End of restoration of the early French computer SEREL ODP-505(germanium CPU) Same message, but cleaned!

What happen? I dont understand! So to have a readable message, i clean the text. I hope to have done that without mistakes... ----- Original Message ----- From: "pichotjm" <pichotjm at free.fr> To: <cctech at classiccmp.org> Sent: Sunday, March 30, 2008 8:20 PM Subject: End of restoration of the early French computer SEREL ODP-505(germanium CPU) |> ECN: Engineering change notice. Indicates revision of PCB. ||Thanks, I did not realise you had already translated it to English. |>|> The ODP-505 is a pure binary machine, not designed for arithmetics, |> statistical purposes or BCD computations.It is a real time (!!!) |> computer. |> Get some datas (voltages, frequencies, switch states, motor speed, |> temperature...) compare to thresholds or values, and speed up / |> slow down motor, or move a cadmium bar (?) in a nuclear reactor... | I see. The Elliott machines were the same (but not the ICT1301), | keeping the gun of a tank level and aimed in the right direction, | accepting data from gyros and inertial sensors and tracking the | position and orientation of an aircraft or ship, cross checking with | compasses and other navigational systems, analysing sonar data and | tracking enemy submarines etc. I think we are right. I never worked with this machine, so i do expectations (with my good knowlegde of micro-processing) | By the time of the 920ATC we even used | a special real time programming language called Coral66, based on | Algol60 syntax but with cut down semantics to give highly efficient | run times. Before that everything was done in assembler (called | S.I.R.) or machine code. I have no information about software. I don't know if they had even an assembler or if they do the job with a pencil and a rubber. (as i explain in my IBM 650 section) Have a look there: http://pichotjm.free.fr/IBM650/fileUS.html |>>| The architecture of your machine reminds me of the first machine I |>>| was allowed to operate. It was an 18 bit binary machine, it had |>>| 8k of memory built in, expandable with external modules and was made |>>| by the Airborne Computing Division (ACD) of Elliotts, it was an Elliott |>>| 920B and was a compact, flyable version of the commercial Elliott 903. |>>| There had been an earlier model 920A which I think would have been |>>| Germanium and was roughly the size and shape of a carpenters work |>>| bench. This may have been contemporary with your earlier machine I |>>| think. Unfortunately I never saw one in the flesh, but there was a |>>| bench in the computer room which I found out later, was the empty |>>| chassis of a 920A. Behind the bench was a large panel full of |>>| electrical 'chocolate strip' connectors where the analogue and |>>| digital input and output signals of the 920A could be connected up. |>>| Apparently some of the analogue outputs had been connected up to an |>>| oscilloscope to provide a visual display unit, though it used a |>>| fair bit of processor time to keep it refreshed, even with the long |>>| persistence phosphor of an oscilloscope. The panel had been covered |>>| with board with pegs to hold mylar paper tapes and until the day it |>>| was scrapped I had not seen what was behind it. |> Very interesting!There are some video connector in the earlier |> machine and the Serel company was specialized in high tech video solutions. | Possible application for computers in video back then were | superimposing images and standards conversion and just possibly | titling and graphics. You seem to have not browsed completly my website. I show about 200 beautiful photos describing these kinds of applications. We did superimposing images and standards conversion for ship, nuclear plants, helicopter tests, meteo ... have a look there http://pichotjm.free.fr/Techno73/Techno73.html I have not translated the pages (need time) but if you click on every links you will see these beautiful pictures i have taken in the 'Le Bourget' aeronautic international fair in 1973.I have no returns about these publishings, so i am not very encouraged to translate them... If you want I will try to take time. I just verify it's translated! http://pichotjm.free.fr/Techno73/Techno73US.html (probably badly, but sufficient) |> I have read |> somewhere in docs, that this computer have a screen output... |If it has two or preferably three digital to analogue converters then | a vector display is fairly easy, but maybe it was more complex than | that, and you only need one digital to analogue converter for a | raster display, provided you have enough CPU power to both drive the | display in real time and do any processing required as well. | |I should explain what I said about two or three A/D converter, you | have one to drive the X plate and one for the Y, ideally one for the | brightness (Z), but with a fast enough converter, you can deflect the | beam so fast so it will not mark the screen significantly and so no | need to turn the beam off before moving to the next item to be | displayed. I have a picture of the display terminal for the earlier one: http://pichotjm.free.fr/Serel/Notices/Serel%20computer%20dept/img_3673.html The following english page indicates 'Oscilloscopic characters display' |> I have found, |> last week a small notice describing microprogramming on |> ODP-505.http://pichotjm.free.fr/Serel/ODP505/MicroProgrammation/ |> MicroProgrammation.html(doc found in photomultiplier doc!)I have found commercial document describing |> displays and analog memories...I don't know the date... (1970?) I |> have to study these documents... | I had a look but my understanding of French is terrible, and | technical details in French are even more difficult. Do you know how I am tired after a translation to publish in english? ... ;-) |>> I have an earlier machine from the same company SEREL, named OA-1001. Built in |>> 1959/1960.I need to restore it. It lays on the floor (horizontal |>> position)... The blue one| |>> here:http://pichotjm.free.fr/Serel/Photos/Photos.htmlI will start |>> restoring|>> next month (with the Sun!) |>| |>| What is involved in the restoration? Do you intend to make it work, |>| this would be very hard without the schematics. |> As you know, i am alone, |> here. I want to make an esthetic restoration:Photos and notes, |> dismantle, photos and notes, wash, dry, fix the rust, protect with Rustol, |> photos and reassemble. I hope to be capable to do that... may be one year, may be |> two... | As I understand it, that is what the big museums call preservation. | There has been a lot of discussion of the relative merits of | preservation and restoration. Restoration is what they call it when | we return a computer to running condition, preferably so it can be | demonstrated. There is some truth that restoring a computer always | destroys some of the original and that preservation is, to the | purist, a better thing to do. I know... but i can use a bad english word... It's why i precised esthetic! | As I have two 1301s, I have the luxury | of being able to restore one and preserve the other, but restoration | is a lot more enjoyable to me. May be, but i know the time needed. And I have many projects for the future.I don't used to watch back. I prefer future [i intend to live in! ;-) ] I do a lot of programs in television, compression, splitting, ECC, FTP, transmission...(I have a son living in the USA, so he can watch french TV in broadcast quality... chutt!) | The other thing of course is simulation. This can be done at various | levels, simple simulation of the instruction set, more complex | simulation at gate level and in theory the ultimate would be to have | a 3D model of the computer you could walk around and inspect the | inside of, and attach a virtual oscilloscope to and watch the | analogue signal levels and be able to do hardware patches and have | simulated hardware faults and diagnose and repair them whilst the | lights and switches all worked like the real thing and the sounds and | 3D models of the peripherals would also be like the real thing. Maybe | I'd leave out the sound of a card deck being mangled and thrown into | the air by the card reader though. Yes i know all these kinds of technics... and can do the work (not 3D) It would probably be better to spend his time in creating a special software to do reverse engineering, (and more exciting for me) with a mixt of taking photos, and piloting schematics and routing simultaneously... |> May be some reverse engineering to get 2 or 3 schematics. (need one |> week for a board! I have 2 boards in process...)JM Pichot |Good, Yes, but i have no courage to finish the task... | are you able to identify the function of each type of board | (such as 'And' gate, Flip-Flop etc) Probably, with the help of my lectures of the sixties... | If not I could perhaps compare with my 1301 schematics and see if | they have any similarities. How difficult is it to identify the power | supply connections? For the small computer, i have to analyse schematics (to be done!)For the earlier, cables are preserved and connected. I have power supplies, so it is allways possible to trace wires... (not really exciting!) | Are you planning on doing schematics for the board interconnections? Not today! or this year. First i plan to preserve the machine and publish photos. | I imagine this would be very difficult, so probably best not to try. There is no national interest, today, in France, to preserve or study old machines... It's a huge task. I can not do that... alone. So i will wait (or my sons..) that French speak, make meetings, makereports, ask for money... may be in 20 years. Museums are for paintings, ceramics, statues... I have to be more positive!!! By the way, if you are not still fed up, i spent last days in publishing an analysis about a display terminal designed in 1969, which use 32 TV delay lines as memory! Have a look there: http://pichotjm.free.fr/DisplayDL/DisplayDLus.html Note. I have not a perfect english, so i can misuse words. I never intend to be agressive, unpolite or to hurt any body. I wish you a good evening. Jean-Marie PICHOT