Need some help with TTL vs CMOS pls.

[Chriss]

Hi!
I have a bit problem with an electronic device what I repair.
It is a BDM interface which have to communicate with the
MPC55xx cpu from Freescale.

Problem with the interface:
- The communication is not stable.
- Hard to enter in the BDM mode.
- Sometime possible to establishing the communication.

I don't have the schematic of the board but I will describe
the part what I'm understand but I will check am I right.

The main interface is based on:
- voltage source 5v.
- voltage stabilizer for 3.3v
- voltage switching mechanism between 3.3v and 5v. ( changeable by software which is Windows based )
- 74HCT245 8-bit transceiver
- 74HCT14 Hex inverting Schmitt trigger

Resolution:
The VCC to the two ic's are powered from the voltage switching mechanism.
It depends what voltage is selected by the software 3.3v or 5v.

By the datasheet of that two ic's a recommended typical VCC voltage should be 5v.
A minimum VCC voltage should not go under 4.5V.

I assume, if the voltage selector is set to delivery 5v programming voltage
than everything should be fine.
But if the voltage selector is set to delivery 3.3v what is used for programming
the MPC55xx cpu and entering into BDM mode then the 74HCT2458 and 74HCT14
are in trouble and can cos a communication problem or any other strange behavior.

If this ic would be a CMOS ic's than it could work but only in 3.3v mode.

Conclusion:
I think, the mentioned problem is because a wrong vcc voltage to that two ic's
wich are actually a TTL ic's and use a higher VCC voltage than 3.3v.
But I can't use 5v because the MPC55xx cpu.
The main problem is the voltage selector, which is wrong wired up to
the main electronic.
The mentioned ic's shouldn't get the VCC from the voltage selector
but from a stable 5v source.

My question:
What wpuld be the behavior of a TTL and CMOS
ic like I described, when I would it power up with a
voltage which is almost half the required voltage?

Actually I know that would lead in a bad working condition
but how bad?

Could it be does an electronic device with this type of voltage setup
should work as excepted?

Thanks in advance for any help and direction.

My best regards.

[Zero999]

My question:
What wpuld be the behavior of a TTL and CMOS
ic like I described, when I would it power up with a
voltage which is almost half the required voltage?

Actually I know that would lead in a bad working condition
but how bad?

Could it be does an electronic device with this type of voltage setup
should work as excepted?

Thanks in advance for any help and direction.

My best regards.

The problem would be intermittency. It may work sometimes but not others. It may work, then become unstable and latch up.

[Ian.M]

74HCT series chips are *NOT* TTL. They are CMOS, but with TTL compatible input thresholds. In addition to interfacing to TTL, they can be used for upwards level translation from 3.3V logic levels to 5V levels, when powered at 5V, as their logic '1' input threshold is only 2V.

However they typically aren't specified to operate below 4.5V so their behaviour at 3.3V will be unpredictable.  Its possible that a particular brand of 74HCT may have been known to work at 3.3V even though not specified to do so, but substituting a different brand, or even a later date code could fail to work.    Substituting 74HC parts may not work either as if the original chips are being used for level translation, 3.3V logic wont reliably reach the higher logic '1' input threshold of a 74HC part powered from 5V.

[Chriss]

The problem would be intermittency. It may work sometimes but not others. It may work, then become unstable and latch up.

This is actually the problem with this interface, as you described.

However they typically aren't specified to operate below 4.5V so their behaviour at 3.3V will be unpredictable.

I assume, if I would disconnect the voltage selector from this vcc line and feed the ic's from a
stable 5v source that wouldn't  harm the electronic, or?
I assume that is a designing mistake (it could be because this is a low cost china interface) of the pcb
manufacturer. The voltage selector voltage goes also from the interface to the BDM port and
powers the CPU with the selected voltage outside the interface.
In that meaning the internal electronic on the interface board has nothing
to do with the vcc voltage of the targeted cpu and the 74HCT ic's vcc.

It's not normal for me to use on the same ic let's talk about the 74HCT two type of vcc.
One 3.3v and than the higher 5v.
I don't think there is any equivalent ic for this two above who can operate in a good condition
if it is feed with 3.3v and with 5v.
Or am I wrong?

One more question pls.
What would the be best way to determine what type of ic should be used on
an electronic device where I know the type of the ic but don't know
should be used a variant of TTL or CMOS ic?

I also assume on this board is something mixed up with TTL and CMOS ic's.
If I would design such of board I would make a decision to use a TTL or CMOS
type of ic's, of course it depend of the behavior of other I/O ports.
But if the devices ( ic's) should communicate between each other on the same pcb
that would be a big mistake to mix them up with TTL and CMOS type and
always correcting the voltage to meet the requirement off the mixed type of ic's.

Thanks.

[schmitt trigger]

I remember the good old, old days, where there were only three types of 74xx logic: regular 74, high speed, 74S and low power 74L.
Later there were Schottky 74S and low power Schottky 74LS.

Now there is a whole alphabet soup.

I believe that you should get a something with 74LV in it, for low voltage. There are several families produced by different vendors, I would encourage you to download and carefully read their datasheets, to determine whether its performance parameters meet your requirements.