panhead sealed beam headlight

[flthead]

turns out my non-charging problems are due to the wrong sealed beam headlight on my '50 panhead drawling about 9 amps . this bulb is the common 6 volt type for cars # 6006 [50/40 watts], this is what's in factory shop book "the headlamp is a sealed -beam type ,specially designed and made for harley-davidson motorcycles. When replacementis required ,use only the prescribed sealed-beam unit . Do not attempt to use an automobile sealed-beam unit because the current requirements for a motorcycle are much less than for an automobile and damage to battery or generator will result" now if you use this #6006 [ common 6 volt replacement lamp] bulb at night or with lights on there isn't enough amps going to charge battery ,so in a short time the battery drains , this is on the 3 brush 32e generator set up . now where can i get or who supplies the right wattage bulb as a replacement to the factory specs so battery stays charged while lights are in use on these panheads using 6 volt sealed beams?

[flthead]

responce from Ronald V. Papasso [seaboard editor]

I am glad you understand the problem and I couldn’t have written it better. I am no expert and am just feeling my way around but I agree with what you wrote. I goofed off this afternoon at work writing this to give you some more information. If this is too long winded skip around to where I defined the electrical characteristics of the 6V head light bulb and what you need to look for in a sealed beam. I apologize ahead of time if I dribbled on too long.

The first thing is how all the terms relate with regards to WATTs, VOLTs, AMPs and CANDLE POWER (i.e. CP). First off I don’t have a clue how to relate CANDLE POWER to WATTs, VOLTs or AMPs but I don’t think we really don't need to know that other to say that the bigger the CANDLE POWER the brighter the light.

WATTs is total power consumed and can directly relate to VOLTs and AMPs. WATTs equals VOLTs multiplied by AMPs. The following examples will better explain this.

A 6-VOLT headlight that draws 8 AMPs equals 48 WATTs. That’s simply 6 times 8, which equals 48. If you know the VOLTs of anything and the AMPs you can easily multiply them together to get the WATTs.

Changing things around a little if you only know the VOLTs and WATTs you can divide WATTs by VOLTs to get the AMPs. The following example will clarify this:

A 6-VOLT, 60-WATT headlight will use 10 AMPs. That’s simply 60 divided by 6, which equals 10.

Now let’s put some real numbers to this. The headlight you have is a 50-WATT high beam and a 40-WATT low beam. The WATTs are rated at 6.2 VOLTs as marked on the headlight. The AMPs that this headlight requires at 6.2 VOLTs is:

50 divided by 6.2, which equals 8.06 AMPs for the high beam

40 divided by 6.2, which equals 6.45 AMPs for the low beam

How this relates to the 32E generator is as follows. The 32E is initially setup with the lights off and the third brush adjusted so the generator output is 4 AMPS. This is the right amount of current output for the generator to provide enough current to the ignition coil and put some charge into the battery. Therefore in order to keep the engine running and put a charge in the battery the generator has to supply at least 4 AMPS for this job.

When the headlights are turned on a second generator field coil is energized boosting the generator output from 4 AMPs to 7 1/2 AMPs. This is a boost of 3 ½ AMPs and the engineers at Harley assumed that for the standard lighting load, i.e. headlight, taillight and speedometer light, that only 3 ½ AMPs more would be needed from the generator for the lights.

With the headlight you currently have it is clear that a 3-½ AMP boost will not keep the battery charged if the headlight needs 6.45 AMPs (low beam). You are 3 AMPs shy for just the headlight, let alone the load of the taillight and the speedometer light. This means that with the lights on the battery is actually discharging at greater than 3 AMPs to make up those extra AMPs for the headlight, taillight and speedometer light that the generator did not. This is even worse with the high beam since that draw is 8.06 AMPs.

I measured the current that the old style headlight bulb draws. The light bulb I measured is marked 32 CP/21 CP. I believe this means that the low beam is 21 CANDLE POWER and the high beam is 32 CANDLE POWER. Since I do not know how to relate CANDLE POWER to VOLTS, AMPs or WATTs let’s ignore that for a minute.

What I could measure is the current the bulb consumes. The low beam draws 2.5 AMPs at 6 VOLTs. In terms of WATTs that is 6 multiplied by 2.5 and equals 15 WATTs. You can see that with the 32E generator the boost of 3 ½ AMPS when the second generator field coil is energized easily covers a headlight draw of 2.5 AMPs (low beam) leaving 1 AMP for the taillight and speedometer light. I did not measure the tail or speedometer light but I think they are around 0.5 AMPs each. If I am correct that would be a total of 1 AMP for the taillight and the speedometer light. Putting it another way the numbers all add up:

4 AMPS to keep the engine running and charge the battery
2.5 AMPS for the headlight (low beam)
0.5 AMPS for the taillight
0.5 AMPS for the speedometer light

Adding it all up equals 7.5 AMPs total needed. That’s exactly what the 32E generator is designed to put out when the lights are turned on.

Of course all these are rough calculations and you can see how the 32 CP/21 CP headlight bulb and its low draw of 2.5 AMPS works great with the 32E generator.

Palmer’s book said the sealed beam 6V Harley light was also 32 CP/21 CP. It is my assumption that if the 32 CP/21 CP bulb draws 2.5 AMPs for the 21 CP low beam then the sealed beam 32 CP/21 CP light would also draw the same to produce an equal amount of light.

This tells me that you need to find a sealed beam light matching the current and power rating of the 6V bulb. Since I did not measure the high beam current draw of the bulb here’s some rough numbers to shoot for based on the bulb’s low beam characteristics:

21 CP (low beam)
2.5 AMPS (low beam)
15 WATTS (low beam)

Get close to any one of these with a 6V sealed beam headlight and we can work with it.

You can also see the dilemma with setting up the 32E generator for the sealed beam headlight you have. If we crank the third brush all the way to maximum generator output you may almost get enough current out of it to keep the lights on but not charge the battery. The penalty is that when you don’t have the lights on you will surely overcharge the battery and boil out the electrolyte.

For the 32E generator to work with your sealed beam headlight the third brush would have to be adjusted to get more AMPs (if at all possible) out of the generator. Working backwards let’s calculate what the generator output would have to be to use the 50W/40W sealed beam headlight:

4 AMPS to keep the engine running and charge the battery
6.45 AMPS for the headlight (low beam)
0.5 AMPS for the taillight
0.5 AMPS for the speedometer light

This equals 11.45 AMPs, rounded to 11.5 AMPs, total generator output with the lights on. Given the second generator field coil boosts the generator output by 3 AMPs when the lights are turned on means the third brush has to be set with the lights off to give a generator output of 8.5 AMPs. Since the proper setting with the lights off is 4 AMPs it is unlikely the third brush set to maximum will more than double the generator output with the lights off from 4 AMPs to 8.5 AMPs. It is also unlikely that the 32E can put out 11 ½ AMPs with the second generator coil energized.

However, if you could achieve 11 ½ AMPs you would have enough for the 40W sealed beam. But, and here’s the catch, if you then drove with the lights off the generator would be putting out 8.5 AMPs, not 4 AMPs. That means you have 4.5 more AMPs going into the battery and this will heat up the battery, boil off the electrolyte and ruin the battery.

I don’t know the way to work this out yet or if there is even a way to use the 50W/40W headlight with the 32E generator. It may be that this is the only 6V sealed beam headlight available now and that’s the limiting factor.

One interesting side note is that when I was working with Dan’s Knuckle his generator seemed to give more than a 3 AMP boost when the second generator field coil was energized when the lights were turned on. I don’t remember exactly what the boost was but what comes to mind was that was 5 AMPs higher than I expected out of a 32E. I think Dan said it was a model 48 generator. I don’t know the differences between the 32E and a model 48 with regards to current output but something was different with his.

With the third brush turned to the minimum setting and the lights off the generator output was per the book at around 4 AMPs. When the lights were turned on, energizing the second generator field coil, the output was around 12 AMPs instead of 7 ½ AMPs, a boost of 8 AMPS instead of the 3 ½ AMPs expected out of a 32E.

This meant that when the lights were on the battery was getting an extra 5 AMP charge that was overheating the battery, boiling out the electrolyte that over time ruined the battery.

To solve this problem I put some resistors in series with the light pole on the ignition switch and the wire going to the light pole on the generator. This reduced the voltage on the second generator field coil from 6 VOLTs to around 4.5 VOLTs when the lights are turned on. Reducing the voltage on this field coil reduces the current through it that reduces the magnetic flux that reduces the generator output. Since this generator field coil is only energized when the lights are turned on it did not affect the third brush setting when the lights are off. Now when the lights are on the generator output is close to the specified 7 ½ AMPs for the standard lighting load.

Before you go searching for model 48 generators there are things I just don’t know:

1) I don’t know for sure if Dan’s generator is a model 48 or not.

2) Even if the generator case or tag says it’s a model 48, does Dan’s have the correct field coils in it?

3) Does the stock model 48 generator have a higher current output capability than the model 32E, if so, how much more? Remember, you need about 11 ½ AMPs plus a little more since you have those bullet lights in the back that I have not included in all the above calculations.

One other solution to buy you some miles at night would be to try a shunt regulator when I build some more (I need to step this project up!). This will not make the generator put out more current but it will allow the third brush to be set to the maximum to get more out of the generator. It may not be enough to charge the battery with the lights on but may just make ends meet to get you home at night. The shunt regulator will prevent the battery from being overcharged when the lights are off.

[rashd]

Incredibly, googled the part # and produced this from Georges World Wide, May '06 (pricey, but interesting):

71. info@ace-ltd-usa.com message: FOR SALE: 50 plus linkert bodies and bowls M14 up to M344 call for your needs - price varies. NOS Guide sealed beams T3 Motorcycle imprinted 6 volts Harley part number 67717-48 - $200 while they last! 1948 Harley EL cases, repaired, matching numbers, good VIN, no title - NH bill of sale, not junk - $2500. 47 WL cases matching numbers - call. "WR" left side case - call. Harley 37-39 UL tanks, no shift gate otherwise nice - $350. Harley 36-39 winter windshield frame complete, no canvas - $250. Guide Cycleray headlamp block lettered, ground screw, near NOS reflector, NOS #9 lens block lettered - $500. Complete 1955 Harley panhead, rebuilt motor, rebuilt trans, perfect speedking bags, correct spots, marker lamps, etc, too much to list - $17K. Harley 1947 only tombstone tail lamp cracked red glass - $200. Call Pete 603 396 8355. See you at Harmony! Thanks George!

I'm curious. What pointed you to the culprit lamp?

RAS

Edit:

http://cgi.ebay.com/ebaymotors/Harle...QQcmdZViewItem

$50 cheaper

[flthead]

jefferson meet .... more homework!

Just got back from the Jefferson meet ..and a lil more homework on these motorcycle sealed beam lights ,turns out the Indian guys have been having the same problem with the sealed beam lights from who I talked to about it. Now what pointed me to .. or should I say who pointed it to me is the seaboards own Ronald. "Mr. gadget" Papasso. turns out the 6vlt #6006 lite draws to many watts with the lights on and will not recharge battery while headlight draws to many amps away from the battery .... so until I find the motorcycle type sealed beam lights, either i keep the ol pan home at night or get some energizer batteries for the flash lite I'll be duck taping to the glide trees ... between the duck tape an that damn energizer bunny I might just keep going an going

[flthead]

responce from Ronald. "Mr. gadget" Papasso

Had a long day fooling around with electrical stuff on my Harley. I also learned something unexpected about my springer light that you may want to post. Get your favorite beverage, this is going to be a long one!
The springer bulb is 32 CP (i.e. candle power) high beam and 21 CP low beam. In order to get this orientation the bulb filaments have to focus correctly which means the bulb has to be oriented in the reflector and the reflector oriented to the headlight bucket. The reflector is oriented to the headlight bucket with a tab on the reflector that fits into a notch cut into the top of the headlight bucket. The pins in the reflector orient the bulb to the reflector. The bulb pins are arranged with 2 pins at the top and one pin at the bottom. With that orientation the 21 CP filament focuses the light low and the 32 CP focuses the light high. Simple.

I have a reproduction reflector and by accident I discovered the low beam is the 32 CP filament and the 21 CP is the high beam. The pins for the bulb are 180 degrees different than the stock reflector. The repro reflector has one pin at the top and 2 pins at the bottom. With this orientation the 32 CP focuses as the low beam and the 21 CP focuses as the high beam. Completely opposite than the stock reflector!

How I found this was while making some amp readings with the springer setup and I noticed that my low beam draws 1 amp more than the high beam, totally opposite than would be expected. This did not make sense since the high beam has to draw more current than low beam to produce more candle power. It seemed to me the bulb was upside down. I have a stock reflector with my basket case and I compared it to the repro reflector. The bulb pins are 180 degrees different. The repro reflector has one pin at the top and two pins at the bottom, 180 degrees out from the stock reflector. This turns the bulb upside down which makes the 32 CP filament focus low and the 21 CP filament focus high.

To correct this I made a cutout at the bottom of the headlight bucket 180 degrees from the cutout at the top of the bucket. This turns the repro reflector 180 degrees and now the 32 CP focuses as the high beam and the 21 CP focus as the low beam. The headlight also had to be adjusted higher since turning the reflector 180 degrees lowered the high and low beams.

Dan also has a repro reflector in his knuckle head and it is the same as mine with the bulb pins 180 degrees from the stock reflector. If you want to post this here’s the summary:

"If you have a repro reflector and the bulb pins with the reflector installed in the headlight are one at the top and two at the bottom, the reflector must be turned 180 degrees to get the 32 CP filament focused high and the 21 CP filament focused low. This can easily be accomplished by cutting a notch for the reflector tab at the bottom of the headlight bucket 180 degrees from the notch in the top."

Now for your problem……..

The best my 32E can muster with the third brush turned all the way up, lighting coil in the generator energized is 10 amps. That’s just 2 ½ amps above the designed setting of 7 ½ amps with the lights on and shy the 13 ½ amps I calculated is needed for your panhead with bullet lights and the 6006.

My first test was to see what the charging current is with no electronic add ons. To do this I disconnected my shunt regulator so I could directly measure the charging current to the battery. I also disconnected my headlight and wired in the 6006. I left my rear bullet lights connected because this is the configuration you have.

With the low beam on there was maybe a trickle charge going into the battery. With the high beam the battery was discharging. This I feel is not adequate to keep the battery from dying with the lights on, even for a short run if you get caught after dark. There is just not enough current to go around as we suspected.

I can make this work but it requires one diode, my shunt regulator and the third brush turned all the way up. The diode is used to reduce the voltage to the headlight enough to reduce the current draw and the shunt regulator is to keep the battery from overcharging when the lights are off. The next experiment I did showed positive results from this configuration.

I wired a diode from the 6006 bulb ground to frame ground. The diode drops the voltage to the 6006 just enough to decrease the current draw by about 1 amp. With the low beam on there was about 2 amps going into the battery. That’s definitely on the good side of charging the battery. With the high beam on the current into the battery dropped to about 1 amp or less. That’s a trickle charge but at least the battery is not discharging.

The problem came with the lights off. With the third brush turned all the way up the current going to the battery was around 6 amps. This quickly brought the battery up to 8 volts and I shut the bike off. Continuing would have heated up my battery and possibly buckling the plates.

At this point I connected the shunt regulator. Now with the lights off the battery is kept to 7 volts and the excess current is shunted around the battery to frame ground and the battery doesn’t overcharge. With the lights on the charging current is the same without the shunt regulator and if the battery starts to creep over 7 volts the shunt regulator kicks in and protects the battery from overcharging.

What I don’t know is if all things are equal. If your 32E does not put out as much as mine then there may not be enough excess current to charge the battery with the lights on. There is some leeway with the low beam as I had 2 amps left over to charge the battery.

Given you are using the small battery the shunt regulator can easily hide the shunt regulator in the battery box as long as there is enough metal to work use a heatsink. The more current the shunt regulator passes to ground the hotter the transistor doing the shunting gets. If there is not adequate metal to help carry away and disperse the heat the transistor will burn out. I don’t see this as a problem because the shunt regulator can be mounted right to the battery box lid. That should be more than enough metal to carry away the heat and keep the transistor from overheating. Since you can weld you may even be able to tack weld a stud to the inside of the battery box lid so there is no mounting screws seen on the outside and the shunt regulator can be mounted on the inside.

Next step if you want is to test this on your panhead. We need to wire a diode in your headlight bucket, turn up your third brush and make sure you are charging with at least the low beam. Then connect up my shunt regulator and make sure the battery is not overcharging with the lights off.

If this works than I can build up another shunt regulator one for you when I order some parts.

[exeric]

I have no technical advice to offer, but I do have a headlight story. About 30 years ago I had a beat up but well loved '37 EL. I had some very good friends that lived in Apopka, Fl. which is about 15-20 miles north of Orlando. In those days Apopka was a very rural town with none of the infrastructure advantages of a city like Orlando. After a late night of wholesome entertainment it was time to go. Like I said, Apopka was way out there and being a moonless night (I assume), I had some trouble finding the old knuck. I felt around and finally found it, got it started and headed down the road. This is the headlight part. I was running the Cycle Ray but it had no rubber gasket in the reflector and the bezel was a bit stretched and mangled so the glass lens was loose but still retained. As I rode down this long pitch black road the lens was rotating in the bucket sending a beam of light into this hypnotising spiral rotation. The light was everywhere but where I needed it. God looks after drunks and children and somehow I made it home. I learned that component sub-assemblies should be reassembled with all of their parts, which is a fundemental restoration cornerstone. Not riding drunk. . . . . or whatever, took longer and more pain to learn.

[Perry Ruiter]

As I understand it, the -48 headlight bulb was a kinda "homemade" sealed beam with, essentially, an old style 21/32CP bulb and reflector fused with a lens to make a sealed beam. This was peculiar to Harley hence the advice from the factory to only use their headlight rather than a jobber 6Volt bulb. Based on the discussion it seems you want to drop the headlight load by about an amp. Maybe try a 4030 headlight bulb. It's a 45W/35W headlight (versus 50W/40W like the 6006). 5Watts at 6Volts is a little less than an amp but maybe it's enough? Especially if you always run with your headlight on low beam to protect your battery?

As an aside, the way I found that bulb number was using Don's Bulbs website. I searched on 6006 and found it had a 3L base and PAR56 glass shape. I then did a search on all 6Volt bulbs with those two characteristics. 4030 is the lowest power one listed. I love the Don's Bulbs site. If I want to know what the 12Volt version of a 6Volt bulb is, it's easy to find. If I'm looking for a bulb with a little more candle power but everything else the same, again it knows. However, the bulb prices are always very high. The 4030 lists at $227. A quick Google search turns up Vintage Auto Parts listing 4030s for a more reasonable $14.75 each.

Finally I think if you want a bike to ride at night with good lighting a 12Volt conversion is the way to go. Your existing 32E can be converted to a 2 brush 12Volt generator and a regulator is available that looks exactly like the existing cutout. So visually it retains the stock look. Should even fool the judges. I'll include a few photos of the regulator and a 32E I converted for someone local ... Perry

[flthead]

4030

before anybody goes and spends 227.00 on the 4030 bulb, i hit up an old junk yard an got about a 1/2 dozen 4030 working bulbs, final amp readings using this bulb is about the same as the 6006 ,[ 4030 bulbs are obsolete ]mind you slightly better but not good enough. also found the 4030 bulbs with a metal backing pressed onto a glass lens .... so i carefully separate the glass lens an metal backing, adapt a springer style bulb and socket. problem is the glass lens focal point doesn't match the new bulb so what you get is a nice bright light but with a "porch light "effect, no beam pointing down the road, an converting to 12 volts is out for me on this one . 'ol Mr. Gadget will whip up his magic in his laboratory I'm sure , but thanks for the info