Full Size Jeep Network

letank wrote:Picts of your soldering artwork.... my CVR just gave up yesterday...

I did mine the same way Michel did. I am positive I took photos of mine when finished but can't find them at the moment. Damn OldTimers kicking in.

The 7806 fastened to the speedo housing with a small self tapping screw, run the wires as shown in the image and schematic diagram he shared at the top of the thread page.

sooo, here is my brand new coolant gauge innards after 500 miles, and you can guess, it failed!



Fortunately, I was able to retrieve an old cluster... and back in business, good to have oil pressure and coolant temp



but I will do the 7806 or 7805, the strange thing, it that I am measuring 8,54 volts at the 5V output

The old cluster had been colonized by mice.... poo all over the inside

Of course after 20miles it appears that temp is never really going up or down, sitting on the cold side of the normal range even in stop and go traffic on a warm day; nor is the oil pressure moving too much, staying at around 40 at idle and may be 50 on the highway.

So off to the LM7806 and may be the capacitors.

found this nice add-on

http://www.jalopyjournal.com/forum/thre ... -3.448589/

but will keep it simple as described at the beginning, the gauge barely draws 0.1 to 0.2 amp

So new year resolution, to have a working temp gauge, did the above listed, except that after a certain age, this is a tough one to solder those tiny legs... I only did one side for ceramic cap using of 0.33 mf on the output leg. Added wires as pict below to the LM7805 . Tested on the bench with a 14.4 v supply and got a 5 volt with no load. Installed and will see in the next trip how the temp is holding up.
I used the mounting screw of the speedo to fix the transistor, tap the hole with a 8x32 and cut the top to allow for a little more room





Removed the filter and slipped a female connector to the post, we can never have too many good ground

letank wrote: ... but I will do the 7806 or 7805, the strange thing, it that I am measuring 8,54 volts at the 5V output ...

Kinda late for this, but I would not worry too much that a multimeter gives you a funny reading. Did you measure on DC or AC? Unless it's a comparatively expensive RMS meter, it will only give you an accurate reading for time-varying voltages that are sinusoidal waveforms. Since this is PWM (pulse width modulated), the voltage in time looks like square-ish wave AC with a constant voltage offset.

Not surprisingly, Fluke has a page about this - http://www.fluke.com/fluke/uses/comunid ... e-rmsfacts - just to point out that it's not a trivial measurement.

tgreese wrote:

letank wrote: ... but I will do the 7806 or 7805, the strange thing, it that I am measuring 8.54 volts at the 5V output ...

Kinda late for this, but I would not worry too much that a multimeter gives you a funny reading. Did you measure on DC or AC? Unless it's a comparatively expensive RMS meter, it will only give you an accurate reading for time-varying voltages that are sinusoidal waveforms. Since this is PWM (pulse width modulated), the voltage in time looks like square-ish wave AC with a constant voltage offset.

Not surprisingly, Fluke has a page about this - http://www.fluke.com/fluke/uses/comunid ... e-rmsfacts - just to point out that it's not a trivial measurement.

Tim, Thank you for the infos.
Never late, I was supposed to do the mod in Oct-Nov... The 8.54 volts were measured with the OEM unit, and I got it, because of the PWM, the meter is measuring a range.
Now that I have the voltage regulator, the hand(s) does not pulse on any of the gauges. I added an extra temperature gauge that I had lying around which as a magnetic sender and it is on the heater hose connector on the back of the intake manifold. Testing in the next few weeks will help to see if the numbers/position make some sense.



The left gauge temp is coolant temp, not trans temp as labelled -recycling stock-

Nice mod. The venerable 780x is a much better solution than the OEM one.

I have been wondering what the 411 was on this mod. Is the 86+ newer cluster regulated similarly?



Sent from my XT1096 using Tapatalk

shimniok wrote:Nice mod. The venerable 780x is a much better solution than the OEM one.

I have been wondering what the 411 was on this mod. Is the 86+ newer cluster regulated similarly?



Sent from my XT1096 using Tapatalk

I believe after 86 the gauges were 12VDC...so no need for a VReg

The word I got from a factory rep years ago regarding the 6v gauges was; The gauge makers in the old days refused to redesign their gauges for 12v so the carmakers added the pulsing VReg to accommodate 12v electrical systems. This seemed to become a standard until the 80s when Stewart-Warner and the others finally extricated their craniums from their respective anal-orifices and began making 12v gauges.
Odd that when the US military chose to switch to 24v systems these yokels had no problem adapting...

tgreese wrote:

letank wrote: ... but I will do the 7806 or 7805, the strange thing, it that I am measuring 8,54 volts at the 5V output ...

Kinda late for this, but I would not worry too much that a multimeter gives you a funny reading. Did you measure on DC or AC? Unless it's a comparatively expensive RMS meter, it will only give you an accurate reading for time-varying voltages that are sinusoidal waveforms. Since this is PWM (pulse width modulated), the voltage in time looks like square-ish wave AC with a constant voltage offset.

Not surprisingly, Fluke has a page about this - http://www.fluke.com/fluke/uses/comunid ... e-rmsfacts - just to point out that it's not a trivial measurement.

Yep! Some digital DVOMs will react differently to a waveform like this. I use an old analog Bird Watt/Volt meter for this sort of thing as it has very good dampening and will show a very accurate reading of a PWM signal.
I've long since forgotten the math behind it all.

I made this mod yesterday on my J20. Works great. I used a 7805 because the TSM calls out 5V as the regulated voltage in the discussion of the CVR, and I had 7805s on hand. I tested the accuracy of the guages with the dummy loads listed in the TSM, and the gauges deflect the correct amount with the 5V regulator.

The 7805 is a bit more common than the 7806 - you should be able to get a 7805 at Radieux Shaque or any other local electronics store.

Pretty neat actually. My CVR was bouncing my voltmeter and fuel gauge needle in time with the pulsing - the gauge needles are now very steady. Recommended!

Fry's has them and they are under 2.00.

http://www.frys.com/product/8052564?sit ... IN_RSLT_PG

shimniok wrote: Is the 86+ newer cluster regulated similarly?
Sent from my XT1096 using Tapatalk

I have been curious too about the 86 and later, so I had to crack one open, yes there is a solid state regulator, that drops the voltage from 12.55 to 9.7 without the sender on the gas gauge, voltage drops to 3.7v with a 40 ohm resistor reading at half and 2.3v with a 20 ohm for give a 1/4 tank.



and testing -no sender, using a resistor-

can I use a gauge that has the vcr already fualty to do this conversion

DARREN LILLY wrote:can I use a gauge that has the vcr already fualty to do this conversion

for the 85 and older, yes you can, you can spot in most of the picts the CVR looks a bit cooked... so it is a tad faulty.

EDIT: my CVR was almost working, but after 30minutes, the oil pressure was wandering to very low, then bounce to High then return to the mid mark, even at slightly above idle which is not correct, the temp gauge was consistant, but not acurate, reading lower than expected.

This is a great write up. I saw it was 5V DC so made me think about USB chargers which are also at 5V. There's a bunch of these ready made 12V - 5V DC converters if you're not up to the soldering and sourcing. I think it would work as well.

https://www.amazon.com/DROK-Waterproof- ... way&sr=8-4

I'm finally getting around to this. I purchased the 5V USB charger from amazon. I want to take the 12V power from the cluster the same way and feed the 5V back to the gauges. Just want to confirm that taking off the RFI shunt is the idea to eliminate the old CVR from working. I will be putting the new 5V positive to where the blue wire is on the Letank pictures and the 12V positive for the USB charger will be from where the red wire is shown with the "shunt" removed?
Also, does anyone have a picture of the PCB board fit to the metal cluster? I've got all my gauges out and the needles seem to be touching the masks with it tightened up.
THanks

JMO - using your USB charger makes this conversion a lot mmm... riskier? The 7805 and 7806 have been tested here and work well. The 78XX chip is basically a transistor that divides the input voltage into the top half, which it throws away as heat, and the regulated lower half that it sends to the output. Very simple and reliable. Your USB supply is a DC-to-DC buck convertor meant for a larger current load than these two gauges. These convertors (which are also known as switching power supplies) are good for larger loads because the internal oscillator converts to AC, steps down with what's essentially a transformer, and rectifies back to DC. This process is more efficient than the 7805, since a smaller portion of the power is thrown away as heat. However, the amount of power we need here is tiny, and these convertors are a lot less reliable, and can throw off a bunch of high frequency noise at the switching frequency. Suit yourself - I expect it will work ok.

When you use the external regulator instead of the CVR, what was the 5v output from the temp gauge, going to the fuel gauge, now becomes the 5V input from the 7805. 12V is no longer connected to the temp gauge, bypassing the step-down of the CVR and connecting directly to the external 5V source. Looks like removing the noise suppressor disconnects terminal I (the ignition) on the temp gauge from terminal E on the big plug. Not sure where I picked up the ignition feed when I did this - I don't think I soldered to the main circuit board. You do need to send the ignition hot wire to your regulator, and then send 5V to either terminal A of the temp or fuel gauge, bypassing the internal CVR.

These parts are cheap and easy to find -
https://www.mouser.com/ProductDetail/ST ... sEyhlP4%3D Datasheet here too.
https://www.mouser.com/ProductDetail/AV ... BJNO9ZI%3D
https://www.allelectronics.com/item/780 ... 05t/1.html
https://www.allelectronics.com/item/104 ... tor/1.html - STM Micro 7805 data sheet calls for 0.33uF on the input and 0.1uF on the output - you could use 3 in parallel for 0.30uF ...

OK-I'm getting confused. I'm assuming that picture is correct. The 12V power source for the CVR or USB will be the red wire coming off the pin from the noise suppressor (now removed). Then the ground is green off the case. The 5V positive supply is going to the nut where the blue wire is attached.

I respect your opinion. I didn't want to re-invent the wheel since 5V power supplies are really cheap right now. I think it was less than $10 for the unit and it puts out almost a perfect 5 volts. If the draw from the senders doesn't require the current, it won't supply it so I think it won't generate heat. I thought it a better idea that driving around shopping for the parts. We don't have any place near to me where you can by off the shelf electronic components and ordering the 7805 online would most likely kill any savings due to shipping charges.

My issue now is that my fuel gauge isn't working but the fuel sender is fine and the temp gauge is fine and the fuel gauge itself moves up and down when you put 5 V to it directly. So my assumption is that there is something wrong on the PCB or the CVR is acting up. I dismantled the whole units. Finely cleaned all the contacts and have it back together. I should have it back in the Jeep by tonight.

Thanks for the reply.

The picture is correct. Look at the letters stamped into the circuit board. The temperature, fuel and oil pressure gauges all have a terminal "A" and they are connected together. This is 5V coming from the original CVR. You could connect your new 5V supply to any of these terminals, since they are all connected together. The blue wire goes to terminal "A" of the temp gauge.

The fuel gauge only has two connections. It gets 5V from the CVR connection, and connects to ground through the fuel sending unit. The sending unit is a variable resistor connected to ground. The value of the variable resistor can range from 10 ohms (full) to 73 ohms (empty). By Ohm's law, this resistance limits current through the gauge, so the maximum current can only be roughly 12V/10ohms = 1.2 amps. Connecting the gauge from your 5V supply to ground provides excess current (limited only by the power supply and the gauge's internal resistance) and could damage the gauge.

Since you have a 5V supply, you could test the gauge on the bench with the resistances given in the TSM. 73 = empty, 23 = half, 10 = full. This means finding some resistors ... less than 15c of parts, but sourcing electronics parts seems to be difficult for you. If you can't find them anywhere, send me a PM and I'll put them in an envelope and send them to you. If it works on the bench, then you could test it in the car by unplugging the wire to the sender (pink wire under the Jeep) and connecting those resistances to ground on the chassis. Keep narrowing it down until you find the fault.

Actually Letank does not need that green wire, since the tab of the regulator is grounded and connects to the cluster chassis and ground there. When you mount the 7805 to the bare steel of the cluster chassis, you don't need the ground wire. However, your power supply won't work unless the ground is connected to the chassis somewhere.