Aquarite generator, only power light is on?

[DeadAquaRite]

Split from This topic. Bama

I am in the same situation. 5yr old unit. Power light on, menu cycles as expected, no other lights. I discovered mine when spring came and the pool turned greenish. Checked the panel, salt level 1200. Not unbelievable as we have had a lot of rain over the winter, but added no salt. So I added salt to raise it to 2800. No change in salt reading. Did some research, played with the panel a bit, and discovered where the instant reading is to find that it reads "- 0". Plugged/unplugged cell, checked contacts, all looks good.

Read on the net that a the varistor likes to cook itself, saw some scorching on the standoffs nearby, static ohms measured 3.7 (vs 2.8 on the tech data sheet), so I replaced it. No joy. Checked the fuse out of circuit with a meter, tests good.

Called tech support today, they said it's probably the board and I need a technician.

All that said, I get no flow light, no low/high salt light, no salt reading. Not even a blip when I turn it on. But the board behaves normally otherwise. I doubt the CPU is smoked, which leaves a handful of discrete components and some wiring. I'm not replacing a $500 board because a 2c resistor has a cracked solder joint. My first guess is a bad ground to the cell, which would kill generation and salt readings. But the flow sensor is separate. Anyone have a schematic?

 

[DeadAquaRite]

Update: salt cell measures 1.8ohms. Flow sensor goes open on no flow, and 2ohms on flow. Starting to work up part of a schematic. I suspect a fault in the salt sensor circuit is making the cpu think there is no sensor and inhibiting all functions. If so, I may force the sensor input "high" and see what happens.

I would kill for a state diagram....

 

[DeadAquaRite]

Took cell to the Leslie and it tested good. Salt level is 4000. Instant level still reads zero.

Cell voltage (31.5v) is good to the relays, but the relays are not engaging. I see 24V to ground at the relay coil, so they should be energized. I would say it's a bad relay, but there are two in parallel? Obviously the cpu isn't calling for it.

Anyone know the prerequisite signals for generation?

How does it detect salt content in the T-CELL-15? Opening it up, there is one sensor (two wires). It intrudes into the flow, but is encased in plastic. I am thinking temperature.

There are also 4 wires (2 white, 2 black) that appear to enter the cell itself through an epoxy plug. Judging from the wiring at the motherboard (2 poles = 4 pins), these are to the generator plates.

It appears that salt detection is either via the plates, or a sensor wired to the plates.

Anyone got a dead Tcell15 they want to donate so I can dissect it for science? I will pay shipping and post pics here.

 

[Bama Rambler]

Sorry you haven't gotten any help on this. I'm not much on salt systems but I'll chunk my 2¢ in here.

If your reading voltage across the relay coils, they should be energized and if they're not then I have to suspect the coils are open. Even though they are two of them.

Those units use conductivity to determine the salt level so there's not actually a salt sensor pre-se.

 

[DeadAquaRite]

Thanks for the help! Both coils have continuity, and continuity from common to each terminal (open/closed) is as expected. I'll double check the resistance today to make sure the coils aren't shorted.

That's what I thought about the salt measurement. So salt levels must be measured only when the relays are engaged. That's important, because it means the relays should engage at least briefly, regardless of salt level.

 

[DeadAquaRite]

Ok, Haven't gotten around to testing the coils as some low life stole my wife's car. So I just tossed in the towel and ordered a replacement board. Before that, I did manage to sketch a schematic of the transformer/rectifier assembly. I hope to get it scanned and posted, but here is the nickel tour in case someone needs it.

Internally the transformer has two loops on the AC side. These are the Grey/Violet and White/Blue terminals respectively. For 120V, terminals 1&2 are shorted and terminals 3&4 are shorted to feed 120V in parallel to both loops. For 240v, terminals 2&3 are shorted to create one large grey-blue loop and 240V is fed to terminals 1 & 4.

Output of the transformer is 24VAC @ 240V. Since the number of input windings is cut in half at 120V, this should not change but I have no way to test it. In my unit, this is the yellow pair of wires

One yellow wire goes to bridge rectifier "A". The other goes to the main board (E11) where it passes through a 20A fuse, the returns on an orange wire (E12) to bridge rectifier "B". The output of those rectifiers is 31.5 VDC, and is connected in parallel and fed back to the main board in the upper right hand corner. These are terminals E10 (black) and E9 (red) IIRC.

Basic power supply testing:
- If your unit is wired to terminals 1 & 4, check for 240VAC there. Low or no voltage = line power problem.
- If your unit is wired to terminals 2 & 3, check for 120VAC there. Low or no voltage = line power problem.
- Check the two yellow wires (rectifier 1 & E10for 24VAC. Improper voltage could indicate a transformer problem.
- Pull the fuse and test it with a meter for continuity. These fuses can blow and appear to be good, so visual is not adequate.
- Check E9 & E10 for ~32VDC. Improper voltage could indicate a problem with one of the rectifiers. If so, pull the red wires from each rectifier and test the voltage at that terminal against common black. Both should show 32VDC independently.
- Check black (E10) against the varisistor (RZ3) leg opposite the red wire (E9). It should be ~32VDC. If it is low or zero, suspect the varisistor.

More pics for those interested:
https://picasaweb.google.com/113610792485688531970/Aquarite?authuser=0&feat=directlink

 

[DeadAquaRite]

Also, while the device says the cell current and salt levels are zero, and the generation light is on, I am not sure this is entirely correct. Here's why:

Last week my chlorine was 0. My test said so, and when I was in Leslie's getting the cell tested I had them confirm. The Leslie's guy said add 4# dichlor to get to it to 4ppm. Yes, I should have checked his math, but since he had tested my water and cell I bought a few bags of shock and headed home. Anyway, I added 2# dichlor, 2# trichlor (thought it was dichlor), and later remembered 2 half used pucks in the skimmers (removed). A little math via poolcalculator.com indicates that I should have been at about 15ppm after additions. Test kit showed >3ppm, and test strips (don't laugh, they were a gift) showed > 10ppm. CYA is 50.

A week passes. The sun shines like crazy, I add 3" of water, it rains (+0.5"), and I mow the grass. Lots of things happen to drop chlorine. But I'm still off the scale on chlorine. So I take the standard FC test and dilute it 2:1, 4:1, 6:1, 8:1, and even 12:1 until it starts to measure on scale. As best I can gather, I'm somewhere between 10-15ppm FC.

Am I wrong in thinking that the generator is actually functioning to some degree? Was my original guesstimate that high? Shouldn't I be seeing a notable drop in chlorine?

 

[DeadAquaRite]

Update: New board arrived today, quick swap, all is now well.

For those dealing with this in the future: On the good board the no flow light flashes for ~60 seconds on power up. This happens whether or not the cell is plugged in. During this time the cell current and instant salt level will be 0. At the end of the time our period, there is an audible click of the relay and the instant salt jumps to 9900 and begins falling to it's stable state. In my case it stabilized at 3500ppm and 5.67A on a T15 cell.

This is very different from the "broken" board where no flow light never activated at all. Earlier I suspected the relay might not be engaging and that was preventing the board from coming alive, but I now suspect a logic failure as the "no flow" functions regardless of the relay or cell status.

 

[DeadAquaRite]

Update: New board arrived today, quick swap, all is now well.

Just thought I'd update this thread. Today I noticed 0.0 FC and my power was light out. Quick inspection revealed a cracked CLR. Changed component, all is well. So FWIW that's one PCB and one resistor in 13 months.