New horns sound great - but latch on!

Second attempt to load the image…

Your wiring is correct, so either your horn button is not disconnecting the ground when you release it, or your relay is bad and not unlatching when it should.

Thanks Carey.  Next I will reattach my old horn to see if the horn button releases ok, and do a bench test on the relay.  

Check to confirm the column isn't grounding against other metal too.

Out of curiosity I was planning to measure resistance to ground at the point where one of the two blue wires connects to the lower end of the steering tube.  However because my old horn worked just fine, I would think the column grounding and horn button work just fine.  I’ll confirm by switching back to the old horn.

So, what happens if you eliminate the relay and wire the horns with the two original blue wires that you started with??  If that makes the horn toot as expected, then the relay is suspect.

Did you use a Hella Relay with the horns? If not get one, the no name relays are junk.

Greg

Gordon I tried your suggestion with one of the new horns, it worked perfectly, as did the old horn when I reconnected it.  This would indicate to me the horn button is OK.

I did attempt to measure continuity between the spot where one blue wire is connected to the steering tube and ground.  My assumption is that the lower end of the steering tube is isolated from ground, otherwise if my sketch is correct, the horn would sound when the ignition key is turned without benefit of pressing the horn button.  Right?

First, understand my limitations. I don’t have a floor jack, so all work under the car is accomplished by me carefully sliding my head beneath the front of the car, and as required wriggling my arm (or arms) up into space available. My old analogue multimeter is held in place between my knees which I can see as I peer through the horn grill.  So I set my meter to x1 scale on resistance, touch my leads together to verify a short.  I then test resistance between the heads of two bolts on the front framework and confirm they are electrically shorted, which tells me either bolt can be used as ground.  I then reach up awkwardly to touch the steering tube with one probe while touching one of the grounded bolts I used on the previous check.  I expected an open measurement but instead measure a short!  What? I tried it several times to make sure I wasn’t accidentally touching the actual steering rod (?), so I’m confident I was only touching the outer tube (I was actually touching the end of the short screw used to attach the blue wire to the steering tube).  By the way the other ends of the two blue wires were disconnected from the relay when I made this measurement.  This makes no sense to me, but I am ignorant on these matters.   Is my assumption that all grounds are equal incorrect?

Bench test of relay to follow.

Yes Greg - I used the relay that came with the Hella horns.

It sounds like you might have the wrong relay. The VW headlight relay has a latching function, take a picture of the schematic on the top of the relay.

After re-reading it sounds like your steering column has a high resistance to ground, which would be a big voltage drop but still allow enough voltage to pull in the relay.

Bench test of relay shows relay is good, at least for the set-up I used.  For a power source I used the DC output of a 12v transformer.  I used an LED light to stand in for a horn.  I realize the LED light draws less current, so perhaps not a perfect bench test.  

Negative lead from the output of the transformer, and negative lead for the LED light were both clamped to a loose ring lug to serve as my common ground.  Positive lead from the transformer was connected to terminal 30 and jumpered to terminal 86 (so same as ignition on).   Negative lead for the light was connected to terminal 87.  When I shorted terminal 85 to my common ground, the relay audibly clicked and the light went on.  When I removed the short, the light went off along with an audible click from the relay.  Ten repeats, same results.

The most logical explanation at this point is the loose nut behind the steering wheel.  Looks like I need recheck all of my work.  

Do you know if the steering column is a VW and kinda-sorta what year that might be?  The three different VW steering column ranges are 52-67, 68-70 and 71-79.  Those three groups wire the horn up differently.

Apart from that, if the new horns work fine when wired without the relay, I would be tempted to just install them that way and call it a day.

What you did to wire in the relay should have worked, but that has proven to have something weird going on.  Just use the new horns without the relay, IMO.

.

Jon, I've been hanging back from commenting here.

The exact same thing happened to me when I changed up the stock horn in my VS to twin Hella horns, and I couldn't figure it out, either.

I've installed lots of other relays in this car and understand the theory of how they work and how to wire them up, but my relay stayed 'latched' on too, after the first button push. I tried other relays and they all did the same, and were OK when bench tested.

As you note, this circuit is already fused and the current draw for the dual horns still isn't enough to endanger the switch contacts or the wiring, so I just installed the dual horns where the old one had been without a relay and all has been peachy ever since. Keep in mind that the duty cycle for the horn is something like a half-second on and six months off, so it's not likely anything will be going up in smoke here.

I had a few theories about this peculiar behavior but never pursued checking out any of them so I'll not muddy the waters with them here. But I wonder if you might do a little test for me while everything's still apart.

Try substituting a 12-volt bulb for the horn and see if that doesn't work as expected - going on when you push the horn button, but going off properly when the button is released. I'm thinking there may be something about the way a mechanical horn works that's causing the relay to stay activated even after power is removed from the primary (trigger) circuit. Thanks.

.

@VSpyder posted:

Did you use a Hella Relay with the horns? If not get one, the no name relays are junk.

Greg

Thanks for that Greg. I’ve got a new $10 harness/relay I’ve yet to install. I’ve also got a Hella relay I didn’t need on my 911 I’ll use instead of the no name chinesium relay.

Well that makes me feel a whole lot better Mitch.  Perhaps I t’s not just me.

My gut tells me the issue lies with what terminals 86 and 85 are exposed to, not so much the type of device connected to 30 and 87.  If there is zero current flow through the relay’s coil, I don’t see how the switch can remain closed no matter the type of device being powered.  One would think that releasing the horn button would cause potential across 85 and 86 to go to zero, but I suspect otherwise.

I imagine the relay switch will only close when a certain level of current increase occurs through the relay’s coil (thus creating a certain magnetic force on the switch) and will only open if the current drops below a certain level. I further imagine the current thresholds for switch closing and opening are not the same.  If for some reason current does flow through the coil when ignition is turned on, but it is not enough to close the switch, but this current is enough to keep the switch closed when the horn button is released, our shared experience could be explained.  

But why would current flow through the coil when ignition is turned on without the horn button depressed, and why wouldn’t this low current cause the old horn to mumble?  The former might be explained by the oddity of the VW path from one side of the horn through the horn button to ground.  Maybe the horn switch changes impedance from one level to another, but even with the horn button untouched there is a path to ground.  The latter may be explained by the old horn needing a minimum potential before it bothers to speak at all.

Unfortunately at this point my garage is getting hot, my back is getting achy, and my day’s honey-do list remains intact.  I may need to set this aside for now.   At least I know I can easily reconnect the old horn of the mood to drive strikes me.

Chris here is a photo of what the diagram on the relay looks like, along with how I wired it.  

This is the relay that came in the box with the Hella horns.  Notice it says ‘made in China’. Maybe a better part of China?

@VSpyder - Greg - does my cartoon schematic fairly represent how you wire the stock horn?  The two leads that attach to the horn are both light blue, I see one light blue wire exit the loom and is attached to the steering tube, and another light blue wire enters the cabin and connects to one side of a 20a fuse.  There are two other wires sharing this 20a fuse, one is green (windshield wipers?) the other is grey, which runs somewhere towards the front underworkings of the chassis (fuel pump???).  I just want to make sure I’m not risking anything important if I skip the relay to solve my problem.

Thanks,

Jon

(By the way, my gear shifting issue seems to be resolved after a bit more fiddling of the stop plate and shifter housing.  I’m loving this car.)

Basic wiring circuit: 12v source , two terminal horn and a ground. Sometimes we all overlook the obvious or over diagnose the problem. I still fall into that trap.  Disconnect everything horning related.....Power the horn direct from the battery then ground the other horn terminal to confirm the horn works even though you know it does- checking that off.....  Then, with the horn still powered test op the with wire than runs up the steering column hollow shaft and ground that to the column to confirm op......  Add in your relay and confirm op..... Doing one small step at a time is the best diagnosis method ....as redundant as it may seem, sometimes the screwy-ist thing pops up that you overlooked.

.

FWIW, I've had very good luck with cheap, Chinese relays from Amazon. I've used lots of them and never had one fail, even in the heat and vibration of an engine compartment. And, as I said, I bench tested the ones I hooked up to my new horns and they all tested OK.

One of my theories was sort of far-fetched, but theoretically possible, so I'm reluctant to bring it up, but...

A car horn works with a little solenoid pulling a metal disc that opens contacts that break the circuit, returning the contacts to the closed position, starting the cycle all over again. This happens several hundred times a second, causing the vibration that we hear as the horn sounding. So, when the horn sounds, the current in the whole horn circuit is switching on and off a few hundred times a second. That circuit includes the secondary half of whatever relay you may have hooked up to the horn.

Inside that rely is a coil in the relay's primary circuit - the circuit you close to activate the relay and sound the horn. It's possible that a rapidly switching current in the relay's secondary could generate a current in the coil of the relay's primary circuit, and that current might be enough to keep the relay activated, even after the horn button is released. The voltage generated would have to be high enough to jump across the (open) contacts inside the (released) horn button, but if that gap were small enough...

Like I said, pretty far-fetched, but the situation is analogous to what happens in the ignition circuit of a car with points. The quick break in the primary circuit of the coil (when the points open) creates a high voltage in the secondary, sufficient to jump the gap in a spark plug.

OK, maybe I've been thinking about this one for too long.

.

true story, my 17 Por CS window switch broke,,,removed it my self, made in China,Por part # on it,,,$36 ebay, dealer wants $300 and its the identical switch..

Well, let's see what I can offer:

A.)  "I just want to make sure I’m not risking anything important if I skip the relay to solve my problem."  No, you're not risking anything.  A 20a fuse running a horn, windshield wipers and fuel pump is overkill, even if all were running (and blowing) at the same time.  You could, however, try moving the horn wire fuse to a different fuse in the panel and see if that improves things (keep reading).

B.) Very nice wiring diagram, btw

C.)  What I suspect is going on, is that you have a "ground loop" somewhere and that is providing just enough ground current through the steering column to hold the relay in once it's been activated.  You tested for this, but were confused by the results:

"I then reach up awkwardly to touch the steering tube with one probe while touching one of the grounded bolts I used on the previous check.  I expected an open measurement but instead measure a short! "

Yup - You should have seen an open there, but I'm thinking you don't have a dead short, just a partial short which is a ground going through some other thing like a light bulb or accessory.  That would give you a very faint ground, maybe not enough to make the horn blow or pull the relay contacts together, but enough to hold the relay closed once the horn button is pushed.  If you were to remove the wires from the relay and attach the two probes, one to ground and one to the stud on the steering column and then have someone push the horn button you should see it go to a dead short.  Letting off the horn button should open it up but you will see some remaining resistance to ground there, instead - Something less than a dead short (try using a different range on your ohmeter to see the difference).  I'm thinking it should always go back to a floating resistance whenever the horn button is released - Not enough to blow the horns continuously.  You could also try using a voltmeter on those probes and I'm thinking you'll maybe see a voltage there when the button is released, too, but that's another story.

At least on my car with a VW column, the column is isolated from ground and any voltage.  The steering shaft (inside the column) electrically connects the horn ground wire to true ground via the top steering column bearing.  True ground is provided by a wire connected to the steering box via a bolt in the rag joint coupler and goes up through the center of the steering shaft to the horn button.  You blow the horn by connecting that ground wire via the horn button to the steering wheel/steering shaft - Beep!  The shaft is connected to the horn wire and the column is totally isolated.

If there are any electrical things in touch with the steering column (not the shaft) that aren't supposed to be be there and are "leaching" a ground through it, that might give the condition you and Mitch are seeing.  It could be as simple as a directional light making a faint contact or something in the ignition circuit.  Your horn wire is trying to find ANY ground, and sees a faint one through a light or some other device.  I only think directional bulbs because they all come in to the directional lever switch in the steering column, but it could totally be elsewhere in the car, too.

These can be a mega-8itch to find and sometimes even harder to cure and that's why I recommend just wiring the horns up directly to the horn button and be done with it.  That's a lot better than tearing your hair out for weeks.

@Sacto Mitch posted:

OK, maybe I've been thinking about this one for too long.

I don't think so.

I like that as an explanation. Induced voltage, EMF, Faraday's Law, etc. - you don't need to understand the exactly what's happening (but Mitch seems to have offered an explanation that's at least possible) to understand that it IS happening.

Your relay is latching for some reason - my money is on Mitch's explanation.

There's nothing wrong with your schematics, Jon (they're beautiful, and you obviously know the electrical symbols). If this was an annunciator in a panel somewhere, it'd work perfectly. It's obvious (from what you are experiencing)  that you can't use any horn with any relay. That other people are using relays with their horns is irrelevant to your situation. Something is different - either the relays or the horns.

If you take the relay out of the circuit, 5 bucks says it works. If it does, you're left with 2 choices:

1. Be happy with no relay in the horn circuit
2. Buy new horns or keep buying different relays. The ones you have don't play nice with each other (my money is on the horns).

There is doubtless a way to wire this so that the coil is not in the same circuit as the vibrating horn itself (probably with a second relay) - but I'd KISS.

Don't be like Stupid Stan - this is no hill to die on.  If you try a Hella relay and get the same thing, just buy new horns or take the relay out.

@barncobob posted:

true story, my 17 Por CS window switch broke,,,removed it my self, made in China,Por part # on it,,,$36 ebay, dealer wants $300 and its the identical switch..

Yep, I replaced 4 of those $300 window switches with cheap China parts. None failed again while I owned the cars.

Here's a couple of long shots:

1) Chasing Mitch's idea, put a capacitor/diode protection circuit on the power lead to the horns. The idea would be to quench the voltage spikes from the horns back into the power supply. (see below for component recommendations)

2) If Gordon is correct and you've got a ground leak somewhere on your steering column, add a resistor inline to the horn button connection that triggers the relay. The idea would be to reduce the current to the relay trigger to the point that the ground leak doesn't latch the relay. Or you could see if you can confirm and eliminate the ground leak.

Components

I used a number of these capacitors to reduce power source interference in the past (eBay):

Panasonic FC 330uF 100v 105c Radial Electrolytic Capacitor Low ESR

Match up the capacitor with a Schottky diode to quench voltage spikes. E.g. https://www.digikey.com/en/pro...497-4548-1-ND/770973

Example circuit:

... and all of the sudden, the _"meep-meep" horn is looking pretty good.

I had an issue where a slight amount of current was going through the rubber donut thing.

Alan - I believe steps 1 and 2 have already been tested, with step two including the horn button.  One new Hella horn work just fine as a direct swap of the old disk horn.  That proved the horn worked, and the horn button worked.  No latching.  (I didn’t add the second horn to see if it would work with two, but we’ll get to that later)

Mitch - I reconnected the relay as described in my first sketch at top of the  thread, but instead of connecting the Hella horns, I connected two 12v LED lights (spotlights used for my Halloween graveyard).  I’m not sure I’m ready to work up a rationale for the results - both lights illuminated with key turned on, no need for horn button depress.  Pressing the horn button had no affect.  Key off, lights off.  Perhaps the voltage needed for enabling the LEDs is less than 12v, and maybe this proves with key-on there is some current running through the relay coil even before the horn button is pressed, and by the magic of old people we name laws after, this sends some juice through the secondary circuit.  I think you were describing something that is the reverse of this (horn actuation sends some juice back to the primary circuit), but the magic is rooted in the same principals.  This amount of juice would not be enough to make the relay go from open to closed, but it is enough to keep them closed once enabled.  Or something like that.

Stan - I did think for about three seconds how a second relay might help, but I stopped that line of thinking cold.  Yup - KISS.

Gordon - you did a much better job saying what I was trying to describe.  I did try one resistance measurement, the results convinced me to put away my meter.  With the two blue wires disconnected (these are the two wires that connected to the old disk horn, and presumably run to either side of the horn button switch for the sole purpose of connecting a ground path for one terminal of the horn), I measured a resistance that was pretty much the same resistance I got with my meter leads shorted.  I then pressed the the horn button and saw an almost unnoticeable change in meter reading.  My meter was on 10 ohm range and I was reading about 2 ohms.  This only proved I really don’t understand how the horn electrical circuit works, and probably don’t understand how to use my meter.  Any further troubleshooting on my part was therefore considered foolish.

So clearly I’m going with plan ‘no relay’.  I made two new leads with twin spades on one end, single on the other and used them to connect directly to the two blue wires that connected to my original disk horn.  Key on, no horn.  Horn press, nice twin trumpets sound off (although I’m biased to think Iv’e lost a few dBs which would be expected skipping the relay).  Horn button release, silence.  Perfect.  I carry spare fuses, so if for some reason I’m running my wipers while honking the horn while doing whatever else that grey wire is for and I blow a fuse, I can quickly recover.

I need to dress my new leads, but for some reason I don’t plan to remove the harnesses I installed and dressed that would be used for the relay option.  At least not yet.  I think I’ll let my knuckle blood dry a bit more first.  And who knows, maybe someone will come forth with a dead simple solution, or I decide to try a different set of horns/relay combination as Stan suggested. I will make sure the end of dangling leads get a squirt of dielectric grease and are then covered, especially the lead I connected to my battery + terminal.

I think I deserve a twilight drive.  Hopefully I don’t piss off too many people as I honk away.

Thanks all,

Jon

Just saw your post Mike.  For another day.  

It is possible that you are getting a residual ground or positive back-feed into the horn circuit. I don't know if I can explain it to you in any sensical way. But that's what I feel is going on.

Try disconnecting the ground lead(86 in your pic) while the horn is blowing. You'll have to isolate this yourself, and the easy fix is to put a diode(1n4001 1 amp standard diode is sufficient) in between terminal 86 and the horn switch. Try it in both directions until it works. Don't overthink this, it doesn't need to be complicated.

These diodes used to be available at Radio Shack, but try a stereo/alarm install store. If not, online is easy...

@Jon T posted:

Just saw your post Mike.  For another day.  

A lot of my stuff has kind of worked until I got around to finally fixing it. Nothing wrong with that.

Based on the low resistance meter readings when the button isn't pushed, and slightly less resistance when it is pushed, you've got  a short to ground. I'm guessing it's in the wire that runs up the steering column or somewhere in the horn button area (connector rubbing against the side maybe). I can't remember if you checked the resistances with the horn button disconnected (might be a partial short inside the horn button).

You may be able to go on forever without doing a fix. However, if it does develop into a total short to ground, your horn would just latch on. If that happens, you could pull the fuse until you could get it home.  Good luck!

This is a diagram of relay wiring alternatives:

This is a link to an article about VW horn wiring through the years appropriately called "Horn Wiring Hell": http://www.thebugshop.org/bsfqhorn.htm

Just got back in the house.  Twin Hella horns now connected directly without a relay and they work.  I’ve kept the wiring for the relay and lead to my battery positive terminal in place, tucked neatly away just in case I feel like revisiting this, after reviewing more carefully above suggestions including linked references - and thanks to everybody for your thoughts.  My tools are stowed, and the impressive amount of debris I created and Ace Hardware receipts I accumulated have been swept away.  I’m done for now.  I’ll walk the dogs, stretch my back then go for a little drive.  

For the record I have not messed with the horn button - I have no experience with that assembly and dared not take it apart, especially since it works perfectly well as long as the relay is not used.

A few of you have described ‘the wire that runs up the steering column’ or something like that.   I don’t see a wire like that, but I just might not be able to see it.  Remember I have no lift or floor jack so my probing is a bit limited.  What I do see is this.  One of the two wires that connected to to the disc horn seems to exit the loom just behind the front right wheel and runs directly to the steering tube/column (the stationary outer shroud, not the actual steering rod).  It appears to me to terminate in a spade or ring lug at the tube/column, and appears to be held in contact with the tube via a short screw.  This is what I assumed when making all above comments.

I tried to take photos by shoving my phone in the wheel well, here is the best shot:

As I type this I realize that I have assumed the wire terminates here, but I suppose it is possible the wire simply passes through the tube at this point and is in fact insulated from the tube.  It could then run up inside the tube toward the horn button bits.  If that is what I have then that would explain away the crazy low resistance to ground reading I measured which assumed this wire is electrically connected to this tube.  If the wire is in fact insulated from this tube then my measurements between this tube and ground were meaningless.  I suppose this would lead to a few other troubleshooting steps I could take…maybe next weekend.

Thanks all!

It's threads like these that interest me. They help me to open doors in my mind that have been closed for decades because of the way I go about doing things. You all think about things differently than I do.

Here are two things I'd try.

First:

1. Leave it wired like you have it, but remove the horn button wire from the 86 terminal
2. Put a jumper from the 86 terminal to the negative terminal of the battery
3. The horn will blow - then remove the wire. If it stops (it will), then you have some issue with the horn button/ground setup (you likely do, and it likely won't be fixable without a ton of work)

Second:

1. Keep the coil side of the equation just like you have it - horn button breaking ground to the relay coil, relay hot all the time
2. Temporarily run a fused power wire to the + side of the horns
3. Remove the ground wires on the horn from the frame.
4. Rather than breaking power with the relay, break the ground. This would eliminate any possibility of induced voltage in the relay.
5. In other words, try grounding the 30 terminal, and connecting the 87 terminal to the negative terminal of the horns.

All of that assumes that the horns have two terminals - a positive and a negative. If they don't - this won't work.

I'm curious now.

.

@Jon T posted:

.

...Mitch - I reconnected the relay as described in my first sketch at top of the  thread, but instead of connecting the Hella horns, I connected two 12v LED lights...

...both lights illuminated with key turned on, no need for horn button depress.  Pressing the horn button had no affect.  Key off, lights off...

OK, thanks for doing the test, Jon. That proves conclusively that my theory is horsefeathers and the problem is what Gordon, Alan, Michael et al are saying. There's some voltage in the horn circuit even with the button not pressed - not enough to make the horn sound, but enough to mess with the trigger circuit of a relay.

Unless you drive a cab in New York, you're not going to use the horn enough to warrant a relay, so I'd toss that and be happy.

Until, of course, you start worrying that your horns don't peek out through the horn grills and look like the original Bosch horns, but obsession with authenticity is a disease that progresses slowly and is relatively benign in the early stages.

For the time being, as we used to say in the '70s, a little toot is a wonderful thing.

.

The wire terminates on the column tube, which is supposedly electrically isolated from the chassis ground. Fiberglass obviously in the front as you can see.

Under the dash, there should be a clamp that goes to a piece of angle iron welded to the frame. On my car, the clamp has a plastic coating on it as does the steel frame, keeping the column isolated from the frame. Your car should be that way too. The horn button itself makes the connection between ground(the steering box/steering shaft) and the outer column, thus completing the circuit of the horn.

Most switched circuits in cars are on the positive side, the horn is simply on the negative or ground side. The horn circuit doesn't care which pole the switch is on.

  One time I had the wiring on the relay wrong, I forget which is was but, it feed either a constant ground or constant 12v to complete the horn circuit.

@Stan Galat posted:

All of that assumes that the horns have two terminals - a positive and a negative. If they don't - this won't work.

By the way Stan, the Hella horns do have two terminals but as far as I can tell they have no polarity.  The ‘helpful’ schematic on the Hella horn package shows a view of the horn from the opposite side of the terminals - you can’t see the terminals and there is zero mention which terminal should be used for ground.  The few Google searches I performed for this specific Hella horn supported the notion that polarity is unimportant.

@Jon T, the wire you show connecting to the tube stops there.  Then, the ground path goes up through the tube itself to the horn button. When you press the horn button a connection is made from the edge of the horn button (with the ground from the tube) to a terminal on the back of the button. A wire connects to the terminal on the horn button and then it goes down through the center of the steering shaft and makes a connection on the lower side of the rubber steering coupler. That allows the ground path to continue to the steering box, front beam, etc.