Hydraulic Brake Light Switch Failure Investigation                             11/07/R. Kwas

For those interested in a detailed explanation of why I recommend upgrading...

Brakes and Brake Lights in General: 

When you apply pressure to the brake pedal, you want the vehicle to slow...sometimes you want this to happen fast (like when at the last moment, you decide against blowing through that yellow traffic light!)...sometimes slowly and gently (like when you’re taking Grandma Bessy and the ladies to the coffee klatch on Sunday afternoon).  In addition to slowing the car in both cases, you want the Brake Lights showing also, to advise those behind of this (speed doesn’t cause accidents, it may contribute to them, but more so do unpredictable driver – and vehicle – actions I believe, and a slowing vehicle without brake lights is definitely doing something unpredictable to those following!).  This is where the Brake Light Switch comes in...it’s supposed to turn ON the Brake Lights when we apply the pedal...to give those following an early warning that we’re slowing, before they take notice of it otherwise (like when they are about to relocate your bumper).  ...trouble is, the hydraulic brake system pressure sensing switches installed on vintage Volvo vehicles up to the late Sixties just seem to not be too reliable...and we're talking about the OE Bosch switches here, not some China trash!



 

After about the fourth time of replacing these switches on various 122 and 1800 vehicles (brake system bleeding required!), I got pretty tired of the action, and started to appreciate why the factory themselves had done away with this unnecessarily complicated design...so developed my own solution based on the much simpler mechanical “Pedal Position Sensing Switch” similar to the later factory designs. 

Theory of Operation of the Hydraulic Switch:  According to the markings found on an OE switch.  These are rated for 1-3 BAR (15-45 PSI), so considering the hundreds of PSI possible in a hydraulic braking system, one would expect the switch should work reliably. 

Considering the range of working conditions, namely everything from slow rising and low pressure, to fast rising and high pressure, suggests that the requirements for a switch to work reliably throughout the possible range of operating conditions could get fairly complicated...and the fact that so many switches fail to satisfy, suggests that maybe it’s tough to meet all of those requirements reliably...

How much engineering and reliability can one expect to be in a $20 switch?  Cutting open a failed switch answers that question with:  Actually quite a bit! 

Pictured is an OE switch, which failed very soon after installation, sectioned with a bandsaw in order to try to figure out how they work, and maybe once and for all, get to the root cause of the problem.  [From a young age, I figured out, that much curiosity could be satisfied, by disassembling stuff...a lot of that stuff didn’t ever get put back together, but understanding how it was assembled and how it worked, or at least was supposed to work I figure, was worth the price of admission!  Now, I leave working equipment alone and try only to disassemble failed things in order to figure out why they failed.] 

In the graphic below, the internal construction and function are clear.  The beefy rubber diaphragm (green) seals against leaks from the hydraulic system (pink)...certainly important.  The diaphragm is allowed to expand out under brakefluid pressure from the “wet side”, and transfer about 0.100” maximum of movement (linearly related to the magnitude of brake pressure), by way of the non-conductive transfer pin (gray), to the free-floating contact plate, (yellow, horizontal), located on the “dry side”, and which is normally held away from the surfaces of the two fixed contacts (yellow, vertical), molded into plastic (blue), by its own return spring. 

One thing I immediately notice is that there is no provision for a snap or wiping action of the contacts in the mechanism.  This standard technique for assuring reliable contact function is conspicuously absent!  


A hydraulic brake light switch in action! 
Source:  Swedish Embassy Analog Technology and Wowie Effects Laboratory Limited (SwEATWELL). 

Intended Normal Function: 

Contact Closure (Making) Action:  As the hydraulic system pressure comes up and overcomes the diaphragm, and the force of return spring, the contact plate is moved towards fixed contacts by way of the transfer pin (typically slowly, if in a normal, non-panic braking situation), finally completing the circuit as contact is made, but not after some amount of contact bounce and resulting arcing (the slower the contact closure, and the more current, the more arcing to be sure). 

Contact Opening (Breaking) Action:  Upon release of the pedal, the diaphragm returns to its rest position, and the return spring pushes the CP away from the contacts, breaking the circuit.  Again, fairly slowly and accompanied again, by a certain amount of arcing. 

Slow Action:  All of this slow action at the contact results in a certain amount of contact bounce and arcing.  This is a distinct disadvantage of this non-snapping action design:  Since both making and breaking actions of the contacts happen relatively slowly, much more contact arcing and carbonization can, and will, take place than if the actions were to take place quickly than with a snap mechanical action (providing a mechanical Hysteresis), or with a wiping action of the contacts (providing a cleaning action).

Microscopic Inspection:  Upon close inspection of the contact surfaces these are both found to be in remarkably good condition indicating extremely little use...in fact, one contact area (left in pix) on the CP and it’s respective mate are virtually virginal, showing zero melted metal or carbonization.  BUT, the story is quite different at the other (right) contact area.  Although it too shows very little use, the supporting plastic of the fixed contact has been heated to the point of melting, and bubbling up to become proud of the contact, preventing the CP it from making contact the next time (indeed, this is further confirmed by holding a straightedge across the fixed terminals...the straightedge is kept raised above the contact surface on the right contact and external switch terminals show no continuity).  The actual amount of contact use, which can be deduced from the carbonization and deformation, is extremely low on this contact also, but melted plastic is still adhered next to the actual contact area, so this proves pretty well, that the CP was closed or close to its two fixed mating contacts as the melting occurred. 


Fixed Contacts.  Virginal on left, overheated, clearly showing bubbled up plastic on right.


Contact Plate CP, Showing virginal contact mating surface on left,
and only the slightest evidence of use on right.


Detailed CP.  Melted Plastic is evident above and below contact line.

 

Switch Failure Mode:  The fixed contact was definitely overheated!  Possible explanations for this are:  Overheating due to prolonged arcing (the most probable scenario, given the possible sloooow contact closing speed).  Or, I2R resistive heating, caused by a poor connection to the spade terminal (unlikely because it is new and clean, and can be assumed to have been even cleaner when the switch was in service some time ago), or by a bad (corroded) crimp at the mating connector (certainly possible with an old vehicle, but unlikely in this case because there is no evidence of melting surrounding the terminal entrance into the plastic from the outside, but only within the switch, and since temperature and resulting melting would certainly be more severe closer to the heat source, it can be deduced, that contact arcing internal to the switch was therefore the heat-source.


Why this switch failed!

...so from the minimal contact use evidence, it would seem that quite soon after installation and use, maybe even the first (sloooow) gentle pedal application, the contact heated enough due to arcing, to cause the surrounding plastic bubble up...the rest, and the switch function(!), are history!  Exacerbating factors causing increased arcing and heat would be particularly slow pedal application and/or having increased the load current by increasing lamp wattage or by adding a third brake light.

Conclusion:  Switch contact failed due to overheating, resulting from arcing due to slow contact closure, which should really have been expected!  The switch may be a quality component, but from an overall engineering standpoint...this is kind-of a sucko system design!

...so the following explanation I posted to the British Forum before doing this post mortem was actually pretty close...and I stand by it.  What I would add at this point is that the hot plasma generated during arcing is causing more damage from heating than just the carbonization I originally gave it credit for, and that aging and fatigue of the diaphram would seem to actually have little to do with the failure mechanism.

--------------------------

Posting Link [ http://www.volvoforums.org.uk/showthread.php?t=45938 ]:

...“The pressure sensing hydraulic switches are just not reliable...the actual electrical contact area closes under slow-speed, low closing force...a perfect recipe for non-function from an engineering design standpoint... it's a poor situation...the pressure comes up very slowly during normal braking, and thus puts the switch diaphragm through a linear (flexible) region which closes the contact quite slowly, typically not breaking through the carbonized surface (which occurred when the contact was last carrying current and opened), when it should ideally snap closed, breaking through this non-conducting layer...add to that aging and relaxation of the diaphragm, and most of the switches quit...some quite soon after installation...add to that the unavoidable delay factor of having to make pressure in the system in order to show brake lights, and all manufacturers including Volvo in about '67, changed over to mechanical switches. 

 Regarding the "sludge" one poster found in the switch...I just can't believe this explanation...IMO, "sludge" is not likely what keeps the switch contacts from closing...in the first place, "sludge" would certainly transfer the pressure to the contact diaphragm just as easily as clean fluid (I guess that means that just as a fluid is non-compressible, so is "sludge")...further, no "sludge" is capable of withstanding the high pressure which occurs in the system. I believe the failure mechanism is as described above. 

 ...bottom line IMO is to just upgrade to a mechanical pedal position sensing switch, loose all of the disadvantages, and get all of the advantages: Reliable Brake Light operation the instant the pedal leaves its rest position. You don't even need to bleed the system!”...

--------------------------

Solution Options:  I’ve seen forum postings where owners have used the hydraulic switch to control a slave relay...believing that the contacts of the switch are overworked from a current-carrying standpoint.  From the current-rating-only standpoint this is not the case.  The contacts are sized to be easily capable of the 3.5Amps = 42W / 12V, load current (including double that for a typical incandescent inrush current), but from a systematic view, which takes into account the slow closing speed, this is actually a reasonable solution in that the current going through the contact and responsible for the heat generating arc would be greatly diminished if a relay, with its milliamps of load current opposed to brake light filaments, were the actual load on the contact.  ...so I would grant, that a relay might be a solution.  I would suggest that a fly-back diode be placed across the coil in order to quench the inductive relay energy which would also cause contact arcing.

Another possible solution would be to convert the brake lights themselves over to LEDs, gaining the advantage of a marked decrease in load current being switched by the contacts. 

It should be noted, that although both of these solutions would seem to successfully reduce circuit current and so address the electrical issue, which would prevent arcing and save the contact, they do not address the delay due to having wait for hydraulic system pressure.  Being an electrical guy, I could have simply also worked up a relay solution, but after recognizing this delay, and wanting to minimize it, and also seeing this as a way to improve the design, I chose a solution similar to the factory, and finally decided on the “Pedal Position Sensing Switch” design for myself, eventually cleaning it up and making it available for the various vehicle models as the LINK: SwEm Brake Switch Upgrade Kit. 

...my advise for those who steadfastly refuse to upgrade to this superior design:  Stomp on those brakes, or there’s a good chance that your brake lights are not lighting!

Silicon Fluid Compatibility Issue?  Finally, at the time of this writing, I can not yet offer any explanations why the switches seem to be even less reliable in DOT5 (silicon) fluid systems (Link to Tech Article:  Amazoning with Silicon Brake Fluid  ).  I would seem that the type of fluid on the other side of the rubber diaphragm should have little if anything to do with the failure mechanism observed and explained here.  What is known, that DOT5 fluid is more compressible (at elevated temperatures)...could it be that the compressibility factor at normal temperatures is somehow contributing to the slow pressure rise and indeed causing even more of the failure mode described here?  Watch this space for additional info on the subject as I develop it.

Comments as well as failed switches (for post-failure examination) are welcome.  In fact, if you send in a failed switch with your order and check for a Brake Switch Upgrade Kit, you can deduct $6 from the price of the kit!  This should get me a batch of dead switches with which to get some more good info.  This offer is good until further notice (I don’t need that many either...I'm pretty much convinced by my findings, but it would be nice to see a similar failure condition on a few more switches...and I wouldn't mind having further support for my conclusions).  Ron

-----------------------------------------------

Legal Notices:  The terms Volvo and Bosch are used here for reference only.  I am not affiliated with Volvo…although I (do also) Roll…or Robert Bosch.  Close cover before striking.  Use your directional indicators….  It’s a car, not a phonebooth!  Your mother wears army boots, blah, blah, blah…

The contents of this article are Copyright © 2007 by Ronald Kwas.  You are welcome to use this article and its contents for non-commercial purposes.  But if you copy and republish it, whole or in part, without giving credit to the author, or linking back to the Sw-Em site as the source, you’re just a lazy, scum sucking plagiarist!  Go write your own damn article!  

Any comments on this information are welcome (and will be given a careful technical consideration, scrutiny and the odd wisea** comment if I can possibly work it in!).  Ron 

B A C K! . . .to Tech Articles...