Carbon Arc Water Cooling Question for old timers

[Tim Reed]

As I sat thinking over the old business, I was reminded of a question I've always had, but never came up with a satisfactory explanation. In a typical carbon arc lamp large enough to be water-cooled, the DC negative was grounded to the lamp frame. The DC positive was attached to the positive carbon contact block, which was electrically isolated from the frame and ground. The water feed was circulated through this energized positive contact block at the arc. Water was often fed from a recirculator/cooler and returned. Although, some installations ran tap water through the contacts and emptied into the soil outside, or in a sink drain.

Lamp manufacturers cautioned against using copper pipe to connect the water to the positive contact block, since it would pose a direct short of the DC current to ground. Instead, the water was connected through supplied rubber hoses. The water feed to and from the rubber hoses could, however, be copper pipes.

My question has always been: why doesn't the DC short by conducting through the water to ground? At first thought, you might think that because recirculators use distilled water it won't conduct current, but that doesn't explain all the installations that used tap water from the city supply. I've asked several old timers over the years this question and never gotten an answer. Usually, I'd be presented with a look that said, "Hummm... I've never thought about that before." Even the late great Jim Ezell of Carbons, Inc. said he didn't know.

Anyone care to postulate?

[Leo Enticknap]

Although, some installations ran tap water through the contacts and emptied into the soil outside, or in a sink drain.

The theater's water bill must have been horrific, especially in desert areas where water is a scarce commodity.

[Josh Jones]

As I sat thinking over the old business, I was reminded of a question I've always had, but never came up with a satisfactory explanation. In a typical carbon arc lamp large enough to be water-cooled, the DC negative was grounded to the lamp frame. The DC positive was attached to the positive carbon contact block, which was electrically isolated from the frame and ground. The water feed was circulated through this energized positive contact block at the arc. Water was often fed from a recirculator/cooler and returned. Although, some installations ran tap water through the contacts and emptied into the soil outside, or in a sink drain.

Lamp manufacturers cautioned against using copper pipe to connect the water to the positive contact block, since it would pose a direct short of the DC current to ground. Instead, the water was connected through supplied rubber hoses. The water feed to and from the rubber hoses could, however, be copper pipes.

My question has always been: why doesn't the DC short by conducting through the water to ground? At first thought, you might think that because recirculators use distilled water it won't conduct current, but that doesn't explain all the installations that used tap water from the city supply. I've asked several old timers over the years this question and never gotten an answer. Usually, I'd be presented with a look that said, "Hummm... I've never thought about that before." Even the late great Jim Ezell of Carbons, Inc. said he didn't know.

Anyone care to postulate?

The conductivity of the water is relative to the amount of dissolved solids and minerals in it. Even the large lamps had less than 100VDC across the arc, so the current flow through the water would be a tiny fraction of total arc current, likely in the tens of milliamps. There simply isn't enough current flow to worry about. However, over the years, the electrolysis can gradually erode the electrodes from the inside out, weakening them and potentially causing leaks.

Old UHF TV transmitters employed klystrons for final stage amplification, which in the old days used distilled water for cooling. It was found that pure distilled water would agresively attack metals, causing pinhole leaks. The solution seemed to be to poison the cooling water with a very small amount of acetic acid.

[Steve Guttag]

Tim,

My question to you is: How sure are you that the negative (or positive) had any ground reference in a rotating carbon lamphouse? I never installed a Super Core Lite or Futura 2 (or Mighty 90)...too young for that, but did run them. Of that size lamphouse, i mostly ran Futura 2. I recall insulators under the negative feed so I'm not convinced that it had a negative ground. In fact, I'm not sure why one would want to ground the arc circuit at all. It is less safe and inherently noisier.

Christie demonstrated this with their xenon systems by grounding their negatives so that you could always hear the arc strike in the sound at the start of each show.

Where I ran Futura 2s, we ran circulated water but had the option of using city water, in an emergency (and it gave a convenient way of draining the circulator. The problem with city water would be the eventual clogging of the silver jaws. They were to be protected due to their extreme expense. If the water flow dropped, they'd melt. They were definitely fed via some form of rubber tubing (I remember the connecting tube being a whitish color connecting the two sides in a "U-loop").

[Tony Bandiera Jr]

The answer can be found in a different industry: electro-plating, where DC current is used to transfer a metal suspended in the plating solution (or sacrificial rod or plate) to the surface being plated.

I don't remember the specific polarities do which direction, but the reason for the rubber hoses (and the prohibition on copper pipe) tells me that the idea of grounding one side would help prevent that transfer of metal from one side to the other.

[Josh Jones]

In plating, material is robbed from the anode and plated to the cathode.

[Tim Reed]

The conductivity of the water is relative to the amount of dissolved solids and minerals in it. Even the large lamps had less than 100VDC across the arc, so the current flow through the water would be a tiny fraction of total arc current, likely in the tens of milliamps. There simply isn't enough current flow to worry about.

That's an excellent explanation! It makes sense. Thanks!

How sure are you that the negative (or positive) had any ground reference

Thanks for pointing that out. I'm writing this thinking of an Ashcraft... the negative was indeed insulated from the frame. However... Ashcraft warns in their manuals:

CAUTION - Under no circumstances connect metal tubing to the lamphouse water system. The CN-4123 water terminal block is energized, being of positive polarity. A metal tubing connected to this block will cause a short circuit to ground, which will damage the lamp.

So it is referenced to ground at some point.

[Steve Guttag]

There is an ashcraft manual in the warehouse. Their schematic, such as it is, doesn't show a tie point to ground on the arc circuit. I think that the designers put effort into floating the arc circuit, they would be correct to be emphatic about an external ground via the cooling circuit since it would be unpredictable where another ground point might be compromised (including how the generator/rectifier might choose to reference their output). It's best to keep it floating.

[Tim Reed]

The theater's water bill must have been horrific, especially in desert areas where water is a scarce commodity.

Most of these in my memory were at drive-ins, where they had well water. I worked one theatre in the city that emptied cooling water into the sink. I was happy to get rid of that arrangement, replacing it with a recirculator.

Thank you, gents, for your insight and thoughts. You've answered a question that's bugged me for a good 40 years.

[Steve Guttag]

Leo, FYI...at the Uptown in DC, our water was always just filtered (sediment, nothing special) city water. It wasn't a large flow..just enough. In the picture below, I've circled the return pipes. They fed into a trough that was the drain. You can see the sediment filter in the picture too:

image.png

I never heard anything about our water bill being crazy high. Then again, seating ~1000 people, compared to the toilet flushes, it is bound to be trivial.

[Dustin Grush]

The theater's water bill must have been horrific, especially in desert areas where water is a scarce commodity.

Without knowing the flow rate it's impossible to estimate, but say it was 3gpm (TIG torch water coolers use 1-2gpm for comparison), that would be about 1000 gallons for a 5 hour drive-in double feature. Back when we had carbons, which weren't water cooled, water was probably $2/1000 gallons. Compare that to the hourly cost of a 6500w xenon lamp and it seems cheap. And as Steve said above, there were nights where we probably used 2-3x that much just for toilets.

[Steve Guttag]

I'd estimate our flow to be below 1gpm per projector. Picture a stream of about ⅛-¼ inch in diameter as it left the pipes. It was a drizzle. Since the water was cool, it didn't take much to keep the gates cool enough. But if the water wasn't on...you'd know it immediately!

Back in the carbon arc days, the flow meter alarm was active to avoid melting the silver jaws but that was before my time at the Uptown. They were already on the Strong X60D by the time I started in 1988 but the make/model would have placed it in the 1980-1983 time frame, I'd think.

[Jim Cassedy]

I ran carbons for over a decade. I worked mostly with Ashcraft Super Core-Lites. I don't recall
the negative carbon contact being 'grounded' to the lamphouse frame, and a quick look at
some old manuals in the warehouse here and in my files, seems to support that they weren't,
although the lamphouse itself, like all the booth equipment ,was connected to a good ground
per the electrical code. Our lamphouse cooling system was 'closed', using the ubiquitous
Ashcraft or Strong water recirculators that were very common in alotta booths back then.

I don't recall having water-flow sensors, but the lamphouses were wired to relays in such a
way that you couldn't start the motor-generators, or the rectifier stacks unless the water
circulators were powered up. During my apprenticeship (necessary to get into the union), I
remember being told that one of the reasons for not being able to start the lamp without the
circulators running was that the positive carbon clamps, which were also water cooled, were
made of (or more likely plated with) silver, and that they were very expensive to replace, and
so melting one due to lack of cooling was one of the big mistakes you never wanted to make.

There was also an air-flow switch that would kill the DC lamp power if the fans in the water
circulators or the booth lamphouse exhaust failed. This, of course, was most mostly to protect
the expensive mirrors in the lamphouses from overheating and cracking.

Our unionguys also operated the spotlights at a very large sports and entertainment complex.
While I never worked that location, I visited their spot-booth on several occasions because one
of my friends worked there. They actually had a pair of STRONG carbon-arc "Super-Trouper"
spotlights that were water cooled, which I'd never seen before, or since. So, you not only had DC
power cables running to the spots, but a couple of flexible hoses too, which hung down from two
water ports in the ceiling. They got their their cold water from the building's air conditioning system.

[Harold Hallikainen]

As others mentioned, the water does not short out the power because it is distilled or deionized resulting in very low conductivity. There was mention of klystrons and water cooling. This also used to be common in high power AM broadcast transmitters. I remember seeing the KRLA transmitter which was water cooled. Plastic pipe carried the water from ground voltage to several kV, then back to ground.

Water conductivity is interesting. I repaired an icemaker ( https://hallikainen.org/hzb-12b/ ) that relies on water conductivity to tell when it is out of water. Water is pumped through a grounded metal collar and a wire probe sticks down into the water stream flowing into the freezing tray. If the probe is not pulled towards ground by the water (due to a lack of water), the machine stops the water pump and lights an Add Water LED. Very clever!

[Tim Reed]

I ran carbons for over a decade. I worked mostly with Ashcraft Super Core-Lites. I don't recall
the negative carbon contact being 'grounded' to the lamphouse frame,.

No, we determined I was mistaken about that. Looking at the parts list shows the negative carriage is insulated from the frame.