Skip to main content

Full text of "StandardMicrosystems-COM20020ULANCUniversalLocalAreaNetworkControllerWith2Kx8On-BoardRAMOCR"

See other formats


STANDARD TECHNOLOGIES CORPORATION 



P.O. BOX 325, CANONSBURG, PA 15317 
(412) 746-8696 FAX 745-8596 



PRODUCT BRIEF 



ST1601 

Serial Controlled Math Processor 





Pin identification 




1 


MCL 


CKPL 


28 


2 


NC1 


NC15 


27 


3 


NC2 


NCI 4 


26 


4 


NC3 


NC13 


25 


5 


NC4 


NC12 


24 


6 


NC5 


NC11 


23 


7 


OE 


BUSY 


22 


8 


VSS 


CS 


21 


9 


OSCl 


VDD 


20 


10 


OSCO 


VSS 


19 


11 


NC6 


NC10 


18 


12 


NC7 


NC9 


17 


13 


NC8 


SDO 


16 


14 


SCL 


SOI 


15 



FLOATING POINT MATH FUNCTIONS 

*Multiply 

*Divide 

*Add 

*Subtract 
*Sine 

•Square root 



CONVERSION UTILITIES 
*Binary I/O 

*Two's compliment I/O 



'BCD I/O 



NC=NO CONNECT 



GENERAL DESCRIPTION 

The ST1601 is a high performance, low cost math processor specifically 
designed to suppliment 3-wire SPI type processors. Use of the ST1601 greatly 
simplifies arithmetic data handling in embedded control applications which would 
otherwise put a heavy burden on the requirements of the host processor. 
Controlled by a high-speed internal RISC type processor core, the ST1601 
quickly calculates floating point results having mantissa's up to 16 bits with 
exponents up to 8 bits (plus signs). A flexible I/O format facilitates data 
translation to and from Binary, Two's compliment, and BCD, so the ST1601 can 
be commanded to accept one type of data format and output the result in 
another format. This feature is very useful for instance in reading data from a 
binary A/D converter, applying a calibration factor, and outputting the result 
directly to a BCD type display driver without requiring any conversion by the host 
processor. While the 3-wire SPI interface is common on many popular 
microcontrollers, if not present on your processor, it is easily simulated in 
software. 



SPECIFICATIONS 

DATA INPUT STRUCTURE 

—--VARIABLE A VARIABLE B 

Command Mantissa Exponent Mantissa Exponent 
1 byte 2[3] bytes 1 [2] byte 2 [3] bytes 1 [2] byte 



Signs 
1 byte 



Command 
1 byte 



Mantissa 
2[3] bytes 



DATA OUTPUT STRUCTURE 

-RESULT 

Exponent Signs 
1 [2] byte 1 byte 



Notes: SIGNS byte is only used for straight binary format. 
[ ] = Number of bytes for BCD format. 



TYPICAL COMPUTING TIMES @ 20MHz 

ADDITION 600uS 
SUBTRACTION 700uS 
MULTIPLICATION 3mS 
DIVISION 2mS 
SQUARE ROOT 8mS 
SINE 500uS 



ACCURACY 

0% error 
0% error 
0.03% max error 
0.04% max error 
0.1% max error 
0.3% max error 



Supply voltage 
Supply current 
Operating speeds 
Operating temps 
Packages 
SPI Clock 



4.5 to 5.5 V 
13mAtyp @ 20MHz 

4MHz, 10MHz, 20MHz (on-chip xtal or ext clock) 
to +70C, -40 to +85C, -40 to +125C 
300 mil 28 pin DIP, 300 mil 28 pin SOIC 
Selectable as normally high or normally low polarity 



TYPICAL APPLICATION 



PROCESSOR 





VDD CKPL 
BUSY 
OSCI cs 
ST1601 

SDI 

osco SD0 

SCL 

VSS 




68HC05 

PA.1 
PA.2 

SDO 

SDI 

SCL 











T 



Specifications subject to change without notice. rev 4/22/96 



(412) 746-8696 FAX 745-8596 



Dear Customer, 



Thank you for your interest in our products! Per your request we are 
submitting information for your evaluation. Enclosed is a short form product brief 
on the ST1601 Serial Math Co-Processor. This information is of a general 
nature , so do not hesitate to contact us for detailed technical or pricing 
information if you have further interest in this or other products. 

The ST1601 is the first in a series of intelligent math processors used to 
suppliment embedded microcontrollers using a standard 3-wire serial interface. 
In addition to the ST1601 , the following products are scheduled for release 
throughout the next year. 



ST1606 



ST1602 
ST1603 
ST1604 
ST1605 

ST1607 

ST1608 



Same as ST1601 but with a programmable 32-bit pulse counter. 
Scheduled for 3rd quarter 1996 

The following are scheduled for 4th quarter 1996. 

Floating point Add, Sub, Mult, Div, Square root, Exponential. 

Floating point Add, Sub, Mult, Div, Square root, base 10 Log. 

Floating point Add, Sub, Mult, Div, Square root, natural Log. 

Floating point Add, Sub, Mult, Div, Square root, and four 8-bit 

A/D inputs which can be integrated into calculations. 

Add, Sub, Mult, Div, Square root. Fixed point 32-bit processor 

for high speed/high resolution. 

Add, sub, mult, div, square root. Fixed point 24-bit processor 
for very high speed/high resolution. 



A complete technical manual is available for the ST1601 and a preliminary 
manual for the ST1606. Product briefs are available for all future products listed 
above. 



All products are available in a wide variety of operating speeds and temperature 
grades with prices ranging from $11.66 to $21.17 each in 100 piece lots. Please 
request our current price schedule for full details.