TR4/4A Fuel and Temp Gauge NightMare! Help!

[jeep937]

So I got a nice driver TR4 a year or more ago. It's an early '65 solid axle car. It drove okay but the dash wasn't fitted properly and some gauges didn't work. I thought it would be a nice quick fix but now after many shop sessions I'm starting to hate the car. I'm having thoughts of destroying it and pulverizing it into very small pieces. I'd gladly used many tanks of torch gas to cut it up and give it back to the earth in a pile of molten rust.
Rant Off
https://www.britishcarforum.com/bcf...with-his-hands-full-in-Ohio!&highlight=lm7810
Here you can see what I had to do in order to fix the dash hack glove box install of a previous owner. I enjoy beating metal into submission but the electronics doesn't seem to want to be fixed. I understand how this show is run. Negative ground. I have pretty much memorized the entire diagram via my trusty TR4 book. Aside from a few hacked wires all the colors seem to be correct. I'm only attempting to get the temp and fuel gauges to operate. I've hot wired the heater and got the lights to work. After the gauges I'll work on getting the wipers running.

First, I found the old voltage stabilizer in the glove and this told me that a recent owner replaced it. I found the new one installed under the dash on the passenger foot well. I ran leads to this with the body grounded and the hot wire from the fuse box connected. I get 12 on the B (IIRC) terminal "in" and 9.XXv on the T terminals "out". But when I hook up the wire that leads to the gauges (which are fully grounded, wired and lights working but off) It reads 0v to .1V. Both gauges don't even twitch when I key it on and they both read all the way left. Empty and cold respectively.

Next I clean all terminals on the Lucas fuse box and make sure every thing is grounded. My trusty HF voltmeter checks out and all is good. So now I think it might be a faulty voltage stabilizer. I found the thread on this board about the 10v chip stabilizer and the 6 volt stabilizer thread on the HAMB board telling the use of the LM7810 and 7806 (for 6 volt) chips. Info is the same in each thread. I order 10 of them since they are so cheap. I solder the in, out and ground to five of the ten LM7810 chips. I buy a 6 fuse block from the parts house. I run the hot from the Lucas fuse box to the main hot on the 6 block. I decide that I can run each of the 3 wires originally coming off the Lucas style voltage stabilizer on their own circuit. Then I can run one 7810 chip on each gauge and on the wiper motor and other accessories as necessary.

Since I have five 7810's ready to go and the panel wired and ready I carefully install each gauge and 7810 on the back. Upon keying on the car I find that the 7810's are heating up and smoking. I check everything and put a new one on. Boom, one explodes in tiny fashion. I've got two more good ones wired and ready and five more that are in the bag. I tested each one after soldering and they all put out just under 10v. I did not use the condensers as the two threads here and the HAMB board mentions as I read they are not needed.

What am I doing wrong here? I've started at this and chased grounds and colored wires for weeks trying to figure this out. Help me get these gauges working please before there is one less TR4 in the states.

 

[jeep937]

I've checked that the wires in question are not shorting out. I ohmed the GB hot wire that goes to the gauges and it's not touching any ground via frey in between the 6 block panel/Lucas VS. I noticed I was popping 5 amp fuses when I had the other green connected that runs to the wiper motor (iirc). I eliminated this from the 6 block and have only been working with the green wire that goes to the gauges.

 

[trrdster2000]

Hi Jeep, did you just give the 2 gauges in question a shot of 12V and see if they move or self destruct?

Wayne

 

[dklawson]

I did not read the other thread but I have read through your post here. Take this one step at a time.

First, only the two electrical gauges need the voltage stabilizer or the 7810 chip. Nothing else needs reduced voltage and nothing needs 6V. If you go with the 7810 chip, you need a TO-220 case type suitable for 1 to 1.5 Amps, not the smaller ones.

I know you have studied the wiring diagram and I understand you are fixing what the previous owner did. However, temporarily separate the gauges from the other wiring issues you are fixing. Get them working by themselves, then connect them to the rest of the wiring you are installing.

Disconnect both ends of both sending unit wires. The green/black wire is for the fuel gauge, the green/blue for the temperature gauge. Make sure neither wire has a short to ground and that they have continuity from end to end.

Connect the output from the 7810 or "I" terminal of the stabilizer to one of the gauges. The wire can go to either gauge terminal. Connect that gauge's sending unit wire to the remaining gauge terminal. Connect the far end of that wire to its sending unit. The temp sender will have a ground connection via the engine. The fuel sender will probably have a black wire that needs to be connected to chassis ground.

So your wiring goes like this:
12V power into the stabilizer "B" terminal, the stabilizer has a ground connection through its case connected to the car, the "I" stabilizer terminal is connected to a gauge terminal and the remaining gauge terminal receives the wire going to the sending unit. The sending unit is grounded. If you use the 7810 chip, make sure you follow the pin out for the chip to achieve the same wiring. The gauge itself is only grounded through its mounting hardware and the gauge ground is only required for the light.

Supply 12V to the stabilizer input. The easiest functional test is to pull the sending unit wire off the sender. The gauge should go to zero or stay at zero. Now hold the sending unit wire to ground and wait a minute. Watch the gauge, it should climb to "full" or "hot".


Hopefully this helps. Also see the following PDF.
https://home.mindspring.com/~purlawson/files/SmithsVoltageStabilizer.pdf

 

[jeep937]

I thought about it but I'm guessing that the inside of these have very very small wires and windings that may be very sensitive to this type of test. I'd hate to fry them and not know what the first problem was as to why they didn't work

 

[TR3driver]

An LM7810 should be able to tolerate a full short on the output. If you are blowing them up, there is something wrong with the way you're using them. Did you use the bypass capacitors as recommended (.33uF input, .10uF output)? I'm going to guess that it is oscillating at high frequency (which it isn't protected against, without the caps).

I assume you realize that a TR4 diagram will be for the original positive ground. Since the 7810 is a positive regulator, it won't work with positive ground.

Use your ohmmeter to check the resistance through the gauges. If hooking them up caused the 7810 output to drop to 0, I think you'll find that something is shorted.

Lucas specifically warned against applying a full 12v to the gauges, but just a second or two won't hurt. If you don't see the needle start to move within a couple of seconds, the gauge is bad.

 

[jeep937]

I did not read the other thread but I have read through your post here. Take this one step at a time.

First, only the two electrical gauges need the voltage stabilizer or the 7810 chip. Nothing else needs reduced voltage and nothing needs 6V. If you go with the 7810 chip, you need a TO-220 case type suitable for 1 to 1.5 Amps, not the smaller ones.

Is the TO-220 case the stock VS case? If not what is it? I'll be googleing in the mean time..

I know you have studied the wiring diagram and I understand you are fixing what the previous owner did. However, temporarily separate the gauges from the other wiring issues you are fixing. Get them working by themselves, then connect them to the rest of the wiring you are installing.

This is my new plan. I was trying to only focus on them but the mess in the car and tools and crap everywhere is killing my shop moral.. Along with those 7810's popping in my face. lol

Disconnect both ends of both sending unit wires. The green/black wire is for the fuel gauge, the green/blue for the temperature gauge. Make sure neither wire has a short to ground and that they have continuity from end to end.

Connect the output from the 7810 or "I" terminal of the stabilizer to one of the gauges. The wire can go to either gauge terminal. Connect that gauge's sending unit wire to the remaining gauge terminal. Connect the far end of that wire to its sending unit. The temp sender will have a ground connection via the engine. The fuel sender will probably have a black wire that needs to be connected to chassis ground.

So your wiring goes like this:
12V power into the stabilizer "B" terminal, the stabilizer has a ground connection through its case connected to the car, the "I" stabilizer terminal is connected to a gauge terminal and the remaining gauge terminal receives the wire going to the sending unit. The sending unit is grounded. If you use the 7810 chip, make sure you follow the pin out for the chip to achieve the same wiring. The gauge itself is only grounded through its mounting hardware and the gauge ground is only required for the light.

Supply 12V to the stabilizer input. The easiest functional test is to pull the sending unit wire off the sender. The gauge should go to zero or stay at zero. Now hold the sending unit wire to ground and wait a minute. Watch the gauge, it should climb to "full" or "hot".


Hopefully this helps. Also see the following PDF.
https://home.mindspring.com/~purlawson/files/SmithsVoltageStabilizer.pdf

This is all exactly what I needed. I appreciate tremendously your long detailed post and I feel it will help me greatly tomorrow when I tackle this problem with fresh eyes and a clean workspace. Tonight I will be pouring over the few links and reads about the 7810 conversion.

 

[jeep937]

An LM7810 should be able to tolerate a full short on the output. If you are blowing them up, there is something wrong with the way you're using them. Did you use the bypass capacitors as recommended (.33uF input, .10uF output)? I'm going to guess that it is oscillating at high frequency (which it isn't protected against, without the caps).

I assume you realize that a TR4 diagram will be for the original positive ground. Since the 7810 is a positive regulator, it won't work with positive ground.

Use your ohmmeter to check the resistance through the gauges. If hooking them up caused the 7810 output to drop to 0, I think you'll find that something is shorted.

Lucas specifically warned against applying a full 12v to the gauges, but just a second or two won't hurt. If you don't see the needle start to move within a couple of seconds, the gauge is bad.

I miss typed in my OP. I know it's a positive ground. I guess I must have spaced that when the 7810's were blowing up and smoke frying. I should have known something was wrong when I was dry run hooking them up via jumpers at the end of the night. 1; Hot (neg) 2; GND (pos) and 3; only getting 4.8v instead of the 10v I was getting earlier when I was hooking the up to a jump pack via traditional neg ground set up. RED FLAG The second I read your post I knew what the wall I was hitting was made out of and it's apparently made from the positive ground set up. I supposed I should go back the the original thread on here about the 7810 and see if it was ever mentioned. I remember my uncle telling me about how he changed his car to neg ground. I think he said he just switched the battery cables on his TR3. But I'll have to do some heavy reading on the subject.

So as it stands I've to some work to do and a direction to move in. I can't thank you guys enough. I was so mad today out there. It seems as if everything I do I'm literally doing it for the first time ever. You'd think I would be used to it by now. In order to know everything there is I should step back and breathe. Clean up my area and take a break. It would be nice if I had an old man neighbor to ask these things but around where I live I am that old man..

 

[TR3driver]

If you want to stay positive ground, then use a LM7910 instead of the 7810. They're a little more expensive, but only a little, and other than the polarity, work the same.
https://www.ebay.com/itm/5Pcs-L7910...401?pt=LH_DefaultDomain_0&hash=item43be86f109

Personally, I'd go for negative ground. The conversion is pretty easy; in fact your uncle's method will work. But, it's better to swap the connections to the ammeter and ignition coil as well. If you don't, the ammeter will read backwards (discharge when it should read charge), and the ignition will be very slightly weaker. And it would be best to re-polarize the generator before starting the engine. Otherwise, it will "wake up" generating -12v (for the old positive ground configuration), and get shorted to +12v when the cutout closes. Probably won't burn anything up, but why take the chance? Just pull the wire off the 'F' terminal and brush it against the 'A' terminal, then put it back where it was. I did my first conversion about 1975 and never regretted it. Although I wanted to keep my current TR3 more or less stock in appearance, there was never any question but it would be negative ground.

As I think I mentioned, a buddy of mine set fire to his MGA trying to run a negative ground CB radio in his positive ground car. Just not worth the hassle, IMO.

TO-220 is the standard "power tab" package for the LM7810 and 7910. AFAIK that is the only way they are available. The gauges don't draw much power though, so IMO you would be ok with an equivalent regulator in a different package. I did all my testing with the tab just sticking up in the air and it didn't even get warm with just a temperature gauge running off it (I have an early TR4 temp gauge in my TR3). And the chip is rated to work up to 125C! I did eventually bolt mine to the can, but more to keep it from bouncing around than for cooling.

 

[dklawson]

As Randall said, if your car is still positive ground, you need the 7910 chip. This is a good time/excuse to change the car to negative ground though. There are other forms availble, not just the TO-220. There are TO-92 and surface mount versions, neither of which will have the current rating you want. The higher the current rating, the better as far as the effect of temperature on life even though the gauges won't draw too much current.

Also as Randall said, supplying 12V directly across the gauge terminals will not hurt them if the time is kept to a minimum. You mentioned the internal wiring of the gauges. The following is just so you will know what's going on and why it's OK to briefly connect 12V.

The gauge system that uses the voltage stabilizer works by resistance heating. The fine wire inside the gauge is wrapped around a bimetallic strip that uses a simple linkage to connect to the gauge needle. When you pass current through the wire it heats up which causes the bimetallic strip to bend and that bending in turn moves the needle. When a car's electrical system is charging the voltage on the system can go up and down and that will affect the bimetallic gauge readings. To address this, Smiths (and others) use a voltage stabilizer (regulator) to provide the gauges with a constant and slightly lower voltage than the car's charging system. In this case, an average 10V. The reason these gauges are not polarity sensitive and do not need a direct ground connection is because current flows from the voltage stabilizer, through the gauge and off to ground via the sending unit. The gauge terminals are not grounded except sometimes during testing.

You mentioned that you measured 9V out of the Smiths voltage stabilizer. Voltage measurements of a stabilizer are "iffy" at best. The voltage stabilizer is an electromechanical voltage regulator. Like the gauges, inside is resistance wire wrapped aroudn a bimetallic strip. Instead of a linkage to move a needle, the free end o fthe bimetallic strip has a set of electrical contacts. As current to the gauges flows through the stabilizer's resistance wire the bimetallic strip is heated causing it to flex. When the bimetallic strip flexes enough the points open and current to the gauges is interrupted. When the current stops, the resistance heating stops, the bimetallic strip cools and relaxes, which in turn causes the points to close again. In effect, the stabilizer is a switch that turns on and off to supply 12V or 0V to the gauge, not a constant 10V like the 7810 chip... but an AVERAGE of 10V over time. If you were to hook the stabilizer output to a scope you would see it turn on and off anywhere from something like 10 times a second to once every few seconds. You can sometimes see this with an analog voltmeter but you will almost never see it on a digital meter. The on/off rate of the stabilizer varies from unit to unit.

If you have a working (original) voltage stabilizer, try that first to get your gauges working since it is NOT polarity sensitive. Later you can switch to the 7810 or 7910 chips if you want to. However, remember what I said about the stabilizer case needing a ground connection. If you leave the ground off the stabilizer case the stabilizer does not cycle the power on and off... its points are always closed. Without a ground connection the stabilizer is just a closed switch that always passes full 12V (14V+ with the engine running and charging). Passing full system voltage through the gauge will cause it to read "high", allow the reading to vary as the charging system voltage fluctuates, and shorten the life of the gauge by passing too much current through it.

In your first or second post you commented that the gauge needles didn't even flicker when you connected power. Do not expect them to. As explained above, the gauges work by resistance heating. These needles are not going to jump, bounce, flutter or do any other quick movement. If you hold the sending unit wire to earth as I mentioned in my first post, you can expect it to take 30 seconds or so for the needle to move from low to high.

 

[TR3driver]

As Randall said, if your car is still positive ground, you need the 7910 chip. This is a good time/excuse to change the car to negative ground though. There are other forms availble, not just the TO-220. There are TO-92 and surface mount versions, neither of which will have the current rating you want.

Do you have a reference, Doug? If it isn't rated to 1 amp, then it's not a LM7910. I'm looking at the Fairchild data sheet, and it only shows the TO220 package (although there are two variants of it).

You mentioned that you measured 9V out of the Smiths voltage stabilizer. Voltage measurements of a stabilizer are "iffy" at best.

If you saw a constant 9v coming out of an original VS that had more than 9v going in, then it is bad. I've tested quite a few of them, and they all switched plenty slow enough for even the slowest DMM to show either full input voltage, or nothing.

In fact, a simple test for whether an original VS is working is to pull off the wire off the sender and hook a test lamp between it and ground. If you can see the lamp flash, the VS is working.

But, you might be looking at an older electronic conversion, trying to supply a constant 10.0v. Lots of folks have done as I did: pry the original can apart, snip out the original mechanism and solder in an IC replacement. (I used a low dropout IC, NTE1953, instead of LM7810) If the input is a bit low (below 12.0v) or your meter is off by a bit, then the 9.xx might not be a problem.

In your first or second post you commented that the gauge needles didn't even flicker when you connected power. Do not expect them to. As explained above, the gauges work by resistance heating. These needles are not going to jump, bounce, flutter or do any other quick movement.

Good point, Doug. These "hot wire" gauges respond very slowly to changes in input (which is why the VS can flash so slowly without changing the gauge reading). When I was having trouble with the Stag overheating (due to a leaking head gasket), the engine could be boiling (steam coming out of the hood) before the gauge even made it past mid-way.

 

[dklawson]

I have tried not to use the prefix "LM" when I wrote as that designates a particular configuration that may or may not be available from different sources. I used the term 7810 which reverts back to just a positive output, solid-state regulator rated for 10V. The link below may or may not take you directly to the correct page at Mouser.com If the direct link does not work, go to www.mouser.com and type 7810 in the search field. From the next screen, select "power management ICs".
https://www.mouser.com/Semiconductors/Power-Management-ICs/_/N-wnwh?Keyword=7810&FS=True
They do not currently show the TO-92 case parts but they do include some other form factors. DigiKey shows only the TO-220, TO-252 and TO-263 for the 7810 family.

In my experience, while you can use a multimeter to verify the output of a solid-state voltage regulator, you cannot use it to accurately measure the output of an electromechanical stabilizer. The electromechanical unit is switching on and off and the digital meter is sampling a few times per second. Typically the meter will jump around and its reading change in an unpredictable manner as the points open and close. Likewise, the flashing light test will tell you if the contacts are opening and closing but not that the electromechanical stabilizer's output voltage is correct. The early stabilizers had an adjusting screw on the PCB that was accessible next to the spade lugs. Typically there was glue on the threads to keep the screw from moving. If you clean the points in an old stabilizer and remove the glue you can tweak that adjustment screw and sometimes bring a dead stabilizer back to life. My method to set the output voltage does not involve a meter. The method I have used with success is shown on pages 3 & 4 of the PDF I linked earlier. However, I have only done this for a few people who insisted they did not want to install a solid state regulator. I like originality but I like working gauges more.

EDIT: Checking on this further, the TO-92 version of the 10V regulator is 78L10 (with an L in the middle) and it is only rated for 100mA. If you search for 7810 or similar you are not likely to find or buy it by accident.

 

[TR3driver]

I have tried not to use the prefix "LM" when I wrote as that designates a particular configuration that may or may not be available from different sources.

Ah, I see. I prefer to specify the prefix, as I learned long ago that "equivalent" parts aren't always equivalent. I once spent several months finding a "software" problem that was actually caused by purchasing having substituted one of the parts with one that responded faster and not having the design requalified with the faster part.

DigiKey shows only the TO-220, TO-252 and TO-263 for the 7810 family.

And only TO-220 for 7910.

In my experience, while you can use a multimeter to verify the output of a solid-state voltage regulator, you cannot use it to accurately measure the output of an electromechanical stabilizer.

Agreed. In fact, I don't know of any meter that will give such a long-term average. But if it is switching on and off, you can be confident that it is working fairly well. The long-term average might not be exactly 10.0 volts, but it isn't going to be very far off either. And, as I said, if you see a consistent 9.xx volts, there is something wrong.

Somewhere, I've got a set of Lucas instructions that show essentially the same method in your PDF.

 

[dklawson]

The LM prefix is used by a few manufacturers with Fairchild being the most common. The NTE low drop out chip you chose is the best choice for cars with a generator. I don't think it will be quite as important for cars with alternators but it certainly won't hurt.

Hopefully some of our past couple of posts will contain something that will help Jeep sort out his gauge issue.

 

[TR3driver]

I don't think it will be quite as important for cars with alternators but it certainly won't hurt.

Depends on which alternator, IMO. The original Lucas alternators don't seem to put out much at idle either. A friend of mine had a TR6 with the stock alternator, where his stereo would quit working at traffic lights. Switching to a Bosch solved the problem. And I've seen a lot of other TR6 where the headlights brighten noticeably as they take off from a light.

 

[dklawson]

You will get no argument from me regarding the deficiencies of some Lucas alternators.

Regardless, I hope that Jeep is still following this thread and can use some of the information to get his gauges up and running. Hopefully this will be a good weekend to work on them.

 

[jeep937]

Well I got it changed over the neg ground. Tomorrow I'll be hooking up the 7810's and hopefully the gauges work.

 

[jeep937]

Wired up the gas gauge. WORKS! Tomorrow I'll fire it up with the temp gauge and see what happens.

 

[dklawson]

Excellent !

 

[jeep937]

So I hooked up the temp gauge and it blew the 7810. I noticed that the input 10v tab on the gauge had been messed with. Looks like they left out the idolater when pulling the tab off the terminal. I'm guessing the gauge is grounding out the 10v causing the chip to blow. I hot wired it so the body wasn't grounded and it all worked. My problem is the gauge on a chily day only got up to 50c maybe a little more. Do these cars run at 125 degrees F? Doesn't seem right. I don't really trust these electric gauges and would like to have a trusty analog gauge. I would like to find a TR3 gauge to run and possible change over the rest of the small glass to curved. I like the look of curved glass much better and am not really concerned with being 100% original. I would much rather have a trusted gauge than a correct one. I see the guy on ebay had the tr3 repli gauges. I don't really want to pay 150$ for a gauge that's not even TR3 correct. I would pay 100 but looking at the completed listings it's not looking to good for me. Is there a better source for a good analog TR temp gauge?

Side note: I of course would respect the originality of the car marking and leaving the sending unit wires in place and keeping all of the original parts to convert back to stock.