• Please review our updated Terms and Rules here

Wang LOCI-2 Calculator Restoration at the RICM

m_thompson

m_thompson

Veteran Member
Joined
Jul 8, 2014
Messages
1,991
Location
Rhode Island, USA
We started working on the RICM's WANG LOCI-2 calculator today. This is a core memory, microcoded, transistor only programmable calculator with some very interesting design features. It does multiply and divide by converting to logarithms adding/subtracting the logarithms, and converting the answer back to decimal. We even have the Teletype interface for it.

It looks like there are some incorrectly performed repairs that we need to fix before we can get started on debugging.

Info is here: https://www.ricomputermuseum.org/collections-gallery/small-systems-at-ricm/wang-loci-2-calculator
 
Texas A&M University had Wang 360SE electronic calculators in the 1970's engineering department "calculator room". I used them for some of the harder Electrical Engineering problems.
This was prior to the introduction of the personal scientific calculators from HP and TI.

attachment.php


photo from https://www.oldcalculatormuseum.com/wang360.html
 
Looks like RCA and Fairchild transistors are used. I have some 1970's Fairchild epoxy trans's (2n3638's and others??) and/or RCA 2N697 (might be guessing wrong on that number) and if correct for restoration can be yours. Let me know the correct 2N numbers and I'll look.
 
Most of the transistors are Germanium 2N404. There are a few on eBay. The original RCA ones are expensive. There are also some 2N3638 and 2N3630.
 
2N404s were very popular back in the day--sort of the modern day equivalent of the 2N2222. There should still be a ton of equivalents around.
 
m_thompson said:
Most of the transistors are Germanium 2N404. There are a few on eBay. The original RCA ones are expensive. There are also some 2N3638 and 2N3630.
Click to expand...

I'm sure I have a few 2N3638's and I might have a couple of 2N404's too. Doubt if I have any 2N3630's, but I'll look. I'll get back to you after Christmas when I have time to dig through my bins.
 
I did a quick look and couldn't find and 2N3630's (not surprised), but I was surprised that I also didn't have any 2N3638's. I guess I used or sold them years ago. I do have 9 pieces of the 2N404A transistors and one 2N404(JAN) mil spec all 10 have full length leads, are TO-5 & unused. After Christmas I will PM you for shipping address.
 
Its Alive! We substituted a failed electrolytic in the +10VDC supply and get it to add and subtract. None of the other math functions yield any results other than zero. All of the broken math functions convert the decimal numbers to logarithms, then add or subtract the logarithms, and then convert the results back to decimal.

We have a complete set of spare boards in unknown condition. We substituted them one at a time, but none of the spare boards made it behave better.

It is going to be challenging to restore this machine because we don't have any high level architecture documentation or a theory of operation. We do have component layout diagrams and schematics, so that is a big help. Time to study the documentation that we have, and ask others for some help.

RICM_LOCI-2_Alive_LR.jpg
 
DeltaDon said:
Do you still think you might need the 2N404s?
Click to expand...

Yes, we will still need some 2N404s.

The LOCI will only add and subtract so far. Anything to do with logarithms is still broken. Since the majority of the transistors in the machine are 2N404s we are likely to need some replacements. One of our spare boards has a 2N404 broken off and we will try to repair the spare boars.
 
It turns out the LOCI-2 isn't as broken as we thought. We don't have a User's Manual and have been guessing about a lot of operations. We found some examples of how to use the registers in the Reference Manual and we were completely wrong about how to operate the machine.

The key presses to multiply and divide are a little arcane. To divide 100 by 20 you press the keys "PRIME", "1", "0", "0", "X", "2", "0", "the divide button", "ANTI LOG", and the result will be displayed in the NIXI tubes.

Pressing the PRIME key clears all of the registers. When you enter a number on the keyboard it goes into the W-register. Pressing the "X" key takes the log of the number in the W-register and puts it in the L-register. Pressing the "divide button" takes the log of the number in the W-register and subtracts it from the log value in the L-register. Then pressing the ANTI LOG button takes the anti-log of the value in the L-register and puts it in the W-register where it is displayed.

We can test the first four of 16 Storage Registers from the keyboard. Three of them change the second digit to a zero, and one changes all digits to zeros. Fixing them will be our next project.

The Reciprocal Square switch on the keyboard isn't working. We have a spare keyboard that has some broken parts. Between the two keyboards we should have enough to make a good one.

We think that you should be able to enter a program from the keyboard, but we haven't figured out how to do that yet.
 
I bought some blank Wang LOCI punch cards on eBay many years ago. After a long hunt I found where I had put them. These cards are printed with information on what the holes do, and the rectangular holes are embossed so you can punch them out with a pen or pencil. There are two rows of instruction codes for a total of 80 instructions. Our LOCI is a model 2A, so it can use 80 instructions instead of the normal 40. There is an option to add a second card reader for a total of 160 instructions. Each instruction equates to a keyboard button press, and there are additional instructions that do not have corresponding keyboard buttons.

Tomorrow we can connect the punched card reader and see if it can read and execute a program. There is no program memory on the LOCI, it uses the punched card as a ROM and executes the program by reading the card. This is probably not the fastest design, but it certainly eliminates a lot of circuitry in the calculator.

The LOCI-2 reference manual has some diagnostic programs and the expected results in the back. We will punch
 
I don't know about the LOCI card reader, but the similar produce for the Wang 600/520 allowed you run run several cards through the reader to create a complete program. you were not limited to the number of instructions on a single card.
 
durgadas311 said:
I don't know about the LOCI card reader, but the similar produce for the Wang 600/520 allowed you run run several cards through the reader to create a complete program. you were not limited to the number of instructions on a single card.
Click to expand...

The LOCI card reader has electrical contacts for each possible hole, so the card doesn't move through the reader. The LOCI doesn't have any program memory, so the card needs to stay in the reader. I think that the 300 series calculators worked the same way.
 
m_thompson said:
The LOCI card reader has electrical contacts for each possible hole, so the card doesn't move through the reader. The LOCI doesn't have any program memory, so the card needs to stay in the reader. I think that the 300 series calculators worked the same way.
Click to expand...

Interesting. They made a lot of advancements in just 5 years. I'd have gone through a lot of cards... can't ever seem to leave my programs well-enough alone...
 
You must log in or register to reply here.
Back
Top