Hi,

I have a 555 outputing to a JK flipflop to generate a 2bit timing circuit (00,01,10,11). The voltages are TTL. However, I am trying to control a switching device (Hittite's HMC344LC3
http://www.hittite.com/content/documents/data_sheet/hmc344lc3.pdf ) which requires 0/-5v control. They recommend a TTL interface circuit (on page 5 of 6), but I have both simulated and built the circuit and can't get either to work.

Any suggestions on how to achieve this would be warmly received.

Many thanks,

Phil

 

That circuit should work. It inverts the signal, however. Why won't it work? What's exactly happening?

 

On the stripboard I am getting a stepping graph going +5v. 0v, -5v, then back up to +5v repeating.

On the simulation I am getting 0v DC

Thanks,

Phil

 

You are likely having problems simulating the circuit because in the simulation model, the74HCT04 has it's ground "hard wired" to node 0 (ground). Unless you use some fancy footwork, you won't get the model to produce an output that is below ground level.

As far as your real-world circuit, you are probably using a typical 500mW or 1W Zener diode. These Zeners require a certain amount of current flowing through them before they will drop their rated voltage; usually in the range of 10mA to 20mA; for example the 1N751's rating is 5.1v @ 20mA current flow. The 10k Ohm pull-down resistor only allows a maximum of 1mA current flow, even if the diode were replaced by a piece of wire.

Decreasing the 10k Ohm resistor to 470 Ohms to 1k Ohms will help the Zener situation, but then you will run into loading problems for the 74xx series IC's and their source/sink capabilities.

I'm attaching a simulation of their circuit using LTSpice and a model from the 74HC.lib available on the Yahoo! LTSpice user group. Note that I modified the "SpiceModel" line from "VCC 0" to "0 VEE" to override the defaults, thus allowing the model to be used with a negative voltage for it's ground node.

Even then, the simulation is not really all that accurate, as in the real world the voltage across D1 would be much lower at the small amounts of current flowing through it (.1mA to .5mA)

 

Just a thought - do you have 0.1uF metal poly or ceramic capacitors across the Vcc/GND or Vdd/Vss pins of ALL of your logic ICs? Without them you will have strange results.

 

I will try to reduce the reistor somewhat, check out the caps across the power & ground IC pins (I assume these are for de-coupling?) and will see if I can respect the zener diode to one with lesser current rating. I will feedback to the group what I find. I assume anyone using these devices will be in the same boat as me.

Thanks for the input!

Phil

You are likely having problems simulating the circuit because in the simulation model, the74HCT04 has it's ground "hard wired" to node 0 (ground). Unless you use some fancy footwork, you won't get the model to produce an output that is below ground level.

As far as your real-world circuit, you are probably using a typical 500mW or 1W Zener diode. These Zeners require a certain amount of current flowing through them before they will drop their rated voltage; usually in the range of 10mA to 20mA; for example the 1N751's rating is 5.1v @ 20mA current flow. The 10k Ohm pull-down resistor only allows a maximum of 1mA current flow, even if the diode were replaced by a piece of wire.

Decreasing the 10k Ohm resistor to 470 Ohms to 1k Ohms will help the Zener situation, but then you will run into loading problems for the 74xx series IC's and their source/sink capabilities.

I'm attaching a simulation of their circuit using LTSpice and a model from the 74HC.lib available on the Yahoo! LTSpice user group. Note that I modified the "SpiceModel" line from "VCC 0" to "0 VEE" to override the defaults, thus allowing the model to be used with a negative voltage for it's ground node.

Even then, the simulation is not really all that accurate, as in the real world the voltage across D1 would be much lower at the small amounts of current flowing through it (.1mA to .5mA)

 

Here's another idea; use a "Rubber Zener" instead. Low-power Zeners usually have to be ordered, and that could take days to get them delivered.

I used a 2N4401 transistor, which has a fairly high gain for a standard NPN. You can even get these at Radio Shack, they come in their 15 NPN transistor bubble-pack assortment; you usually get 4 or 5 of them in the assortment. A 2N3904 should work OK too; you'll get several of those in the assortment as well.

Have a look at the attached schematic.

 

That's true about the zeners. I tend to forget that.

What about powering the 555 and flip-flop with 0v Vcc and -5v Vee? Or you need that signals for something else?

 

That's true about the zeners. I tend to forget that.

What about powering the 555 and flip-flop with 0v Vcc and -5v Vee? Or you need that signals for something else?

Don't need it for anything else, but didn't thing that would work. Might be worth a try though!!!

 

You'll really to use a CMOS 555 timer, like a TLC555. A standard BJT 555 timer doesn't work very well at low voltages, and the output doesn't go higher than it's Vcc-1.3v even under light load.

You can somewhat compensate for that somewhat by using a pull-up resistor of around 300-330 Ohms, but that uses a fair bit of power.

The trouble with the CMOS 555 is that it has very low current source ability; even at 15v it's a wimpy 1mA or so. At low Vdd, it's far less.

Here's yet another revision - this one gets rid of the 74HC04 entirely, and the output drive current is increased to normal TTL levels.