manual mode shifting an electronic automatic transmission

[alec_t]

Ah, that's because LTS doesn't ship with the component models used here and that other forum members use. Yahoo has an LTS User Group which is worth joining to get free access to third party models. You need to register, then you can download the CD4000.lib and 74HC.lib libraries which contain the missing models (among many others). There are other threads on this (ETO) site which deal with how/where to install the library files, but if you get stuck just ask. Happy simming!

 

[Millennium]

thanks. Will start researching and playing.

now about programming the code. what products do i need to be looking into? Is there code in your circuit or .....?


on a side note, some of the car/engines I tune utilize rom chips using 27sf512 and flash burn software with a Burn 2 burner. Are those products in any way remotely connected to what i want to do?

 

[Millennium]

nvm found files

 

[alec_t]

Is there code in your circuit

Not in the circuit I posted. Hardware logic provides the A and B relay drive signals.

By way of explanation of the circuit, gates U2a and U2b together act as a monostable circuit with a period of about 40mS for 'debouncing' the gear-switching signal from the 'up' switch contact or the 'down' switch contact. The switch can be a single-lever single pole momentary type with a centre-off position (i.e. mom-off-mom).
After the 40mS delay the BCD counter U1 is clocked by the rising edge of the pulse from the monostable. Whether the count is up or down is determined by the state of a set-reset latch formed by gates U5b, U5c.
The count value at outputs Q1, Q2 is monitored by gates U3a, U5d. Gate U3a detects a count of '0' and sets the latch, which enables up-counting by U1, enables (via gate U5a) the 'up' switch to clock the counter, and blocks (via gate U2d) the 'down' switch action. Gate U5d detects a count of '3' and resets the latch, which prevents further up-counting, switches the counter mode to down-counting, enables the 'down' switch action and blocks the 'up' switch action.
The counter outputs control the multiplexer U4 such that a count of '0' (= 1st gear) connects I0a to Ya and I0b to Yb, thus putting +5V on Ya and on Yb, signifying both relays 'on'. Similarly, count values '1' etc put other appropriate outputs (+5V or 0V) on Ya and Yb as per your table.
C2, R4 provide a power-on-reset function to ensure the counter starts at zero (1st gear). That may or may not suit your needs.

If you decide at a later stage that you wish the counter to count from 0 to 4 (or more) instead of from 0 to 3, the count monitoring logic can be changed and another multiplexer added.

 

[Millennium]

Okay looking at the run program and what i'd call a datalog.....


It looks like you press one of the buttons once it will go through a sequence of changing states of a and b outputs. Am I using/interpreting this right?

If so i need the circuit to increment or decrement each time you press the switch.

and the circuit after a power loss will always start in gear1 so a and b on.

is there a reset that can be pressed to make it act like a power off then power on and it will be a and b outputs on?



so operation of circuit:
power on a and b on
upshift button pressed a off b on
stays in this state until reset, upshift, downshift button pressed
upshift button pressed a off b off
stays in this state until reset, upshift, downshift button pressed
down shift button pressed a off b on
stays in this state until reset, upshift, downshift button pressed
upshift button pressed A off b off
tays in this state until reset, upshift, downshift button pressed
upshift button pressed A on B off
tays in this state until reset, upshift, downshift button pressed
downshift button pressed A off B off
tays in this state until reset, upshift, downshift button pressed
downshift button pressed A off B on
tays in this state until reset, upshift, downshift button pressed
downshift button pressed A on B on

 

[alec_t]

cant seem to find dview.lib

It should have been in the CD4000 zip file. It needs to go in the sub folder, and there's a DVIEW5.asy file which needs to go in the sym/cd4000 folder.
Congratulations on progress so far with LTS. Whenever you want to use a model from these non-native libraries you place a directive .inc CD4000.lib or .inc dview.lib on the schematic.
That dview gizmo is handy for displaying multiple waveforms vertically spaced on one screen (but the Y-axis scale has to be ignored).

 

[alec_t]

Posts 24 and 25 obviously collided.
Re the simulation, it is not intended to represent real time switch actions (but could be modified to do so) and is just for testing the logic of the switching/counting.
It starts with power up, counter value 0.
V2 simulates a sequence of 5 'up' switch presses (each lasting a non-real 1mS) at (non-real) 150mS intervals starting after a 100mS delay.
The 0 count is detected and sets the counter to up-counting. So at 140mS, 290mS, 440mS (bear in mind the 40mS monostable delay) the counter is clocked up. But when count 3 is detected further up-counting is prevented and the counter is switched to down-counting ('up' presses 4 and 5 are ignored).
Likewise V3 simulates a subsequent sequence of 5 'down' switch presses (each lasting 1mS) at 150mS intervals starting after a 900mS delay. When count 0 is reached 'down' presses 4 and 5 are ignored.

 

[Millennium]

ah!...so i was playing with Ncycles on V2 V3. Not getting what i would expect but i'll figure it out. I assume you entered values to see the circuit work overall and i'm expecting it to work as i would want to actually operate it.


pressing on the v1 does that turn the power supply on and off each press?

 

[Millennium]

Alec,
I've built the circuit you posted and having some issues. Some questions

Does it matter what gates i choose inside the cd4093's to use as long as they go to the right side of the gate? Am i asking the right questions? I assume that U5A, U5B, U5C, U5D are all in the same 4093 chip and that U2A, U2B, U2C, U2D are in another.

I followed the diagram very closely as each gate (4093s') appeared to have a "polarity", that the symbol was important the direction of each connection.

I've connected LED's to the Ya and Yb outputs and touch and release ground the upshift switch it turns both LED's on, I touch and release again and it turns both off. I repeat the process again and i get the same result rather than the table listed earlier in the thread. I'm sure its something i've done wrong. What do you need from me to try to get this working.


Thanks for your time.

 

[KeepItSimpleStupid]

You can mix and match the 4093's as your layout dictates. Don't forget to tie all unused inputs of CMOS chips to some logic level. Real world also suggests bypass caps 0.1 uf ceramic or so close to the IC power pins. Don't forget power and ground.

 

[alec_t]

Wow! 3.7 years since your previous post. Lucky I'm still around to reply .

each gate (4093s') appeared to have a "polarity", that the symbol was important the direction of each connection.

Correct. For example, pins 1 and 2 are both inputs and pin 3 is the output for a specific one of the gates, but you can refer to that particular gate as U2c, or U2-whatever-you-choose. To minimise wire lengths (and potential problems from interference pickup etc) it is usual to use the gates within one IC for closely associated circuit functions. Btw, I've just noticed that apart from the 0.1uF caps very close to the power supply pins (7 and 14 in the case of the 4093) that KISS mentioned, there should be a resistor (at least 10k) connected between the top of R1 and the input pins of U2b.
The unused gate in U3 (CD4001) should have its inputs (not output) tied to ground (as per KISS' post).

I've connected LED's to the Ya and Yb outputs

What value current-limiting resistors did you use?

 

[Millennium]

i'll review at the shop tomarrow. Basically i searched "resistor values to use with led's" and it gave me a calc formula i entered supply, amperage rating of led and something else.

what i meant about the polarity is one input has a dash pointing at the input and the other input its pointing away.

 

[KeepItSimpleStupid]

LEDs: A typical LED might have a Max Vf of 2.4V and a suggested operating current of 20 mA.
You have a supply voltage to the LED and we might call it Vdd.
Transistor switches drop about 0.6-0.7V or lower and FET switches much less.

So, the super simple calculator is R <= (Vdd-Vf(led)-V(CEsat))/20 mA ; 20 mA is 20E-3 Amps.

==

You may find circles at inputs or outputs or a line over top of a letter of even ~A. It means the outputs or inputs need to be thought of as inverted. When you follow all of the rules, you can build any logic circuit for all NAND gates or all NOR Gates.

So a NAND gate is a NOT AND an therefore the circle appears at the output of the AND gate symbol, turning it into a NAND gate.

If you take the two inputs of a 2-input NAND gate and tie them together, you have a NOT gate or inverter.

If you find a C with a bar over it, it may mean that the input is active low. The ~C is Not Clear or a low clears a counter or whatever.

 

[alec_t]

what i meant about the polarity is one input has a dash pointing at the input and the other input its pointing away.

Those dashes don't relate to the inputs. They are the top and bottom of a 'square loop' symbol denoting a Schmitt trigger (a circuit which has 'hysteresis').

 

[Tony Stewart]

To protect the engine and transmission, safety logic should include RPM feedback and a table of gear ratios to prevent unwanted downshifts.
Consumer vehicles ignore manual downshift if RPM is too high and automatically downshift if RPM is too low, but never upshift automatically in manual mode.
Naturally the design prevents skipping a gear with only 1 solenoid changing at a time sequentially, but safety logic can ignore false commands to skip gear.
Then Auto transmission often has two modes for comfort and speed to shift at different torque threshold settings for smooth or optimum RPM for max power transfer. Then there must be a delay to ignore inputs until shift has occurred and only respond to momentary edge ON and not steady ON for Up/Down.

If you wish any of these protection or features, its not that hard , but more input needed.

Gear....................1.....2.....3......4
Shift solenoid A......1.....0.....0......1......where On=1, Off=0 momentary thrust ON then hold with reduced power or?
Shift solenoid B......1.....1.....0......0
Up........................!A...!B....!A.. .x ......... where ! means toggle current Solenoid indicated only and x means do nothing.
Down....................x.....!A...!B....!A ......... toggle in logic means Exclusive OR (XOR)

This method requires only a dual FF and an XOR IC with gates to decode present gear and input command.

Then initial power on conditions and neutral clutch needs to be defined as well as engine normal/ emergency off. Solenoid must have some time limit, I expect and then delay before accepting next command even if just in xx milliseconds.

From these type of specs but expanded in detail (e.g. solenoid current) for the designer , after agreed perfect, a simple design can be made.

 

[Millennium]

No protection is needed. No comfortable shift is desired. I have all those issues handled. This is not a stock vehicle.

Currently transmission is controlled by same ecu that controls the engine. I have all bases covered from those commenting on possibilities.

This is a modified valve body trans that has no accumulators and greatly enlarged feed holes to clutch circuits. The slowest shift i've seen(start of engine deceleration to shift completion and engine accell) in datalogs is .21 seconds. Commanded shift to start of engine deceleration is .1 seconds.

Once i get the circuit to work like the table, i'll notice any operation that will concern me and i'll post questions.


Again thanks for the help.

 

[Millennium]

resistors used for led's are 180 ohms.

 

[Millennium]

bypass caps 0.1 uf ceramic or so close to the IC power pins.

unfortunately i dont understand what that means. obviously one leg of cap goes to power/ground pins, where does other leg connect? 5v supply ? ground?

 

[KeepItSimpleStupid]

For the DC4001, Vdd is pin 14. Vss is +Power
and pin 7 is labeled Vss. Vss is ground.

I wanted to make sure that:
a) You connected power to the IC's used.
b) You connected bypass caps. Ceramic capacitors have no polarity, so they would be connected between pin 7 and pin 14 near the IC.
c) You connected unused inputs of the gates to either ground or power.

CMOS devices can appear to work without power connected, unfortunately.

 

[Millennium]

HOLY **** IT WORKS!!!!!!!! i had q1 and q2 going to the same pin on one of the 4093's


You guys have NO idea what this is going to do for me and enjoying driving the car.


Some impresions that i LIKE:

it wont restart counting past 4th (back to 1st)if i was to press the upshift again.
it wont shift to 4th if i press downshift if its in 1st.
It wont count through if i was to hold either switch.

OMG. Im so excited.


Now i need to figure out how to get the outputs to drive the coils for 12 volt automotive relays. So i need to find an amplifier circuit to amplify 5 volts low current to 12 volts ~~2 amps should be more than enough.