diff --git a/examples/keybrd_shift_register/keybrd_shift_register.ino b/examples/keybrd_shift_register/keybrd_shift_register.ino
index 16bea9d..b331696 100644
--- a/examples/keybrd_shift_register/keybrd_shift_register.ino
+++ b/examples/keybrd_shift_register/keybrd_shift_register.ino
@@ -5,21 +5,23 @@ The keyboard hardware for this sketch has 4 shift registers,
  with every 4th input pin connected to a pull-down resistor and matrix column, also the 31st key.
 Unused input pins are not grounded, so add this line to RowScanner_SPIShiftRegisters::scan():
     //clear unpowered pins (for testing on breadboard)
-    rowState &= 0b01010001000100010001000100010001;
+    rowState &= 0b11110001000100010001000100010001; //todo
 
 Layout
-| Left  |**0**|**1**| | Right |**0**|**1**|**2**|**3**|**4**|**5**|**6**|**7**|**8**|
-|:-----:|-----|-----| |:-----:|-----|-----|-----|-----|-----|-----|-----|-----|-----|
-| **0** |  a  |  b  | | **0** |  0  |  1  |  2  |  3  |  4  |  5  |  6  |  7  |  8  |
-| **1** |  c  |  d  | | **1** |  a  |  b  |  c  |  d  |  e  |  f  |  g  |  h  |  i  |
+| Left  | **0**|**1**| | Right |**0**|**1**|**2**|**3**|**4**|**5**|**6**|**7**|**8**|
+|:-----:|------|-----| |:-----:|-----|-----|-----|-----|-----|-----|-----|-----|-----|
+| **0** |capLck|  a  | | **0** |  0  |  1  |  2  |  3  |  4  |  5  |  6  |  7  |  8  |
+| **1** |   b  |  c  | | **1** |  a  |  b  |  c  |  d  |  e  |  f  |  g  |  h  |  i  |
 */
 // ################## GLOBAL ###################
 // ================= INCLUDES ==================
 #include <Debug.h>
 #include <ScanDelay.h>
+#include <LED_PinNumber.h>
 
 //Codes
 #include <Code_Sc.h>
+#include <Code_LEDLock.h>
 
 //Matrix
 #include <Row_uC.h>
@@ -35,8 +37,27 @@ Debug debug;
 
 // ================= LEFT PINS =================
 uint8_t readPins[] = {14, 15};
-uint8_t READ_PIN_COUNT = sizeof(readPins)/sizeof(*readPins);
+uint8_t KEY_COUNT = sizeof(readPins)/sizeof(*readPins);
 
+// ==================== LEDs ===================
+LED_PinNumber LED1(16);
+
+//sometimes OS takes 6 seconds to recongnize keyboard, LED blinks from the begining
+void wait()
+{
+    for (uint8_t count = 0; count < 6; count++)
+    {
+        //print count
+        Keyboard.print(count);
+        Keyboard.print(F(" "));
+
+        //blink LED
+        LED1.on();
+        delay(500);
+        LED1.off();
+        delay(500);
+    }
+}
 // =================== CODES ===================
 Code_Sc s_a(KEY_A);
 Code_Sc s_b(KEY_B);
@@ -48,6 +69,13 @@ Code_Sc s_g(KEY_G);
 Code_Sc s_h(KEY_H);
 Code_Sc s_i(KEY_I);
 
+Code_Sc s_u(KEY_U);
+Code_Sc s_v(KEY_V);
+Code_Sc s_w(KEY_W);
+Code_Sc s_x(KEY_X);
+
+Code_Sc s_z(KEY_Z);
+
 Code_Sc s_0(KEY_0);
 Code_Sc s_1(KEY_1);
 Code_Sc s_2(KEY_2);
@@ -57,14 +85,15 @@ Code_Sc s_5(KEY_5);
 Code_Sc s_6(KEY_6);
 Code_Sc s_7(KEY_7);
 Code_Sc s_8(KEY_8);
-Code_Sc s_z(KEY_Z);
+
+Code_LEDLock o_capsLock(KEY_CAPS_LOCK, LED1);
 
 // ================= LEFT ROWS =================
-Key* ptrsKeys_L0[] = { &s_a, &s_b };
-Row_uC row_L0(0, readPins, READ_PIN_COUNT, ptrsKeys_L0);
+Key* ptrsKeys_L0[] = { &o_capsLock, &s_a };
+Row_uC row_L0(0, readPins, ptrsKeys_L0, KEY_COUNT);
 
-Key* ptrsKeys_L1[] = { &s_c, &s_d };
-Row_uC row_L1(1, readPins, READ_PIN_COUNT, ptrsKeys_L1);
+Key* ptrsKeys_L1[] = { &s_b, &s_c };
+Row_uC row_L1(1, readPins, ptrsKeys_L1, KEY_COUNT);
 
 // ================= RIGHT ROWS ================
 //typedef should be large in /home/wolfv/Documents/Arduino/keybrd_proj/keybrd/src/config_keybrd.h
@@ -105,44 +134,25 @@ const uint8_t KEY_R0_COUNT = sizeof(ptrsKeys_R0)/sizeof(*ptrsKeys_R0);
 Row_ShiftRegisters row_R0(8, ptrsKeys_R0, KEY_R0_COUNT);
 */
 
-//prints 0 1 2 3 4 5 6 7 8, microseconds_per_scan=87 with SAMPLE_COUNT 4
+//prints 0 1 2 3 3 4 5 6, microseconds_per_scan=87 with SAMPLE_COUNT 4
 Key* ptrsKeys_R0[] = { &s_0, &s_z, &s_z, &s_z, &s_1, &s_z, &s_z, &s_z,
                        &s_2, &s_z, &s_z, &s_z, &s_3, &s_z, &s_z, &s_z,
                        &s_4, &s_z, &s_z, &s_z, &s_5, &s_z, &s_z, &s_z,
-                       &s_6, &s_z, &s_z, &s_z, &s_7, &s_z, &s_8 }; //31-key limit because endRow
+                       &s_6, &s_z, &s_z, &s_z, &s_3, &s_4, &s_5, &s_6 };
 const uint8_t KEY_R0_COUNT = sizeof(ptrsKeys_R0)/sizeof(*ptrsKeys_R0);
 Row_ShiftRegisters row_R0(8, ptrsKeys_R0, KEY_R0_COUNT);
 
-//prints a b c d e f g h i
+//prints a b c d u v w x
 Key* ptrsKeys_R1[] = { &s_a, &s_z, &s_z, &s_z, &s_b, &s_z, &s_z, &s_z,
                        &s_c, &s_z, &s_z, &s_z, &s_d, &s_z, &s_z, &s_z,
                        &s_e, &s_z, &s_z, &s_z, &s_f, &s_z, &s_z, &s_z,
-                       &s_g, &s_z, &s_z, &s_z, &s_h, &s_z, &s_i }; //31-key limit because endRow
+                       &s_g, &s_z, &s_z, &s_z, &s_u, &s_v, &s_w, &s_x };
 const uint8_t KEY_R1_COUNT = sizeof(ptrsKeys_R1)/sizeof(*ptrsKeys_R1);
 Row_ShiftRegisters row_R1(9, ptrsKeys_R1, KEY_R1_COUNT);
 
-const uint8_t LED_PIN = 16;                     //indicates wait
-
-//sometimes OS takes 6 seconds to recongnize keyboard, LED blinks from the begining
-void wait()
-{
-    for (uint8_t count = 0; count < 6; count++)
-    {
-        //print count
-        Keyboard.print(count);
-        Keyboard.print(F(" "));
-
-        //blink LED
-        digitalWrite(LED_PIN, HIGH);
-        delay(900);
-        digitalWrite(LED_PIN, LOW);
-        delay(100);
-    }
-}
 // ################### MAIN ####################
 void setup()
 {
-    pinMode (LED_PIN, OUTPUT);
     Keyboard.begin();
 
     wait();
@@ -150,8 +160,7 @@ void setup()
     row_R0.begin();
     row_R1.begin();
 
-    Keyboard.print(F("keybrd_shift_reg.ino "));
-    debug.print_free_RAM();
+    Keyboard.println(F("keybrd_shift_reg.ino"));
 }
 
 void loop()
diff --git a/src/Code_LEDLock.cpp b/src/Code_LEDLock.cpp
index c5c502d..6a91533 100644
--- a/src/Code_LEDLock.cpp
+++ b/src/Code_LEDLock.cpp
@@ -59,7 +59,7 @@ This debug code prints "keyboard_leds=0" when scrollLock is pressed:
 */
     if (keyboard_leds & USB_LED_bit)            //if USB_LED_bit is set
     {
-        refLED.off();                      //LED on-off seem inverted, but it works for active high
+        refLED.off();       //LED on-off seem inverted, but it works for active low and active high
     }
     else
     {
diff --git a/src/ColPort_PCA9655E.cpp b/src/ColPort_PCA9655E.cpp
index ed71571..19317d6 100644
--- a/src/ColPort_PCA9655E.cpp
+++ b/src/ColPort_PCA9655E.cpp
@@ -3,8 +3,7 @@
 /*
 configures column port's configuration, input, and pins.
 */
-ColPort_PCA9655E::ColPort_PCA9655E
-(IOExpanderPort& port, const uint8_t colPins)
+ColPort_PCA9655E::ColPort_PCA9655E (IOExpanderPort& port, const uint8_t colPins)
     : ColPort(colPins), port(port), configurationByteCommand(port.num + 6), inputByteCommand(port.num)
 {}
 
diff --git a/src/Debug.cpp b/src/Debug.cpp
index e2e558b..62ea166 100644
--- a/src/Debug.cpp
+++ b/src/Debug.cpp
@@ -1,12 +1,4 @@
 #include "Debug.h"
-#include "getFreeSRAM.h"
-
-void Debug::print_free_RAM()
-{
-    delay(1000);                                //give OS time to find USB
-    Keyboard.print(F("Free SRAM = "));
-    Keyboard.println( getFreeSRAM() );
-}
 
 void Debug::print_microseconds_per_scan()
 {
diff --git a/src/Debug.h b/src/Debug.h
index f047f12..4afe0d3 100644
--- a/src/Debug.h
+++ b/src/Debug.h
@@ -9,7 +9,6 @@ class Debug
         unsigned int scanCount = 0;
 
     public:
-        void print_free_RAM();                  //print free SRAM, call this from setup()
         void print_microseconds_per_scan();     //print microseconds per scan every second
         void print_scans_per_second();          //print scans per second every second
 };
diff --git a/src/RowBase.cpp b/src/RowBase.cpp
index 7b03575..be4690d 100644
--- a/src/RowBase.cpp
+++ b/src/RowBase.cpp
@@ -2,13 +2,12 @@
 /*
 pressRelease() calls key's press() or release() function if it was pressed or released.
 Both parameters are bitwise.
-rowEnd bit marks positioned immediatly after last key of row.
 */
 void RowBase::pressRelease(const uint8_t KEY_COUNT, const read_pins_t debouncedChanged)
 {
     read_pins_t isFallingEdge;                  //bitwise, 1 means falling edge
     read_pins_t isRisingEdge;                   //bitwise, 1 means rising edge
-    read_pins_mask_t rowMask;                   //bitwise, active col bit is 1
+    read_pins_t rowMask;                        //bitwise, active col bit is 1
     uint8_t col;                                //index for ptrsKeys[col] array
 
     //bit=1 if last debounced changed from 1 to 0, else bit=0
diff --git a/src/RowBase.h b/src/RowBase.h
index 9a70dd4..e0102fc 100644
--- a/src/RowBase.h
+++ b/src/RowBase.h
@@ -14,7 +14,7 @@ class RowBase
         virtual void keyWasPressed();
     protected:
         read_pins_t debounced;                  //bitwise, 1 means pressed, 0 means released
-        void pressRelease(const read_pins_mask_t rowEnd, const read_pins_t debouncedChanged);
+        void pressRelease(const uint8_t KEY_COUNT, const read_pins_t debouncedChanged);
     public:
         RowBase(Key *const ptrsKeys[]) : ptrsKeys(ptrsKeys), debounced(0) { }
         virtual void process()=0;
diff --git a/src/RowScanner_PinsArray.cpp b/src/RowScanner_PinsArray.cpp
index d51ce46..19742bf 100644
--- a/src/RowScanner_PinsArray.cpp
+++ b/src/RowScanner_PinsArray.cpp
@@ -28,10 +28,7 @@ RowScanner_PinsArray::RowScanner_PinsArray(const uint8_t STROBE_PIN,
 }
 
 /* scan() Strobes the row and reads the columns.
-Sets rowEnd and returns rowState.
-rowEnd is a bitwise row mask, one col per bit, where active col bit is 1.
-At end of function, 1 bit marks place immediatly after last key of row.
-rowEnd is a larger type than portMask so that it can not overflow.
+Sets KEY_COUNT and returns rowState.
 
 https://www.arduino.cc/en/Tutorial/DigitalPins
 https://www.arduino.cc/en/Reference/PinMode
@@ -39,10 +36,10 @@ https://www.arduino.cc/en/Reference/DigitalWrite
 https://www.arduino.cc/en/Reference/DigitalRead
 https://www.arduino.cc/en/Reference/Constants > Digital Pins modes: INPUT, INPUT_PULLUP, and OUTPUT
 */
-read_pins_t RowScanner_PinsArray::scan(read_pins_mask_t& rowEnd)
+read_pins_t RowScanner_PinsArray::scan(uint8_t& keyCount)
 {
-    read_pins_t rowState = 0;                   //bitwise
-    rowEnd = 1;
+    read_pins_t rowState = 0;                   //bitwise, one col per bit, 1 means key is pressed
+    read_pins_t rowMask = 1;                    //bitwise, one col per bit, active col bit is 1
 
     //strobe row on
     if (ACTIVE_HIGH)
@@ -60,9 +57,9 @@ read_pins_t RowScanner_PinsArray::scan(read_pins_mask_t& rowEnd)
     {
         if ( digitalRead(READ_PINS[i]) == ACTIVE_HIGH )
         {
-            rowState |= rowEnd;
+            rowState |= rowMask;
         }
-        rowEnd <<= 1;
+        rowMask <<= 1;
     }
 
     //strobe row off
@@ -75,5 +72,6 @@ read_pins_t RowScanner_PinsArray::scan(read_pins_mask_t& rowEnd)
         digitalWrite(STROBE_PIN, HIGH);
     }
 
+    keyCount = READ_PIN_COUNT;
     return rowState;
 }
diff --git a/src/RowScanner_PinsArray.h b/src/RowScanner_PinsArray.h
index ddb7e04..09f5a75 100644
--- a/src/RowScanner_PinsArray.h
+++ b/src/RowScanner_PinsArray.h
@@ -7,7 +7,6 @@
 #include <ColPort.h>
 
 /* RowScanner_PinsArray class uses Arduino pin numbers (not port pin numbers).
-The maximum keys per row is 31, because Arduino's largest type is 32 bits and rowEnd consumes the last bit.
 Constructor is in RowScanner_PinsArray.cpp
 */
 class RowScanner_PinsArray
@@ -19,9 +18,8 @@ class RowScanner_PinsArray
         const uint8_t READ_PIN_COUNT;         //number of read pins
     public:
         RowScanner_PinsArray(const uint8_t STROBE_PIN,
-                const uint8_t READ_PINS[], const uint8_t READ_PIN_COUNT);
-        virtual read_pins_t scan(read_pins_mask_t& rowEnd);
-        //read_pins_t getRowState(read_pins_mask_t& rowEnd);
+                const uint8_t READ_PINS[], const uint8_t READ_PIN_COUNT); //todo rename READ_PIN_COUNT to KEY_COUNT ??
+        virtual read_pins_t scan(uint8_t& KEY_COUNT);
 };
 #endif
 
diff --git a/src/RowScanner_PinsBitwise.cpp b/src/RowScanner_PinsBitwise.cpp
index 6976f5e..b00a052 100644
--- a/src/RowScanner_PinsBitwise.cpp
+++ b/src/RowScanner_PinsBitwise.cpp
@@ -1,7 +1,6 @@
 #include "RowScanner_PinsBitwise.h"
 /*
 Strobes the row and reads the columns.
-Sets rowEnd and returns rowState.
 */
 ColPort* const RowScanner_PinsBitwise::scan()
 {
@@ -29,6 +28,5 @@ ColPort* const RowScanner_PinsBitwise::scan()
         refRowPort.setActivePinHigh(strobePin);
     }
    
-//    return getRowState(refColPort, rowEnd);
     return &refColPort;
 }
diff --git a/src/RowScanner_SPIShiftRegisters.cpp b/src/RowScanner_SPIShiftRegisters.cpp
index 730a259..994ebd1 100644
--- a/src/RowScanner_SPIShiftRegisters.cpp
+++ b/src/RowScanner_SPIShiftRegisters.cpp
@@ -2,7 +2,7 @@
 
 //constructor
 RowScanner_SPIShiftRegisters::RowScanner_SPIShiftRegisters(const uint8_t STROBE_PIN, uint8_t KEY_COUNT)
-    : STROBE_PIN(STROBE_PIN), ROW_END(1 << KEY_COUNT), BYTE_COUNT(ceil (float(KEY_COUNT)/8))
+    : STROBE_PIN(STROBE_PIN), BYTE_COUNT(ceil (float(KEY_COUNT)/8)), KEY_COUNT(KEY_COUNT)
 {
     //configure controller to communicate with shift register matrix
     pinMode(STROBE_PIN, OUTPUT);
@@ -16,9 +16,9 @@ void RowScanner_SPIShiftRegisters::begin()
 }
 
 /*
-Sets rowEnd and returns rowState.
+Sets keyCount and returns rowState.
 */
-read_pins_t RowScanner_SPIShiftRegisters::scan(read_pins_mask_t& rowEnd)
+read_pins_t RowScanner_SPIShiftRegisters::scan(uint8_t& keyCount)
 {
     read_pins_t rowState = 0;
 
@@ -34,10 +34,10 @@ read_pins_t RowScanner_SPIShiftRegisters::scan(read_pins_mask_t& rowEnd)
     //strobe row off
     digitalWrite(STROBE_PIN, LOW);
 
-    rowEnd = ROW_END;
+    keyCount = KEY_COUNT;
 
-    //for testing breadboard, clear unpowered pins
-    rowState &= 0b01010001000100010001000100010001; //todo
+    //for testing on breadboard, clear unpowered pins
+    rowState &= 0b11110001000100010001000100010001; //todo
 
     return rowState;
 }
diff --git a/src/RowScanner_SPIShiftRegisters.h b/src/RowScanner_SPIShiftRegisters.h
index b6e9139..6b106eb 100644
--- a/src/RowScanner_SPIShiftRegisters.h
+++ b/src/RowScanner_SPIShiftRegisters.h
@@ -15,7 +15,6 @@ in sketch:
     call begin() from setup()
 
 Upto 4 shift registers can be in a daisy chained.
-The maximum keys per row is 31, because Arduino's largest type is 32 bits and rowEnd consumes the last bit.
 
 The shift registers are active high:
     10k pull-down resistors are grounded
@@ -33,11 +32,11 @@ class RowScanner_SPIShiftRegisters
     private:
         static const uint8_t SHIFT_LOAD;        //controller's pin number that is connected to shift register's SHIFT_LOAD pin
         const uint8_t STROBE_PIN;               //Arduino pin number connected to this row
-        const read_pins_mask_t ROW_END;         //bitwise, 1 bit marks positioned after last key of row
         const uint8_t BYTE_COUNT;               //number of bytes to read from shift registers
+        uint8_t KEY_COUNT;
     public:
         RowScanner_SPIShiftRegisters(const uint8_t STROBE_PIN, uint8_t KEY_COUNT);
-        virtual read_pins_t scan(read_pins_mask_t& rowEnd);
+        virtual read_pins_t scan(uint8_t& KEY_COUNT);
         void begin();
 };
 #endif
diff --git a/src/Row_ShiftRegisters.cpp b/src/Row_ShiftRegisters.cpp
index a0dc898..0e60540 100644
--- a/src/Row_ShiftRegisters.cpp
+++ b/src/Row_ShiftRegisters.cpp
@@ -4,12 +4,12 @@ void Row_ShiftRegisters::process()
 {
     //these variables are all bitwise, one bit per key
     read_pins_t rowState;                       //1 means pressed, 0 means released
-    read_pins_mask_t rowEnd;                    //1 bit marks positioned after last key of row
+    uint8_t keyCount;
     read_pins_t debouncedChanged;               //1 means debounced changed
 
-    rowState = scanner.scan(rowEnd);
+    rowState = scanner.scan(keyCount);
     debouncedChanged = debouncer.debounce(rowState, debounced);
-    pressRelease(rowEnd, debouncedChanged);
+    pressRelease(keyCount, debouncedChanged);
 }
 
 void Row_ShiftRegisters::begin()
diff --git a/src/Row_uC.cpp b/src/Row_uC.cpp
index a147f06..bef7362 100644
--- a/src/Row_uC.cpp
+++ b/src/Row_uC.cpp
@@ -7,10 +7,20 @@ void Row_uC::process()
 {
     //these variables are all bitwise, one bit per key
     read_pins_t rowState;                       //1 means pressed, 0 means released
-    read_pins_mask_t rowEnd;                    //1 bit marks positioned after last key of row
+    uint8_t keyCount;
     read_pins_t debouncedChanged;               //1 means debounced changed
 
-    rowState = scanner.scan(rowEnd);
+    rowState = scanner.scan(keyCount);
+/*
+Keyboard.print(" keyCount=");
+Keyboard.print(keyCount);
+Keyboard.print(" rowState=");
+Keyboard.print(rowState);
+*/
     debouncedChanged = debouncer.debounce(rowState, debounced);
-    pressRelease(rowEnd, debouncedChanged);
+/*
+Keyboard.print(" debounced=");
+Keyboard.print(debounced);
+*/
+    pressRelease(keyCount, debouncedChanged);
 }
diff --git a/src/config_keybrd.h b/src/config_keybrd.h
index f3d84cd..30f2c8f 100644
--- a/src/config_keybrd.h
+++ b/src/config_keybrd.h
@@ -2,8 +2,8 @@
 #define CONFIG_KEYBRD_H
 #include <inttypes.h>
 
-/* size of read_pins_t and read_pins_mask_t depends on the maximum number of pins scanned by RowScanner.
-By default, read_pins_t and read_pins_mask_t are set to the largest type.
+/* size of read_pins_t depends on the maximum number of pins scanned by RowScanner.
+By default, read_pins_t is set to the largest type.
 If your 8-bit AVR is running low on memory, using a smaller type saves SRAM.
 Using smaller types on a 32-bit uC (Teensy LC) would accomplish nothing.
 */
@@ -13,17 +13,9 @@ Using smaller types on a 32-bit uC (Teensy LC) would accomplish nothing.
     For RowScanner_SPIShiftRegisters, RowScanner_SPIShiftRegisters::KEY_COUNT
     For RowScanner_PinsBitwise, cover the last 1 bit in RowScanner_PinsBitwise::strobePin
 */
-typedef uint8_t read_pins_t;
+//typedef uint8_t read_pins_t;
 //typedef uint16_t read_pins_t;
-//typedef uint32_t read_pins_t;
-
-/* read_pins_mask_t is only used for rowMask and rowEnd, which extends one bit beyond the last col pin.
-uncomment typedef that covers one bit beyond the last col pin.
-This could be the same typedef as read_pins_t, or the next larger typedef.
-*/
-typedef uint8_t read_pins_mask_t;
-//typedef uint16_t read_pins_mask_t;
-//typedef uint32_t read_pins_mask_t;
+typedef uint32_t read_pins_t;
 
 /* SAMPLE_COUNT = 4 is very reliable for a keyboard.
 Split keyboards with a long connecting wire or in environment with
diff --git a/tutorials/images/bend_diodes_en_masse2.JPG b/tutorials/images/bend_diodes_en_masse2.JPG
new file mode 100644
index 0000000..bcaded4
Binary files /dev/null and b/tutorials/images/bend_diodes_en_masse2.JPG differ
diff --git a/tutorials/images/bend_diodes_en_masse3.JPG b/tutorials/images/bend_diodes_en_masse3.JPG
new file mode 100644
index 0000000..52f4fba
Binary files /dev/null and b/tutorials/images/bend_diodes_en_masse3.JPG differ
diff --git a/tutorials/images/breadboard_keyboard_2x2_crossColumns.jpg b/tutorials/images/breadboard_keyboard_2x2_crossColumns.jpg
new file mode 100644
index 0000000..2258983
Binary files /dev/null and b/tutorials/images/breadboard_keyboard_2x2_crossColumns.jpg differ