New ESC48 Salt Cell Underproducing - SOLVED!

[jzelno]

First, some history: We have a 23K gallon pool in southern CA that came with the house we bought a year ago. The SWG is an ESC48, and it was not working when we moved in. After getting the pool chemistry right, and checking things out, we found that the cell was clogged with calcium. I cleaned the cell in an acid bath, probably around 1:6 - 1:8 acid:water. I had to soak it overnight to clean all the deposits off, but afterward, it worked beautifully, producing Chlorine with about a 60% SWG on-time for a 10-14 hour pump time (I don't recall how long we ran the pump back in August/September 2013). A few weeks later, it quit working. Cleaning didn't help. So, we switched to regular chlorine (Leslie's Chlor-Brite) for the winter.

Fast forward to late April 2014. We bit the bullet and bought a $400 CompuPool GRC/EM replacement cell, calculating we would save at least that much in Chlorine over the next few years. We plugged it in, turned it on, and the ESC control panel was very happy... it said that it was at +/- 100, and all green lights. But, we couldn't keep the chlorine level up, even with the dial set to 100% SWG time. We tested the pool chemistry, which showed high alkalinity (~180), slightly high pH (~7.8), high calcium hardness (~500... this is normal for our area... the tap water isn't much softer, so it doesn't do any good to drain and refill), high phosphates (~200). CYA was a little high at 100, and the salt level was good at 5000 (everyone tells me this is high, but the ESC48 manual recommends 4500-5000, and higher isn't a problem for Chlorine production). We also had an algae problem. We added muriatic acid to the deep end overnight to bring down the alkalinity, which is now around 120. The pH had also improved to 7.6, and CYA was still good at 80. CH didn't change, as expected. First, we treated the phosphorus with Pool Magic + Phos Free, then a few days later, we shocked the pool with Chlor-Brite, and brought the chlorine level up over 10ppm, which got rid of all the algae (I still don't know the new phosphate level). We figured we had it all worked out.

Problem is, we still aren't maintaining a Chlorine level, even with a 10.5 hour filter time, and 100% SWG time. I know the cell works, because I put it in a 5gal bucket of pool water, and after 2 minutes, the 5gal of water was over 20ppm of Chlorine. Also, the other night, we had just the spa running for a few hours, and the Chlorine in the spa spiked up to around 5ppm.

Note that we also have a solar water heater for the pool, and the pool water is up around 85 degrees F.

Any ideas on what can be going wrong? Has anyone else had underperforming ESC48 cells? What can I try, what can I check?

Thank you.

 

[jblizzle]

Re: New ESC48 Salt Cell Underproducing

Welcome to TFP!!!

While I only briefly skimmed, I am guessing that you still have algae in the pool that is consuming the FC as fast as the SWG in making it. You need to follow the ShockLevelAndMAINTAIN Process.

 

[jzelno]

Re: New ESC48 Salt Cell Underproducing

Thanks for the quick reply, jbizzle. We did shock the pool for several days last week, maintaining a chlorine level above 5ppm (in some cases above 10ppm) for at least 5 days. We have not had any visible algae in the past week (we've been maintaining Chlorine level between 1-4 with Chlor-Brite and liquid Chlorine since then). I suppose it could still be algae, but I would have expected to see it...

 

[Richard320]

Re: New ESC48 Salt Cell Underproducing

You're at a stalemate. The algae is eating the chlorine as fast as it can generate it. Follow the link Jason posted above ^^ and SLAM it.All that Chlor-brite (Dichlor) that you fed the pool has probably driven the CYA to astronomical levels, which could make it impractical to SLAM without replacing some water first. Start with a proper CYA test. If it reads over 90, repeat it using the dilution method.

PS. I see you replied again. Maintaining 4-5 FC is useless at 80+ CYA. Have you seen the chart in Pool School? http://www.troublefreepool.com/content/128-chlorine-cya-chart-slam-shock

 

[jzelno]

Re: New ESC48 Salt Cell Underproducing

I'll get a water test done, and look into SLAMing. After I get CYA to a useful level for SLAMing, what's the cheapest way to add chlorine? Sounds like I should stay away from Chlor-brite for this (will raise CYA), and I'd rather not use calcium hypochlorite (with a "locally" high CH, I'd rather stay away from products containing Calcium). Is there anything cheaper-per-potency than sodium hypochlorite?

Also, just for clarification... just because I don't SEE any algae, doesn't mean it's not there?

And for the record... to attempt to quantify the "algae problem" I mentioned earlier... the water was clear, and the walls/floor of the pool (gunite) were mostly clean, just a few large (maybe up to 1 foot diameter), but thin, spots in various locations. I would say maybe 5% of the underwater surfaces had some green algae growing (no other colors), but not so much that you couldn't see the color of the pool wall through it. It all brushed off very easily.

Thanks again.

 

[jblizzle]

Re: New ESC48 Salt Cell Underproducing

Don't "get a water test done". Invest in one of the Recommended Test Kits.

Generally using liquid chlorine / bleach is the cheapest per ppm of FC out there ... and there are no side-effects. Use this for the SLAM and for any supplemental FC you need in the future.

The algae does not need to be visible for it to be there and consuming you FC. You can test this by Performing the Overnight Chlorine Loss Test, but you need the FAS-DPD chlorine test which is included in the recommended test kits to do this and the SLAM properly.

 

[jzelno]

Re: New ESC48 Salt Cell Underproducing

OK... we drained the pool to lower the CYA, and it's now at 70. The salt level was measured at 5500, but I don't really believe that it's that high (based on what it was before, how much water we drained, and how much salt we added). At any rate, the SWG is not complaining about it. We also SLAMed the pool, and now pass an overnight chlorine loss test (we lost about 0.5 ppm the other night, all of the chlorine is free, and the water is crystal clear).

However, the salt cell is still not able to keep the chlorine level up, with 10.5 hours of on time. So, back to my original question... has anyone had a new replacement salt cell that did not produce enough chlorine, and what did you do about it? What are the chances that it's defective to the point that it underperforms, but not totally broken (since it can produce chlorine in a bucket of pool water)?

Thanks again.

 

[JasonLion]

Re: New ESC48 Salt Cell Underproducing

The odds of a SWG being broken in the way you describe are just about zero. In just about every instance they either work, or they don't work at all.

Is there any chance that you have the SWG setup for 230 volts, but are actually supplying 115 volts?

Have you done a visual examination of the cell plates to make sure there isn't any calcium scaling on the cell?

What are you using to measure the CYA level?

 

[jzelno]

Re: New ESC48 Salt Cell Underproducing

The SWG was working fine last year with a different cell, which eventually failed. No electrical changes were made, so I have no reason to believe that the supply voltage is wrong.

I have looked at the cell plates, and there's no sign of any scaling.

CYA was most recently measured at Leslies (black dot in the vial method), but we also have the aquacheck strips that suggest CYA is in the 60 range.

 

[JasonLion]

Re: New ESC48 Salt Cell Underproducing

I am included to think the CYA level is actually much lower than you think it is. CYA tests are by least reliable for test strips and pool stores. We can get a general sense of the CYA level to double check this theory by measuring the starting and ending FC level over the course of a sunny day. If CYA is low I would expect you to lose perhaps 2/3rds of the starting FC level (with the SWG off). If you lose less than 1/2 then I am wrong and your CYA results are probably right.

 

[jzelno]

Re: New ESC48 Salt Cell Underproducing

According to other information on this site, the CYA measurement seems critical to know, so that you can know what your target FC levels are, so that you can know what the proper shock level is, etc. So, then, there MUST be a way to accurately measure CYA levels. If it's not the strips, and if it's not the black-dot-in-a-vial method by the pool store, what is it? If the "see how much burns off during the day" method suggests CYA is lower than measured, how am I ever going to find out what the starting level is in order to know how much to add without making it go "astronomical"?

 

[jblizzle]

Re: New ESC48 Salt Cell Underproducing

The best CYA test is the black dot in tube test performed by you in the correct light. The stores are really bad at this test for some reason, likely some due to the lighting.

 

[jzelno]

Re: New ESC48 Salt Cell Underproducing

Back to this one again...

I just got a new warranty replacement salt cell, and installed it. Everything seemed fine... for a few minutes. I checked the water being injected back into the pool, expecting to see high chlorine levels, but it was just at the same level as the rest of the pool water. So then, I went and pulled one of the wires to the cell, expecting the readout to drop to 0... but it stayed at 100. I then went to get my multi-meter, and measured 0V at the cell terminals... AC and DC. I also noticed that the controller readout was dimmer than usual. I thought maybe this was just due to the bright sunlight, or maybe the controller dims the LEDs after a while, but now it's apparent that the controller has some problem. None of the fault LEDs are lit, the whole display looks normal (showing 100%, 2 green status LEDs, no standby or flow LEDs), just dim.

Does anyone have a schematic for the ESC48 controller? Or any ideas on what might be wrong, or what to test?

Thanks.

 

[jzelno]

Re: New ESC48 Salt Cell Underproducing

So far, I've traced the problem to the fact that the circuit board 5V regulator will eventually cut out, due to it's input voltage (looks like 12V nominal) drops to around 5V. This causes the on-board processor to sit in reset. This may be heat related, as the problem happens faster when the control unit is fully closed up and during the day. It seems to take longer to cutout if the unit is open and in the shade or at night.

If anyone has any details on this circuit that would help, particularly what part generates the 12V, that would be most helpful. Thank you.

 

[jzelno]

Solved the problem. In a nutshell, one of the components on the circuit board was causing the controller to get stuck in a reset state, such that it would stop generating the required voltage to the cell. The explanation below gets pretty geeky into the electronics details, but included in case anyone else experiences similar problems with any of the ESC series generators.

The ESC48 is controlled by a microprocessor that requires a 5V power supply. The 5V supply is monitored to make sure it stays at or around 5V. The 5V is generated by a switching DC-DC converter that expects a nominal ~16V input, but seems to operate fine down to around 8-10V. The 16V input is generated from ~48V by a series of 3 reverse biased zener diodes, totaling just over 30V. The DC-DC converter draws less than 0.2A, so these diodes are generating somewhere around 10W of heat. Just before the 48V input gets to these series diodes, there is a PTC resettable fuse, which appears to have a nominal, room temperature trip point of 0.3A (if I interpret the markings correctly). This unit is quite old (I have no idea how old, since it came with the house when we bought it over a year ago), so its possible that the wear and tear on the PTC fuse decreased it's performance over the years.

For a little background on PTC fuses, they're basically temperature dependent resistors, where if they stay cool enough, the resistance is fairly low. As the current through them increases, their temperature also increases, and the material reaches a point where the resistance increases exponentially. This has the effect of creating a pseudo open circuit, cutting off the current flow. Eventually the PTC cools off to the point that the resistance decreases back to normal, and current is allowed to flow again.

My controller is mounted to a block wall, and it does have a small air gap behind it to allow airflow to cool the unit. I don't think that this air gap is big enough however. But recall from above, the unit did work fine at the end of last summer. In addition to that, this wall faces East, and gets beat on by the southern CA sun through about 1pm, when the sun is behind the wall, and the controller is in the shade of the wall. However, being a block wall, it's already absorbed a lot of heat, and will continue to radiate. In addition to that, the controller box has a large transformer that puts off a lot of heat. Don't forget about the 10W of diode heat.

Now, it just so happens that the original designers placed the PTC fuse (remember... temperature dependent resistor/fuse), within a few inches of those diodes. I found that when I left the circuit board exposed (i.e. out of the box), the unit ran fine all day, even though the circuit board was in full sun. When the unit was in the "failed" state (after the circuit board was closed up in the box), I measured the voltage drop across the PTC fuse up over 10V. This, along with the ~30V or more across the diodes, starved the DC-DC converter, such that it was only putting out around 3V. The voltage monitor would then issue a reset to the processor, preventing it from doing anything. This would persist all day, since the PTC would never cool down enough, and the processor voltage would never return to 5V.

The fix was to remove the PTC fuse, and replace it with a 250mA slow blow fuse in an inline fuse holder. The fuse holder was routed to a small compartment at the bottom of the box that was a little more accessible than the circuit board, but still protected from the elements. The unit has been working fine and generating Chlorine for 3 days straight now, with no problems!