Victor running underwater!

[Adam Y.]

Originally Posted by sanddrag

I bet you could run it under DI water without any waterproofing.

Technically, DI water has ions in it. There is a class of purity above that water called reagent grade or nanopure water in which you measure purity in resistance.

[Al Skierkiewicz]

De-ionized water (sometimes referred to as "distilled" by users) is used for cooling water in some high power transmitters. Water flows over high voltage and low voltage parts and current probes are used to determine the conductivity of the water. High voltage power supply trips at a few milliamps are common with poorly maintained equipment. In some cases, designers actually use china pipes to keep the water insulated from sensitive (read 'human') objects just in case some contamination enters the system and causes conduction to skyrocket. The UHF transmitter I worked on in Peoria many years ago had 25,000 volts across the output tubes and the cooling water. As I remember it had both a ceramic filter and a de-ionizer in the circulating water. The collector (plate/anode) was vapor cooled. This means that cold water is just poured onto the structure and cools while turning to steam vapor. The steam is carried off to a heat exchanger where it is turned back into water. Regular maintenance includes scheduled water changes.

[MrForbes]

I was lucky Al, I got to work on installing satellite ground stations after the cryogenic cooled receivers and water cooled transmitters had run their course....

[PICgnosis]

The Victor is sealed in Max GFE A/B epoxy from Polymer Composites, Inc. It has a thermal conductivity of 0.682 W/mK, which is about the same as a thermal grease like Wakefield 120 (0.735 W/mK).

Here's the closest thing I can find to an online datasheet for this product:

http://cgi.cafr.ebay.ca/EPOXY-RESIN-...7 C240%3A1318

It also provides information on their recommended mixing and casting procedures.

[falconmaster]

Originally Posted by PICgnosis

The Victor is sealed in Max GFE A/B epoxy from Polymer Composites, Inc. It has a thermal conductivity of 0.682 W/mK, which is about the same as a thermal grease like Wakefield 120 (0.735 W/mK).

Here's the closest thing I can find to an online datasheet for this product:

http://cgi.cafr.ebay.ca/EPOXY-RESIN-...7 C240%3A1318

It also provides information on their recommended mixing and casting procedures.

Thanks Karen!

[yodameister]

Originally Posted by Al Skierkiewicz

De-ionized water (sometimes referred to as "distilled" by users) is used for cooling water in some high power transmitters. Water flows over high voltage and low voltage parts and current probes are used to determine the conductivity of the water. High voltage power supply trips at a few milliamps are common with poorly maintained equipment. In some cases, designers actually use china pipes to keep the water insulated from sensitive (read 'human') objects just in case some contamination enters the system and causes conduction to skyrocket. The UHF transmitter I worked on in Peoria many years ago had 25,000 volts across the output tubes and the cooling water. As I remember it had both a ceramic filter and a de-ionizer in the circulating water. The collector (plate/anode) was vapor cooled. This means that cold water is just poured onto the structure and cools while turning to steam vapor. The steam is carried off to a heat exchanger where it is turned back into water. Regular maintenance includes scheduled water changes.

*sigh* I thought I sounded intelligent...*sigh*

[MrForbes]

That's hard to do in a thread where Al says something

[BitTwiddler]

*sigh* I thought I sounded intelligent...*sigh*

A few years back when they made me a manager and the radar guys said they needed some deionized water for the transmitter tube, I had never heard of the stuff. They said to make sure it was reagent grade.
Boy did I have a hard time finding that stuff! Now I know why.

[Al Skierkiewicz]

Originally Posted by yodameister

*sigh* I thought I sounded intelligent...*sigh*

You are!!!! People like you are my heros.

[Jeff Pahl]

Originally Posted by Al Skierkiewicz

Fredi,
At high current, I would be concerned with expansion and contraction coefficients of the material and the electrical components. I think the epoxy can do a great job as a heatsink when exposed to water. What I might suggest is a double dip such that the tabs of the transistors only are embedded in the epoxy while the circuitry remains open. Then encase the assembly as you have shown. This would allow the components on the board to move around with temperature without being broken off as material expands and contracts.

If you find the expansion / contraction to be an issue, another option is to pot the power devices with a layer of a thermally conductive RTV to provide a "cushion" layer between the packages and the epoxy. We use a material from NuSil at work that has a thermal conductivity of 0.81 W/mK to bond high power RF terminations to mounting surfaces. This is for a vacuum application where there is no other way to get the heat out besides conduction. The part number we use is CV-2960. I doubt it is cheap, but you could probably talk to them and get them to sample you enough to coat several units.

[EricVanWyk]

I think it would be to your advantage to reduce the amount of potting material you use, for both cost and thermal reasons. Perhaps place a few inserts in your mold?


As a point of reference, Aluminum is about 250 W/mK and stationary water is about 0.58W/mK.

[Andy A.]

It's been a few years since I opened up a victor, but I recall that they have an optical isolation on the PWM input from the controller.

Is it safe to assume that thats accomplished with a 'closed' unit, and there are no concerns about getting something opaque in there?

[Al Skierkiewicz]

Andy,
Yes, that is a sealed opto-isolator, one of the 4N series as I remember.

[EricVanWyk]

Originally Posted by Andy A.

It's been a few years since I opened up a victor, but I recall that they have an optical isolation on the PWM input from the controller.

Is it safe to assume that thats accomplished with a 'closed' unit, and there are no concerns about getting something opaque in there?

The optical isolation is accomplished within the confines of a single component on the board. There is no risk of getting something opaque in the way unless they break open that chip.

If you do break a victor open and take a look, it is the tall-ish 4 or 6 pin* device near the connector.


* (I can't remember: Only 4 pins are used, but often these devices are in a 6-er to match the 5-pin devices in the family.)

[JesseK]

If you want to go with high current, you will need some sort of heatsink. Probably the easiest is to solder some copper tubes coming off the metal transistor tabs on the Victor that poke out of the compound. Why use a 40A speed controller if you're never going to pull it? I think this is eluding to a new possible design from these guys