ozone control for geothermal dehumidifier

[alec_t]

The new diode measures as you'd expect.
Looks like D1 is leaky; that could affect the resetting.
D2 is the LED. Did you mean D3? Whatever, it's dead. By rights the reverse relay coil current at switch-off shouldn't be enough to kill a 1N4148 so perhaps D3 was dead from the off.

 

[salty joe]

Since both diodes need replaced, could we get beefy and use either 1N4007 (1 Amp rectifier) or 1N5819 (1 amp schotkey)? I have them handy.
Also, should I put a 4u7 back at C2?
Thanks!

 

[alec_t]

I'd keep D1 as 1N4148 (check the replacement with your meter!), but D3 could be a 1N4007. Yes, C2 would be better as 4u7 again.

 

[salty joe]

After installing the new 1N4148 at D1 (I checked it first), with the probes on the DMM reversed I got the same kind of number as the one I removed. I checked the 1N4148 I cut out and it was normal!??? When it was installed I got the bad number, but on the bench it's normal!

Stranger still, the 1N4007 I installed at D3 shows 72 either way the probes from the DMM are, pretty much the same as when the 1N4148 was at D3.

Cooked ICs?

 

[KeepItSimpleStupid]

I'm late getting into this, but there are delay on break timers. I use one for the bathroom fan.

Nice to see you back, Joe.

 

[salty joe]

Hi KISS, very nice to see you. Thanks for chiming in!

 

[alec_t]

Stranger still, the 1N4007 I installed at D3 shows 72 either way

It occurs to me (rather late in the day ), did you do the D3 diode test with the relay coil still in parallel with it? If so, what you measured is primarily a function of the coil resistance rather than the diode.
As for the new D1, if you're getting similar results to the old one then that points to the 40106 being flakey.
Try monitoring the voltage at the D1/R2 junction. It should be ~ 4.3V, and stay at that, when the timer times out.

Edit:
Get out the magnifier and check there is no solder whisker or flux residue anywhere around the 40106 pins.

 

[salty joe]

I got 2.35V from the cathode of D1 when the LED is on and 2.32V when the LED is off. This is with the radio, variac and blower unplugged.

I'm glad you were able to make sense from the D3 numbers.

I need to hit the jobsite and see if I can make a buck or two. Thanks!

 

[alec_t]

I got 2.35V from the cathode of D1 when the LED is on and 2.32V when the LED is off.

2.35V when the LED is on is probably ok. It's ~ the expected average voltage while the input of U1a is toggling. However, 2.32V with LED off isn't right, suggesting gate U1a is suspect. You could try disabling U1a by disconnecting pin 12, then using U1f as a substitute by disconnecting pins 10,11 from their present positions and instead connecting them to R2,R1 respectively. Remember that the 40106 is static-sensitive.

This is with the radio, variac and blower unplugged.

As they should be, for now.

 

[salty joe]

I am not sure if this is what you mean but I think you are saying to connect pin 10 at R2/D1 and put pin 11 where pin 12 was.
But that reverses the gate which doesn't seem right. Does pin 13 stay hooked up? I just want to be sure before making changes.
I appreciate it, thanks.

 

[alec_t]

But that reverses the gate which doesn't seem right.

Ah, you spotted my deliberate mistake . I should have said "disconnecting pins 10,11 from their present positions and instead connecting them to R1,R2 respectively." Sorry about that.
I think it best to connect pin 13 to ground instead of to D1/R2. That will isolate gate U1a better.

Looking again at the whole ozoniser system and bearing in mind that at least the relay switching the variac is handling an inductive load (not sure about the ozoniser), it would be advisable to add a respective snubber circuit across each of the relay switches. Like this:

 

[salty joe]

Thanks for clearing that up. Thanks especially for making sense of the bad numbers coming from D1 as well as D3. If U1f is not cooked, might this be the final repair? Just curios-with U1b-U1f all lined up are they all doing the same thing so we can divert U1f and be done?

The AC caps are on the way-thank you for the snubber refinement. We have plenty of room on that board.
It will probably be the weekend before I can fire up the soldering iron. As soon as I get the IC re-arranged, I’ll let you know.

 

[KeepItSimpleStupid]

Hey Joe:

Are you making videos to be able to show to the critters entitled something like: "My new home - a work in progress (20xx to 20xx)" or maybe some instructional onesnamed " Learning how to solder".

Remember, the other thread was inspirational in allowing me to use Aquarium parts to make a laundry lint filter. Now, the design is "perfect". I used a Sock-it filter https://www.drsfostersmith.com/product/prod_display.cfm?pcatid=16746, but changed to a 400 um Nylon bag,

Initially, I used a hose clamp and a thick washer below the plate and now, Now I use a Ruland Quick Clamp from **broken link removed** bolted to the top of the top plate with a 1/4-20 screw.

I had to use a friend's lathe to take of 0.015" off of the radius (reduce the diameter by 0.030") of the plastic pipe to make the clamp fit.

Now, it's cake to empty the filter.

It sits at the front corner of the Laundry sink. I had ideas of moving it to the rear corner. Doing that creates other issues like the need for front dumping on a temporary basis. I'd have to shorten the discharge and hard pipe it along the wall and then attach with a union in the horizontal direction. I would not need the clamp in this configuration.

Without a clamp, there is a possibility for the force of the water with an unemptied filter to dislodge the pipe.

So, thanks.

 

[salty joe]

Glad the fish tank stuff helped you solve a problem KISS.

Well I grounded pin 13 and disconnected pin 12 then connected U1f in place of U1a and the same darn thing is happening. I guess I could have fried U1f as well as U1a. I forgot to use a damp sponge to keep the IC cool during soldering. Nothing got hot enough to smell funny....IDK.

I do have another 40106, do you think I should cut the old IC loose and wire in a new IC?

 

[KeepItSimpleStupid]

Joe:

Not really sure that Ozone is your best choice: **broken link removed**

Then again, not sure how you might implememnt a proven, EXPENSIVE, technology that does work. UVC bulbs. See **broken link removed**
Poven in HVAC design. but the bulbs need replacing after so many hours of operation.

 

[salty joe]

What I'm looking to do is use O3 to keep a non-porous surface (ductwork) disinfected. It will be used in an isolated manner. Further, the O3 will be used for a short time. I plan to start out with 1 minute three times a week. The half life of O3 is as short as 10 minutes in warm, humid conditions with air movement.

I understand the risks of O3. They can be managed just like the risks of gasoline, natural gas, electricity, chlorine bleach, driving a car etc. can be managed. BTW, UV is also dangerous. Back in the day when I bought into the sterile approach to reefkeeping, I used UV. I appreciate the concern and the link. I am not using the ‘shotgun’ approach. This is a pinpoint application and it will be contained.

 

[alec_t]

Just curios-with U1b-U1f all lined up are they all doing the same thing

Yes. Probably one gate would have sufficed, but by having the 'spare' gates in parallel they at least earn their keep and share the current-providing task for switching Q1. I figured it was worth a try to 'borrow' one as a sub for U1a.

so we can divert U1f and be done?

That was the hope. However, you've shown that didn't work. So yes, replace the 40106. I've run and re-run the sim in LTspice, trying various tweaks, but it's still telling me the circuit should work as designed. We've eliminated D1 as a suspect, so it looks increasingly as though there's something weird with the 40106. CMOS ICs do need careful handling. (I recall a replacement was needed for a damaged 4093 in the early days of getting a PDM to work correctly.)

 

[salty joe]

(I recall a replacement was needed for a damaged 4093 in the early days of getting a PDM to work correctly.)

Or two.

I must have cooked it to death. This time I'll keep it cool. I'll try to get it done this weekend. Thanks.

 

[salty joe]

I think I found why the timer was acting up-I did not supply 5V to the 4020. Sorry for making such a vacant mistake. I'm embarrassed and a little mad at myself… I didn’t find it until the 40106 was removed. That board got a bit mauled while trying to replace the 40106 so I ended up making a new one. It works perfectly with the LED. The shortest run time is 48 sec. which is what I’ll start out with. A very nice circuit, amazing really. Thank you Alec.

Should we wait until the snubbers are installed to test run the radio (in place of the ozonizer) the variac and blower?

BTW, I want to mention this because you might find it interesting Alec-D1 still shows 1976 or so with the DMM probes reversed.
Thanks again for that super schematic, you’re the best!

 

[KeepItSimpleStupid]

The snubbers protect the relay contacts. Your choice. They are not likely to fail in week used 1x/day.

The beefier the contacts ], the less likely failure.

I dinlt find what relay your using or the fan specs, but a "definite purpose contactor" MAY be in order. You typically find these in furnaces. https://www.mscdirect.com/product/d...__15557577904_c_S&026=-99&025=c&item=07428022

They aren't called relays.

I didn't read the entire thread. Just stuck my nose in.