I'm working on developing a custom SSR circuit for a very specific application. In an effort to better understand what I'm doing, I've been doing lots of reading, lots of experimenting and simulation, and some reverse engineering.

Most recently, I've reverse engineered a circuit that my boss is especially fond of because it was designed (by someone we no longer have access to) for a machine he designed many years ago that has proven exceptionally reliable. I feel like I have a general understanding of most of what's going on in this circuit, but I don't understand the purpose of the two diodes. As far as I can tell, they'll pass both AC and DC, only adding a small voltage drop corresponding to their forward voltages. I can't think of any purpose for the tiny voltage drop, unless it acts as a noise filter of sorts, insuring that very small voltage fluctuations don't make it to the gate of the triac.

I made a rough simulation of the circuit in LTSpice, and the circuit seems to behave essentially the same with or without the diodes. The simulation could be somewhat off though; I didn't have the appropriate models for the opto nor the triac, so I used models for other units I'm considering in my own SSR design.

If anyone can shed some light on the value of the diodes in this circuit, I'd appreciate it!

 

Its probably acting like a Diac, and it increases the gate firing voltage.

 

Its probably acting like a Diac, and it increases the gate firing voltage.

Ahhh... I had seen Diacs used a lot in triac circuits, but didn't realize one could be constructed with parallel diodes like this. After reading your post, I read a quick lesson on Shockley diodes, which I was previously unaware of.

Other circuits I've seen used the Diac as part of a wave chopping SCR-dimmer style circuit. In this case however, the goal is to start a motor and run it at full power. I don't see the benefit of the Diac here. Wouldn't a gradual increase in voltage provide cleaner switching? The opto is zero-crossing, which would seem to go along with the clean switching mindset, but wouldn't a Diac create a phase delay and a noisier circuit?

 

A Diac has a negative resistance region, so it can't be exactly simulated by regular diodes.
But Diacs are used in phase-control circuits and, if you circuit is only going to be on or off, I agree with you, there seems to be no need for the diodes.
(But you may have to put them in to keep your boss happy).

 

Interestingly, I just tried using the diode test function on my DMM on one of the diodes, and it shows a forward voltage of just under 0.5V. Based on the description of how Shockley diodes work, I would've expected them to require far more voltage than the DMM provides, and therefore to look like an open circuit to the DMM. Does this reading indicate that they are not actually Shockley diodes? I also just read the Wikipedia article on tunnel diodes. Could it be one of those? I'm really curious as to what these things are and what purpose they serve. It seems like any oddball kind of diode that would do something significant here is obsolete.

Just to be clear here, I don't have a schematic - I just traced a circuit. These parts look exactly diodes, and have a forward voltage measurement that seems to fit with that, but I don't really know what they are, only what they look like! If they are just standard diodes, I really can't understand their purpose here in the circuit.

 

Why would you expect Schottky diodes to have a higher voltage drop than 0.5V? They have a lower forward drop then junction diodes, which is upward of 0.6V.

They are not likely tunnel diodes. Those are used in high frequency circuits.

 

They could also just be there to help prevent spurious noise from false triggering the triac.

 

Why would you expect Schottky diodes to have a higher voltage drop than 0.5V? They have a lower forward drop then junction diodes, which is upward of 0.6V.

They are not likely tunnel diodes. Those are used in high frequency circuits.

Not Schottky, Shockley.

Based on Dodgydave's comment that a pair of diodes could be acting as a Diac, I did some quick reading on Shockley diodes, which apparently aren't made anymore. A pair of Shockley diodes in parallel, with reversed polarity, would act like a Diac, so I thought maybe these diodes were Shockleys in that configuration, since Diacs are often used in Triac control circuits (even though I wouldn't see the benefit of the Diac in this particular application.)

Doh! I did some more twisting, turning, cutting, and aiming flashlights... and finally found a part number I can lookup. It's marked as a 1N4005, which would be a 600V, 1A standard diode with a Vf of 1.1V. It's a little disconcerting that my Vf measurement is basically half of that, but the part number is a standard one now that I was finally able to read it, so I've got to assume it's a standard diode. That brings me back to the question of what purpose a pair of standard diodes in this configuration would serve.

 

They could also just be there to help prevent spurious noise from false triggering the triac.

That was my original thought too. I've never seen it done before, so I doubted my guesswork. Is this a common technique? Does it seem like a worthwhile addition that I should consider when designing my own circuits in the future?

 

That was my original thought too. I've never seen it done before, so I doubted my guesswork. Is this a common technique? Does it seem like a worthwhile addition that I should consider when designing my own circuits in the future?

It your circuit, the control Triac is in series with the output Triac gate, so I don't see how the diodes can have any significant effect on the noise immunity.
A Triac is triggered by a gate current and since the control Triac is basically an open circuit until it fires, then there's no path for any significant noise current to trigger the output Triac.

If you design a circuit where the circuit operates on a voltage threshold then the diodes could help with noise immunity, but they don't significantly help where the circuit operates with a current threshold, such as here.

 

It your circuit, the control Triac is in series with the output Triac gate, so I don't see how the diodes can have any significant effect on the noise immunity.
A Triac is triggered by a gate current and since the control Triac is basically an open circuit until it fires, then there's no path for any significant noise current to trigger the output Triac.

If you design a circuit where the circuit operates on a voltage threshold then the diodes could help with noise immunity, but they don't significantly help where the circuit operates with a current threshold, such as here.

As usual, I'm learning a lot here. Thanks!

Do you think that these diodes might have been included in a misguided attempt to improve noise immunity, or is there still some other potential purpose I'm overlooking?

 

Take them out of the circuit and try it, and then report back what the difference is...

 

They look like they are there to create a "dead band" - a region where nothing happens. Probably for noise, like recklessrog suggested.

 

They look like they are there to create a "dead band" - a region where nothing happens. Probably for noise, like recklessrog suggested.

That would be true for a voltage-sensitive node but the Triac gate is basically a low-impedance current-sensitive node.