ZXD2400 modification guild 3Kw 0-60V 0-50A

[Burgerman]

I have another still to do. And 2 more BT conversions to fit. And still stuck on bed mostly... And a BM3 to fix. And my ex gf single, fit younger sister is coming for 14 days from belgrade, and I have to chat here up and take her to the pub too... So much to do.

[shirley_hkg]

I might get one or two Shirley How much in GBP£'s including shipping?

CNY￥up a bit , PSU without new outfit @$1,220.00HKD .

[Scooterman]

Arrived today, thank you very much Shirley, with a bonus of no customs charges.

Same here for the cells Shirley, I'm so pleased about that. .

[steves1977uk]

I might get one or two Shirley How much in GBP£'s including shipping?

CNY￥up a bit , PSU without new outfit @$1,220.00HKD .

How much is it with the new outfit already done? Think Shirley's again!

Steve

[shirley_hkg]

Have to check .

[shirley_hkg]

CNY￥up a bit , PSU without new outfit @$1,220.00HKD .

How much is it with the new outfit already done? Think Shirley's again!

Steve

$1750 HKD with everything installed and shipping included .

[steves1977uk]

Thanks Shirley So roughly about £172 which is a very good price!

Steve

[Gnomatic]

That is a good price. Just a bit more than I paid buying my 2nd PSU + fancy case separately.

I haven't quite finished installing my second PSU in its aluminum case yet. Just recently cut the notch on the rear panel with a dremel to make room for the Anderson. Thin, noisy stock fans have been replaced by 25mm thick fans that still move air with far less noise. Its coming along.

[shirley_hkg]

Gonna get one for Scooterman.

Anyone else wants your ZXD2400 , please let me have your PM, and tell me which one you want .

[Williamclark77]

PM sent. I don't need the fancy suit. I just want another for backup and to keep in my shop to use on other projects. Thanks Shirley!

[shirley_hkg]

Warning :
Don't try to modify it by yourself to save that labour fee. Won't do a third one .

It just don't justify .

[steves1977uk]

PM sent Shirley!

Steve

[Scooterman]

Gonna get one for Scooterman.

.

[LROBBINS]

FAN NOISE
Someone here (but I don't remember in which thread) suggested I look for a low current, 25 mm thick fan to replace the noise front fan. He (or she) suggested a particular one, but the price here in Italy was out of reason, so I searched e-bay and came across this one that has its own temperature sensor https://www.ebay.it/itm/VENTOLA-SILENZIOSA-per-CASE-PC-80mm-GELID-SILENT-TC-8-FAN-80-3-PIN-GOMMINI-12V/191374336023?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2649. Price in the U.S. is a lot less, but still this is reasonable, so I bought one to see if the built in thermal control can overcome the lousy programming of the fans.

I put a piece of fiberglass insect screen in place of the perforated plastic to improve airflow, but installing the new fan did require one "kludge". The power supply monitors current through the fan to tell if it's been disconnected, and this quiet fan draws too little current to let the power supply actually turn on. A little experimenting with some resistors showed that one actually needs ca. 0.5A (which is what the noise maker draws) to get the PS to actually turn on - a 310 ohm, 1/2 Watt resistor in parallel with the fan does the trick, 410 ohm does not - still too little current. The odd values are because I actually paralleled two 620 or 820 ohm 1/4 watt resistors because that's what I had. I then attached the temp sensor to one side of the heat sink, and a thermocouple to the other, buttoned it up and started testing. With 22oC ambient temperature and 24V 4.8 Amp load (the biggest dummy load I have), temp barely rises to 25oC and the new fan is turning quite slowly. It does speed up, but is still very quiet, if I hit the fan's temp sensor with hot air. I do wish I had one of Burgerman's room-heating dummy loads for testing at high current, but so far so good. As heat transfer improves with source vs. sink temperature difference, this should still run pretty cool at full current unless, perhaps, the ambient temp is very high (our member from Viet Nam might have some trouble with that).

I've now ordered a second one of these fans to replace the rear one. If it stays cools with that, the PS will be close to silent at low current, and not at all annoying even at high output.

[Gnomatic]

Good information.

The stock fans are RATED to draw 0.27A, but I didn't actually measure. The 25mm thick Rosewill fans I used to replace the stock ones are only rated to draw 0.18A. And the PSU works fine with them. Again I did not measure the current they actually draw.

I'll remember your little trick with the resistor if my second PSU doesn't turn on when install the low current SilenX fans I purchased for it.

[LROBBINS]

That's odd as the stock fans in my unit say 0.5A. Probably used different fans in different lots. I'll bet the current sensing is trimmable, but I have no idea where the trimmers (probably separate ones for front and rear) are on that board. I probably could figure that out if I could read the Chinese in the images in the instructions, but I can't.

[Gnomatic]

I never bothered to check the stock fans on my second PSU before, I just assumed they were the same that came with my first. But now that I checked, they are indeed different!

My first PSU came with this 0.27A fan.



My second unit used this 0.40A fan.



So yeah they are using different fans in these. Interesting.

[Burgerman]

My first PSU came with this 0.27A fan.

Both my power supplies have the same. Still noisy. But the resistor cable extentions slowed them a lot. Still cold to the touch at 50A...

[shirley_hkg]

Another three ZXD2400 arrived .

[steves1977uk]

Looking forward to receiving my new fancy-cased ZXD2400

Thanks Shirley

Steve

[LROBBINS]

The reason that the supply detected your 0.18A fan, but not my 0.14A fan, is that the 0.14A is at sensor temp >= 40o and RPM = 2000. At cold startup this fan is throttled to 900 RPM (which means ca. 5.5V) and though they don't say what the current draw is, it must be a lot lower or there's a lot of PWM "off" time that tells the PS "no fan connected". At 5.5V the 310 ohm resistor drains 0.02A (0.1 W) and that's enough to say "fan is connected". At 12V the shunting resistor adds a drain of 0.04A (0.46W). So this kludge is probably only needed because the fan has its own temp control. BTW, the PS still has it's own temp sensor and has an over-temp shutdown at 60o heatsink temperature. It controls fan speed at low current (EXCEPT in charge mode float when it just doesn't) by switching in a 2:1 voltage divider made up of two 51 ohm resistors. So, half speed cold, full speed hot (or in float) and nothing in between.

[Gnomatic]

More good info, thanks for sharing. Did the new fans quiet down the PSU as much as you expected?

[LROBBINS]

I was so pleased with the first one that I've bought and installed a second. So far, at 24V 5A, the unit stays so cool that they hardly speed up at all and I can't hear it at all over normal background noise. The aluminum case in contact with the heatsink on the left side and right about where I put the temp sensors does get barely warm though I haven't actually measured the temp yet. There's probably a good bit of heat being gotten rid of through the case, especially since I put a thin layer of arctic silver on the heatsink. Next step will be to wire up some more auto lamps as dummy loads, but I'll probably only get to about 24V 15A that way, and will probably nearly blind myself in the process.

[Burgerman]

Cant you just charge a battery at 50A for a few mins (anderson)?

[LROBBINS]

Yes, I can and will, but need to clear some space of other stuff before I can do so. I do have a pair of old Odysseys in the garage and they haven't been charged in a while so should draw a fair amount for at least a few minutes. If they go to CV too quickly, I'll have to discharge them some and try again. Can't decide whether it would be better to charge 2 in parallel or 2 in series - watts through the power supply would be the same, but I wonder whether higher voltage or higher current is more stressful.

[Burgerman]

Amps makes heat. So 50A at 12V (14.4 TO 14.7V) is harder on the supply and should make it twice as hot. But on a bench you can do 24V AND 50A. But you will need todischarge battery to about 12V or lower first. In any case thats where we would be using it to charge lead or rather thats what I do as both get topped up equally.

[LROBBINS]

I bought a second one of those extra quiet, temperature-controlled fans, replaced the rear one and did some testing. I'll show the results below, but will try to tell you the take home right at the top.

Even a pair of these going full speed are very, very quiet. I think you could put the ps next to your pillow and hardly notice the sound.

With some tweaking of the EEPROM temperature settings, cooling is adequate for lead acid batteries with ambient temperatures <= 23o. High summer temperatures will probably cause some current limiting.

Cooling may not be adequate for full 50A current with LiFePO4 batteries because current flow stays high during Vcc. Current may drop to 40A for extended periods, and you might want a stronger, noisier rear fan.

Now, the actual data. All testing was done with a pair of very old Odyssey P1500 batteries that probably have less than 50% of their original capacity. The temp sensor for the front fan was put on the top (left side) heat sink, and that for the rear fan was put between the fins on the rear heat sink at the hottest spot - close to the MOSFETS. I started testing with the two batteries in parallel and with the factory settings for when fan power goes from 6V to 12V and for the 5 overtemp stages:

For most of the time before current drops during Vcc temps reach stage 1 and current is throttled to 40A.

The quiet fans getting only 6V turn very slowly even after their temp sensors warm up. So my first tweak was to make sure they always get 12V by lowering EEPROM 59 to its minimum of 18o. I also raised the temps for stage 1 and stage 2 limits slightly:

With each test, I got better at recording relevant information; here I monitored case temp with a thermocouple over the hot spot of the rear heatskink. It's better - at least it ran at 50A for a few minutes before going to level 1 overtemp - and it reached Vcc sooner so naturally started dropping current flow. But this is not good enough.

So, raised stage 1 and 2 temps a bit more (it never goes to stage 2 in any case, so irrelevant, but I just figured that I might as well have all 5 steps):

Again, a bit better with less current throttling and transition to solid Vcc at just over 15 minutes. Further tweaking, and data for series connected batteries will be in the next message.

[LROBBINS]

Continuing from where I left off in the last message, here are the results for the two batteries wired in series and with a slight further increase in overtemp setting:

With the two batteries in series, and 50A current through the pair rather than divided between the two, the transition to Vcc comes quite quickly and there's never any throttling of the output.

I've calculated the AHr returned and it's quite obvious that these batteries are well past their prime, though they probably do have somewhat more capacity than the 41% measured return. For one thing, the starting voltage (22V = nearly completely discharged) was measured only 15' after draining them so under-estimated. Had I left them disconnected longer before charging, the apparent SOC would not have been near 0, but perhaps 10 or even 20%. Nevertheless, these batteries are far from healthy, and with good ones the time at 50A would certainly have been longer. Nevertheless, I think that these almost-silent fans are giving enough airflow for recharging lead acids.

With lithium, however, it would probably spend a lot of time at stage 1 (40 Amp output) with these fans. It would probably be safe to make the temperature transitions a bit higher AS LONG AS ONE DOESN"T GO OVER 65o for shutting output to 0 Amps. Alternatively, and probably my choice, would be to keep the temperatures the same, but have stage 1 limit current to 45A (instead of 40A), stage 2 drop to 40A, and stage 3 (which it will likely never enter) drop to 30 A. 65o is not a very high temp even for consumer-level components, but do remember that this is not being measured at the chips but at the heatsink, so some of the MOSFETS could get a good bit hotter, though I think it would take quite a bit more than 65o at the heatsink to reach a die temperature >125o (the lowest temp rating commonly available).

OR, use a rear fan that's somewhere between the 0.13A, 27 CFM (at max speed) of this fan and the hyper noisy 0.5A fan my unit came with. If doing that, I'd keep the temp controlled front fan, but just use a constant speed fan at the rear.

[Burgerman]

A complete research paper no less! But interesting. I kept the original rear fan, and a slower unknown front fan with resistor. From my draw. And no filter thingy. Seems to say at full 50A for as long as it needs. I tested with that dump thingly. Excelent room heater. Maybe not as quiet as yours though. But I am not hyper sensitive to noise. A note. That filter in the front does reduce more than half the airflow. It seems to stop 2/3rds of it... Unscientific guestimate testing. But also makes it quieter!

I think, a 1 inch thick fan at the rear only may work well. Just a resister on the front.

[shirley_hkg]

Sat in my kitchen. Single. No kids. 50kg, architect who designed a few hotels and stuff for the socci olympics recently.

Too young for you BM, send her to 51 Deanside.

@Scooterman . What you've been longing for will soon knocking at the door .

Enjoy it .