Home Brew Test Equipment: High Voltage Supply

[KD1SH]

   In addition to ham gear, I love building my own test equipment for the bench. Here's one I just finished today: an adjustable 0-1200 VDC bench power supply. The project was actually inspired by my acquiring nice empty cabinet at the last Nearfest. What to build into it? I've got a nice old Heathkit high voltage bench supply, but it only goes up to 400v, and quickly runs out of gumption above 100ma. So, empty cabinet, plenty of transformers, chokes, caps, and meters handy; it's going to be a power supply!
  Nothing complicated: a Hammond 1250v transformer, Triad swinging choke, bridge rectifier, a variac, a couple of old panel meters, and a can of green spray paint. The output connector is an SHV type.
  The mismatched panel meters wasn't intended. Always verify the functionality of old panel meters before cutting the hole in the panel, not after! (Ask me how I know this.) I was lucky to dig up another meter that fit the same hole/screw pattern.
  Also note, in the pictures of the uncovered chassis, how close the filter cap bank comes to the rear of the panel meter. The cap is actually touching the meter! Angels watch over fools and builders who don't measure before drilling holes!
  

[W7TFO]

Nice job.  Like the color, too.

73DG

[KD1SH]

  It's just green Rustoleum, but it matches the paint I used on a couple of other DC bench supplies that I restored, and the case has just about exactly the same dimensions as those, so it'll look fine sitting beside them.

Nice job.  Like the color, too.

73DG

[KD1SH]

  The fun part was making nice compound radius cuts on the corners of the panel, to fit the enclosure. My milling machine isn't CNC. I've got a rotary table for the mill, which would make very nice circular radius cuts, but not compound. So it's the old fashioned way: trace out lines on the panel; make rough initial cuts, and then remove tiny bits at a time on the bench sander, check the fit; another touch on the sander; check the fit; another touch on the sander...
  One day I'm going to get myself a CNC router, robust enough to cut at least 1/8" or even 3/8" aluminum.

[KD6VXI]

A cnc mill / router is cheap now.  I bought a kit and built it, use it to make pc boards, labels, etc.

Think it was 300 dollars shipped, took a weekend to build.  Love it, also put a laser in it when I want to etch stainless, etc.

--Shane
WP2ASS / ex KD6VXI

[KD1SH]

   Yes indeed, they're getting very attractive. I'd want one capable of handling the standard rack panel width of 19", and probably 14" (8U) high, for homebrew stuff, so it would probably get well above $300, but still, compared to what things like that cost ten or twenty years ago, or the cost of a full sized CNC standup milling machine, it's dirt cheap.

A cnc mill / router is cheap now.  I bought a kit and built it, use it to make pc boards, labels, etc.

Think it was 300 dollars shipped, took a weekend to build.  Love it, also put a laser in it when I want to etch stainless, etc.

--Shane
WP2ASS / ex KD6VXI

[W7TFO]

My chassis/panel/metalwork is not so modern.

I get by with a DeWalt table saw, just about every round Greenlee punch ever sold, and a 30-ton hydraulic press.

73DG

[KD1SH]

  Yes, table saws work quite nicely for cutting metal, with the proper blade. Of my two table saws, I usually keep a wood-cutting blade on one, but for the other I keep several special blades handy, for cutting metal, fiberglass (with a face mask—nasty business) or plastics. It's an old belt-drive saw, and I changed the pulley ratio for a slower blade RPM. As long as I feed the work slowly, it'll cut 1/4" flat plate stock, and even thicker bar and rod stock.
  Still, a band-saw is best for that stuff. I'd love to have a full sized stand-up metal cutting band saw, but they're pricey, and big.

My chassis/panel/metalwork is not so modern.

I get by with a DeWalt table saw, just about every round Greenlee punch ever sold, and a 30-ton hydraulic press.

73DG

[KD1SH]

   Happily chugging out 150ma at 600v, with ripple less than 0.2%. For low level audio stages I'd prefer it just a tad smoother, but not too shabby. I didn't have the resistors handy to load it to 200ma at 600v, but with the resistors I had it hits 200ma at 850v.

[Opcom]

Beautiful job!

[KD1SH]

Thanks. Honestly, I don't consider myself nearly in the same league as builders like Tom, K1JJ, or Rick, W8KHK, and a host of others. Sometimes I build deliberately fugly stuff, like my "Hail Mary" transmitter, but I'm sort of fussy about the test equipment that sits on my bench, for whatever reason.

Beautiful job!

[w8khk]

Thanks. Honestly, I don't consider myself nearly in the same league as builders like Tom, K1JJ, or Rick, W8KHK, and a host of others. Sometimes I build deliberately fugly stuff, like my "Hail Mary" transmitter, but I'm sort of fussy about the test equipment that sits on my bench, for whatever reason.

Beautiful job!

Bill, that is a very nice compliment, thank you so much!  Your power supply project looks every bit as good as anything I could do, but when I am in a hurry to get something working quickly, I probably wouldn't post any photos, (unless it was a bunch of big triodes with the plates glowing brightly).  Your projects can be as great as anyone else's, it is just a matter of taking your time and attention to detail.  A career at Hewlett Packard engrained that in my memory.

Now that WE have reached our destination, adding every last desirable feature to our pet project, I am pretty much free to work on a few hollow-state transmitters that have been in limbo for almost five years.  We shall see what develops!  I might even get motivated to finish my sand-state class-E rig, with a class-H modulator.  I created all the printed circuit boards for those over two years ago, and here they sit.

[KD1SH]

  Rick, the compliment is well deserved; I've seen pictures of some of your work, and of course there's the Max Processor.
Thanks for the encouragement. There's certainly a time and place for stuff with a buzzardly look, and even an outright fugly look, but I like my test equipment, whether it's commercially made or home-brew, to look presentable.
  At my last—and final— workplace, I had to build a lot of test equipment. Most of it would go from my lab bench to the production floor, where everyone down there would have their hands on it, including the occasional customer rep or ISO auditor, so it had to look good.

Thanks. Honestly, I don't consider myself nearly in the same league as builders like Tom, K1JJ, or Rick, W8KHK, and a host of others. Sometimes I build deliberately fugly stuff, like my "Hail Mary" transmitter, but I'm sort of fussy about the test equipment that sits on my bench, for whatever reason.

Beautiful job!

Bill, that is a very nice compliment, thank you so much!  Your power supply project looks every bit as good as anything I could do, but when I am in a hurry to get something working quickly, I probably wouldn't post any photos, (unless it was a bunch of big triodes with the plates glowing brightly).  Your projects can be as great as anyone else's, it is just a matter of taking your time and attention to detail.  A career at Hewlett Packard engrained that in my memory.

Now that WE have reached our destination, adding every last desirable feature to our pet project, I am pretty much free to work on a few hollow-state transmitters that have been in limbo for almost five years.  We shall see what develops!  I might even get motivated to finish my sand-state class-E rig, with a class-H modulator.  I created all the printed circuit boards for those over two years ago, and here they sit.

[w8khk]

 Rick, the compliment is well deserved; I've seen pictures of some of your work, and of course there's the Max Processor.
Thanks for the encouragement. There's certainly a time and place for stuff with a buzzardly look, and even an outright fugly look, but I like my test equipment, whether it's commercially made or home-brew, to look presentable.
  At my last—and final— workplace, I had to build a lot of test equipment. Most of it would go from my lab bench to the production floor, where everyone down there would have their hands on it, including the occasional customer rep or ISO auditor, so it had to look good.
.....snip..........

Bill, there is a lot more going on than just the "looks" of equipment going to the production floor.  Appearance is trumped by performance and reliability, at which I am quite sure you excelled.  And I expect that your homebrew rigs do not have the word "zorch" in their vocabulary.  I am sure that was a rather high-pressure position you were in to deliver on all the required categories.

Ahhh, you mentioned The Max Processor...    and unfortunately I am not quite so proud of the appearance of some of the earlier versions.  While they performed well, they required a great deal of hand-wiring to the controls and indicators on the front panel, which ended up appearing like the venerable rats-nest of wires when completed.  That is all gone now, the face-lift is complete.  While the kits do require soldering a large number of components to a PC board, there is almost no hand-wiring required on the current versions, as evidenced by the recently updated website photos and documents.  I could not stop working on it until I resolved that issue to my own satisfaction.  Now time is available to focus on my hollow-state projects!

[KD1SH]

  Indeed, functionality and reliability aren't optional. My test equipment was quite valued by the assembly floor, if I may say so in all humility. In fact, some of my test gear was clandestinely referred to as "Bill Boxes." That was actually a sort of code-speak. The issue was that—and I'm sure that you're very familiar with all this—all production floor test equipment was, officially, subject to a rigid set of ISO doctrines and processes: all parts had to be acquired through ISO registered vendors and be traceable to the nth degree; there were numerous documents—always identified by cryptic acronyms—that had to be laboriously filled out and approved, documenting the processes used in the testing, validation, and qualification of the device, and a safety analysis had to be completed (no sharp edges or pinch-points, for example).
  In short, when the production floor requested a test box through the official process, it easily took months for them to actually get one unit in their hands, much less a number of them. So, there was a less official channel, typically beginning with me getting a phone call or an email from one of the production managers, asking for a "Bill Box," with the understanding that it would mysterious disappear from sight whenever an ISO inspector or customer rep was near. To them, my impromptu boxes didn't exist, but the production managers and expediters loved them.
  I just came from looking at the Max site—very impressive! Awesome work!

 Rick, the compliment is well deserved; I've seen pictures of some of your work, and of course there's the Max Processor.
Thanks for the encouragement. There's certainly a time and place for stuff with a buzzardly look, and even an outright fugly look, but I like my test equipment, whether it's commercially made or home-brew, to look presentable.
  At my last—and final— workplace, I had to build a lot of test equipment. Most of it would go from my lab bench to the production floor, where everyone down there would have their hands on it, including the occasional customer rep or ISO auditor, so it had to look good.
.....snip..........

Bill, there is a lot more going on than just the "looks" of equipment going to the production floor.  Appearance is trumped by performance and reliability, at which I am quite sure you excelled.  And I expect that your homebrew rigs do not have the word "zorch" in their vocabulary.  I am sure that was a rather high-pressure position you were in to deliver on all the required categories.

Ahhh, you mentioned The Max Processor...    and unfortunately I am not quite so proud of the appearance of some of the earlier versions.  While they performed well, they required a great deal of hand-wiring to the controls and indicators on the front panel, which ended up appearing like the venerable rats-nest of wires when completed.  That is all gone now, the face-lift is complete.  While the kits do require soldering a large number of components to a PC board, there is almost no hand-wiring required on the current versions, as evidenced by the recently updated website photos and documents.  I could not stop working on it until I resolved that issue to my own satisfaction.  Now time is available to focus on my hollow-state projects!

[w8khk]

Bill, from what you describe, the floor managers and expediters had great faith and trust in your deliverable to bypass the standard process, so Bill Boxes are quite a feather in your cap, so to speak.  I am sure they were every bit as good and safe as those that had such a long and expensive paper trail holding them up.

Your job makes my last 5 years or so at HP and Agilent seem like a walk in the park.  Not wanting to hijack your thread, we will need to chat about that sometime.  Environment was much more relaxed, delivery expectations were very high, it just "had to work", or I had to make it right while under the gun.  I got worldwide recognition, both good and sometimes not so great, though the latter was a very rare occurrence.  Contact me off-thread sometime and I will share some interesting stories.

Thanks again for the Max comments.  In addition to all the hardware work, PCB architecture, and parts management, I handled the assembly instructions, parts, BOM, availability, and contact data for the web. But Clark did a massive facelift to support multiple versions of the processor, such that each one got "individual" attention.  He handled all the other data on the website, no small task indeed.  Gee, we have been partnering on this project for many years, we need to set up an eyeball QSO at some point in the future!