Voltage ranges and logic levels 1130

[Roland Langfeld]

Hello,

What are the exact voltage ranges associated with the logic levels (TRUE, FALSE) in a 1130 system ? Is it possible to use standard TTL-circuits for interfacing ??

Roland

[John Doty]

On Jan 25, 2018, at 1:50 PM, Roland Langfeld <rolandl...@gmail.com> wrote:

Hello,

What are the exact voltage ranges associated with the logic levels (TRUE, FALSE) in a 1130 system ? Is it possible to use standard TTL-circuits for interfacing ??

I don’t know the details, but I also don’t think there was a single standard. The 1130 used a mix of technologies. For some of the older IBM tech, the logic levels alternated: PNP modules could drive NPN modules and vice-versa, but modules couldn’t drive their own type!

Roland

--
You received this message because you are subscribed to the Google Groups "IBM1130" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ibm1130+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

John Doty              Noqsi Aerospace, Ltd.

http://www.noqsi.com/

j...@noqsi.com

[John Pierce]

SLT is essentially RTL (resistor-transistor logic), and they used -3V and +6 or +12 V supplies for the STL logic, and depending on the circuit 'high' could be as high as 12V although a lot of it was 3V for 'high'.   

mostly SLT inputs want see < 0.29V for low, and > 2.0V for high.

Roland

--
You received this message because you are subscribed to the Google Groups "IBM1130" group.

To unsubscribe from this group and stop receiving emails from it, send an email to ibm1130+unsubscribe@googlegroups.com.

[Carl Claunch]

SLT uses 0 for logic low, +3 for logic high. It is DTL logic, using diodes to form the logic functions such as NAND and the transistor to invert and re-drive the output. 

Interfacing with current technology chips is similar to kind to the issue of interfacing CMOS and various TTL families together. 

Logic families have specifications for maximum voltage for a low input, minimum voltage for a high input - what matters is whether your output signals connected to the SLT logic will satisfy those specs. In the same way, SLT has a minimum high output and a maximum low output, which you must ensure will meet the TTL or CMOS chip input specs. 

For my FPGA based box that connected to the 1130 Storage Access Channel to extend functionality, I built level shifter transistor circuits to exactly match the 1130 requirements and on the other side the 3.3V LVCMOS requirements of the FPGA. Works perfectly. 

Be aware that any device connected to the 1130 with relatively long wiring will also have to accommodate the cumulative voltage drop across the wire and other effects of the transmission line you will have created simply by the cabling length. IBM used a special SLT circuit to drive and decode signals across transmission lines such as the IO device cables. 

Carl

[Carl Claunch]

Hi John

The circuits make internal use of the other voltages but inputs and outputs are nominally 0 and +3, with the max/min values you specified. 

SLT is DTL, not RTL. It uses diodes to form the AND, OR etc functions and a transistor to invert and redrive the output. 

Carl

[Carl Claunch]

The prior SMS technology (1401, 1620, 7090) did use two sets of levels, something they called Complementary (as in Complementary Diode Transistor Logic or CDTL), so that a gate taking a U level voltage input might create a T level output, which therefore has to be connected to a different type of gate accepting a T level input. 

The 1130 logic itself is all SLT, but some peripherals had controller logic built with SMS. Most significant was the control logic for a 1403 printer, which was a gate full of SMS cards that sat in the 1131 expansion box. There are very small amounts of SMS in devices such as the paper tape punch, but by and large the 1130 is SLT. SLT has only one set of voltages, 0 and +3, unlike the SMS generation. 

In fact, SMS generation had many voltage levels, such as U, T, S, P, N, . .. and it mixed DTL type (CDTL), RTL type (CRTL) and ECL type (current switching) plus used two types of germnanium transistor technology, speedy and cheap. A nightmare compared to SLT which had only 0, +3 or current switching. SLT did have several speed grades to the gates - the 1130 used the medium speed type while a 360 high end mainframe model needed the fast type. 

Carl

On Thursday, January 25, 2018 at 11:12:53 AM UTC-8, John Doty wrote:

I don’t know the details, but I also don’t think there was a single standard. The 1130 used a mix of technologies. For some of the older IBM tech, the logic levels alternated: PNP modules could drive NPN modules and vice-versa, but modules couldn’t drive their own type!

[Bob Flanders]

Thanks for this information, Carl. Interesting and cool.

Bob

--
You received this message because you are subscribed to the Google Groups "IBM1130" group.

To unsubscribe from this group and stop receiving emails from it, send an email to ibm1130+unsubscribe@googlegroups.com.

[de...@laptop4hire.com]

On 2018-01-26 10:57, Bob Flanders wrote:

Thanks for this information, Carl. Interesting and cool.

 

For your next IBM 1130 Trivial Pursuit game:

IBM designed and manufactured SLT modules. With the advent of the S/360, they knew they needed millions of logic IC's and they weren't absolutely sure third parties could supply enough of them. So, they decided to make their own and created the Semiconductor Division to do it. The majority of devices were DTL chips on ceramic substrates. When IBM manufactured the transistors and diodes, the tiny semiconductor chips look the same. A transistor is essentially back to back diodes with the base junction being super narrow so minority carriers live long enough to cross it to the collector. However, if you widen the junction enough, the transistor function vanishes and you end up with two back to back diodes. The little tiny chips on an SLT substrate look identical, but some are transistors, some are diodes.

The rumor I heard when SLT chips started showing up in San Jose was that the first transistor off the line cost $10 million to manufacture. But when production ramped up they'd cost a penny or less.

Actually the stories about IBM semiconductor procurement are hilarious and legendary. The best story I heard was that IBM guaranteed the success of Texas Instruments because of GE's "value engineering."

-- Dean

[Bob Flanders]

Ok... I shouldn't send it, but...

Cool story, bro!

(Really. Enjoy hearing these tidbits.)

Thanks,

Bob

--

[Roland Langfeld]

Thankyou to everybody for answering

for the purpose to create a plotterinterface with discrete logic, I found out that TTL can read the IBM-signals, but TTL-out „low“ will not be recognized as „low“ by 1130, whereas TTL „high“ will be accepted correctly. I understand for a „low“ U has to be <0.29 V, which may not be the case. I have to chek with a scope and have to use level-shifters

Regards
Roland

[Carl Claunch]

In most cases if you are going to drive a signal into the 1130, you will be connecting to a wired-OR and thus will use an open collector gate to pull the junction down to ground if your signal is activated. If you can keep the wire and connection resistance low enough, and can sink enough current with the open collector gate, it should satisfy the voltage requirement. 

Carl

[Roland Langfeld]

Thankyou Carl ! I will report on the progress

Roland