Cluster inverter reverse engineering project

98 svt

Well-Known Member
Established Member
Premium Member
Joined
May 16, 2005
Messages
24,005
Location
Massachoooosetts
No luck with @vrt03cobra

Shit, if I had one it would be all yours, Mark. You've been a huge asset to the SVT community over the last 20 years or so.
 

rymax101

Active Member
Established Member
Joined
Jan 17, 2016
Messages
432
Location
USA, USA, USA
Any pic of the part you are looking for? I have a plan to go to my local pick and pull for some odds and ends, I can add this to the list.
 

mwolson

Gray beard
Established Member
Joined
Mar 28, 2004
Messages
3,411
Location
San Jose, CA
The inverter is only in the 03 and 04 Cobra or 02 to 03 or 04 Lightning clusters as far as I know. It isn't in any other vehicles that I know of. You have to take the cluster apart and look for about a 1" cube, open on 3 sides with a circuit board that looks like this:
1664137424610.jpeg
 
Last edited:

mwolson

Gray beard
Established Member
Joined
Mar 28, 2004
Messages
3,411
Location
San Jose, CA
I just finished working on a couple of clusters that were sent to me, a cluster from the guy's car and a donor cluster plus a spare Glowhut inverter. It turned out that all 3 inverters were fully functional. The EL panel from his car was dead, but the EL panel from the donor cluster was functional. I put the functional panel into his car's cluster and it is now working again. The dead EL panel read about 15K ohms when unplugged from the inverter, so that seems to be a pretty good diagnostic to check the EL panels.

At this point, I have seen two failed EL panels and only one failed inverter so far. It isn't enough data to decide anything, but I am starting to wonder if the EL panels are less reliable than the inverters are.

Still looking for failed inverters to diagnose...
 

wckdvnm

Well-Known Member
Established Member
Joined
Aug 10, 2015
Messages
1,337
Location
Kingman, AZ
I just finished working on a couple of clusters that were sent to me, a cluster from the guy's car and a donor cluster plus a spare Glowhut inverter. It turned out that all 3 inverters were fully functional. The EL panel from his car was dead, but the EL panel from the donor cluster was functional. I put the functional panel into his car's cluster and it is now working again. The dead EL panel read about 15K ohms when unplugged from the inverter, so that seems to be a pretty good diagnostic to check the EL panels.

At this point, I have seen two failed EL panels and only one failed inverter so far. It isn't enough data to decide anything, but I am starting to wonder if the EL panels are less reliable than the inverters are.

Still looking for failed inverters to diagnose...

Have you figured out what would cause the backlit gauges to be less bright? It has been a few years since mine have been on but they were definitely not as bright as some other clusters I have seen. If that has to do with the inverter I could send you mine when I finally tear into the interior of the car.
 

SID297

OWNER/ADMIN
Administrator
Joined
Mar 27, 2003
Messages
55,746
Location
Myrtle Beach, SC
I just finished working on a couple of clusters that were sent to me, a cluster from the guy's car and a donor cluster plus a spare Glowhut inverter. It turned out that all 3 inverters were fully functional. The EL panel from his car was dead, but the EL panel from the donor cluster was functional. I put the functional panel into his car's cluster and it is now working again. The dead EL panel read about 15K ohms when unplugged from the inverter, so that seems to be a pretty good diagnostic to check the EL panels.

At this point, I have seen two failed EL panels and only one failed inverter so far. It isn't enough data to decide anything, but I am starting to wonder if the EL panels are less reliable than the inverters are.

Still looking for failed inverters to diagnose...

03-04 Jeep Grand Cherokees have the same type of cluster. They may be a source for some parts in the future.
 

mwolson

Gray beard
Established Member
Joined
Mar 28, 2004
Messages
3,411
Location
San Jose, CA
Unfortunately, they simply get dimmer as they age. This is what I got from an EL supplier's web site:

Unlike conventional lamps, EL lamps will not burn out. There is no filament to break or gas to escape as with LED’s. They do lose some of their brightness over time. Lamp life is related to how much voltage and frequency the EL Panel is operated at, the cycles that it is going through and the amount of light being put out for the specific application. The lamp does not “burn out”; as long as there is conductivity in place, the lamp will still light up. It will simply become dim enough to require replacement at some point in time.

Putting in an inverter with a lower AC voltage such as a Glowhut inverter will also cause it to be dimmer. Too bad.
 

mwolson

Gray beard
Established Member
Joined
Mar 28, 2004
Messages
3,411
Location
San Jose, CA
I have a spare cluster I have never try more for decorating, can we make the investor fail ?
I want to learn what naturally causes them to fail. Forcing a failure wouldn't tell us anyting. I was suspecting that the transistor would be the thing to fail, but so far, I have only seen one transformer fail. Unfortunately, I have been unable to source a compatible transformer. You should hang on to that spare cluster for parts.
 

Murderedsvt

Well-Known Member
Established Member
Joined
Dec 30, 2012
Messages
358
Location
Florida
I want to learn what naturally causes them to fail. Forcing a failure wouldn't tell us anyting. I was suspecting that the transistor would be the thing to fail, but so far, I have only seen one transformer fail. Unfortunately, I have been unable to source a compatible transformer. You should hang on to that spare cluster for parts.
Are you able to tell if it’s blown visually?
 

mwolson

Gray beard
Established Member
Joined
Mar 28, 2004
Messages
3,411
Location
San Jose, CA
Are you able to tell if it’s blown visually?
I have only seen one bad one so far. It had an open circuit in one of the transformer windings. There was no external signs that anything had failed. Other failures may be visible, I don't know yet. Another reason I want to see some more failed units.
 

mwolson

Gray beard
Established Member
Joined
Mar 28, 2004
Messages
3,411
Location
San Jose, CA
Time for an update. I have actually been doing a lot of work on this project in spite of not having any more failed inverters to analyze. To recap, I have been able to diagnose 3 bad clusters. Two had bad EL panels and one had a blown inverter.

It is easy to diagnose a bad panel. If you disconnect the inverter from the panel and put an ohmmeter across the panel pins, you should see a really high resistance and see it go up until it is off the scale if the panel is good. If the resistance is lower and is a fixed resistance, the panel is bad. The two failed panels I saw were in the 10s of K ohms. There is no way to fix a blown EL panel that I know of.

The one bad inverter I have has a blown (open circuit) primary winding. There are no off-the-shelf transformers that I have found that will work in this circuit. We will need to have custom transformers made if we want to replace blown transformers. I had expected that the transistor would be the culprit, but not in this one case.

The transistor that Ford used is a cheap Chinese transistor made by a company called Zetex. There are no specs available for that transistor, but the curve traces I made from the one in the failed inverter shows that it is pretty much a standard NPN small signal transistor.

I measured the transformer to figure out the spec as best that I could. I got good inductance measurements for the secondary winding and the feedback winding from the bad transformer. I didn't want to risk the one working loaner inverter that I have so I didn't take the good transformer out of the circuit on that one to read the primary winding. I measured it in the circuit, and I hope that it is close enough.

I put the entire circuit into the LTSPICE simulator and got it to work. I tried a bunch of different small signal NPN transistors and they all worked. This implies that just about any small signal NPN transistor with the correct pinout will probably work in the circuit, just need to check the voltage and current specs to make sure that the max values aren't exceeded.

The transformer inductances I measured also worked in the simulation which implies that we could get custom transformers made using those specs to replace any blown transformers.

It is also possible to design a new inverter circuit to replace the entire inverter. I am also looking into that, which could use a cheaper, currently available off-the-shelf transformer.

While looking into all of this, I learned that EL panels are pretty much like a capacitor that emits light whenever the cap is being charged. The higher the voltage and/or the higher the frequency, the brighter the EL panel will be. However, the higher the voltage and the higher the frequency is, the shorter the life of the panel will be.

The SpeedHut inverter puts out a lower peak-to-peak voltage than the OEM inverter does, which is why I assume that the panels are dimer when using the SpeedHut inverter compared to using the OEM inverter. However, the SpeedHut inverter works at about 1.5KHz vs the OEM inverter that works at about 500Hz. I don't know if voltage or frequency has more impact on EL panel longevity. It is possible that the SpeedHut inverters will lead to a longer life for the EL panels or a shorter life.

If I wind up designing a replacement inverter for these vehicles, I will target 300V P-P and 500Hz to keep it as close to the OEM inverter as possible.

I am not interested in making a big investment in either replacement transformers or a new inverter if there aren't many failures. To date, I have only seen 3 failed systems.

I will keep playing with the new inverter circuit from time to time, but this project will not be on the front burner until I see at least a few more failed inverters.
 

mwolson

Gray beard
Established Member
Joined
Mar 28, 2004
Messages
3,411
Location
San Jose, CA


el panel source?
Those panels were custom made for Ford. I can't see how anyone could adapt a generic EL panel to work. :(
 

bull_dog190

Well-Known Member
Established Member
Joined
Jun 27, 2007
Messages
187
Location
U.P. Michigan
Unfortunately, they simply get dimmer as they age. This is what I got from an EL supplier's web site:

Unlike conventional lamps, EL lamps will not burn out. There is no filament to break or gas to escape as with LED’s. They do lose some of their brightness over time. Lamp life is related to how much voltage and frequency the EL Panel is operated at, the cycles that it is going through and the amount of light being put out for the specific application. The lamp does not “burn out”; as long as there is conductivity in place, the lamp will still light up. It will simply become dim enough to require replacement at some point in time.

Putting in an inverter with a lower AC voltage such as a Glowhut inverter will also cause it to be dimmer. Too bad.
So using this knowledge.....could one possibly extend the life of the OEM unit by slightly dimming the cluster? (manually with the knob on the dash)
reducing the voltage and extending the life?
 

mwolson

Gray beard
Established Member
Joined
Mar 28, 2004
Messages
3,411
Location
San Jose, CA
So using this knowledge.....could one possibly extend the life of the OEM unit by slightly dimming the cluster? (manually with the knob on the dash)
reducing the voltage and extending the life?
That's a good point. Dimming the cluster will reduce the peak-to-peak voltage which should help to extend the life of your cluster. How much it would help, I have no idea, but it sure can't hurt.
 

Users who are viewing this thread



Top