Peltier element and power supply needed.

[wwdwgs]

Hello, everyone.
first, I'm not an electrician, therefore I'd like to get answers in plain terms.
Problem: solid-state dehumidifier stopped performing and stopped extracting/collecting moisture.

Construction: two 30mm Peltier units (probably rated 12V and 3-4A), one 120mm fan (12v, ?A). a few led's, main switch and two contact switches (onw of them shut-of power).

Observations: power input 12V 5A, power on the board - 2.8-2.65V.
LED's are powered thru 20KOhm resistors (seems too much).
Fan has a hard time starting after power is on. Flipping power switch a few times pushes fan and it rotates.
Fan (finally) has burnt and smelt like burnt plastic.
heat sinks on "sold" side are barely cold - the temp difference between hot and cold sink is about 5-11 F.
Hooking up 9V battery of one Peltier unit resulted in greater temp. difference.
Power supply doesn't supply constant V and it varies from 11.5 to 12.2V oscillating every second.

Experiments: I measured Voltage at main switch - 12+_V, measured at another contact switch - 2.65V.
Bypassed contact switches and install regular 2pole rocker switch - voltage is now 12V
Bypassed LED's and connected 2 Peltier's and fan - nothing works well, fan doesn't spin, heat sinks are barely colder.
hooked up fan alone - works fine. Hooked up one Pletier unit - works well - cold and "sweaty".

My conclusion - power supply is weak. It probably needs to be about 9A or more (assuming Peltier units are 4A each and fan - 1A).

Remedy:  get more powerful power supply with 10-15A

Am I correct?

Thanks in advance.

[Ian.M]

No.  It stopped working not it never worked, therefore 12V, 5A must have been enough to run it originally unless the nameplate rating is a pure lie.  Figure 0.3A for the fan, and the Peltier modules must either be 6V <4.7A. or thereabouts, in series or 12V <2.35A in parallel.   Repeatedly thermally cycled Pelter modules tend to go high resistance so the best bet is to test them individually on your bench supply and see what current they draw, and how much temperature differential they maintain, while letting the old PSU power the fan.  If you don't have a bench supply, 12V from a PC power supply or a car battery (not running), with a 10A fuse in series for safety will do to test them, individually if originally wired in parallel and together in series if originally wired that way.   If they are in series, note the midpoint voltage.   If they draw very different currents, or one is far less effective, or for a series pair if the midpoint voltage is far from 6V, you've got at least one bad module.

Once you've found out out if they are OK and how much current they actually draw, you can spec a power supply, on the basis that one from a recognizable 1st world brand manufacturer will probably survive running at 80% of its nominal rating (even if its actually made for them in China) but an off-brand 'Chinesium' one is unlikely to last running over 60%.  Derate both further if its got to run 24/7 or in a hot environment.

If the Peltier Elements aren't OK you'll probably need to replace the pair, preferably with ones you can find a datasheet for.   Caution: if the new modules are more powerful it may ice up on the cold side.  This is self-perpetuating as the ice blocks the air flow and significantly insulates the cold plate.   If your unit has a thermistor or other temperature sensor in contact with the cold plate it may already have an anti-icing circuit.  If not, you'll need to add one - a temperature sensing relay board set to activate just above freezing will do the job, wired to cut power to the elements, so residual heat on the hot side can warm the cold side enough to melt any ice that's starting to form.  It should be set for a fairly wide hysteresis range (at least 5 deg C) so it doesn't short-cycle which would be very hard on the elements.

I'd also want to replace the cooked fan  . . .

It *MAY* be cheaper to simply get another complete dehumidifier as good PSUs aren't cheap, then either keep this one for spares or until you can scrounge up the parts cheap to fix it as a backup unit.

[wwdwgs]

I measured amperage. Results as follows:
only Peltier #1 = drawing 2.3A (supply 2.8A, hooked up after peltier = 0.44A) Voltage dropped from 12.2V to 2.8V
only Peltier #2 = same.
fan = draws 1.8A (sounds high!) separate power 1.9A (9V), drops to 0.07A.
temp. difference after 5-6 minutes - 5.5F (hot 80F, cold 75.5F)

[Ian.M]

I suspect the fan failure let the modules cook themselves to death and they are internally shorted due to the low temperature solder connecting the Bismuth Telluride pellets melting, hence the low voltage with relatively high current.  The max. rated operating temperature for consumer Peltier modules can be as low as 100 deg C!

The fan ever pulling 1.8A, except possibly transiently for a fraction of a second during startup is also a bad sign.

It also seems like the PSU has cooked itself till nearly dead as well.  I suspect its secondary side electrolytic capacitors have degraded to the point its output ripple has increased to a significant proportion of its output voltage, and also its pulsing because the primary side. electrolytics are also bad.  That, you can check - load it up to approx. 80% of its rated amps with a high power resistive load (power resistors if you've got suitable ones, or car bulbs), and if it can't deliver a steady voltage of at least 90% of its nominal output, its bad for sure.

Odds are: "Its dead Jim!" 

[wwdwgs]

Hm. Capacitors...I'll open power supply and check caps. Now I can recall a similar problem with capacitors when my monitor started to flicker and turn itself on and off - the capacitors were "exploded" and, when I ordered the new ones, I checked specs for failed ones. The old ones were rated at 500 hours.... I bought new caps with 5000-hour rating.
I'll post back after dealing with power supply.
Thanks.

+++Update+++
"secondary" capacitors are bad. There are two of them 1000Uf, 25V. i don't know hour rating, but assume it to be 500 hours.

[perieanuo]

I measured amperage. Results as follows:
only Peltier #1 = drawing 2.3A (supply 2.8A, hooked up after peltier = 0.44A) Voltage dropped from 12.2V to 2.8V
only Peltier #2 = same.
fan = draws 1.8A (sounds high!) separate power 1.9A (9V), drops to 0.07A.
temp. difference after 5-6 minutes - 5.5F (hot 80F, cold 75.5F)

the fan 1.8A, that's only because your 12V drops when the fan starts
your supply is a joke, test with a real one, at least 9V one. a good one
or a car battery
the fan isn't a 20-25W one, so it should take like Ian said 300mA or so
obvioud the peltiers are ok, fan almost sure ok, ps bad
bon courage

[wwdwgs]

I've already looked at caps on Newark.com. 2000 hours at 105F would do, I think. But.... thinking about power supply output... only 5A to power 2 Peltier units and a fan.. just barely enough.
I will also entertain an idea of purchasing a stronger power supply, perhaps 8-10A.

[Ian.M]

You are putting the cart before the horse looking at getting PSU parts unless you are 100% certain your Peltiers are good, because you have tested them on a 'stiff' 12V supply fused or otherwise limited at 10A.  AFAIK you haven't done that . . .

I believe the Peltiers are bad - partially shorted internally.  You said the voltage dropped to 2.8V, at a current of 2.3A, powering one Peltier.   That means the Peltier is going to initially draw something like 9.8A at 12V, and the two together just short of 20A, which is patently ridiculous as its extremely improbable that a consumer dehumidifier would contain 120W Peltiers only 30mm square, and impossible that it would have two 120W Peltiers used to their full capacity with only a 60W nameplate rating!   Therefore your Peltiers are dead, deceased, goneski, pushing up daisies, or even 'pining for the fjords!'

Normally the current would drop some from its initial value as the temperature differential increased and the resulting opposing Seeback EMF built up, so originally that 5A PSU probably had a bit of 'headroom' after it had been running a while, though if the design had been Muntzed, it may have relied on it current limiting during startup

[perieanuo]

Ian, you missed something: he said the ps output dropped drastically (2.8V) when peltier was powered
here, you suspect peltier is bad? no way, first the ps should work, then the loads will be validated
no disrespect here

[Ian.M]

IMHO The Peltiers are pulling far too much current at far too low a voltage to be good.  They certainly aren't constant current devices.  e.g. see this TES1-12074 34W 30mm Peltier module datasheet which is fairly typical of the breed.  On page 4 it has an I vs V graph, which shows a straight line through the origin for zero temperature differential, so its purely resistive. 

If the PSU is good, its in foldback limiting.  However as he described its output as pulsing:

Power supply doesn't supply constant V and it varies from 11.5 to 12.2V oscillating every second.

I'd bet its gone bad as well. 

Also would you trust a fan that smells "burnt"?

[wwdwgs]

update:
I hooked up new fan separately with 12V battery. works fine.
I hooked up existing Peltier unit with 12V 5A battery - temp difference - 29F (hot 80F, cold 51F)
...another existing Peltier unit - temp. difference 22F (hot 79F, cold 54F)
Hooked up both Peltier units in parallel - temp diff - 34F  (hot 88F, cold 54F)
One Peltier unit is less effective that the other. They are tec1-12702 manuf. date 04/2019). Peltier units work, condensation forms on the cold side heatsink.

conclusion - replace capacitors in power supply (1000Uf, 25V) or replace power supply with a stronger one, ie 8A of 10A.   

Thank you all of you, who helped me.

[Ian.M]

The fact the two Peltier modules are not operating equally well indicates at least one is damaged.  Can they reliably get below the dew point on the cold side? If not the dehumidifier isn't going to work.   You said the cold side is getting down to 54 deg F.   What was the ambient temperature at that time (if you didn't note it, even an approximate temperature is better than nothing).   Also what is the typical relative humidity (R.H.) of the space you want to dehumidify, and what's the highest R.H you can tolerate after dehumidifying. 

[wwdwgs]

IanM.
Ambient temp was about 80-82F. Humidity - I don't know, over 60-70-80%, I guess.... I was raining for a few days on East Coast. This dehumidifier works. Perhaps, not as good as a new one. During last test, cold heatsinks did get "sweaty" and dehumidifier, re-assembled, does extract water - not much - about 1/8 of a cup within 3 hours or so.

[Ian.M]

Rather than paying higher power bills for the foreseeable future, it may be worth looking for a replacement Peltier and swapping out the worse of the two modules.  Chinese 'off-brand' 30mm Peltier modules are reasonably cheap.  If you are good at scrounging stuff its also worth keeping an eye out for 12V fridges and coolboxes on trash pickup day, which often have salvageable Peltiers.

[perieanuo]

IanM.
Ambient temp was about 80-82F. Humidity - I don't know, over 60-70-80%, I guess.... I was raining for a few days on East Coast. This dehumidifier works. Perhaps, not as good as a new one. During last test, cold heatsinks did get "sweaty" and dehumidifier, re-assembled, does extract water - not much - about 1/8 of a cup within 3 hours or so.

gooood