Charge voltage discrepancy on a Midnite Classic 150

God knows I've scoured the Midnite fora for information on this but come up with nothing. Anyway, here it goes:
I have two different houses with similar systems. Equipment is otherwise identical. Configuration is not (24v and 48V). The PV panels, batteries and charge controllers are the same. Inverters are both the same make and 3kW but one is 24V, the other 48V. Recently changed one of the charge controllers from an EP Solar iTracer IT6415ND 60A to Midnite Classic 150. Everything appears to work fine and I have to admit, I'm more than impressed with the Chinese controller. Given, I had some misgivings earlier but that's history now. I swapped the controllers round after a month, just for kicks.
Here's the confounder: I live smack in the tropics and it's an oven out here right now. Set absorb to 28.8V to keep my batteries from deteriortating rapidly in the heat (Trojan T105-RE 6V). The classic shows 28.8V at absorb and so does every other piece of equipment. I transplant the Midnite to the 48V system. Double the absorb to 57.6V. A walkover, presumably. How wrong! From that point on, everything I do to make it jump to 57.6V fails. It reaches a peak of 57.0V and NEVER exceeds it. Every equipment tells me that's the exact voltage I'm reaching and not a system quirk.
Funny thing, my inverter can push it all the way up to 57.6V but I start smelling sulfur fumes. Not good for 6 month old batteries. I go back to charging with solar only. I have fairly reasonable grid supply so I never have to charge using generator and cut back on use on days when there's no grid power and the sun isn't out. On a typical day, both systems are at float by midday to 3pm (depending on overnight discharge and presence/absence of grid supply the previous day).
The question: is this behaviour of my 48V system normal or have I gone horribly wrong somewhere (the 24V system works flawlessly). If I have gone wrong, what is the remedy?
I have two different houses with similar systems. Equipment is otherwise identical. Configuration is not (24v and 48V). The PV panels, batteries and charge controllers are the same. Inverters are both the same make and 3kW but one is 24V, the other 48V. Recently changed one of the charge controllers from an EP Solar iTracer IT6415ND 60A to Midnite Classic 150. Everything appears to work fine and I have to admit, I'm more than impressed with the Chinese controller. Given, I had some misgivings earlier but that's history now. I swapped the controllers round after a month, just for kicks.
Here's the confounder: I live smack in the tropics and it's an oven out here right now. Set absorb to 28.8V to keep my batteries from deteriortating rapidly in the heat (Trojan T105-RE 6V). The classic shows 28.8V at absorb and so does every other piece of equipment. I transplant the Midnite to the 48V system. Double the absorb to 57.6V. A walkover, presumably. How wrong! From that point on, everything I do to make it jump to 57.6V fails. It reaches a peak of 57.0V and NEVER exceeds it. Every equipment tells me that's the exact voltage I'm reaching and not a system quirk.
Funny thing, my inverter can push it all the way up to 57.6V but I start smelling sulfur fumes. Not good for 6 month old batteries. I go back to charging with solar only. I have fairly reasonable grid supply so I never have to charge using generator and cut back on use on days when there's no grid power and the sun isn't out. On a typical day, both systems are at float by midday to 3pm (depending on overnight discharge and presence/absence of grid supply the previous day).
The question: is this behaviour of my 48V system normal or have I gone horribly wrong somewhere (the 24V system works flawlessly). If I have gone wrong, what is the remedy?
Semi off-grid
255W Canadian Solar × 12, 200AH 48V US 185 XC2 bank, Victron Bluesolar MPPT 150/85, Victron CCGX, Victron MultiPlus 48V/5kVA/70A inverter (primary system) Victron Phoenix 48V/375VA inverter (backup for critical loads)
300W Yingli × 2, Midnite Brat, 200AH 24V bank (powers DC LED security lights)
Comments
btw, this post shows up as 3 separate posts.
--vtMaps
http://zoneblue.org/cms/page.php?view=off-grid-solar
Sorry for the triple post (I hope the moderator deletes the replicates). Poor internet connectivity when I was making the post. Funny thing is, I clicked just twice. First time I got no repsonse so after about 5 minutes, I tried again and it notified me of admin ratification being required before it goes live.
VoC is 37V and VmP is 29.4V
Array is in three strings of three. 9 panels of 240W to make 2,160W raw, attained on unusually clear and cool mornings. Seen 2,200+W on an occassion or two. Typically about 1,400W on the hottest of days to about 1,800+W on a typical day. Combined VoC varies with the day. Usually between 105 and 110V. VmP depends on SoC but with a discharged battery can drop to 69V but typically is within 72V and 80V. Everything else is as previously stated.
Temperature compensation settings are standard. And are as stated on the panel nametag.
255W Canadian Solar × 12, 200AH 48V US 185 XC2 bank, Victron Bluesolar MPPT 150/85, Victron CCGX, Victron MultiPlus 48V/5kVA/70A inverter (primary system) Victron Phoenix 48V/375VA inverter (backup for critical loads)
300W Yingli × 2, Midnite Brat, 200AH 24V bank (powers DC LED security lights)
The Spam filter is working very well--But I need to figure out how to make it not put long term posters into the moderation queue. It can be done, but I don't have a lot of free time at the moment--Sorry.
-Bill
255W Canadian Solar × 12, 200AH 48V US 185 XC2 bank, Victron Bluesolar MPPT 150/85, Victron CCGX, Victron MultiPlus 48V/5kVA/70A inverter (primary system) Victron Phoenix 48V/375VA inverter (backup for critical loads)
300W Yingli × 2, Midnite Brat, 200AH 24V bank (powers DC LED security lights)
-Bill
--vtMaps
The Trojan guide says 2.35 - 2.45V/cell for absorb. The 56.8V I see at absorb is still sufficient for getting my cells to 2.366 at absorb. Considering I'm in the hot tropics, I believe I'm still okay. Nonetheless, Why am I getting 56.8V when my settings are clearly at 57.6V. I went one step further on two occassions and did not discharge the batteries, letting them go into absorb as early as possible the next morning. It's flattens out the voltage at 56.8V still.
By the way, when I initially set it to 56.8V (my original intention), I was getting 55.9V - 56.1V hence the need to bump it up in the first instance. Thoroughly baffling
255W Canadian Solar × 12, 200AH 48V US 185 XC2 bank, Victron Bluesolar MPPT 150/85, Victron CCGX, Victron MultiPlus 48V/5kVA/70A inverter (primary system) Victron Phoenix 48V/375VA inverter (backup for critical loads)
300W Yingli × 2, Midnite Brat, 200AH 24V bank (powers DC LED security lights)
Some CCs display the uncorrected voltage while delivering the corrected voltage. I think that yours is showing you what actual voltage it is delivering based on applying the temperature correction to the set point.
--vtMaps
255W Canadian Solar × 12, 200AH 48V US 185 XC2 bank, Victron Bluesolar MPPT 150/85, Victron CCGX, Victron MultiPlus 48V/5kVA/70A inverter (primary system) Victron Phoenix 48V/375VA inverter (backup for critical loads)
300W Yingli × 2, Midnite Brat, 200AH 24V bank (powers DC LED security lights)
--vtMaps
(1) absorb charge duration (I use 3 hours and have a daily DoD between 20% and 40%)?
(2) absorb voltage limits per cell (manual states 2.35V - 2.45V but I'm thinking bigger batteries with thicker plates like the L16-RE or the IND17 can't possibly share the same absorb voltages as the comparatively tiny T105-RE even though the manuals state the same voltage limits for them all)?
255W Canadian Solar × 12, 200AH 48V US 185 XC2 bank, Victron Bluesolar MPPT 150/85, Victron CCGX, Victron MultiPlus 48V/5kVA/70A inverter (primary system) Victron Phoenix 48V/375VA inverter (backup for critical loads)
300W Yingli × 2, Midnite Brat, 200AH 24V bank (powers DC LED security lights)
You are fortunate to have flooded batteries. That means you can check your SG. Your goal should be to discover the minimum absorb voltage and absorb time that will get the batteries charged. If your SG is good, then try lowering the absorb voltage by 0.1 or 0.2 volts. See if the SG suffers. If not, then lower it some more. Find out what works for your batteries in your environment.
--vtMaps
255W Canadian Solar × 12, 200AH 48V US 185 XC2 bank, Victron Bluesolar MPPT 150/85, Victron CCGX, Victron MultiPlus 48V/5kVA/70A inverter (primary system) Victron Phoenix 48V/375VA inverter (backup for critical loads)
300W Yingli × 2, Midnite Brat, 200AH 24V bank (powers DC LED security lights)
The important thing is that the cells are actually gettting to 100% regularly. A little over for FLA is ok, a little under is not. (Hard sulphation and difficult to reverse capacity walkdown).To some extent a really decent float after the absorb will compensate for a prematurely terminated absorb. But again as VT said SG is the acid test. Pun is included free of charge.
http://zoneblue.org/cms/page.php?view=off-grid-solar
Pun ...... free of charge
That's a double pun, if I ever saw one! I'll whip out my hydrometer and start taking serial SG measurements next week. I'll give feedback as appropriate (mind you, that's checking SG measurements twice daily for 24 cells).
255W Canadian Solar × 12, 200AH 48V US 185 XC2 bank, Victron Bluesolar MPPT 150/85, Victron CCGX, Victron MultiPlus 48V/5kVA/70A inverter (primary system) Victron Phoenix 48V/375VA inverter (backup for critical loads)
300W Yingli × 2, Midnite Brat, 200AH 24V bank (powers DC LED security lights)
boB
Have you established a "standard" - that is if you are relying solely on the controllers display voltage, have you doublechecked that against your own voltmeter measured directly across the battery terminals?
Slight discrepencies like 0.3, 0.6v differences or so can also mean that the measurement is being taken after a diode voltage drop. Or, like mentioned above, the display value is slightly different from the real values.
I'd vet out the accuracy of BOTH controllers with your own measurements at the terminals. Even a low-end Fluke like a 11xx series should do fine, min-max hold and all that. It doesn't have to be Fluke of course, and I don't want to turn this into a multimeter thread - but consider using something as a standard against which all others are compared to.