but not too accurate if the spark is smaller than you notice.

Put a amp meter inline between a battery post and the cable to that post. DO NOT USE ANY ELECTRICITY while doing this test that uses more amperage than your meter can measure. The meter will measure any amp flow and you can connect/disconnect wires and remove bulbs and such while the meter is inline and see immediately what the results are.

'50 Champion, 1 family owner
http://i251.photobucket.com/albums/gg316/studebakerbob/SDC%20avatar/Studebakerstuff019.jpghttp://i251.photobucket.com/albums/gg316/studebakerbob/SDC%20avatar/Studebakerstuff018.jpg

mdelapp
07-28-2008, 05:36 PM
does it make a difference whether you do the meter between the positive or negative terminals?
Not using electricity, I assume you mean to not attempt to start the car. what about using the headlight. Of course that is always hot.

http://i81.photobucket.com/albums/j231/druefer/Stude4.jpg

Jeff_H
07-28-2008, 05:59 PM
Should not matter if you get the meter in series with the battery post and cable. No other connection to the battery post except the meter. If you use the + post on the battery, you will want the meter leads + to post and - to cable. On the - post of the battery, the meter - is to the post and + to the cable. This will give you positive current readings for current drain from the battery. I think you mentioned using a analog meter with a needle?? Important to know the expected direction of current then or it may peg the meter "hard" <0 if backwards and may damage it. A digital meter can read + or - current without that concern. Start out with the current range on its max setting to see where you are and then use more sensitive ones as needed.

Jeff in ND
http://i156.photobucket.com/albums/t5/ee-engineer/53byalaketiny.jpg
'53 Champion Hardtop