manual mode shifting an electronic automatic transmission

[Tony Stewart]

Don't guess, find out the Solenoid resistance and if active low side or high
Is solenoid always active?

 

[Millennium]

The computer i use to control all the relays right now have a max output current of 1.5 amps per instructions.

Duty cycle of the solenoid in the relay will depend on the gear i'm in. So 4th gear atleast one relay will be 100% for the majority of a drive, and my current trans/engine control unit has been fine with it for several years.

 

[Tony Stewart]

Can you answer Rs and hi/lo side?

 

[Millennium]

Well letting it sit turned on for about 10 mins it stopped responding to inputs. I unplugged power supply and back in no fix. Flipped polarity of c100 and it started working again.

is that why i need those small caps between supply and ground pins?

 

[Millennium]

Can you answer Rs and hi/lo side?

not right now. All testing equipment at work.

I understand resistance but not the high low side you speak of


unless you mean what i'm going to use to trigger the relay wheather it be +volts or ground signal. I have the ability to do either.

 

[Tony Stewart]

not right now. All testing equipment at work.

I understand resistance but not the high low side you speak of


unless you mean what i'm going to use to trigger the relay wheather it be +volts or ground signal. I have the ability to do either.

yes Hi side switch means +12,, Low side switch means 0V. Solenoid current is never constant and max current is due to Rs of coil but steady reading would be useful.

 

[Millennium]

82 ohms for the relay

In all reality we could make it a whole solid state circuit, instead of using the relays to send 12 volts to the solenoids in the trans. Lemme go measure those coils.

25 ohms for the shift solenoids.

If you guys recommend all solid state stuff with transistors and such make current capability at least 5 amps. 12 supply to that circuit is from battery and current is abundant. I would need to be sending 12 volts to the solenoids in the trans for activation, they are grounded to the valve body inside the trans

 

[Millennium]

Having some problems with the circuit.

It will count correctly if you wanted to complete the process, gear 1 2 3 4 and then back down if you wanted gear 4 3 2 1. It will not react or count correctly say if i wanted to select gear 1 then 2, then back down to 1 (as an example)

Would incorrect C1 values cause this? After 2 trips to radio shack I dont know why i keep choosing something other than 100nf caps.

Thanks

 

[alec_t]

It will count correctly if you wanted to complete the process, gear 1 2 3 4 and then back down if you wanted gear 4 3 2 1. It will not react or count correctly say if i wanted to select gear 1 then 2, then back down to 1 (as an example)

If I recall correctly the original circuit design (4 years ago!) didn't allow for random shifts like that so would need modification. I'll look into it.

 

[Tony Stewart]

Design Logic
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)
LED 1 ..................1........0........0........0
LED 2...................0........1........0........0
LED 3...................0........0........1........0
LED 4...................0........0........0........1

This can drive individual LEDs with numeric lens cover.
Very simple design to complete, if needed.

 

[Millennium]

If I recall correctly the original circuit design (4 years ago!) didn't allow for random shifts like that so would need modification. I'll look into it.

THANKS!!!!!

 

[alec_t]

Is one of your 4 gear positions Neutral, or are there 4 gears plus Neutral to consider? And what about reverse? KISS mentioned this in post #4.

 

[Millennium]

neutral, park, or reverse need no consideration; these are controlled by the factory shifter location and its resulting piston location in the valve body sending oil pressure to those components. These positions are not controlled by any solenoid, only mechanically by the above mentioned method.

 

[alec_t]

Ok. This revised circuit should enable random ups and downs through the gears.

R1-R4,R7,R8,C1,C2,C4 take care of debouncing the Up button and Down button switch contacts. C6,R11,R12 provide a power-on reset of counter U2. When the count reaches 3 (= gear 4) further up-counting is blocked by gates U1b,U1d. When the count reaches 0 (= gear 1), further down-counting is blocked by gates U1a,U3a,U3b. U4 decodes the count to give the required logic output signals A,B as per the table.

For automotive use the circuit will require beefing up. Components used should be automotive grade. Each IC should have a 100nF ceramic cap connected directly across its supply pins, and the 5V supply itself should have adequate decoupling and transient spike suppression. I give no guarantee that the circuit will behave as expected when in situ, so its use is entirely at your risk. Its operation has been tested by simulation only. You should ensure that its use is acceptable to your insurer and in compliance with any applicable vehicle construction regulations.

 

[Millennium]

I like your style. I accept all responsibilities.

5v will be supplied by my ECU(which supplies voltage to sensors on the car that need 5vdc). I'll test current draw of circuit before implementation. Now I just need to get Ya and Yb beefed up to handle sending 12vdc to a 25ohm solenoid. I'll test and report back.

THANK YOU!!!

 

[Millennium]

All right, got some updates.

Added a couple transistors to make the YA and YB outputs power the solenoids directly.
Fabbed up the shifter mechanism with detents for positive feel.


So i'm almost there, have one last request. Currently the circuit reacts and changes gear on the "rising edge" of the switch, or as you are releasing the shifter handle. I would like the circuit reaction to be at the very moment the switch is closed or pressed. Is this possible? I've been thinking about playing with normally closed switches, 555 timed normally open switches.....(or would that even work)


Thanks again.

 

[alec_t]

Currently the circuit reacts and changes gear on the "rising edge" of the switch

That's to allow ~150mS for simple debouncing of the switch contacts.

I would like the circuit reaction to be at the very moment the switch is closed or pressed. Is this possible?

Not at the very moment. The Up/Down logic level has to be established shortly before the counter (U2) is clocked. How much delay can you tolerate?

 

[alec_t]

Post deleted.

 

[Millennium]

The delay in the de-bounce isn't noticeable right now. If you could advise tuning theory of specific components related to de-bounce i'll play with that later.


Is it possible to have a timed output from a momentary switch? Meaning if you press a momentary to ground, it only outputs a ground pulse for lets say 100ms; no matter when you release the switch.

Is that getting into 555 timer circuits?

 

[Millennium]

alright i think i came up with something.


two resistors pulled to 5 volts on each leg of a 1 uf cap. switch output connects to one leg of cap, other leg of cap goes to shift circuit. Switch grounded results in about .1ms of full ground and then takes about 30ms to charge back up to 5 volts. Circuit reacts at exact moment of momentary switch connecting to ground.