Fixing a laboratory waterbath

[eccerr0r]

Re: Fixing a laboratory waterbath

The reasons why I'm doubting relay:

- Voltage of the relays are typically printed on the relay
- Switch characteristics are typically printed on the relay
- It appears to be controlling the triac if that really is a triac there - versus a load. Again I don't see the connections so have to make assumptions with the incomplete data here.
- tracing the input path, what a mess, the device is being driven with very little current, not enough for a relay due to a resistor in series with the current path.

My current guess: "optoisolator" in quotes as a guess someone potted a LED and some optical sensors in that plastic.

But this is still just a guess. But if that device feels like it has metal in it then fine it can be a relay and will need to retrace the circuit more carefully. Really would like a picture with all the boards assembled together to make sure things are the way I thought they are connected.

[vrasp]

Re: Fixing a laboratory waterbath

You need to post a clear (in focus) high res, picture of the trace side of the board the relay is mounted on.
Looking at the top suggests the small bridge rectifier and the 470µf capacitor are the supply for the relay, the to220 transistor controls the relay coil, The small glass diode next to the relay will be across the relay coil. I suspect the relay coil is either 12 or 24v based on the voltages marked on the transformer

This is a picture with diagram from the very beginning of this "journey". Someone from reddit made this for me. I think it s useful in this case. You can see the connections where the relay is, if it really is one. I think it is though, but that's my non professional opinion.

It would make sense that it is. The lower current side is connected to the pins (here labelled heater control signal) that go to the control PCB (not seen here) and activates the high current side which is connected to the gate. When the relay closes the circuit, the gate recieves a current and the triac conducts current to the heating element. Again, thats my non professional understanding of this circuit.

Edit: there is a component (looks like a zener but I ll have to check) that is not seen on this picture. It s above the red capacitor and in front of the relay. It connects the relay to the gate of the triac. Sorry if this is confusing. I m gonna take better pictures and get back to you.

[vrasp]

Re: Fixing a laboratory waterbath

The reasons why I'm doubting relay:

- Voltage of the relays are typically printed on the relay
- Switch characteristics are typically printed on the relay
- It appears to be controlling the triac if that really is a triac there - versus a load. Again I don't see the connections so have to make assumptions with the incomplete data here.
- tracing the input path, what a mess, the device is being driven with very little current, not enough for a relay due to a resistor in series with the current path.

My current guess: "optoisolator" in quotes as a guess someone potted a LED and some optical sensors in that plastic.

But this is still just a guess. But if that device feels like it has metal in it then fine it can be a relay and will need to retrace the circuit more carefully. Really would like a picture with all the boards assembled together to make sure things are the way I thought they are connected.

It is really messy. It doesn't help that I replaced a bunch of stuff and sometimes on the other side of the boards to make it work with the space I have. I ve also blown a few things because of things touching the grounded case again due to lack of space. But these things were replaced and it has worked fine since. Until this recent issue.

I ll try to tske some more pictures with more details.

The triac is a NTE56010.

[R_J]

Re: Fixing a laboratory waterbath

Now that I see the under side of the board, that to220 is likely a +5volt regulator ic and not the relay driver. The relay is driven by the heater control board so that would need to be traced to determine the relay coil voltage.

[vrasp]

Re: Fixing a laboratory waterbath

Here are some more pictures and a video. Hopefully, that helps.

The red capacitor and the "relay", or whatever it is, have been removed.

https://vimeo.com/575955271

The burned trace next to the triac is from the previous triac. This one is new. There are some cables hanging that I used to make it easier to measure pins I couldn't get to. These will be removed once I put everything back together.

[eccerr0r]

Re: Fixing a laboratory waterbath

oh gawd now I see that the two boards go together component side to component side, I thought they were stacked... this makes a huge difference, and previous assumptions made earlier on may need to be invalidated... ugh...

1. How did you get away without an insulator on the TRIAC? Was this repaired before you worked on it? If so I'm very sorry. If you did replace it, you missed an insulator.
2. It looks like the 5V from the transformer board is actually the LED power supply after all, and the transformer may be connected wrong.
3. What gave you the impression the relay or whatever it is ... is bad? Also relays are denser than optoisolators, does it actually feel dense? Does a magnet stick to it?

Sigh. I hate remote debug. About ready to give up on this.

[vrasp]

Re: Fixing a laboratory waterbath

Oh man. I had no idea I needed an insulator (The previous triac had what looks like an insulator already on it). Maybe that explains the short. The triac did come with a sticky plastic part which must be it. I guess I might have to replace it again.

Everything was working fine with the transformer set up that way. The 5v from the 7805 do go to the LED.

The relay is somewhat heavy. I ll try the magnet test. I thought it was bad because the high current side is always closed and so has continuity. Also the trace that blew was where the relay was (you can see on the photo I put some copper wire with solder on it).

(Please dont give up haha)

[eccerr0r]

Re: Fixing a laboratory waterbath

That trace blew because you didn't put the insulator on. You need both the kapton/mica/... insulator with heatsink compound, plus the plastic shoulder washer. If the old screw doesn't fit with the shoulder washer, you need to find a thinner screw, usually #4 or M2.5 screws are thin enough.
Since you blew that trace, yes you may have blew the (relay). You might be able to get away with an arbitrary 12V relay but I still think it's optical because relays are not fast enough.

Then again perhaps the original designer was an idiot and designed it this way... *facepalm*

BTW, the "input" side should be quite conductive if it's a relay, this would be a good test and shouldn't get damaged by the shorting TRIAC. If it were a true optoisolator it needs to be an LED input, and these would be conductive in only one direction... again if it really is a relay, this is a very poor design.

Oh and another thought... it could be a transformer. Now discerning a damaged transformer and a relay with simple tools...

[vrasp]

Re: Fixing a laboratory waterbath

What about the triac? Can I assume that since the trace next to the "relay" (or let's call the the black thing) blew, the triac didn't suffer any damage. Do you think I should replace it?

Both sides of the black thing have continuity (between pins of the same side) in both directions. A fridge magnet sticks to both sides shown on the pictures but not the other 2 sides.

The writing says:

1/0,66
01/L06

I m trying to think what other info I could provide. If the signal that comes from the green/blue PCB is a DC signal then maybe we can rule out a transformer since they work with AC only.

[eccerr0r]

Re: Fixing a laboratory waterbath

Aha... yeah that looks like transformer, that bottom side shows a lot. It's probably is fine, all the better - finding a replacement of this one would be tough. If your triac doesn't appear shorted it may still be fine, else you may need to replace it again.

[R_J]

Re: Fixing a laboratory waterbath

I was also thinking about this and I suspect it is a pulse transformer and not a relay. Likely similar to a Schurter IL series
What is your plan for the transformer with the leads ripped out?

[vrasp]

Re: Fixing a laboratory waterbath

That's great! Very helpful. Do you know why they put s transformer there? What is its purpose?

Thank you!

[R_J]

Re: Fixing a laboratory waterbath

The transformer may step up the pulse from the drive circuit, but it could also be for isolation, isolating the driver board from the triac which would be connected to the ac line.

[eccerr0r]

Re: Fixing a laboratory waterbath

yep same reason for putting a relay or optoisolator there, transformers also provide isolation. Of the three, relay is the dumbest (why have a triac then?) Transformer is "ok" but trickier to use, and optoisolator would be ideal.

Now the question is whether you ever heard it "click" ... a transformer making clicking sounds is not normal.

[vrasp]

Re: Fixing a laboratory waterbath

I never thought of a transformer that way..to isolate circuit. I always just saw it as voltage step up / step down. That s so useful.

No I never heard a click. I tried to listen when I was testing it. Nothing. I believe you guys are right and it s a transformer.

Thinking further about this, if you have a short on one side which drives up the current, the voltage will go up too and have repercussions on the other side of the transformer. So the circuits are physically isolated but still reactive to one another/energically interconnected. What benefit does the physical isolation provide?

[eccerr0r]

Re: Fixing a laboratory waterbath

Isolation is to make sure two circuits cannot interact with each other with a single additional connection. They are mainly for making two circuits that can be connected together without unexpected interactions, and the more important reason: protection from electrocuting people using the circuit.

[vrasp]

Re: Fixing a laboratory waterbath

I was also thinking about this and I suspect it is a pulse transformer and not a relay. Likely similar to a Schurter IL series
What is your plan for the transformer with the leads ripped out?

The ripped out leads are from the previous transformer that has been replaced since.

That trace blew because you didn't put the insulator on. You need both the kapton/mica/... insulator with heatsink compound, plus the plastic shoulder washer. If the old screw doesn't fit with the shoulder washer, you need to find a thinner screw, usually #4 or M2.5 screws are thin enough.
Since you blew that trace, yes you may have blew the (relay). You might be able to get away with an arbitrary 12V relay but I still think it's optical because relays are not fast enough.

Then again perhaps the original designer was an idiot and designed it this way... *facepalm*

BTW, the "input" side should be quite conductive if it's a relay, this would be a good test and shouldn't get damaged by the shorting TRIAC. If it were a true optoisolator it needs to be an LED input, and these would be conductive in only one direction... again if it really is a relay, this is a very poor design.

Oh and another thought... it could be a transformer. Now discerning a damaged transformer and a relay with simple tools...

Should the thermal paste be between the triac and mica or mica and heatsink? or both? Also, what is the triac metal back plate connected to inside the triac? I thought the issue was that the triac metal part was touching the heatsink and was therefore grounded, but I dont get continuity between that metal part and the heatsink.

EDIT:MT1, MT2, and the gate all have continuity with eachother (even when the triac is not on the heatsink) which I think means it's damaged.

[eccerr0r]

Re: Fixing a laboratory waterbath

Thermal paste (use VERY little just like on a CPU) should be on all surfaces that sit between a heat source and the sink where it could have touched air.... so both sides of a mica or kapton insulator.

If you have a silicone insulator, which feels rubbery, you usually don't need the paste - only the hard stuff.

In a properly designed circuit especially with TRIACs you may not actually need the paste as the dissipation shouldn't be as high as semiconductors running half on or at high frequencies.

The standard TO-220, the metal tab is connected to the center terminal of the three pins in general. If your center pin is connected to hot on your AC input... BANG dead short to ground.

However if you dirty the contact from the tab to the heatsink you might get high resistance, though you don't want to leave that to chance - use the insulator and the shoulder washer if it was insulated before.

There are all plastic coated TO-220FP that are isolated from the center pin.

[vrasp]

Re: Fixing a laboratory waterbath

Thermal paste (use VERY little just like on a CPU) should be on all surfaces that sit between a heat source and the sink where it could have touched air.... so both sides of a mica or kapton insulator.

If you have a silicone insulator, which feels rubbery, you usually don't need the paste - only the hard stuff.

In a properly designed circuit especially with TRIACs you may not actually need the paste as the dissipation shouldn't be as high as semiconductors running half on or at high frequencies.

The standard TO-220, the metal tab is connected to the center terminal of the three pins in general. If your center pin is connected to hot on your AC input... BANG dead short to ground.

However if you dirty the contact from the tab to the heatsink you might get high resistance, though you don't want to leave that to chance - use the insulator and the shoulder washer if it was insulated before.

There are all plastic coated TO-220FP that are isolated from the center pin.

Got it. Thank you!

Just got this stuff in the photo. Hopefully I m all set. I got a capacitor to replace the red one in case it broke when thebshort happened. I have to wait until next week to be in the lab.

I get an extra insulator on top of the one provided with the triac. Which one should I use?

Any tips on how to solder the triac pins to a wire? I just put flux and solder on both ends and melt it together for a few seconds then add a few layers of shrink tube (Like in the previous photo of the triac). Maybe there is a "cleaner" way to do it - like some kind of specialty connectors I can get with a female end to insert the pins.

[eccerr0r]

Re: Fixing a laboratory waterbath

capacitor shouldn't be harmed unless the desolder job broke it.
You should use one or the other insulator, just not both.
Make sure you use that shoulder washer that came with the TRIAC. The screw must not touch the TRIAC either.

Without "the helping hands" tool, probably just mount the TRIAC and then solder to the pins. Careful not to pull on the wire after it's been soldered. Should make it a bit easier to solder. Definitely tin the wires first as you were doing.

You don't want a socket here.