Failed project / eBay TEC1-12706 fail
Sulaiman, Sat Nov 24 2018, 06:29PMFor my chemistry hobby I want a recirculating water cooler/chiller to cool condensers.
I'd previously experimented with a dual TEC1-12706 @ 12V cooler,
but the performance was inadequate,
so I built a new water cooler using four TEC1-12706 @ 6V each for better efficiency hence performance,
(I assumed that the performance of the heatsinks (AK-786) would be the limiting factor)
At 6V per TEC I was getting 6.24 Watts total effective cooling power.
I was expecting more like 100 Watts.
So I rewired the module with all TECs in parallel connected to 12.0V,
the module performed even more poorly.
(I've requested a refund from the eBay seller)
Here is the module just after I tested it.
(Testing was at 10.5 Celcius ambient, water was at 10.0 Celcius, water flow was adequate, HY810 heatsink compound)
I damaged my 'old' TECs removing them so I can't do a comparison now)
Two questions ;
1) Is it 'normal' to get seriously sub-standard TECs if from China via eeBay ?
2) Could anyone point me (in UK) to a reasonably priced supplier for genuine/original 40mm x 40mm TECs ?
Re: Failed project / eBay TEC1-12706 fail
Plasma, Sat Nov 24 2018, 10:39PM
Alexnld.com
Tegmart is only questable based on max tempture, I buy off alexnld ,with heatsink compounds and All or copper they will probably reach specs
Plasma, Sat Nov 24 2018, 10:39PM
Alexnld.com
Tegmart is only questable based on max tempture, I buy off alexnld ,with heatsink compounds and All or copper they will probably reach specs
Re: Failed project / eBay TEC1-12706 fail
iJim, Sat Dec 01 2018, 07:58PM
Sulaiman, exactly the same thing happened to me. Picked up 5x 50W TECs from EBay for about £10, wired two in series and had terrible performance. Annoyingly I used thermal epoxy to secure on the heatsink so that will also probably go in the bin. Ended up buying 2x TECs from RS for £35 each and they work fine.
iJim, Sat Dec 01 2018, 07:58PM
Sulaiman, exactly the same thing happened to me. Picked up 5x 50W TECs from EBay for about £10, wired two in series and had terrible performance. Annoyingly I used thermal epoxy to secure on the heatsink so that will also probably go in the bin. Ended up buying 2x TECs from RS for £35 each and they work fine.
Re: Failed project / eBay TEC1-12706 fail
klugesmith, Sun Dec 02 2018, 01:18AM
At my workplace, thermally-epoxied heatsinks are routinely taken off in the course of equipment repair.
You soften it by getting it plenty hot.
I bet that with heat applied from the heatsink side, there's a point where you can separate TEC ceramic from the heatsink with a thin knife, and not melt soldered joints inside the no-good TEC. Heat pumped toward heatsink by energizing the no-good TEC probably wouldn't hurt.
klugesmith, Sun Dec 02 2018, 01:18AM
At my workplace, thermally-epoxied heatsinks are routinely taken off in the course of equipment repair.
You soften it by getting it plenty hot.
I bet that with heat applied from the heatsink side, there's a point where you can separate TEC ceramic from the heatsink with a thin knife, and not melt soldered joints inside the no-good TEC. Heat pumped toward heatsink by energizing the no-good TEC probably wouldn't hurt.
Re: Failed project / eBay TEC1-12706 fail
Plasma, Mon Dec 03 2018, 02:44AM
I changed the heatsink ability, but I brought 3 different batch, two matach with on 0.3volt with five ,five times as much, they also have refreagtion tec, the spelling mistice with MA is a launwage barrier, I don't want to add heat sink but it changed from 50*50mm block of steel to a transformer core.
Alexnld.com for 6 times as cheap you have to budget for the quality, but it might make sense
Plasma, Mon Dec 03 2018, 02:44AM
I changed the heatsink ability, but I brought 3 different batch, two matach with on 0.3volt with five ,five times as much, they also have refreagtion tec, the spelling mistice with MA is a launwage barrier, I don't want to add heat sink but it changed from 50*50mm block of steel to a transformer core.
Alexnld.com for 6 times as cheap you have to budget for the quality, but it might make sense
Re: Failed project / eBay TEC1-12706 fail
Sulaiman, Mon Dec 03 2018, 09:48AM
I just asked eBay resolution centre to step in.
I can see that buying from RS would give me a working solution,
but 4x TEC at RS prices make the project more expensive than using conventional refrigeration.
For my first prototype I used epoxy to glue Tank:TEC and TEC:Heatsink rather than using mechanical clamping,
a very poor choice because;
Force on the heatsinks will tear the TECs apart, damaging ingot:pad connections
There is negligible clamping force on the TECs, which should be under compression
Epoxy resin adhesive has poor thermal conductivity compared to decent heatsink compound, 10x poorer
Replacement of a faulty TEC is difficult.
EDIT: In the time that it took to write this post, eBay have refunded my payment in full.
Now that I'm in a better mood I'll start looking for cheap genuine TECs.
Sulaiman, Mon Dec 03 2018, 09:48AM
I just asked eBay resolution centre to step in.
I can see that buying from RS would give me a working solution,
but 4x TEC at RS prices make the project more expensive than using conventional refrigeration.
For my first prototype I used epoxy to glue Tank:TEC and TEC:Heatsink rather than using mechanical clamping,
a very poor choice because;
Force on the heatsinks will tear the TECs apart, damaging ingot:pad connections
There is negligible clamping force on the TECs, which should be under compression
Epoxy resin adhesive has poor thermal conductivity compared to decent heatsink compound, 10x poorer
Replacement of a faulty TEC is difficult.
EDIT: In the time that it took to write this post, eBay have refunded my payment in full.
Now that I'm in a better mood I'll start looking for cheap genuine TECs.
Re: Failed project / eBay TEC1-12706 fail
Shrad, Wed Dec 05 2018, 08:01AM
For small prices on electronics I would have a look at banggood or similar sites, for all the items I bought from them I have never really had any issue knowing the quality of what I bought
I think the main difference with random ebay sellers is that at least china is trying to raise the bar on expectations via official channels, and for what I saw it seems to be OK.. just maybe delivery delays and politics for US guys
I'll try alibaba as soon as they will have their new delivery centre in Belgium, waiting time should be way less
Shrad, Wed Dec 05 2018, 08:01AM
For small prices on electronics I would have a look at banggood or similar sites, for all the items I bought from them I have never really had any issue knowing the quality of what I bought
I think the main difference with random ebay sellers is that at least china is trying to raise the bar on expectations via official channels, and for what I saw it seems to be OK.. just maybe delivery delays and politics for US guys
I'll try alibaba as soon as they will have their new delivery centre in Belgium, waiting time should be way less
Re: Failed project / eBay TEC1-12706 fail
Inducktion, Wed Dec 05 2018, 08:15PM
Yeah, honestly based off what I've been hearing about more and more lately, Chinese sourced semiconductor goods are very suspect and ... it'd probably be a better idea to just go with a reputable supplier in the first place, so you don't
1. Waste time trying to diagnose the problem
2. Waste money on something counterfeit in the first place.
Especially, since a lot of semiconductor goods are hard to verify as legit in the first place too. Unless you have specialized equipment.
It's just hard to justify saving a few dollars when you might end up getting something that's ...basically not what you ordered.
Inducktion, Wed Dec 05 2018, 08:15PM
Yeah, honestly based off what I've been hearing about more and more lately, Chinese sourced semiconductor goods are very suspect and ... it'd probably be a better idea to just go with a reputable supplier in the first place, so you don't
1. Waste time trying to diagnose the problem
2. Waste money on something counterfeit in the first place.
Especially, since a lot of semiconductor goods are hard to verify as legit in the first place too. Unless you have specialized equipment.
It's just hard to justify saving a few dollars when you might end up getting something that's ...basically not what you ordered.
Re: Failed project / eBay TEC1-12706 fail
teravolt, Thu Dec 06 2018, 08:40PM
usaly those pelter devices are power hungry. did the heat sinks get warm and hiw many amps did you supply when I tried this I need a 12v 10amp supply
teravolt, Thu Dec 06 2018, 08:40PM
usaly those pelter devices are power hungry. did the heat sinks get warm and hiw many amps did you supply when I tried this I need a 12v 10amp supply
Re: Failed project / eBay TEC1-12706 fail
Sulaiman, Mon Dec 31 2018, 03:02PM
Good News :
I bought four TEC1-12715 for £12.68 total from here
and they perform close to specifications.
I tried 2x TEC1-12715 per heatsink
(AK-786 cpu heatsink/fan assemblies, 0.25 C/W per heatsink, measured)
... fail
With One TEC1-12715 per heatsink performance is tolerable;
2x TEC wired in series, 15Vdc @ 6.3 Adc,
Tmin= -10 C, about 40W heat pumping power at 20 C.
0.25 C/W is too great, something more like 0.1 C/W is required ... per TEC1-12715
Sulaiman, Mon Dec 31 2018, 03:02PM
Good News :
I bought four TEC1-12715 for £12.68 total from here
and they perform close to specifications.
I tried 2x TEC1-12715 per heatsink
(AK-786 cpu heatsink/fan assemblies, 0.25 C/W per heatsink, measured)
... fail
With One TEC1-12715 per heatsink performance is tolerable;
2x TEC wired in series, 15Vdc @ 6.3 Adc,
Tmin= -10 C, about 40W heat pumping power at 20 C.
0.25 C/W is too great, something more like 0.1 C/W is required ... per TEC1-12715
Re: Failed project / eBay TEC1-12706 fail
klugesmith, Mon Dec 31 2018, 05:32PM
Nice way to end the year, Sulaiman.
Your project inspires me to get some TECs and make an absolute humidity sensor, based on dew point temperature.
Would cool a small shiny surface exposed to gas under test, and optically detect the onset of fogging.
Ever seen anything like that in DIY land, or in a catalog at a hobby-friendly price?
klugesmith, Mon Dec 31 2018, 05:32PM
Nice way to end the year, Sulaiman.
Your project inspires me to get some TECs and make an absolute humidity sensor, based on dew point temperature.
Would cool a small shiny surface exposed to gas under test, and optically detect the onset of fogging.
Ever seen anything like that in DIY land, or in a catalog at a hobby-friendly price?
Re: Failed project / eBay TEC1-12706 fail
Conundrum, Thu Jan 10 2019, 07:34PM
Also have some tiny TECs here.
These ones verifably work: another good idea is to salvage units from c@s!0 projectors.
They are used to cool the photonic lattice red LED and are very efficient BUT:
Do not forget the 1:4:12 rule: 1 Watt on unit A means 4 Watts on unit B (1+1+2) and 12 Watts of heat on unit 3's hot side to get rid of because you are dumping both ohmic and transferred heat. So scale the current accordingly for best efficiency.
Putting units in parallel is an acceptable workaround and does also yield better overall efficiency if they are electrically in series.
My tiny units are fine for say cooling the front end on a radio but worthless for serious cooling.
I believe they are but not absolutely sure.
Conundrum, Thu Jan 10 2019, 07:34PM
Also have some tiny TECs here.
These ones verifably work: another good idea is to salvage units from c@s!0 projectors.
They are used to cool the photonic lattice red LED and are very efficient BUT:
Do not forget the 1:4:12 rule: 1 Watt on unit A means 4 Watts on unit B (1+1+2) and 12 Watts of heat on unit 3's hot side to get rid of because you are dumping both ohmic and transferred heat. So scale the current accordingly for best efficiency.
Putting units in parallel is an acceptable workaround and does also yield better overall efficiency if they are electrically in series.
My tiny units are fine for say cooling the front end on a radio but worthless for serious cooling.
I believe they are
but not absolutely sure.Re: Failed project / eBay TEC1-12706 fail
Sulaiman, Sun May 05 2019, 02:13PM
To end on a positive note;
I purchased two more fan/heatsink CPU coolers (£4.50 ea. w/p&p)
So now I have two similar modules, each module has ;
(1x) 40 x 40 mm water block
(2x) TEC1-12715 with one fan/heatsink per TEC
At moderate to high water flow rates it makes little difference if they are plumbed in series or parallel.
For my application the TECs work best at around 7.5V each with these heatsinks.
Sulaiman, Sun May 05 2019, 02:13PM
To end on a positive note;
I purchased two more fan/heatsink CPU coolers (£4.50 ea. w/p&p)
So now I have two similar modules, each module has ;
(1x) 40 x 40 mm water block
(2x) TEC1-12715 with one fan/heatsink per TEC
At moderate to high water flow rates it makes little difference if they are plumbed in series or parallel.
For my application the TECs work best at around 7.5V each with these heatsinks.
Print this page