1937 Buick 46s headlight blinking

[John Jacques]

Hello again, this weeks problem: as I'm in the final stages of restoration, when the headlamps are on for about 15 seconds they start flashing, as if I had 4 way flashers turned on. Not only do the headlamps flash, the fender lamps and tail lamps also flash. I double checked if all the lamps were correctly wired and that they had all good grounds. Don't know what else to check. Any suggestions????? Thanks for any help that you can provide.

Jakes

[Rock10]

Do you have turn signals installed?

[Ben Bruce aka First Born]

The headlight switch probably has a resettable breaker built in. It is overloading. Either a short or too many bulbs on the circuit. Unhook one thing at a time until breaker stays on.

 

  Ben

[John Jacques]

Rock10 no, I do not have turn signals installed.

Ben, I will do as you suggested. Is it better that I try taking out each bulb separately or disconnect each wire separately from the headlight switch? I don't understand "until breaker stays on".

[Gary W]

The 1937 Light switch has a "thermo circuit breaker" mounted to it. Current in excess of the normal lighting current causes it to heat up, which pulls the points apart, thereby reducing the current load.

In the manual, it states the points will stay closed indefinitely at 25 amps.  But at 38 amps, it will stay closed only one minute and then open.

Check out the manual pages 210 and 211.  It goes through three diagnostic steps to figure where your short is.

 

"Until the breaker stays on" or in other words, "Keep the points closed"

 

Pg. 209.  Top of right column details the detachable thermo circuit breaker

 

 

 

Here's the detail blown up

 

 

In real life, its that extension part on my fingers

 

 

Page 210 of the service manual.  All about the lighting

 

 

Pg. 211.  Finishes the diagnostic tests.

 

Hope it helps!

Gary

[Bloo]

Ben Bruce nailed it. There is a thermal circuit breaker. Either the breaker is too weak or the current is too high. They will trip, reset, trip, reset all day long. Either the bulbs are way too big or there is a short circuit somewhere. The breaker is at the headlight switch on a 37.

 

Don't rule out the dimmer switch as a possible cause. On some dimmer switches water damage has caused them to start leaking current to ground.

 

 

 

 

Edited August 21, 2021 by Bloo (see edit history)

[MCHinson]

Recent personal experience tells me that you should start by checking the headlight bulb socket wiring. It is easy to have the springs in the sockets shorting the wiring to ground causing the circuit to pull too many amps and cause the circuit breaker to cycle on and off flashing things.

 

If you happened to install new reproduction headlight terminal blocks https://bobsautomobilia.com/shop/electrical/headlight-terminal-block-6-terminal-1936-39-ht-369/ I would check them very closely. I recently had a problem with one of the screws on one of those cutting through the insulating tape that I put under it as a precaution whenever I bolted the headlight onto the body. That one had me scratching my head for a few days before I figured it out. 

[John Jacques]

Thanks to all for your suggestions. Later this morning I'll start checking it out and let you know the status. It's guys like you that make repairing/restoring these old gems rewarding and fun. Keep up your good work!!!!

[Jack Worstell]

The circuit breaker in our 1937 Special had weakened over time

 

So we by-passed it   and added in a new breaker

 

Jack Worstell

[John Jacques]

Well I checked the socket springs in the headlight and those were fine. Here's whats happening: the first position of the headlight switch (pushed in all the way) nothing lights up or blinks (as it should), the second position (pulled out one notch) the fender lights and the tail light/license plate light comes as it should. In the third position all the lights blink, this is with the high beams off. With the high beams on only the tail lights come on, no headlights, fender lights or license plate lights. In the 4th position with the high beams off, all the lights work fine, but with the high beams on, the left headlight and tail lights blink, no right headlight or fender lights. What do thing the problems is ?????

[Bloo]

Does this car have a sealed beam or some other headlight conversion? If so, how many candlepower or watts?

 

Try it with the headlight bulbs removed and see what if anything changes compared to your last post.

 

 

[Gary W]

So rule out fender lights, tail lights and license plate wiring as they all light up and do not overload the system.

You may have a bad floor switch, it may be shorting out when depressed.

Other than that, go over every inch of headlight wiring, from the bulb, through the socket to the terminal block and be sure there are no shorts, nothing touching the frame or fenders...

 

These headlight switches are complicated and if they are not assembled / restored properly (the internal "triangle" contacts), they can cause all kinds of havoc.

[dibarlaw]

Yes Gary as we well know!

[John Jacques]

Well, I checked out all the wiring for shorts and found all was OK. It seems that the floor switch maybe the culprit here. I am going to order that switch and go from there. Thanks, again for your inputs.

[EmTee]

If you have an Ohm meter, check for continuity between the terminals and the switch body.  If so, that's the short-circuit culprit.

[John Jacques]

I opened up the floor dimmer switch to see what it looked like. It was pretty clean inside with no evidence of rust or corrosion. I did check it with a Ohm meter and it operated as it should. Therefore, I did not order a replacement floor switch and I am back to scratching my head. I'll probably open the headlamp switch (again) and check out those three triangle once more. Originally, I did not see any difference between the three triangle, but did notice that they all had different numbers on the bottoms. Does anyone know the correct location of this triangle, or does it make a difference?  I also saw that there were three springs. One of the springs was a little shorter than the other two. I installed that shorter spring in the center position, as that position had a brass insert with a ball bearing. I have to admit in attempt to correct this headlight problem and prior to disassembling the headlamp switch the first time, I ran 600 grit wet/dry sand paper through the points on the thermo circuit breaker, thinking that would clean them up.

[MCHinson]

12 hours ago, EmTee said:

If you have an Ohm meter, check for continuity between the terminals and the switch body.  If so, that's the short-circuit culprit.

 

Did you try this? This would be your best diagnostic test at this point. If none of them are shorted to the headlight switch case, then maybe your circuit breaker is too weak. If one of them is shorted, you can then disconnect the wire attached to it. If the switch terminal screw is shorted to the switch case, the problem is in the light switch. If the disconnected wire terminal is shorted to ground, the problem is in that circuit.

[Ben Bruce aka First Born]

Perhaps I missed something.  Have you tried the headlight switch ON with the wire from headlight switch to dimmer [ floor ] switch unhooked?  Joy or no joy?  Hook the wire back to the dimmer switch. Now what?  change the dimmer from high to low beam. What now?

 

  Ben

[Bloo]

^^Do this.

[EmTee]

...or try with the dimmer switch: (1) bolted to the floorboard and (2) isolated from the floor/chassis.

[Bloo]

And/or that ^^

 

These can be difficult to troubleshoot over the internet. The breaker is simply a bi-metallic strip and a set of points. Construction varies, and I don't remember exactly how the 30s Buick on is made, but there is some electrical resistance. They might have just run the current right through the bimetal, or there could be some resistance wire wrapped around it. Whatever they did, it is series resistance and it gets HOT, for the same reason a bad battery connection gets hot. When it gets hot enough, the bimetal goes "PLINK!!" and opens the points. The headlights go off. This basic idea was used clear into the 80s at least on the headlights on some american cars.

 

The idea is you have just enough resistance that the rated current of the breaker makes just enough heat to push it over the edge and open the points. They can get weak from age (like Jack Worstell's), but that doesn't usually happen. Dirty points could also contribute more resistance and heat, tripping the breaker early. That happens more often, but the most likely scenario is that it is actually overloaded. You need to figure out why.

 

Now you might be thinking, series resistance? That sounds like a horrible design. Isn't that going to make the headlights dimmer than they need to be? The answer is yes, but in a 1930s context it is probably the most reliable thing they could have done, the alternative being glass fuses. If the breaker kicks out for some reason it will TRY to come back on. You might appreciate it if the headlights go out in the middle of a corner at 60MPH. It has happened to me. Glass fuses sometimes melt because of resistive connections, it is super common in 60s American cars. A melted glass fuse won't TRY to come back on, but on those cars the headlights are on a circuit breaker, just like this Buick. The glass fuses are for everything else.

 

Stuff that could be causing this if the breaker isn't bad (and the breaker probably isn't bad):

 

1) Wire bare, shorting to ground somewhere

2) Bulbs too big, too high wattage. Sealed beam conversions, H4 conversions, extra-bright taillight bulbs, etc. or some combination of the above.

3) Corrosion causing electrical leakage to ground somewhere, usually the dimmer switch, because they get wet. Offhand, 37 Buicks have junction blocks up at the headlights that probably get wet too.

 

So, if there are no oversize bulbs or accessories present that would trip it, you probably need to disconnect things selectively until you find which wire draws too much current. This can be fiddly because it is the TOTAL current that trips the breaker, and disconnecting something might make the breaker hold even if it isn't the right thing, and when you disconnect something you remove part of the current draw.

 

Fortunately, something like a shorted wire or a dimmer switch leaking to ground will usually trip the breaker pretty definitively.

 

 

Keep trying. Good luck.

 

 

Edited August 25, 2021 by Bloo (see edit history)

[EmTee]

Ideally you'd have an ammeter that could be used to measure the actual current at the time the breaker trips.  Then you could compare that to the breaker's rated trip current.  That would tell you whether it's a faulty headlight circuit breaker or an overloaded circuit.  Another option would be to buy a heavy duty fuse holder and an assortment of fuses (e.g., https://www.amazon.com/MCIGICM-Inline-Fuse-Holder-Blade/dp/B081DHT8Y7/ref=asc_df_B081DHT8Y7/?tag=hyprod-20&linkCode=df0&hvadid=430996963124&hvpos=&hvnetw=g&hvrand=15934156162568889549&hvpone=&hvptwo=&hvqmt=&hvdev=c&hvdvcmdl=&hvlocint=&hvlocphy=1023225&hvtargid=pla-902657587165&psc=1&tag=&ref=&adgrpid=103875137927&hvpone=&hvptwo=&hvadid=430996963124&hvpos=&hvnetw=g&hvrand=15934156162568889549&hvqmt=&hvdev=c&hvdvcmdl=&hvlocint=&hvlocphy=1023225&hvtargid=pla-902657587165).

 

You could temporarily wire this in series and like an ammeter, use different fuses to estimate the amount of current flowing.  In other words, if a 25A fuse blows and a 30A does not, then you know the current is between 25 and 30 Amps.  Compare that value to the max current for the breaker, or the value based on the sum of the specified lamps in the headlight circuit ( I=P/V).

[Ray34]

If you suspect the cct breaker on the light switch is faulty or weakened these units are readily available in different amperage ratings.

[John Jacques]

Sorry that I have not gotten back to this topic as I had to take a break, due to other issues going on. There is sooooo much information that was provided from all of you that it will take me some time to perform these tasks. Please be patient and I'll reply with my findings as soon as I can. I really do appreciate all of the responses.

Jakes

[John Jacques]

Well, here it is nearly 6 months since my last entry. On and off during this period I attempted to resolve my light flickering problem. After taking everything apart and checking all connections I located the problem. It turned out to be the wiring within the right head lamp socket. The springs and wiring in that socket corroded. I replaced the wiring and cleaned the springs. Viola, everything is back working as it should. I apologize for taking so long, and hope I never have to go though this again.

Thanks for everyone's help.

[Jolly_John]

And, the trophy goes to Matt Hinson!! See his post quite a ways up on this thread. John