EMRFD Message Archive 12289

Message Date From Subject
12289 2016-02-07 20:28:33 Ashhar Farhan 74L04 as an amplifier
for a long time, i have wanted to evaluate 74L04 as an amplifier. finally, this morning i did it. here are the results.

the setup is as follows : A 74L04 was strapped to a copper clab board all inputs but one were grounded. A 78L05 with a 0.1 fed it 5 volts. 
upon soldering a resistor of any value from 2.2K to 220 ohms, the input always showed around 1v. this is an emitter input, hence, a 2.2K resistor was added between the input and the ground. 
the circuit around the gate is like this : a 470 ohms feedback resistor between the gate's input and output. a 2.2K from the input to the ground. the input is fed through a 0.1uf resistor.The output is taken through a 0.1uf with a 47 ohms resistor in series.

the gate showed 16db gain across the HF spectrum and dipped by 3 db at 60 MHz. beyond 60 MHz, the gain drops rapidly, reaching unity at 100 MHz. 

the input return loss was 11 db. fair, but not good enough. nothing seem to move this. i tried different feedbacks resistors and different biasing to the ground in place of the 2.2K.

the output return loss was very good. over 22db. probably the output has really low impedance and the 47 ohms resistor is pushing it up to 50 ohms or thereabouts.

the IIP3 is pretty measly, at -21dbm.

it still has it's uses. for instance, it can form a long IF chain or an IF limiter. using it in the front end will make it as weak as an NE612. 

- f
12291 2016-02-07 20:56:28 DuWayne Schmidlko... Re: 74L04 as an amplifier
Never tried this with a 74L device, but a long time ago I had a project
that needed a little extra gain in the amplifier chain for one of the
sensors. Because of the size of the board I had to work with, I
couldn't add another op=amp. I had a couple unused gates in a cmos hex
inverter. I added some bias and feedback resistors to get one to work
as an amplifier. This took care of the gain problem, but the added
power consumption was higher than if I could have added a low power op amp.
DuWayne
KV4QB

12292 2016-02-08 00:35:56 Jim Davey Re: 74L04 as an amplifier
Hi Farhan:

Back in 1975, G4ADJ had a similar idea with the 7400 series logic available at the time.  He built a sideband transceiver around the idea and published it in Nov 1975 Ham Radio Magazine.  Ironically, he measured 15 dB gain at 8 MHz with this logic family.  While you could certainly do better in 1975, the novelty of the approach stuck with me for over 40 years and I recently went looking for it.  

I can post the article if there is wide interest, or send you a copy offline if you want.

Jim  K8RZ


12293 2016-02-08 00:40:51 Ashhar Farhan Re: 74L04 as an amplifier
vy intersting, jim and duwayne,
one could do a full transceiver, easily. i might even try it for laughs. the only trouble is that others might take it seriously!

- f

12294 2016-02-08 06:39:18 timshoppa Re: 74L04 as an amplifier
The CD4007 has far lower bandwidth but is much easier to characterize for amplifier/oscillator use.

Obligatory reference: What's All This CD4007 Stuff, Anyhow?

 

Tim N3QE

 
12295 2016-02-08 08:30:34 DuWayne Schmidlko... Re: 74L04 as an amplifier
There is an old app note from National, and updated by Fairchild about using cmos in the linear mode.
http://www.datasheetarchive.com/AN88-datasheet.html
I think this is the one i used as a reference back in 1980 when I was working on that instrument project.
The  mention of the  CD4007 is interesting, because you can use the individual FETs in the package to build an op-amp,
or just use one as a FET.
http://cmosedu.com/jbaker/courses/ee420L/s14/students/ocampor5/Project/Project.html
here is one project using it as a Audio oscillator
http://hobimikro.blogspot.com/2012/10/open-7400-logic-competition-entry-using.html

I think it wouldn't be to hard to make a simple crystal controlled CW transceiver just using  cmos logic ICs for all the active components.
DuWayne
JV4QB

12296 2016-02-08 09:25:01 vasilyivanenko Re: 74L04 as an amplifier
Please post it Jim
12297 2016-02-08 09:32:22 Dana Myers Re: 74L04 as an amplifier
12299 2016-02-08 11:51:54 Leinweber, Glen Re: 74L04 as an amplifier
Have tried ECL, TTL, HCMOS logic families this way.
Had some success with an ECL XOR gate as a double-balanced
mixer, but I wouldn't recommend it.
For 74HCU04 input Z and output Z is somewhat high. A linearizing
feedback resistor will lower the input Z. You might think that these
would work as buffer stages (they give an open-loop voltage gain
of about 10) but the reverse gain (S12) isn't wonderful. Perhaps
useful as FM I.F. limiter preceding FM detector?
2-port parameters also vary with DC supply.
I have found TTL logic family is pretty non-linear for analog use.
Logic gates aren't optimized for low-noise analog performance.


12300 2016-02-08 16:31:30 iam74@rocketmail.... Re: 74L04 as an amplifier
Yeah, which is the whole point. Digital logic is supposed to be either zero or one with minimal interference in achieving either one.

I suppose if one were to try to adapt, the best choices might be 74F (for 5v systems) or 74AC ( for 12-15v systems). These are probably the fastest.

ECL is absolutely great both for ruggedness and adaptability, but it is pretty useless above 1 MHz or so. OTOH, they are virtually indestructable.

Maybe it would be more fruitful to look more closely at MOSFETS...

john
AD5YE

---In emrfd@yahoogroups.com, wrote :

Have tried ECL, TTL, HCMOS logic families this way.
Had some success with an ECL XOR gate as a double-balanced
mixer, but I wouldn't recommend it.
For 74HCU04 input Z and output Z is somewhat high. A linearizing
feedback resistor will lower the input Z. You might think that these
would work as buffer stages (they give an open-loop voltage gain
of about 10) but the reverse gain (S12) isn't wonderful. Perhaps
useful as FM I.F. limiter preceding FM detector?
2-port parameters also vary with DC supply.
I have found TTL logic family is pretty non-linear for analog use.
Logic gates aren't optimized for low-noise analog performance.


12301 2016-02-08 16:47:10 Dana Myers Re: 74L04 as an amplifier
12302 2016-02-08 17:59:04 kb1gmx Re: 74L04 as an amplifier
ECL is the fastest logic going as its non-saturating.  ECL100K nor gates as amplifiers run well 
into the several hundred mhz range (typical is 1ghz).  Its far from slow even the old MECL 1K
(Ca 1969) ran at 100mhz easily.  And they do make good amplifiers the NOR gate or  the OR/Nor
(complementary output) parts are ideal as the structure is basically a dif-amp.  The big limitation 
is high power drain when biased as linear and they are not usually cheap.

TTL is a poor choice as its noisy and eats power.  Cmos is ok but again high power as
biasing them has both output devices turn partially on.  Without feedback the gain is 
very high so they saturate with small inputs.   The common TTL linear ap was a crystal 
oscillator and as they were used it was only ok and tended to be hard to reproduce and 
tended to take off at harmonics or overtones.   These days CMOS gates are often used 
for oscillators, for non critical uses they work fair.

An old logic family RTL was popular as linear devices and the uL914 was commonly used as Dif-amp
and comparator with decent gain to 10mhz or more (back then that was really good, as in 1968).
biggest limitation was low operating voltages (under 5V).  They were widely available then and 
cheap relative to common silicon transistors.  

Generally when one wants gain cheap, MMICs are a good choice and some can easily develop 
power in the 10s of milliwatts.  Gain is predictable and bandwidth is usually 1ghz or more some 
exceed 10ghz.  Price is dependent on the part but can be very cheap.


Allison
12303 2016-02-08 18:17:36 Will Re: 74L04 as an amplifier
Hi Alison,

Thanks, a good round up.

I recall (think I recall) that I saw way back 1980's that if using CMOS
logic in linear mode to use the *U version. U=unbuffered. Maybe in an
RCA handbook

Cheeers,
Will

12304 2016-02-08 23:01:10 iam74@rocketmail.... Re: 74L04 as an amplifier

You are right, I spoke in error. I was thinking of the ancient diode-transistor logic,

---In emrfd@yahoogroups.com, wrote :

12305 2016-02-09 00:10:38 John Re: 74L04 as an amplifier
ECL is pretty fast. I was using the Motorola MC10116 triple line
receiver back in the '60's as a amplifier at 50 MHZ. The MECL3 family
had flip flops with a 300 MHz toggle rate.

John
KK6IL

12306 2016-02-09 01:40:13 Dave Re: 74L04 as an amplifier
  I use the SY10EP52 LVPECL flip flops to generate the I and Q LO signals in the UHFSDR.  The data sheet says they will clock at a minimum maximun rate of 3GHz!  They also drive a 7 level dBm through a 3dB pad very nicely.

Dave - WB6DHW


12307 2016-02-09 04:11:07 i7swx Re: 74L04 as an amplifier
Hi all,

atbeginning of 70s, when I was working with DEC in Milan, I was able to try all the digital ICs for amplifier use.
I started with the TTL, then tried the RTL and DTL. The latter ones had a sporty linearity, better were the RTL. The easiest where the CMOS. The ECL were interesting being linear logic. As already mentioned by other group members theresults were frequency limited, linearity limited and the difficulties was high noise, not too good for low signals applications. The ECL Flip-flop where good for mixing two signals (VCO/VFO and LO).

73

Gian
I7SWX
12340 2016-02-11 11:53:29 Leinweber, Glen Re: 74L04 as an amplifier
Have measured a small signal "linear" model of a 74HCU04. Note that this is an unbuffered version of the more common 74HC04. It consists of a single PMOS and NMOS totem-pole. Static protection (consisting of diodes, resistors) is present too. It was characterized @ 1 KHz. so doesn't include high-frequency effects. Vdd was +5v, and its input was biased at half-Vcc so that static Vin = static Vout.

Small signal transconductance was measured to be 0.0224 A/V. Philips data sheet suggests that it should be about 0.037 A/V.
Open-loop voltage gain measured 12.5, suggesting that equivalent output resistance is 557 ohms.
These data will change with different Vdd supply.

The 74HC04 will have higher gain (likely to oscillate) because each of its 6 inverters has three cascaded stages. Probably not useful to try a linear model.
I'd say that the 74HCU04 is one of the more useful logic gates for pseudo-linear, if you really must. It CAN make a fairly decent oscillator.
-Glen