Swiching level voltage values for TTL and CMOS.

Thread Starter

k.vinay.nayak

Joined Dec 20, 2007
7
Is VIH: [Voltage Input High],VIL: [Voltage Input Low], VOL: [Voltage Output Low], VOH: [Voltage Output High] ,VT: [Threshold Voltage] values are same for 5V TTL and 3.3V TTL or not?

Also these [Voltage Input High],VIL: [Voltage Input Low], VOL: [Voltage Output Low], VOH: [Voltage Output High] ,VT: [Threshold Voltage] values are different for CMOS Gates compared to TTL Gates?

Please help me to get clarified.

Thanks
 

bloguetronica

Joined Apr 27, 2007
1,541
I don't know about the 3.3V TTL, but logic levels are different when comparing TTL with CMOS. Normally a TTL gate needs a lower voltage than a CMOS gate to assume logic high in that input. The output voltage of a TTL gate driven high might be lower than the output voltage of a CMOS gate in the same situation. It even might be lower than the required voltage to drive a CMOS input high. That is why a TTL gate wont reliably drive a CMOS gate without a pull-up resistor.

Here is detailed information about the logic levels of the different families:
http://www.interfacebus.com/voltage_threshold.html

Notice that the 74HCT family is not a TTL family. It is a CMOS family with TTL compatible inputs. It is the only CMOS family that can be reliably driven by TTL chips.
 

Papabravo

Joined Feb 24, 2006
21,158
There is no such thing as 3.3 Volt TTL. TTL requires that Vcc be in the range 4.5 V to 5.5V

3.3Volt operation is a CMOS thing
 

Papabravo

Joined Feb 24, 2006
21,158
Does it have any advantages? Is there a full complement of parts in the family?
Apparently not. It consists of interface buffers, some octal flip flops and latches, no gates or decoders or multiplexers at least none in the TI lineup. Hardly worth a second thought.
 

Distort10n

Joined Dec 25, 2006
429
...and clock buffers/generators, transcievers, some gates. Many of the newer logic families do not have a full compliment of devices; e.g., AUC logic.
 

Papabravo

Joined Feb 24, 2006
21,158
So why would you use these parts given the other choices? Just because they are there seems somehow less than sufficient.
 

Ron H

Joined Apr 14, 2005
7,063
You guys apparently know of companies that actually make LVTTL products. Who are they? My searches found lots of devices with LVTTL level compatibility, and LVTTL to (insert logic family) translators, etc., but I didn't find any that were LVTTL in and LVTTL out, made with a bipolar process.
 

Papabravo

Joined Feb 24, 2006
21,158
That might be correct -- at least on the TI website. Apparently making input circuits tolerate higher voltages is getting to be pretty common among logic families. If there are no strictly LVTTL input and output parts I'm thinking it means that they are only used for interfacacing to ASICS and other low voltage circuits. The basis of my original skepticism on the viability of such parts is that CMOS does it better and more flexibly, so why bother with a bipolar tecnology.
 

Distort10n

Joined Dec 25, 2006
429
I would have to go through the selection guide again on TI's website, but TI is a large semiconductor manufacturer of a large portfolio of logic. National, ONSemi, ST Microelectronics would be others. Pericom will also have options but mostly in the clocking solutions.

Papabravo is correct that purpose of LVTTL is to exist in a design where the rail voltage is 3.3V, yet the TTL thresholds remain the same while still able to handle 5V on the inputs. This is because there is no ESD protection diode to the positive rail.

If you look at any TI logic datasheet you will usually see a negative clamp current specific in the ABS MAX table, but not positive clamp current. This means there is no ESD diode to the positive rail, so higher voltages can be used without worry of damage.

As far as I know LVTTL is still based on CMOS technology (LV family), but the thresholds are TTL on a 3.3V rail.

As to bipolar, newer logic technologies like ABT or CBT are BiCMOS. Bipolar gives you the speed man!
 

Papabravo

Joined Feb 24, 2006
21,158
The speed abvantage of bipolar is hardly worth writing home about. Given the vastly superior power consumption of CMOS it seems hardly worth the effort to use anything else. Real speed daemons use PECL any way...lol
 

bloguetronica

Joined Apr 27, 2007
1,541
There is no such thing as 3.3 Volt TTL. TTL requires that Vcc be in the range 4.5 V to 5.5V

3.3Volt operation is a CMOS thing
Yes there is. Just see the graph inside the page I've indicated.

Why bother with old bipolar TTL technology? Because electronics schools use them.
Some TTL chips are very handy. Also, the CMOS families are incomplete. An example is the 7447/74LS47 chip that doesn't have a CMOS equivalent. A 7-segment CMOS driver with leading/trailing zero blanking would be handy. The lamp test of the 4511 serves no purpose to me.
 

hgmjr

Joined Jan 28, 2005
9,027
Greetings cumesoftware,

I took a look at the link you provided in an earlier post and the way I interpret the graph is that TTL requires +5V power to operate but that its is able to accept 3.3 volt logic signals at its input. The output will still swing between 0V an 5V.

What I believe papabravo is saying is that there is no standard TTL that is powered by 3.3V. To implement 3.3V input and output signal levels involves the use of LVTTL.

At least that is the way I interpret the graphs provided.

hgmjr
 
Top