My workbench still isn't right, or how to kill 3 days.

I wanted to call this post 'lack of power, that was our dilemma' but that is an expression that no-one here will have heard before. This will be a long post, not because the subject matter deserves it, but because of how I let the project consume several days!

So, as we have seen previously, I have started to setup my workbench. As part of that, I bought this old laboratory bench power supply for cheap. Another image:

This is a perfectly wonderful piece of equipment. It supplies 3 different adjustable voltages, which allows you to build a mixed analog and digital circuit on your workbench to do almost anything a labrat would want. 0-30 volts (2 of 'em), and one (4 - 6.5) volt 5 amp output supply for the digital side.

The problem is the digital side. It was designed to allow some variation in output, to cover a number of early logic families. (Uh-oh, more history). We could go really back, a really early digital computer was made with electro-mechanical relays (see pic) (if you were near it, you would here them clicking back and forth, clackity clackity, clack) ,  and then they were made using vacuum tube technology (see pic). This was a tremendous increase in speed, approaching some semblance of modern technology, but tubes were still too big, occupying far too much space, and produced a large amount of heat.

Well ok, my time doesn't go back that far, but I do have one piece of old tech (which was already surplus when I was in high school) which had a bunch if tiny little tubes on a circuit board. Anyway, as logic transitioned to the solid state world, many different supply voltages were used. There were discrete diode and transistor boards (see pic) which might use any voltage the designer cared for. Some used negative power (DEC), etc. We were making progress. We could now get one WHOLE logic gate on a board!

At some point there could be more tutorial on digital logic, but I'm not sure about that. While I love the role of teacher, my audience probably has a handle on the basics. For now, see my favorite Dilbert cartoon.

This was followed by the introduction of a few different forms of integrated circuit (IC) logic. In the early days of RTL and DTL, the power ranges narrowed, but not completely.

Then we get to TTL logic. (transistor-transistor-logic aka 74XXX). This was becoming the main stream as I really started working, and it also forms the basis for the KI-10 CPU (which was where we started right?) TTL was much more rigidly defined which was a boon to the digital engineer. He could concentrate on the logic design, because way more analog variables had been removed. (We'll talk about output fan-out later). TTL logic required a 5 volt power supply. Technically the range was 4.5 to 5.5 volts (because they had to allow for some slop), but the truth is it was optimized and happiest at 5 volts. This was widely accepted and in fact following technologies (CMOS NMOS, etc) continued to use this voltage and still today it is commonly in use. Very high density memory and CPU chips now use 3.3 volts (to reduce power consumption), but 5 volts is the general standard.
  OK, I need to talk about TTL more, but not now. I want to get back on track...

5 volts. Not 4 volts. Certainly not 6.5 volts. (I smell smoke!). 5 volts. So, there is a problem with my venerable old bench supply. IT CAN supply the 5 volts needed, but what if someone accidentally turns the adjustment know, or bumps into it? Oops. This is a problem I am not prepared to allow. I don't want to fry any of my chips because of an accident. What I really need is 5 volt fixed power supply.

So, my pretty lab supply is not really what I want. I need to a simple single fixed 5 volt power supply.
So, I get this cheap new internal 5 volt 30 amp fixed supply for $35.00. $35.00? It continues to amaze me how cheap most of this stuff is.

But wait, there is a problem! What? Really, it's internal. I could just screw a power cord into it, but I would have live exposed 120 vac power a bit to close to possibly clumsy me on the workbench.

It needs to be in a box.
Wait, I have one. It's been on my shelf for a hundred years, waiting for me to have a project.

A perfect little box to hold my power supply.

A box to cover it up and make it pretty. Uh-oh, strap in for the rollercoaster of my brain.
It should have a plug in power cord, like my computer.
It should have a nice on/off switch.
It should have pretty binding posts for the output.
It should have a light indicating it's on.
No, better yet, it should have a digital display, showing that it's doing its job! Why? Because they are CHEAP! A BIG 3.5 digit LED volt measuring panel meter is $12.00! Back in the day, each digit would have cost $12.00, forget the volt meter. The panel would have cost more like $150.00.
Hey, while I'm at it, let's have a switch on it to show Volts or Amps in use.
I buy most of these extra parts online.

Were you buckled in? Did you survive the wave of my GREAT ideas for spending time on this? I figure I can put this together in a day. So, we have labor day weekend and I am stuck here minding the dog. If I take a day to throw this together, I'll still have 2 days to go out and do something. hahahahahahaha!

Well, this tale is getting there. I'm taking a work break now, more later.
-------------------------------------------------------

So ok, I have the design in my mind, now I have to start laying it out (mostly) on the aluminum front panel of the box. Now resurrecting long unused tools, I have more shopping to do. Simple issues like rust that need to be dealt with. So off to home depot for 3-in-one oil, new drill bits, etc..

Measure parts, scribe, square, scribe some more. One day gone.

2nd day, do the metal working. I don't have a shop so I have to proceed slowly to avoid having aluminum shards f*#k-up my floors. Have we talked about 'the nibbler' yet? I didn't think so. You see those NOT round holes? The cheapest way to make those for low volume use, when you don't have a machine shop is to use a device called a nibbler.



This evil device takes a very small (1/16" by 1/4") bite out of a piece of sheet aluminum. It operate by hand strength. As you can see by the size of those 1 by 2 inch sized holes, that is a lot of little bites. It's hard to use. It hurts my hand. I am a pussy not used to manual labor. With every bite I dream of having my own end mill.  (I went online to look for a pic of the nibbler and found possible powered ones for future endeavors) Notice the lopsided holes where my measurements went awry and further action was needed.  Two days gone.




Day 3. Wire it up and finish as seen in the below pic. There are of course mechanical issues getting the parts mounted. Last minute cuts drills and files...  I start wiring with the power cord/switch because I have to calibrate the actual 5 volt supply before I connect anything to it. I'd like to state up front that I cross wire the switch and instead of applying power it blows the internal fuse. I need to test it twice before I realize this idiot mistake.

Finally power is applied, now I don't like the wire stock I have on-hand to do the actual 5 volt wiring. Another trip to home-depot to buy more wire.

How about that cool 3 1/2 digit panel display?  As mentioned earlier, it was cool cheap. What does it do? It is a simple volt meter. It measures 0-199 millivolts to display 0-1999. There are no ranges. You have to externally adjust your input signal to fit the range of the panel. So, I need 2 resistors to make a voltage divider. A 10kohm followed by a 100ohm will be a 100/1 divider and keep my signal in range. (should be 5 volts divided by 100 = 50 millivolts which the panel can handle.)

At some point in this journey I decide that the panel should have a switch so I can display how much current is being used (amps)  so I can see how my test circuits is doing power wise.

An ammeter for those who are not trained, is different. We don't actually measure amps. Instead we use ohm's law.
                
We only know how to measure voltage, so we get a known resistance and measure the voltage over it. By ohm's law voltage/resistance=current(I).

So, to set this up, I need to calculate backwards. My maximum output is 30 Amps. I need that to be measured as 30 millivolts (because my panel tops out at 200 and I need a value less than).
.03 volts / 30 amps = .001 ohm. Crap, do .001 ohm resistors exist? Yes, it turns out I have NOT invented the wheel. These are available to measure current (exactly what we are doing)

I buy one of these and it will arrive in a few days. I will put it in later.  Turns out these are relatively expensive. Where a 'normal' resistor might be 25 cents, the shunt is about $25.00




The ALMOST finished project. It looks fantastic from the outside, where you can't see the flaws.
The panel is too close to the edge. I had to cut away some inside stuff to get the box to close. I have ordered yet more wire. I had to use some that is too small for part of the wiring. I will rewire with some 18 AWG solid to better carry the current.



So, what's the point? There are several, but we'll see if I learn from any of them.

1. It takes much longer to actually do something, than to think it.
        a This could have severe consequences for the larger project
        b remember to review cost/benefit. Should I have looked harder for one to buy?

2. I need to have a system to keep my life balanced and out of 'the project sinkhole'

3. I am proud that I got this done. An (almost) completed electronic project. woohoo! Obviously the processor is far larger, but now I can physically start to work on test and development.

If you got this far, you may have less life than me.....