Friday, July 29, 2022

Why I write in all caps

 I just write everything in capital letters.

My parents are adamant that my 2nd grade teacher ruined my handwriting, and that it was very good before that.

You know how some kids say their grades are bad because their teacher hates them, Well, my =parents= said that about my 2nd grade teacher. She found some fault in everything I did. I got so I didn’t know what was right in my handwriting, and got locked in analysis paralysis while writing, especially longhand but printing, too.

She had the typical alphabet examples on the wall around the room. But they showed more than one way to do some letters in longhand, like G, S, P, and R. Then she’d mark me down if I chose the wrong example, but she would not unequivocally state which was the one I was supposed to follow. Nor would she cover the ones she did not want us to follow.

I also get impatient when writing. My thoughts flow much faster than I can write.

^My handwriting in the mid-80s. Originally drawn only for my own use.

^ Same schematic, redrawn this year. Note that 12 years ago, I had manual drafting as part of an EET AT degree that included a lot of practice lettering all in caps.

^ My handwriting earlier this year. As you can see, I now write almost exclusively in all caps.

^ A couple of years ago, prior to chemo. If I don’t like how I’ve written a letter, I will erase and rewrite it. I have a particular problem with K and R.

^ Even when I put a fair amount of effort into it, my writing isn’t that great. Everything just wants to get out so fast.

Steve’s Hackaday.io Projects

Wednesday, July 27, 2022

Digital Photography and Composites on an Amiga computer

 Back in the mid '90s, I built a digital photography/compositing business around an Amiga 3000. There was even an interview with me published in Amiga Format magazine.

The Amiga 3000 was a great machine. I was using a desktop A3000 with a MicroniK case that gave me a lot more Zorro III and bridgeboard slots. At first, I used a video camera to capture still images, but that just wasn't high enough resolution for print. So as what I was doing expanded, I eventually ended up with a Phase 5 68060 accelerator and Cybervision 64 graphics card and a Polaroid digital camera that captured 1600x1200 tethered via the SCSI bus. Massive 128MB of RAM, a 1G Seagate SCSI, a 4G Fast SCSI 2 Micropolis AV drive on the Amiga SCSI bus, and a 4G Wide SCSI 3 Micropolis AV drive on the accelerator SCSI bus.

No Amiga drivers for the digital camera I was using (Polaroid PDC-2000), so I ran Shapeshifter and captured with Apple's Mac OS, System 6 (or 7, don't recall now). Since no Mac ever had a 680x0 greater than a 68040 and the PowerPC Macs at that time emulated the 68k chip, my Amiga was faster as a Mac than any Mac. How much faster? When I had the camera demo'd to me in Seattle on a PowerPC Mac, it took nearly 10 minutes to transfer an image from camera to computer. But my 68060 Amiga 3000 took about 15 seconds.

We went to sci fi conventions and composited people into custom backdrops in their costumes. I'd have an inkjet printing out a high res glossy, which took about 15 minutes then, while I had Real3D (a 3D modeling/rendering program) rendering several scenes or elements in the background, while working on a multiple layered graphic at 2400x3000 pixels in ImageFX, then I'd pop over to Shapeshifter to snap someone in their costume while my wife posed them. Printing and rendering continued uninterrupted, and everything ran smooth as silk.

The first Norwescon that we took this to, I had six different Microsoft programmers watch me and talk to me while I was working, and they'd suddenly say "Bill Gates ruined software!" after finding out that I was doing all of this on one computer running at 60MHz.

Between conventions, we'd work from home. I also repaired computers, especially Amigas, and other electronics. For a few years, we also had a science fiction collectibles store.

I had a Primera Pro dye sub printer that required an absolutely uninterrupted stream of data from the computer or the print would be ruined. If you were running Windows, it was best to make sure nothing was running but the print driver, and don't touch the mouse or keyboard. I talked to other people using this printer with Windows 95 on Pentium 100 and up who had prints ruined because they moved the mouse, and it might take 20 minutes to over an hour for the computer to process the file for printing.

It took my Amiga 3000 almost exactly 60 seconds between when I hit Print and the printer started. I tested it really hard one day - I hit Print, then quickly started Real3D and set three different high res 2400x3000 pixel images rendering, dialed into my ISP, started my web browser and opened a dozen tabs on different websites. The result? It took 2 minutes before it started printing, and the print itself was flawless. At no time did my typing or mouse pointer get jerky.

My old website is still up, although the digital photography business is gone. I never took orders online, so no shopping cart. Before you look at it, keep in mind that I never claimed to be an artist.

http://www.polyphoto.com

Hello, Hackaday viewers! If the above website doesn't link, please try again in a bit. Turns out there was a problem with the nameservers. I've fixed it, but it may take a few hours for the changes to propagate.

http://hackaday.com/2016/06/05/how-an-amiga-graphics-business-ran-in-the-1990s/

Update:

I’m inordinately excited – I was cleaning out my storage, and discovered that a friend of mine had given me an A3000, still in the original box! It needs a SCSI hard drive, and I think I have some RAM ZIP chips to max it out (it has 12M now, I think), and I am sure it could use a recapping… but… the battery DID NOT LEAK!

I have a 1G SCSI drive in a removable drive to install in it. I also found I have Kickstart 3.1 chips for it, and both Workbench 3.1 floppies and Workbench 3.9 on CD. I have a portable SCSI CDRom drive. I still have CD reader/burner software. Most of the other software is gone, however I kept a disk image of my original hard drive installation including Directory Opus and ImageFX 3.something.

I find I’m ridiculously excited about this.

What was owning a Commodore Amiga like?

 It was wonderful! The Amiga could multitask like no other consumer computer.

I had a digital photography business using an Amiga for photo compositing (like Photoshop), 3D rendering, special effects, and desktop publishing. My now rather old website is still online. Not up to today's standards, I know, but we did these compositions live while the customer waited, something you really could not do on a Windows computer in 1995:
Polymorph Digital Photography- Digital Portrait Photos and Composites

It had a 68060, which was a Motorola RISC processor, running at 50MHz with 128MB of RAM. We'd take it to science fiction conventions where we'd take pictures of people in their costumes and composite them into scenes and deliver the prints a bit later that day.

I ran a program called Directory Opus, or DOpus for short, as a Workbench replacement. It made the Amiga operating system, Workbench, even more of a joy to use.

I would typically have an image processing program called ImageFX running with between 6 and 20 layers, with the previous image printing on an inkjet (which at the time took 15 minutes to print an 8x10 photo), with a 3D modeling/rendering program called Real3D rendering things to composite into the image, and a Mac emulator running with the software for the tethered SCSI camera and the software to access Corel stock images CDs. All running at the same time.

The first convention we sold our services at was Norwescon, and 6 separate times over the weekend while chatting with different Microsoft programmers, they'd burst out with "Bill Gates ruined software" from watching all I was doing.

And while all this was going on, the printer did not stutter. My mouse did not freeze or stutter. Drawing selections and such in ImageFX was smooth, even with all that stuff going on at the same time....

When we had the equipment back home (home office, then shared space inside a sci fi collectibles store), I'd also have my web browser running with a dozen or so tabs open at the same time, often while running the Mac emulator and printing to a dye sub printer while making coffee cup prints, and running ImageFX, Real3D, and a desktop publishing program whose name escapes me.

And the mouse was so smooth... no hesitations. Typing was smooth. I could have Real3D rendering several high resolution images (2400x3000 pixels) while printing, -and- start other programs with no perceptible effect.

Nothing like Windows. I also owned a Windows 3.1 computer, and then a Windows 95 system so that I could run Pagemaker (desktop publishing). Even when using a Pentium 333MHz, Windows 95 stunk. I could not render in 3D and do anything else. If it was printing to the inkjet or rendering in 3D, I could -not- start another program without the whole thing becoming nonresponsive for several minutes, during which printing and/or rendering would stop and both the mouse and keyboard would stop responding.

In Windows 95, even if I started Pagemaker before beginning to print or render, typing or using the mouse would become a nightmare.

It wasn't until PCs started to get well over 1GHz clock speed before the interface started to feel as smooth as that 50MHz Amiga.

PCs are now really stinking fast, with more RAM than all the hard drives I had connected to my Amiga 3000. If only Commodore hadn't been gutted by Medhi Ali and Irving Gould, who knows what we'd have now?

Funny thing- it is not the hardware. I still have an old Pentium 200MMX with Windows 95 on it. It even has the same graphic chip on the video card that my Amiga had. But it stinks on ice. Slow, jerky... but I also have an Amiga emulator, Amiga Forever, running on it. Windows 95 stinks at multitasking, but it sees Amiga Forever as a single task. So within the Amiga emulator, it runs like the fastest Amiga ever!

I upgrade my Amiga Forever version every few years, and I still have the hard drive image from my now-defunct Amiga 3000. Once in a while I start it up and play around in it, and remember...

Amiga Emulation, Games, History and Support Since 1997

We even had a 3 page article done about my business in Amiga Format in 1998:

Sunday, July 24, 2022

Repurposing/Rewinding a Microwave Oven Transformer

 Transformers are funny things. Those magnetic shunts are there to limit the power available to the magnetron. They add something called leakage inductance. It is as if you’ve added an inductor in series with the primary winding.

So you are rewinding a microwave oven transformer to repurpose it? There is something very important you must understand: Microwave oven transformers are designed to use the least copper and iron possible. Drawing power from a transformer actually reduces the magnetic field strength in the core over no load conditions. So a microwave oven is designed so that the primary is only powered up to run the magnetron, and the magnetron always presents a load to the transformer.

Remove the load and power it up, excess magnetic field causes the insufficient iron core to saturate, causing the inductance to go down, causing excess current and therefore a lot of heat. There are thermal switches on the transformer to protect against it burning up if the magnetron malfunctions. A fuse won’t do it, because the transformer is still drawing less current than when the magnetron is working, but all of the power is dissipating as heat in the transformer.

Note that if you will be using the transformer in such a way that it will =always= have a large load when powered up and running it for a short time, you do not really need to add turns to the primary, but then I’d strongly suggest leaving the magnetic shunts in place. For instance, if you are building a spot welder, you have a few turns on the secondary and you are shorting them out. Then the magnetic shunts limit the short circuit current so you won’t burn up wires or pop your house breaker. In that case, you should be applying voltage to the primary ONLY when the electrodes are firmly in place on the metal being spot welded and ONLY for a short time.

If you look carefully between the windings, you can see several layers of grey metal. Those are the magnetic shunts.

To use as a more conventional power supply transformer, you must do a bit more than just replace the secondary. You can remove the shunts, then you must add more windings to the primary so the iron core doesn’t go into saturation when not under full load. They don’t have to go over the primary winding, but there is no reason not to if they’ll fit without damaging the wire.

Here, the shunts are being removed after a different secondary had been added. You can see the layers of steel in the shunts.

Here is one with the extra windings (white wire) wound over the primary and where the magnetic shunts used to be and the green/yellow wires are the new secondary:

We need to figure out how many turns we need to add. You could just measure the current draw on 120V, then add 20 turns, make sure of the phasing so it is in phase, add in series, and measure the current again at 120V.

It would be MUCH easier to find the inflection point with a variac, aka a variable transformer. Then you can monitor the current draw (no load) as you increase the voltage into the primary.

This test should be done after you have removed the secondary and the shunts, before adding a new secondary. This will alter the transformer turns ratio if you have already added a new secondary.

You are looking for that inflection point where the current/power draw starts going up steeply. The desired end result is to not go above that point. So on the above chart, let’s call it at 100V. Now we need to find out how many extra turns we need to add so that the original primary has no more than 100V with 120V line voltage coming in.

Wrap 10 turns around, fire up the variac and bring it to 100V again. Measure the AC voltage on the 10 turns. That is your voltage per 10 turns, divide by 10 to get volts per turn. You want 20V, since the incoming voltage is 120V.

Calculate how many turns you need, add them, and put them in series with the primary. Connect the variac up again, and confirm that the inflection point is now at or above 120V. If you find the current going up more than before when bringing the variac up, you have the phasing wrong. Just reverse the new winding hookup and try again. It does not hurt to put a few more windings and go lower on the loss scale.

After you have the extra primary windings installed to your satisfaction. add 10 windings as a temporary secondary, measure, divide by 10 to get the volts per turn. Now just divide the desired output voltage by the volts-per-turn to get the number of windings needed. Use the largest wire that will fit that number of windings in the space available, as the size of wire limits the current and therefore power.

Since this is relatively low voltage, thinner insulation is better, but make sure all sharp edges of the transformer core are covered. I strongly suggest using stranded wire, as large solid wire is just too hard to bend and you’ll never thread more than one turn through with solid wire.

I strongly suggest using the thermal switch that was on the transformer when it was in the microwave as a safety feature.

And as with any circuit, a fuse is a good idea. All wiring must be able to withstand in excess of the fuse rating for an indefinite period of time. The maximum expected current draw should be about 80% of the fuse rating.

Edit: Another method of doing this without adding primary windings when you have two microwave oven transformers of the same capacity:

Wire the primary windings of the two separate transformers in series. In this way, both transformers have 1/2 of the line voltage and so should be well below the current inflection point, well below saturation with no load.

To keep the load on each transformer as equal as possible, wind the same number of secondary turns, and wire them in series with each other. With turns the same, in series the current is the same, so the load will be the same. This is also a handy way to get a virtual centertapped transformer, which with a full bridge rectifier can give you equal positive and negative polarity sources. Great if you are building a custom audio amp, or a variable power supply for circuit testing.

Microwave Oven Transformers on QSL.net

Originally posted as Steven J Greenfield's answer to What is the purpose of the magnetic shunt that is in most microwave power transformers? Can the shunt be safely removed when winding a single replacement secondary?