Spectra Physics 270-1 power supply

[ARG]

Found this on an electronics forum few days ago, thought you guys may enjoy the pictures, I could stare at the beauty of the TO-3's all day

This is the PSU for a 45W Argon

 

[Livinloud]

holy wiring/troubleshooting nightmare if one of those components crap the bed! really interesting though, thanks for the share

 

[weeba2kv]

HOLY SHIT that is some power supp. haha

 

[Sigurthr]

Holy Power Electronics Pr0n, Batman!

 

[The_LED_Museum]

husoos cristo en una muleta!!!

 

[djQUAN]

And the transistors are liquid cooled. I wonder how much heat has to be dissipated on that.

 

[ARG]

And the transistors are liquid cooled. I wonder how much heat has to be dissipated on that.

Didn't notice that!

I'm more worried about how much power it must draw

 

[ped]

Now THAT is a power supply haha.

Thanks for the share, I could look at stuff like that all day long.

 

[Meatball]

It looks a like an array of ~166 passbank transistors.. though, those first 3 might be setup just to bias all the others.

To get a stable current through a 45W Argon tube, you need some large components.

I count 6 transformers - one of them is probably just a 1:1 iso. I enjoyed looking at the bus bars on the electrolytics! Thanks for sharing!

 

[BShanahan14rulz]

No PnP machines here, that's some repetitive soldering!

 

[djQUAN]

Reminded me of the time I had not much to do and decided to build a 31 band real time analyzer from scratch. Took a bit of work to complete mounting all those SMD parts by hand.
Build your own RTA

 

[vk2fro]

Now THAT is a passbank - yep sure are a helluva lot of TO-3's there

 

[ARG]

Wonder what it would have cost when new

Most if not all was done by hand I think.

 

[djQUAN]

Yup, definitely made by hand.

Just noticed that the date codes on the transistors are 1980 and 1981 so it must have been built around '81-'82.

 

[LSRFAQ]

Having repaired them, they are easier to repair then you think. Each transistor has its own fuse. If not, worse that happens is a whole string of seven pops.
They fail shorted, most times, so you find them with a Ohm meter.

If you can not find them with a Ohm meter, you hook up a lab grade current limited power supply at set at 10-20 amps limiting and find the transistor with the least voltage across it.

There are other techniques, but a good tech can fix a passbank in less then two hours, often in an hour.

That is hooked to a 3030 or 3050 ceramic based head. SP used the 270 until they could get the more modern switched resistor and SCR based power supplies working for their metal ceramic tubes.

It needed 440 to 480 volt three phase to run, resulting in 35 amps or more down the tube bore. The tube had 670V open circuit across it before the ignite pulse fires.

A 741D 14 pin ceramic version of the classic LM741 op-amp drives the passbank.

Steve