Waveforms in a 74LS04 ring oscillator.

[dannyf]

For hobby uses, it seems to be OK - whenever ebay parts are OK, those potato chips should be ok too.

Anything above that, I would think twice about using them.

[joeqsmith]

74HC132 -> 120Mhz for single gate and 40Mhz+ for 3 gates, all at 5v. 3 gates work down to 2Mhz @ 1v and 60Mhz+ at 10v (yes, 10v).

1/2*120M*1 = 4.2nS = 1/2*40M*3.  Yea, the math works.      Sounds like your numbers are on par with mine.   I am a bit surprised it would run at 10 volts.  What is the exact part you used?   I have a few duplicate parts I could run up to failure and see what I get. 

The OP had started out with 5 gates, then dropped it to 3.   I am not sure why but guess they wanted it running slow enough that they could make the measurement with the equipment they have available.   

With JoeN trying ECL, we could see some faster numbers.  It will be interesting to see what they get working. 

[dannyf]

The HC132 I used earlier was actually an AHC132 (SN74AHC132 by TI).

I drop'd in a HC00 (Philips) and got similar results (95Mhz vs. 120Mhz, one gate only).

[dannyf]

Tried some MM74HC00 from ST. 75Mhz tops (1 gate, 10v).

This thing can be made into local oscillators. With a mixer + LPF, you get a direct conversion, pot-turned radio.

[joeqsmith]

Tried some MM74HC00 from ST. 75Mhz tops (1 gate, 10v).

This thing can be made into local oscillators. With a mixer + LPF, you get a direct conversion, pot-turned radio.

This would make IMO an interesting thread.     

[dannyf]

current consumption for the MM74HC00 part:

12ma @ 5v;
56ma @ 12v (yes, 12v)
250ma @ 20v (yes, 20v, for a 5v part).

[Zero999]

At what voltage does it release magic smoke?

[joeqsmith]

 

When you measure the current, are you using all four gates?    Curious how much is the loading and how much is the part. 

[dannyf]

All four gates were used.

[joeqsmith]

I set the jig to sweep from 2.6 to 24 volts, with the current limit set to 6 amps.   It did not provide me with the excitement  I had hoped for.   The counter's input impedance was set to 1Meg rather than 50 ohms to limit the load current and prevent damage to the front end.  Nothing else was changed in the test setup. 

Looking at the 4069, you can see how the frequency actually starts to fold back.  This is very reproducible.   This is the only part the survived. 

The 74HC04 didn't just fail.  It sort of slowly died after 11.2 volts.

The 74AC14 died real quick after 8.4 volts.   You can see it trying to stay alive for a few more cycles..   

Damage2 is zoomed into the area where the HC and AC devices failed.


No smoke, no plastic fragments.....

[joeqsmith]

Found an duplicate Harris HC part which ran all the way out to 15.2 volts.  Notice again how the frequency was falling off just prior to failure.   Again, no real excitement. 

Grabbed an old Fairchild TTL part thinking that would certainly actually blow up...  It did!   

What's real interesting is how the TTL parts frequency fell way off, then it takes such a radical jump, rolls off again, then dies a flaming smoke filled room kinda death.

If you found an HC part that runs to 20 volts as you suggest and is linear, stick with that part.   

[joeqsmith]

An old National 54F04 military device.  Not as fast as the Fairchild F04.   Also ran it up to 110 and down to -12C. 

[David Hess]

In the past when building these types of test circuits, I usually included a fast discrete JFET or bipolar buffer to drive a 50 ohm transmission line directly to the oscilloscope.  I guess now there are some integrated buffers or operational amplifiers which could do the same thing.