diff --git a/Makefile b/Makefile
index 5e85620..72141ec 100644
--- a/Makefile
+++ b/Makefile
@@ -1,7 +1,9 @@
 SDCC ?= sdcc
-STCCODESIZE ?= 13000
-SDCCOPTS ?= --code-size $(STCCODESIZE) --xram-size 0 --stack-auto
+STCCODESIZE ?= 13312
+SDCCOPTS ?= --code-size $(STCCODESIZE) --xram-size 256 --stack-auto --idata-loc 0x80
+#SDCCOPTS ?= --code-size $(STCCODESIZE) --xram-size 256 --stack-auto --model-large
 FLASHFILE ?= main.hex
+LARGE_LDFLAGS += -L/usr/share/sdcc/lib/large/
 
 SRC = src/lcd.c src/key.c src/utils.c src/decn/decn.c src/calc.c
 
@@ -15,6 +17,7 @@ build/%.rel: src/%.c src/%.h
 
 main: $(OBJ)
 	$(SDCC) -o build/ src/$@.c $(SDCCOPTS) $(CFLAGS) $^
+#	$(SDCC) -o build/ src/$@.c $(SDCCOPTS) $(CFLAGS) $(LARGE_LDFLAGS) $^
 	@ tail -n 5 build/main.mem | head -n 2
 	@ tail -n 1 build/main.mem
 	cp build/$@.ihx $@.hex
diff --git a/src/decn/decn.c b/src/decn/decn.c
index d76945e..246ea6f 100644
--- a/src/decn/decn.c
+++ b/src/decn/decn.c
@@ -7,13 +7,16 @@
 
 #include "../utils.h"
 
+#define EXTRA_CHECKS
 //#define DEBUG
 //#define DEBUG_COMPARE_MAGN
 //#define DEBUG_ADD
 //#define DEBUG_MULT
+//#define DEBUG_MULT_ALL //even more verbose
 //#define DEBUG_DIV
 
 #ifndef DESKTOP
+//#undef EXTRA_CHECKS
 #undef DEBUG
 #undef DEBUG_COMPARE_MAGN
 #undef DEBUG_ADD
@@ -34,9 +37,9 @@
 #include "decn.h"
 
 #ifdef DESKTOP
-uint8_t num_digits_display = DEC80_NUM_LSU*2;
+static const uint8_t num_digits_display = DEC80_NUM_LSU*2;
 #else
-uint8_t num_digits_display = 16;
+static const uint8_t num_digits_display = 16;
 #endif
 
 
@@ -168,7 +171,9 @@ void build_dec80(dec80* dest, const char* signif_str, int16_t exponent){
 			if (!SEEN_POINT(curr_sign)){
 				//begin tracking number of digits to right of decimal point
 				curr_sign |= 1; //seen point
-			} else {
+			}
+#ifdef EXTRA_CHECKS
+			else {
 				//multiple '.'s in string
 #ifdef DEBUG
 				printf("  ERROR: multiple '.'s in string\n");
@@ -176,6 +181,7 @@ void build_dec80(dec80* dest, const char* signif_str, int16_t exponent){
 				set_dec80_NaN(dest);
 				return;
 			}
+#endif
 		} else if (signif_str[i] >= '1' && signif_str[i] <= '9'){
 			if (nibble_i < DEC80_NUM_LSU*2){
 				if (nibble_i & 1) { //odd
@@ -245,6 +251,7 @@ void build_dec80(dec80* dest, const char* signif_str, int16_t exponent){
 					assert(DEC80_NUM_LSU*2 > num_lr_points);
 					new_exponent = exponent + (num_lr_points - 1); //1 digit left of implicit point
 					//check for overflow
+#ifdef EXTRA_CHECKS
 					if (new_exponent < exponent || exponent > DEC80_MAX_EXP){
 	#ifdef DEBUG
 						printf("   overflow (new_exp, exp)=(%d,%d)\n",
@@ -253,12 +260,14 @@ void build_dec80(dec80* dest, const char* signif_str, int16_t exponent){
 						set_dec80_NaN(dest);
 						return;
 					}
+#endif
 				} else if (num_lr_points < 0) { //right count exists
 					// (-num_past_point represents #0s right of decimal)
 					// (this ends up being a subtraction)
 					new_exponent = exponent + num_lr_points;
 					new_exponent -= 1; //decimal point after 1st non-zero number
 					//check for underflow
+#ifdef EXTRA_CHECKS
 					if (new_exponent > exponent || exponent < DEC80_MIN_EXP){
 	#ifdef DEBUG
 						printf("   underflow (new_exp, exp)=(%d,%d)\n",
@@ -267,6 +276,7 @@ void build_dec80(dec80* dest, const char* signif_str, int16_t exponent){
 						set_dec80_NaN(dest);
 						return;
 					}
+#endif
 				} else {
 					//no change
 					new_exponent = exponent;
@@ -319,6 +329,7 @@ static uint8_t decn_is_zero(const dec80* x){
 	return 1;
 }
 
+#ifdef EXTRA_CHECKS
 void set_dec80_NaN(dec80* dest){
 	uint8_t i;
 
@@ -329,6 +340,7 @@ void set_dec80_NaN(dec80* dest){
 		dest->lsu[i] = 0;
 	}
 }
+#endif
 
 void negate_decn(dec80* x){
 	static const int16_t xor_val = -(0x7fff) - 1;
@@ -339,7 +351,7 @@ int8_t compare_magn(const dec80* a, const dec80* b){ //a<b: -1, a==b: 0, a>b: 1
 	uint8_t a_i, b_i;
 	int16_t a_exp=0, b_exp=0;
 	int8_t a_signif_b = 0; //a<b: -1, a==b: 0, a>b: 1
-	dec80 a_tmp, b_tmp;
+	static __xdata dec80 a_tmp, b_tmp;
 	//copy
 	copy_decn(&a_tmp, a);
 	copy_decn(&b_tmp, b);
@@ -448,7 +460,7 @@ static void _incr_exp(dec80* acc, int16_t exponent){
 static void sub_mag(dec80* acc, const dec80* x){
 	int8_t i;
 	uint8_t carry = 0;
-	dec80 tmp;
+	static __xdata dec80 tmp;
 	copy_decn(&tmp, x);
 	//normalize
 	remove_leading_zeros(acc);
@@ -478,7 +490,7 @@ static void sub_mag(dec80* acc, const dec80* x){
 }
 
 void add_decn(dec80* acc, const dec80* x){
-	dec80 tmp;
+	static __xdata dec80 tmp;
 	int8_t rel;
 	uint8_t carry = 0;
 	int8_t i;
@@ -599,8 +611,8 @@ void add_decn(dec80* acc, const dec80* x){
 }
 
 void mult_decn(dec80* acc, const dec80* x){
-	dec80 tmp; //copy of x
-	dec80 acc_tmp; //holds sum
+	static __xdata dec80 tmp; //copy of x
+	static __xdata dec80 acc_tmp; //holds sum
 	int8_t i, j;
 	uint8_t carry = 0;
 	uint8_t is_neg;
@@ -631,7 +643,7 @@ void mult_decn(dec80* acc, const dec80* x){
 			carry = digit100 / 100;
 			assert(carry < 100);
 		}
-#ifdef DEBUG_MULT
+#ifdef DEBUG_MULT_ALL
 		printf("\n%d:", i);
 		printf("\n      acc:");
 		for (j = 0; j < DEC80_NUM_LSU; j++){
@@ -644,6 +656,8 @@ void mult_decn(dec80* acc, const dec80* x){
 			else
 				printf("    ");
 		}
+#endif
+#ifdef DEBUG_MULT
 		printf("\n  acc_tmp:");
 		for (j = 0; j < DEC80_NUM_LSU; j++){
 			printf(" %3d", acc_tmp.lsu[j]);
@@ -681,14 +695,16 @@ void mult_decn(dec80* acc, const dec80* x){
 }
 
 void div_decn(dec80* acc, const dec80* x){
-	dec80 tmp; //copy of x, holds current 1/x estimate
-	dec80 acc_copy; //holds copy of original acc
+	static __xdata dec80 tmp; //copy of x, holds current 1/x estimate
+	static __xdata dec80 acc_copy; //holds copy of original acc
 	uint8_t i;
 	//check divide by zero
+#ifdef EXTRA_CHECKS
 	if (decn_is_zero(x)){
 		set_dec80_NaN(acc);
 		return;
 	}
+#endif
 	//store copy of acc for final multiply by 1/x
 	copy_decn(&acc_copy, acc);
 	//get initial estimate for 1/x, by negating exponent, and setting signif. to 1
@@ -702,24 +718,24 @@ void div_decn(dec80* acc, const dec80* x){
 	for (i = 0; i < 20; i++){ //just fix number of iterations for now
 #ifdef DEBUG_DIV
 		extern char Buf[80];
-		dec80_to_str(buf, &tmp);
-		printf("%2d: %s\n", i, buf);
+		dec80_to_str(Buf, &tmp);
+		printf("%2d: %s\n", i, Buf);
 #endif
 		mult_decn(acc, x);
 #ifdef DEBUG_DIV
-		dec80_to_str(buf, acc);
-		printf("  %20s: %s\n", "recip*x", buf);
+		dec80_to_str(Buf, acc);
+		printf("  %20s: %s\n", "recip*x", Buf);
 #endif
 		negate_decn(acc);
 		add_decn(acc, &DECN_1);
 #ifdef DEBUG_DIV
-		dec80_to_str(buf, acc);
-		printf("  %20s: %s\n", "(1-recip*x)", buf);
+		dec80_to_str(Buf, acc);
+		printf("  %20s: %s\n", "(1-recip*x)", Buf);
 #endif
 		mult_decn(acc, &tmp);
 #ifdef DEBUG_DIV
-		dec80_to_str(buf, acc);
-		printf("  %20s: %s\n", "recip * (1-recip*x)", buf);
+		dec80_to_str(Buf, acc);
+		printf("  %20s: %s\n", "recip * (1-recip*x)", Buf);
 #endif
 		add_decn(acc, &tmp);
 		//new_est(acc) = recip + (1 - recip*x)*recip, where tmp is current recip estimate
@@ -737,7 +753,7 @@ void dec80_to_str(char* buf, const dec80* x){
 	int16_t exponent = 0;
 	uint8_t trailing_zeros = 0;
 	uint8_t use_sci = 0;
-	dec80 tmp;
+	static __xdata dec80 tmp;
 
 	//copy and normalize
 	copy_decn(&tmp, x);
diff --git a/src/key.c b/src/key.c
index f7b0013..d4cf0ff 100644
--- a/src/key.c
+++ b/src/key.c
@@ -7,16 +7,16 @@
 
 #include <stdint.h>
 #include "utils.h"
+#include "key.h"
 
 #ifndef DESKTOP
 #include "stc15.h"
 #endif
 
 
-#define TOTAL_ROWS 5
 #define M_COLS 4
 
-static uint8_t keys[TOTAL_ROWS]; //only bottom nibbles get set
+uint8_t Keys[TOTAL_ROWS]; //only bottom nibbles get set
 static uint32_t new_keys_pressed; //bottom 20 bits get set
 static uint8_t last_state[TOTAL_ROWS]; //all used, 2 bits per key
 static int8_t last_count[TOTAL_ROWS][M_COLS]; //all used
@@ -81,7 +81,7 @@ static void raw_scan(void){
 	uint8_t i, j;
 
 	//scan top row
-	keys[0] = 0; //initially
+	Keys[0] = 0; //initially
 	//temporarily set P3[1:0] push pull output M[1:0]=b01
 	P3M0 |= 0x3;
 	P3 &= ~(0x3);
@@ -89,19 +89,19 @@ static void raw_scan(void){
 	//back to quasi-bidirectional w/ pullup M[1:0]=b00
 	P3M0 &= ~(0x3);
 	//read inputs
-	keys[0] += (P3_0 ? 0 : (1<<0)); //on
-	keys[0] += (P3_1 ? 0 : (1<<1));
-	keys[0] += (P5_4 ? 0 : (1<<2));
-	keys[0] += (P5_5 ? 0 : (1<<3));
+	Keys[0] += (P3_0 ? 0 : (1<<0)); //on
+	Keys[0] += (P3_1 ? 0 : (1<<1));
+	Keys[0] += (P5_4 ? 0 : (1<<2));
+	Keys[0] += (P5_5 ? 0 : (1<<3));
 	//scan matrix
 	for (i = 0; i < M_ROWS; i++){
-		keys[i+1] = 0; //initially
+		Keys[i+1] = 0; //initially
 		for (j = 0; j < M_COLS; j++){
 			uint8_t read;
 			P1 = (~(1 << (j+4))); //pulldown column (all others pull up)
 			read = (P1 & (1<<i)); //read button
 			read = (read == 0 ? 1 : 0); //pressed if grounded by column driver
-			keys[i+1] += (read << j);
+			Keys[i+1] += (read << j);
 			P1 = 0xff; //reset all as pullups
 		}
 	}
@@ -123,7 +123,7 @@ static void debounce(void){
 			new_count = curr_count;
 			new_state = curr_state;
 			//update count
-			if (keys[i] & (1 << j)){
+			if (Keys[i] & (1 << j)){
 				if (curr_count < COUNT_LIM_HIGH){
 					new_count = curr_count + 1;
 				}
@@ -198,9 +198,11 @@ void KeyScan(void){
 	debounce();
 }
 
+#ifdef DEBUG_KEYS
 const uint8_t* DebugGetKeys(void){
 	return keys;
 }
+#endif
 
 const uint32_t GetNewKeys(void){
 	return new_keys_pressed;
diff --git a/src/key.h b/src/key.h
index f487852..8523fc7 100644
--- a/src/key.h
+++ b/src/key.h
@@ -7,10 +7,20 @@
 #ifndef SRC_KEY_H_
 #define SRC_KEY_H_
 
+//#define DEBUG_KEYS
+
 void KeyInit(void);
 void KeyScan(void);
-const uint8_t* DebugGetKeys(void);
 const uint32_t GetNewKeys(void);
 
+#ifdef DEBUG_KEYS
+const uint8_t* DebugGetKeys(void);
+#endif
+
+//definition included for determining if multiple keys are pressed
+//prefer using GetNewKeys();
+#define TOTAL_ROWS 5
+extern uint8_t Keys[TOTAL_ROWS]; //only bottom nibbles get set
+
 
 #endif /* SRC_KEY_H_ */
diff --git a/src/lcd.c b/src/lcd.c
index 6c6ebae..ac8b7fb 100644
--- a/src/lcd.c
+++ b/src/lcd.c
@@ -29,9 +29,6 @@
 
 
 static int row, col;
-static int OpenFlag = 0;
-static int ErrorTimeout = 0;
-static int Error = 0;
 
 #define CLEAR_BIT(port, bit) (port &= ~(_BV(bit)))
 #define CLEAR_BITS(port, bits) (port &= ~(bits))
@@ -112,7 +109,6 @@ static char readBusy() {
 	return busy;
 }
 
-//sets ErrorTimeout flag if busy signal not cleared in time
 static void wait_busy() {
 	unsigned int i;
 	for (i = 0; i < 100; i++){
@@ -122,7 +118,6 @@ static void wait_busy() {
 		_delay_ms(1);
 	}
 
-	ErrorTimeout = 1;
 	return;
 }
 
@@ -148,49 +143,39 @@ static void LCD_OutChar(unsigned char c) {
 }
 
 void LCD_Open(void) {
-	static int open_error = 0;
-	if (!OpenFlag) {
-		//set ports to push-pull output M[1:0] = b01
-		//P2 entire port
-		P2M1 = 0;
-		P2M0 = 0xff;
-		//P3 pins 7:4
-		P3M1 &= ~(0xf0);
-		P3M0 |= (0xf0);
+	//set ports to push-pull output M[1:0] = b01
+	//P2 entire port
+	P2M1 = 0;
+	P2M0 = 0xff;
+	//P3 pins 7:4
+	P3M1 &= ~(0xf0);
+	P3M0 |= (0xf0);
 
-		_delay_ms(30); // to allow LCD powerup
-		outCsrBlindNibble(0x03); // (DL=1 8-bit mode)
-		_delay_ms(5); //  blind cycle 5ms wait
-		outCsrBlindNibble(0x03); // (DL=1 8-bit mode)
-		_delay_us(100); // blind cycle 100us wait
-		outCsrBlindNibble(0x03); // (DL=1 8-bit mode)
-		_delay_us(100); //  blind cycle 100us wait (not called for, but do it anyway)
-		outCsrBlindNibble(0x02); // DL=1 switch to 4 bit mode (only high 4 bits are sent)
-		_delay_us(100); // blind cycle 100 us wait
-		//set increment, no shift
-		outCsrBlind(0x4 + LCDINC + LCDNOSHIFT);
+	_delay_ms(30); // to allow LCD powerup
+	outCsrBlindNibble(0x03); // (DL=1 8-bit mode)
+	_delay_ms(5); //  blind cycle 5ms wait
+	outCsrBlindNibble(0x03); // (DL=1 8-bit mode)
+	_delay_us(100); // blind cycle 100us wait
+	outCsrBlindNibble(0x03); // (DL=1 8-bit mode)
+	_delay_us(100); //  blind cycle 100us wait (not called for, but do it anyway)
+	outCsrBlindNibble(0x02); // DL=1 switch to 4 bit mode (only high 4 bits are sent)
+	_delay_us(100); // blind cycle 100 us wait
+	//set increment, no shift
+	outCsrBlind(0x4 + LCDINC + LCDNOSHIFT);
 #ifdef DEBUG
-		//set display on, cursor on, blink on:
-		outCsrBlind(0x0c + LCDCURSOR + LCDBLINK);
+	//set display on, cursor on, blink on:
+	outCsrBlind(0x0c + LCDCURSOR + LCDBLINK);
 #else
-		//set display on, cursor and blink off:
-		outCsrBlind(0x0c + LCDNOCURSOR + LCDNOBLINK);
+	//set display on, cursor and blink off:
+	outCsrBlind(0x0c + LCDNOCURSOR + LCDNOBLINK);
 #endif
-		//set display shift on and to the right
-		outCsrBlind(0x10 + LCDNOSCROLL + LCDRIGHT);
-		//set 4-bit mode, 2 line, 5x7 display:
-		outCsrBlind(0x20 + LCD2LINE + LCD7DOT);
-		//clear display
-		LCD_Clear();
-		_delay_ms(50);
-		if (!readBusy()) {
-			OpenFlag = 1;
-		} else {
-			open_error = 1;
-		}
-	} else { //display already open
-		open_error = 1;
-	}
+	//set display shift on and to the right
+	outCsrBlind(0x10 + LCDNOSCROLL + LCDRIGHT);
+	//set 4-bit mode, 2 line, 5x7 display:
+	outCsrBlind(0x20 + LCD2LINE + LCD7DOT);
+	//clear display
+//	LCD_Clear();
+//	_delay_ms(50);
 }
 
 //row and columns indexed from 0
@@ -199,8 +184,6 @@ void LCD_GoTo(unsigned int row_to, unsigned int col_to) {
 		outCsr(0x80 + 0x40 * row_to + col_to); //set ddram address to position
 		row = row_to;
 		col = col_to;
-	} else {
-		Error = 1;
 	}
 }
 
@@ -267,19 +250,8 @@ void LCD_OutNibble(uint8_t x){
 }
 
 void LCD_Clear() {
-	if (OpenFlag) {
-		outCsr(0x01);
-	} else {
-		outCsrBlind(0x01);
-	}
+	outCsr(0x01);
 	row = 0;
 	col = 0;
 }
 
-unsigned char LCD_Timeout_Error(void) {
-	if (ErrorTimeout != 0) {
-		return 1;
-	}
-	return 0;
-}
-
diff --git a/src/main.c b/src/main.c
index 490b8bf..264066c 100644
--- a/src/main.c
+++ b/src/main.c
@@ -25,14 +25,17 @@ static const char KEY_MAP[20] = {
 uint32_t NewKeyBuf[4];
 volatile uint8_t new_key_write_i;
 volatile uint8_t new_key_read_i;
-volatile uint8_t new_key_empty;
+volatile uint8_t NewKeyEmpty;
 #define INCR_NEW_KEY_I(i) i = (i + 1) & 3
 
 volatile uint8_t SecCount;
 void timer0_isr() __interrupt 1 __using 1
+//#define TRACK_TIME
 {
+#ifdef TRACK_TIME
 	static uint8_t count = 0;
 	static uint8_t min_count = 0, hour_count = 0;
+#endif
 
 	uint32_t new_keys;
 
@@ -40,16 +43,17 @@ void timer0_isr() __interrupt 1 __using 1
 	KeyScan();
 	new_keys = GetNewKeys();
 	if (new_keys != 0){
-		if (!new_key_empty && (new_key_write_i == new_key_read_i)){
+		if (!NewKeyEmpty && (new_key_write_i == new_key_read_i)){
 			//do not overwrite keymap currently being processed
 			INCR_NEW_KEY_I(new_key_write_i);
 		}
 		NewKeyBuf[new_key_write_i] = new_keys;
 		INCR_NEW_KEY_I(new_key_write_i);
-		new_key_empty = 0;
+		NewKeyEmpty = 0;
 	}
 
 	//track time
+#ifdef TRACK_TIME
 	count++;
 	if (count == 200){
 		count = 0;
@@ -63,6 +67,7 @@ void timer0_isr() __interrupt 1 __using 1
 			}
 		}
 	}
+#endif
 }
 
 
@@ -104,8 +109,15 @@ int main()
 	uint8_t entering_exp = ENTERING_DONE;
 	uint8_t no_lift = 0;
 	uint8_t exp_i = 0;
+#ifdef DEBUG_KEYS
+	uint8_t j = 0;
 	const uint8_t* keys;
 	uint8_t key_i;
+#endif
+
+#ifdef DEBUG_UPTIME
+	uint32_t i;
+#endif
 	Timer0Init(); // display refresh & switch read
 	LCD_Open();
 	KeyInit();
@@ -116,12 +128,17 @@ int main()
 	P3M1 &= ~(0x4);
 	P3M0 |= (0x4);
 
+#ifdef DEBUG_UPTIME
 	i = 0;
-	j = 0;
+#endif
 	// LOOP
 	while (1)
 	{
 		LCD_GoTo(0,0);
+#ifdef DEBUG_UPTIME
+		u32str(i++, Buf, 10);
+		LCD_OutString(Buf, MAX_CHARS_PER_LINE);
+#else
 		//display y register on first line
 		if (entering_exp == ENTERING_DONE){
 			dec80_to_str(Buf, get_y());
@@ -130,6 +147,9 @@ int main()
 			dec80_to_str(Buf, get_x());
 		}
 		LCD_OutString(Buf, MAX_CHARS_PER_LINE);
+#endif //DEBUG_UPTIME
+
+#ifdef DEBUG_KEYS
 		//keyboard debug
 		keys = DebugGetKeys();
 		for (key_i = 0; key_i < 5; key_i++){
@@ -150,14 +170,16 @@ int main()
 		} else {
 			LCD_OutString(u32str(SecCount, Buf, 10));
 		}
+#endif //DEBUG_KEYS
 
-		///new keys
-		if (!new_key_empty){
+
+		///get new keys
+		if (!NewKeyEmpty){
 			uint8_t i_key;
 			uint32_t new_keys = NewKeyBuf[new_key_read_i];
 			INCR_NEW_KEY_I(new_key_read_i);
 			if (new_key_read_i == new_key_write_i){
-				new_key_empty = 1;
+				NewKeyEmpty = 1;
 			}
 
 #ifdef DEBUG_KEYS
diff --git a/src/utils.c b/src/utils.c
index 948a1dc..aa38cc5 100644
--- a/src/utils.c
+++ b/src/utils.c
@@ -5,11 +5,12 @@
 void _delay_ms(uint8_t ms){
 	//TODO:
 }
+#ifdef ACCURATE_DELAY_US
 void _delay_us(uint8_t us)
 {
 	//TODO:
 }
-#else
+#endif
 void _delay_ms(uint8_t ms)
 {
 	// delay function, tuned for 11.583 MHz clock
@@ -28,6 +29,7 @@ void _delay_ms(uint8_t ms)
 	__endasm;
 }
 
+#ifdef ACCURATE_DELAY_US
 void _delay_us(uint8_t us)
 {
 	// delay function, tuned for 11.583 MHz clock
diff --git a/src/utils.h b/src/utils.h
index c2c8808..0c4d1fc 100644
--- a/src/utils.h
+++ b/src/utils.h
@@ -11,7 +11,13 @@
 
 void _delay_ms(uint8_t ms);
 
+#define ACCURATE_DELAY_US
+#ifdef ACCURATE_DELAY_US
 void _delay_us(uint8_t us);
+#else
+#define _delay_us(x) _delay_ms(1)
+#endif
+
 
 char* u32str(uint32_t x, char* buf, uint8_t base);