Help Debug Circuit

rjenkinsgb said:

Replace Q2 with a 47K resistor and see what happens??

Click to expand...

No, leave it alone.

I went through the old thread, and understand most of what he is doing. Q2 is being used as a zener diode to set the capacitor C1 voltage level at which the fan comes on. 99 other people would use a zener diode, but it's a personal design choice. Could be his version of a watermark, or a trick he learned from a mentor, or ...

The part I don't understand is using a relay to drive a relay. I don't see any reason not to rearrange the circuit around Q1 to eliminate the small relay and have Q1 drive the SSR directly. In fact, I would put the output LED3 in series with the SSR rather than in parallel with it, so it gives a true indication of the output condition. But that's just me.

Also, because he is driving the small relay with an emitter follower rather than a saturated switch, the voltage across the relay coil is going to decrease very slowly. There is an excellent chance that its contacts will "chatter" before opening completely, a not-good thing for whatever is downstream. In your case, the drive circuit inside the SSR might have a little de-bouncing capability to keep the output from producing a small bunch of single-cycle AC bursts to the fan. And, the reed relay coil in his original schemati probably requires more current than the SSR input. To be clear, there is nothing inherently wrong with the double-relay approach, but it adds nothing to the safety of the circuit or any other operational aspects that I can see.

To your original question (no, I didn't forget), there is nothing in your schematic that would make the circuit not work for a while after manufacture, then start working.

ak

AnalogKid said:

No, leave it alone.

Click to expand...

What an amazing diagnosis. Thank you.
I tested all 5 boards and none of them trigger the relay. only the modified one does but with just a minimal 1.5 second max delay. Ill replace it with this BAT46.
My knowledge is limited but perhaps its the way or order in which i am applying power to the components.

AnalogKid said:

other people would use a zener diode

Click to expand...

Could you tell me what direction to insert the diode? Cathode toward the resistor or the base of the transistor.

Is it possible I have my transistors in backwards? Images like this confuse me so maybe there's an orientation problem.

ThomsCircuit said:

Could you tell me what direction to insert the diode? Cathode toward the resistor or the base of the transistor.

Click to expand...

Cathode to the Q3 base, cathode to R1.

There is a trade-off here. The larger the zener value, the more "stretch" you will get out of the timing capacitor. In round numbers, Q3 will turn on when the voltage across C1 is greater than the zener voltage plus 0.6 V. -ish. The larger the zener voltage, the longer it takes for C1 to charge up (down?) to that value. The tradeoff is that the closer the magnitude of the C1 voltage gets to the value of the power rail, the more slowly it changes, leading to a very slow turn-on of Q3. Not a problem in this particular circuit, but a real issue in other R-C timing topologies.

A common design point is to select the zener such that the transistor comes on when the cap is charged up to 0.63 times Vcc. This makes the timing math very easy, as the timer period now equals the R-C time constant.

ThomsCircuit said:

Is it possible I have my transistors in backwards? Images like this confuse me so maybe there's an orientation problem.

View attachment 134646View attachment 134645

Click to expand...

Transistor pinouts are all over the place. You need the correct datasheet from the manufacturer of the exact part you have.

ak

AnalogKid said:

Transistor pinouts are all over the place

Click to expand...

I understand. I dont have a manufacturer. But I got this super close up image. I think Q3 is backwards. Could you help me verify?
Let's say that pin 1 here is the emitter on the 3906. While Q2 points correctly it looks like Q3 is in upside down.

Actually it appears Q2 is backwards.

AnalogKid said:

Cathode to the Q3 base, cathode to R1.

Click to expand...

So which is it. The transistors are all correct.

AnalogKid said:

select the zener such that the transistor comes on when the cap is charged up to 0.63 times Vcc

Click to expand...

So how do I tell the voltage of a diode? Is it the forward voltage? And although the information you shared about calculating what size diode i would need it interesting it made no sense to me. At least not at this time. I also feel like I'm putting a bandaid on this. If all the components are correct and my schematic matches the original why are these not working?

rjenkinsgb said:

I meant purely as a diagnostic test

Click to expand...

I tried your resistor suggestion and it worked. I also tried reversing q2. No change. I also tried a bat46 diode. No change. Can you provide insight as to why?

AnalogKid said:

would put the output LED3 in series with the SSR rather than in parallel with it, so it gives a true indication of the output condition

Click to expand...

The SSR has its own light.

ThomsCircuit said:

So which is it. The transistors are all correct.

Click to expand...

Typo. Anode to R1.

AnalogKid said:

Typo.

Click to expand...

Thank you. That did not work. Possibly the bat46 is not right for the circuit. I can order them if you suggest one. I apologize for being such a noob. All the info you have provided is helpful. I'm Google-ing everything trying to make the most of this.
So bothered that this does not work. Ive made plenty of projects all successful. Why not this.

Ok found a 5v diode.
1N4733A. Can you verify that im getting the correct component to replace Q2

I must have missed something. I thought you were going to replace the transistor with a zener. For a signal diode, reverse the connections - anode to the transistor, cathode to the timing capacitor. Because the forward voltage of a Shottkey diode is way less than the reverse voltage of a transistor base-emitter junction, the turn-on timing delay will be much shorter. It sill be a bit longer with a regular diode such as a 1N4148 or 1N4004.

Basically, Q3 is acting as a comparator. When the base is low enough for the transistor to turn on, it does. "Low enough" changes if there is something in series with the base that has its own turn-on voltage, like a zener diode.

ak

ThomsCircuit said:

Ok found a 5v diode.
1N4733A. Can you verify that im getting the correct component to replace Q2

Click to expand...

Sounds good to me. Connect as in #32.

ak

AnalogKid said:

I must have missed something. I thought you were going to replace the transistor with a zener. For a signal diode

Click to expand...

Nope. I just couldn't tell what diode to use. I must be told exactly what to do. Otherwise I'm just guessing. Another viewer suggested a 5.1 void diode. That's what I thought I should get.

Good news!
I tried a BC327 and it worked. Ive updated 3 boards and they all worked. So being that the 3906 & the 327 have an emitter voltage of 5 would that make them similar? Whats the difference here. Why do you think this transistor made all the difference?

A ZENER diode - somewhere from 4.7V to 6.8V range.
It needs the capacitor to be able to charge more before Q3 starts to turn on.

The reverse transistor junction works a bit like a zener but the voltage is somewhat variable.

Edit - sorry, I did not see that there was another page! - That was relating to you mentioning BAT 46 & Bat 85

rjenkinsgb said:

A ZENER diode - somewhere from 4.7V to 6.8V range.

Click to expand...

Absolutely. Ive ordered two dozen 1N4733A. Future builds of this project will have the zener diode. I mean the BC327 is ok correct? It wont cause any problems.
Solving this has been extremely educational. Again I am grateful for the help and this group. I still dont know why the 3906 did not work but pleased nevertheless.