Link: http://www.maxim-ic.com/app-notes/index.mvp/id/4414

Check out figure 4. Seems to do the trick but have to find a faster comparator. What do you guys think?

 

As Bill mentioned, a discrete diff amp should work, but its performance depends on the capacitive load. That's what I've been working on for you. It will require more parts than an IC, so it might not be suitable.

 

For a 2 year old datasheet, it is piss-poor. I couldn't find anything specifying the control voltage loading - no capacitance spec, no current spec. How do the expect you to drive it?
I think you need to call their apps guys and get more info.

 

As Bill mentioned, a discrete diff amp should work, but its performance depends on the capacitive load. That's what I've been working on for you. It will require more parts than an IC, so it might not be suitable.

Hi Ron,

It is driven with differential inputs. Can I get help with maybe a totem pole configuration with Small Signal MOSFET? A circuit topolgy taking 0 to 5V and outputting -5V it would help. Then use another totem pole to do 0 to 5V and output 0V. I can then use it to drive the switch.

Thank you for the help.

 

Hi Ron,

It is driven with differential inputs. Can I get help with maybe a totem pole configuration with Small Signal MOSFET? A circuit topolgy taking 0 to 5V and outputting -5V it would help. Then use another totem pole to do 0 to 5V and output 0V. I can then use it to drive the switch.

Thank you for the help.

I don't see how you would do that.
I have a circuit that give fast differential outputs, 0 to -5V, with little delay, but I need to make sure I can drive it with your TTL. What is the part number of the TTL device that you are interfacing with?
Why are you using TTL instead of CMOS?

 

I don't see how you would do that.
I have a circuit that give fast differential outputs, 0 to -5V, with little delay, but I need to make sure I can drive it with your TTL. What is the part number of the TTL device that you are interfacing with?
Why are you using TTL instead of CMOS?

I don't have a part number for the TTL and I think they are using it because that is what is available.

Can you post the the circuit that you have? It would be really helpful. Thanks.

 

I don't have a part number for the TTL and I think they are using it because that is what is available.

Can you post the the circuit that you have? It would be really helpful. Thanks.

Who is "they" is this a school project? What's it supposed to do?

 

Who is "they" is this a school project? What's it supposed to do?

Yes it's a school project and it is suppose to drive a SPDT switch that has differential control lines.

 

I don't see how you would do that.

Are there circuit schematics that do level shifting with MOS?

 

Post what schematics you have. How do you control an SPST switch?

 

Post what schematics you have. How do you control an SPST switch?

The SPDT switch has a differential input (2 lines). The datasheet was posted.

Below are two schematics that I thought could control it. It was taken from an application note. Output should be -0.5V or -5V where the complementary will be on the other output.

I did it in orcad capture but they don't switch when I use the pulse. kind of odd. Haven't been able to figure it out.

I'm also investigating using MOSFETs to do level shifting and providing complementary outputs but no luck yet. I know small signal FETs with low capacitance are available.

 

You'll need a split supply for starters.

 

Here's the circuit I have been working on. I doubt that TTL can drive it (TTL sourcing current is weak). CMOS could do it.
The 2V source can be replaced by a voltage divider and an emitter follower, or an LM 317, or... It just needs to be able to source and sink more than 2mA. It doesn't have to be exactly 2V.
The MPSH81s will oscillate if you turn your back on them. Q3 is the most susceptible. A small chip capacitor (e.g. 10nF) from the base of Q3 to the GND PLANE might be helpful in preventing oscillation. A small series base resistor (10Ω) might also help. A good ground plane is essential for making your entire circuit work. Power supply decoupling caps are also essential, but are not shown.

 

Here's the circuit I have been working on. I doubt that TTL can drive it (TTL sourcing current is weak). CMOS could do it.
The 2V source can be replaced by a voltage divider and an emitter follower, or an LM 317, or... It just needs to be able to source and sink more than 2mA. It doesn't have to be exactly 2V.
The MPSH81s will oscillate if you turn your back on them. Q3 is the most susceptible. A small chip capacitor (e.g. 10nF) from the base of Q3 to the GND PLANE might be helpful in preventing oscillation. A small series base resistor (10Ω) might also help. A good ground plane is essential for making your entire circuit work. Power supply decoupling caps are also essential, but are not shown.

Ron this is awesome. Can I ask a few questions?

How did you come up with this design? Is that transistor topology standard or something else? Sorry for the ignorance. I'm just trying to understand your design methodology.

 

This is fairly basic electronics, though I doubt I could have designed it without agonizing over it. Ron has a strong engineering background in this specific subject.

It resembles ECL very strongly. The reason ECL isn't used more is it is an energy hog. It's speed is why it isn't used less.

 

This is fairly basic electronics, though I doubt I could have designed it without agonizing over it. Ron has a strong engineering background in this specific subject.

It resembles ECL very strongly. The reason ECL isn't used more is it is an energy hog. It's speed is why it isn't used less.

Probably basic for you guys but I always had trouble wrapping my brain around this stuff. I'm going to try in the morning to explain how this works and you guys can correct me if I'm wrong. That way I'll learn from my mistakes. I just don't understand want spawns a design like this from scratch.

 

Here's the circuit I have been working on. I doubt that TTL can drive it (TTL sourcing current is weak). CMOS could do it.
The 2V source can be replaced by a voltage divider and an emitter follower, or an LM 317, or... It just needs to be able to source and sink more than 2mA. It doesn't have to be exactly 2V.
The MPSH81s will oscillate if you turn your back on them. Q3 is the most susceptible. A small chip capacitor (e.g. 10nF) from the base of Q3 to the GND PLANE might be helpful in preventing oscillation. A small series base resistor (10Ω) might also help. A good ground plane is essential for making your entire circuit work. Power supply decoupling caps are also essential, but are not shown.

Also, is it possible to do this with MOSFET's? The same circuit.

 

Using iCircuit, I built the circuit and ran it. It shows that the high is 1V as oppose to 0V. When the source is 4V it switches but then the green is 1V instead of 0V. Did I do something wrong?

Picture attached.

 

Your results will only be as accurate as your models. I used a spice model for MPSH81. What transistor model did you use?
Even if you built this in hardware, your results would probably be different from the simulation. At the very least, you would probably have to change the value of R7 in my schematic.I simulated the circuit using LTspice, and included the .ASC file, which will run in LTspice. LTspice is free, and is not a twinky toy like some simulators.

 

Haha, twinkly toy. That's a good one.

Your right about the model. I didn't take that into account. I did it on an iPad while lying in my bed. It helps me to understand general functionality. We have OrCAD available and will try it there.

If you have the desire and/or time I asked a few more questions above. I appreciate the help big time.