What to do whilst FerrariChat is down for a few days. Hmmm....might as well do some running repairs. If you have a Testarossa or F40, I strongly suggest you have a look at the following link. I was chasing an air con problem in one of my cars and found a bit of a problem with the wires going into the fuse panel. I fixed it up okay, then thought that I had better check my other Testarossa. Have a look and you will see what I found. http://www.ferrariaustralia.com/Technical/fuse_panel_repairs.htm Given that both of my cars had similar issues, and that it is so easy to check your own cars, I really think it would be good preventative maintenance to pull each plug out of your own fuse panels and check them out. If faulty, fix them, of have them fixed immediately. Last thing anyone wants to see is another Testarossa or F40 with a front end fire. I haven't finished the repairs yet, so the write up isn't complete either, but I'll keep you all posted.
Good day Robert, Very interesting subject and kudos for you on documenting the procedure. Just for clarification, does each board have printed circuits on both the top and bottom sides? I ask, as I have never seen multiple individual printed circuit boards (PCBs) sandwiched together. Usually multi-layer PCBs are made by the lamination of printed circuit layers with insulator layers. Sadly, the inner printed circuit traces do not have the same current carrying capability as those on the outer (top and bottom) layers. I wonder if this is the reason why this PCB delaminated the way it did, as normally one cannot separate the layers as you have done. I am also curious as to how common these PCB failures are on the Testarossas? I ask, as it is quite straight-forward to create a more robust (higher current capability) PCB design. Sadly the setup costs ($2500 or so) for this type of PCB fabrication make it potentially too expensive for a single PCB. If this was a somewhat common problem one could amortize the setup costs over multiple units to make the per unit cost more reasonable. It is also interesting that Ferrari has so many issues with their Fuse blocks. From 308s to BBs to Mondials, Ferrari Fuse blocks do not seem to be robust. In all the fuse blocks I have seen, the problems stem from inadequate design. After reviewing the issues with my BB fuse block, I decided to design a significantly superior replacement (waiting for protoypes as I type) so that I do not have to worry about it. Cheers, Sam
TR, Mond, and 328 are all prone to it on high frequency cycle/high current circuts like cooling fans and fuel pumps. It is not only Ferrari, the Germans went through it too. In the CIS days VW did quite a business in replacement boards but no one notices when the board is $100. Nissan really had trouble for a few years too. When the part costs $1500 everyone screams Ferrari builds junk parts. Same junk everyone else uses, Ferrari just charges more. You should see what I am paying for a new bezel on a Rolex. Quite a bit more than my Timex and doesn't work any better.
Sam the boards are definatley built as single sided seperate boards. There was no adhesion between the layers at all. The only thing that holds them is the soldered pins to that individual board. Thats why the pins also wobble around, as they are only supported by the little holes that they pass through in the front plastic sheet, and then by the thin layer of board. Very weak. As I think this is going to be a bit of an 'ongoing' issue, I spent nearly as much time on the repair as I did documenting what was going on, and where each wire went. As I finished a layer, I traced and documented the wires that were on that layer. That way if I ever get another burnt plug, I can look at the drawing and know instantly which relay or fuse pin that wire goes to, effect the repair with a jumper wire, and move on. Not real keen to go through pulling one fully apart again I have to say. Here is the PDF I created. Image Unavailable, Please Login
Good day Robert, Thank you for the information and the documentation! You have done a lot of work documenting this! I am still amazed that the design uses multiple separate boards... Cheers, Sam
Robert - Rather than bending down the printed circuit board cladding (conductor) that came loose from heat, did you consider simply soldering a wire with greater cross-sectional diameter (gauge) from point to point? Without having the assembly in front of me, I cannot tell whether there is sufficient space to allow for a wire, or even run the wire outside of the board. I have entertained attacking mine, and I commend you on your documentation and presentation. I always throught that I would improve the electrical characteristics by simply using wire where high current flows on the board. Jim S. Edit - just saw your last photos. Sorry for asking a question that was answered in your photos and presentation. Good job.
Good day Brian, Thanks for the info! I was unaware that this issue affected other car manufacturers as well. Indeed, large replacement part prices certainly brings issues like these to everyone's attention. Given that this is obviously a design issue, it is too bad that Ferrari (including other manufacturers too) did not offer an inexpensively priced replacement part... I know in my industry this approach goes a long way in maintaining customer loyalty and longevity. Cheers, Sam
Well how many Testa and TR owners have had a look at the white plugs in their fuse panels? Seems weird that of my two cars, both have had issues that required surgery. Let us know if yours is okay.......or not!
I am not sure how long you have been dealing with Ferrari's but it has never been and I doubt it will ever be their interest to supply low priced fixes for their cars. In fact in many ways the opposite is true. Many of the boards that were a problem are no longer available at any price. If, and that is a big if, and when they are made available again the price will be astronimical. As bad a customer service or relation topic it may be, the only gauge they ever pay attention to is the line down the block for people clamoring to buy a new car. As long as that situation exists their attitude will not change.
They will all be in need of repair or replacement at some point. They are just not robust enough to handle the current loads that are run through them.
The PC board itself is not the real culprit. The poor quality connectors are. The damage always starts there. Once those degrade and resistance goes up the current load goes with it and the board and connector experience China Syndrome. It always follows the same pattern. We have been repairing them since the 80's.
From what I've seen I agree with Brian. The connectors are very poor with only a small two point contact to the terminal pins. The damaged areas on mine are obviously showing me which ones are experiencing high amp current draw for some reason or another. All of these will be replaced with female spade crimp connectors which will have a far greater contact point than the originals. Regarding the cheap replacement part, as Brian points out, not going to happen. With cars like the BB, Testarossa, 348, 308 etc, we are now at a stage when we are not repairing them, but restoring them. It's like having a Fiat Spyder, or an Alpha Montreal, we all own classic Italian motoring masterpieces. Maybe a little more top end than a Fiat or Alpha, but the theory is the same. It won't be too long before parts availability makes things even more difficult and we certainly can't rely on Ferrari to fill the void here and continue to make the older parts. It is more unrealistic to consider that they make re-make a dodgy part with a new one for a 20 y/o car. That said, isn't that what we are all about here? For me, FerrariChat isn't a place to carry on about seeing an F430 or Enzo on the highway, or whatever, it's a place to share and glean genuine information about these fantastic cars. I put the effort into my initial post on this topic so as to help others to identify and rectify their own faults. I hope that as time goes along, I will be able to pick up pearls of wisdom from those of you who also get your hands dirty, or are possesed of specific and enlightening details. To me, that's what a club is all about.
Great write up Bob and the pix tell it all. Sure like to have a look at mine, but 'if it aint broke' and so on. What a job. Thanks
Wish I had seen this thread at the beginning of last summer before my fuse box suffered a complete melt down, literally melting the backside of the fuse panel. Maybe now I'll fix my old one for next time it happens... Ferrari OE replacement panels are still around, but they are not all good. The first one I got was poor and all of the contacts did not work properly, so my mechanic sent it back. Price for the box was ~$2000
Steve, Given that you can pull the plugs on the fuse panel in less than two minutes, I figure a little preventative maintenance may be better than a meltdown. If your plugs are showing some brown discoloration around any of the pins, then you know you have something to remedy before potential 'issues', or worse case, a fire! Really easy to jump a wire across the back or even change a connector as against a rebuild like mine. I'm all for the 'if it aint broke' story, but of the twelve burnt Testarossa images on Wrecked Exotics website, 8 have front fires. This can only be from serious fuse panel meltdowns as there isn't much of anything else up front to cause a fire. A few other guys on a UK forum located burnt and brown wires when they checked, and they too thought that everything was fine. Worth a look!
Ya know you are absolutely right. Will venture into the pannel this weekend. Hope I dont screw anything up.
This brings up a good point. Instead of paying 2k for a board, why not pull the pins from the board, xerox the circuit tracing and contact some of the places that make and drill boards. They advertize in the back of Nuts and Volts magazine, probably $200 for three! We should pull the connectors and spray contact cleaner as a winter project. And maybe carry larger fire extinguishers. YMMV. Jeff Pintler 89 348tb, 86tr
Good day Jeff, To make a quality printed circuit board (PCB) of this type will be a lot more than $200. The setup costs (tooling, artwork, electrical test, etc) alone for a 6 layer (the design is 5, but PCBs are made in 2 layer increments) protoype PCB (proto runs are usually 2-10 pieces) of this size is approx $2000 - $2500 (depends on the number of proto boards you wish, panelization, etc) ... and this does not include the design time (transposing/reverse engineering the physical board into a suitable PCB layout program). If it was a simple one or two layer design, you could use the xerox method to retrieve the artwork, transpose copy to copper clad PCB, manually etch, and then manually drill... However, the results would be home-brew (read this as poor) at best. Not too mention very time consuming, as you would have to do this process for every single board. Cheers, Sam
After reading the previous post, I thought that further research was needed. Perhaps I really don't know what I'm talking about. So for sport I ripped out the circuit board and indeed found two cooked pins....more about that later. The board is made by layering mylar backed boards stacked (four of them) onto an epoxy board. In studying the outermost board, it looks like 20 solder connections will allow the first layer to be removed. Something very simple to do with a solder-sucker. The pins for the relays and connectors are mechanically retained by the epoxy base so they won't fall out or mis-align. Each successive layer is registered by fuduciary (sp?) notches in the epoxy, or base, circuit board. And each mylar board has a large number etched into the copper so it is very hard to mix them up, 5...4...3...2...and epoxy base. There are companies that make these mylar boards very cheap and you just retrieve the art work with a xerox. It would also be easy to sort through the layers and replace only the bad layer...so you might be able to repair the assembly for lass than $200....alot less. Having considered the construction of the part, it is easy to understand why this board fails. I had assumed that the failure mode was high resistance in the connector contact causing heating which cooks the board. Not the case. It looks like the solder connection to a buried layer got hot enough to reflow the solder of the joint. That process could make a short to an adjacent copper trace or layer and burn to the water line. A modern five layer epoxy board would be alot safer, but you not have to buy the production equipment. Too bad an electrical engineer with access with the gear couldn't make a small run of boards. The hardest part would be finding all new pins to mount in the board. YMMV. Jeff Pintler 89 348tb, 86tr
Also, one more thought, the companies that produce boards will drill and through plate holes to connect layers. And pardon me for saying this but some of us "home brew" mechanics can do some stuff better than the factory. We have access to technology that didn't exist when Ferrari built the cars. FWIW Jeff Pintler 89 348tb, 86tr
Good day Jeff, Excellent investigative work on the TR fuse panel! Sadly the issue of generating a PCB is a bit more complex than you present. I am a Professional Electrical Engineer and I design PCBs as part of my daily work. When I was younger (read this as less grey hair), I used to fabricate my own PCBs and so I am quite familiar with the process, costs, and issues surrounding PCB design/fabrication. That being said, the issue of transposing the existing layers via xerox/scanning, etc is possible but has a variety of issues: 1. Xeroxing/scanning each layer will require some level of touching-up, as the xerox/scan is not perfect. This can be a time consuming and tedious process. 2. Maintaining accurate layer and hole registration (i.e) ensuring all layers/holes "lineup" during fabrication? 3. Manual editing of xerox/scan of layer for inner layer hole thermal reliefs. Thermal reliefs are necessary so that the solder will flow through the hole and contact the inner layers. If you did not have thermal reliefs, there would be a distinct possibility that the inner layer holes would have poor contact with the part's pin. 4. Any design improvement such as increasing the spacing between tracks, etc has to be done manually on the artwork... This would be very time consuming and increases the risk that one could introduce an error (shorted tracks, disconnected tracks, etc). 5. How does one accomodate hole drilling? The xerox/scan of the layer only gives track information... Once could manually drill, however, if the layers are not prefectly registered, problems result. Extracting hole data electronically from the scan is possible, but again this is time consuming. 6. DRC (design rule checking) has to be done manually. DRC is a combination of electrical connectivity checking (ensuring all traces are connected) and design violations (too little spacing between traces, shorted traces, etc). As I said it is possible to do the above, but given the time and effort and the probabilty of error it is certainly not the direction I would go (personally or professionally). If you wish to pursue this method, by all means do so, but be aware of the issues you face. My suggestion would be to transpose the design into a standard electrical enginneering PCB layout program. In a nutshell, one starts with an electrical schematic (since Robert has done the ground work on this already, it would be straight forward to transpose his info into a schematic). Next one acquires or extracts physical dimensions of the connectors (hole sizes, etc), other parts, and the original PCB with its mounting holes, etc. Next one generates a PCB file from the schematic and the physical part information. Finally the PCB layout software generates the necessary files for the PCB fabrication. Going this route, one can automatically: a. adjust/improve the design to accomdate higher current capabilities of the layers b. adjust thermal reliefs c. adjust hole and annular ring sizes d. perform automated DRC checking (ensuring the design is correct) e. extract industry standard automated drill information f. create industry standard layer (gerber) files g. extract solder mask files h. generate test point data for PCB testing after manufacturing i. panelize PCBs for ease of assembly (manual or automated) This is the professional direction to go and will yield far superior/repeatable results... Time wise, if one is doing only one PCB, perhaps the method you suggest may be shorter. However, the quality, testability, reliability will not be comparable. Cost wise... again if one is making only one PCB, then your method may be cheaper. However, if one would increase the volume to 10, 20, or more PCBs, the method I (and Industry) present would be extremely cost effective. One could easily absorb the setup/design costs on as a little as 20 PCBs ($2000 PCB setup + $2000 engineering costs = $4000 Amortized over 20 PCBs = $200 each) After the initial run, the cost per bare PCB could drop to anywhere between $10 - $30 per PCB (price depends upon the number of boards ordered). Cheers, Sam
To me the big flaw with replicating the boards has is that the design flaw that is causing the problem in the first place is not being addressed. The connectors. The panel itself is designed and built to accomodate a specific connector and no matter the quality of the PCB's they are attached to the connection will break down and resistance go up to the point that something will give and cause an electrical failure. Since we are talking about a total of four circuts I think it far easier, better and long term far more reliable a fix to just by-pass those four circuts. It has the added attraction of being a cheaper remedy as well.
