Okay, have got two panels up and running, with the new regulator ticking away just fine.
In at Altronics yesterday, and when I asked the sales person what I wanted, he did a few calculations, and advised that the diodes (2x) for the two 160w panels (4 diodes all up) should be put into a metal enclosure, that would act as a heat sink, and after pressing the diodes against the side of the metal case, use heat transfer paste to get the conductivity for the heat. Also advised to use 2 of the 15 amp diodes for each panel in parallel to reduce heat and voltage drop. he did ask where we may be going and I replied all over Australia. To day I went there again to get more stuff, and the sales person on duty assured me that that person would really knew his stuff. OK that job is done.
The third panel (the one I am working on now) has a plastic box mounted on the roof of the van, that allows me to connect the solar panel wire to the house cable, with room enough to mount the diodes, OK I'll use two as I have all up 6 (thanks Terry and Tony), So here is my question, How much heat can these diode generate? is a plastic box about 75 x 45 x 45mm big enough to handle the heat, can I just join two diodes between the solar output and the house positive, or do I need to have a terminal (tag strip) and hard mount them? and or in a metal box, or one with more air space.
I need advise as soon as possible.
T1 Terry said
10:09 AM Oct 20, 2018
To calculate how much energy will be available for heating, or wasted as heat depending which way you look at it, the math is fairly simple .... sort of :lol:
The closer to the max. current rating of the diode the higher the voltage drop, it isn't linear but for simplicity we will assume it is. This means, a standard 15 amp diode with 15 amps being passed through it will have a 0.7v drop between the in and the out. The Schottky diode that I sent you has half that voltage drop, so at 15 amps current the voltage drop would be 0.35v.
OK, now to put the numbers into a formula, the 160w panel you have can produce roughly 160w x 70% divided by 12v = 9.33 amps. Let's round that up to 10 amps. 10 amps is roughly 2/3 of 15 amps or 66%, 0.35v x 66% = 0.23v. We know the figure for amps is 10 amps in this case and we now know the figure for the voltage is 0.23v. 10a x 0.23v = 2.3watts and this is the max energy available for heating. With a standard diode that figure would double to 4.6w and over time that could build up some heat, but with access to air cooling the 2.3w of heat will dissipate quickly so no heat issues would occur. If you were to double up the diodes, or two diodes in parallel on the positive cable, then the heat generation would be halved across each diode, or 1.15w from each diode and that would not be enough energy to heat anything because the heat dissipation would be greater than that.
Next time you are in Altronics, run the numbers past your helper there and see what he thinks.
T1 Terry
dorian said
06:26 PM Oct 20, 2018
Has anyone considered using one of these "ideal diode" chips, or similar?
Interesting chip but I can get 50 Shottky diodes for that price.
iana said
10:39 AM Oct 21, 2018
Thanks again Terry, no I won't be asking at Altronics, I just want to get the job done. I had a small problem with fitting the diode (s) to the last panel, and your answer, and a bit of thinking outside the square has hopefully solved it.
I have also been working on the Daughters home irrigation system, so was too busy to answer the post yesterday.
Tony Bev said
12:22 PM Oct 21, 2018
Hi Ian
Sorry that I did not see this post earlier I have read that some silicon types diodes get hot, but that Schottky diodes stay cool
After also hearing that diodes get hot, I measured mine with a thermal heat gun
With 10 amps going through the 15SQ045 HV Schottky diodes, with an ambient temperature of 21°C, I measured 38°C at one diode, and just over ambient temperature at the second diode
My understanding of reading the Internet, (which may or may not be true), is that the diodes must be in perfect length/temperature/resistance
If not in perfect harmony, I am led to believe that the current will flow through the diode with the least resistance, when they are in parallel
The feedback, through the diode at night time, when the solar panel was not producing anything, was 160 millivolt, 0.16 volts, for the diodes we received from Terry I measured this just to make sure that, I had not damaged a diode while fitting them, or if I have damaged one, it is in the open position, with no current flowing I have no way of measuring the amps at such low voltage
The DC/DC regulator I have also stops any feedback to the panel/s I had to connect a positive wire from the battery to the outlet of the diodes, and measure from the inlet of the diodes to the negative of the battery
Below is a picture showing my diodes in parallel, and not in perfect harmony, or perfect soldering, I had to pull them upside down to take a picture They will live in a plastic Jiffy box
Do not let the different coloured shrink wraps confuse the issue, I was running short of the red ones Also the bent wire is from my handling, and not from any type of heat
dorian said
07:32 AM Oct 22, 2018
A diode's forward voltage (Vf) has a temperature coefficient of appproximately -2mV per deg C.
As has already been stated, a parallel pair of diodes will divert the majority of the current load to the one with least resistance. This mismatch in the load sharing is exacerbated if the diodes are thermally isolated. This is because, as a diode becomes hotter, its Vf drops, thereby diverting even more of the current to itself, resulting in even more heating ...
There are some I-V characteristic curves in the following document.
ISTM that if one insists on paralleling two diodes, then they should come from the same batch and be mounted on the same heatsink. A better solution would be a dual diode in a single package (3-terminal) that can be screwed down.
DeBe said
05:10 PM Oct 22, 2018
Interesting point, might give these a try as I have several. Will probably use the panel frame as a heatsink.
Great idea dorian. Just fitted one to the frame of the 160W solar panel. Tested with a continous 8A draw on a 25degC amb temp, works a treat & doesn't get hot at all. Will be doing the other 2 panels now.
You can salvage these dual diodes, including mounting hardware, from old AT/ATX PSUs. Check the current rating of the +12V or +5V rails. Also be aware of the reverse voltage ratings for these Schottky rectifiers.
Okay, have got two panels up and running, with the new regulator ticking away just fine.
In at Altronics yesterday, and when I asked the sales person what I wanted, he did a few calculations, and advised that the diodes (2x) for the two 160w panels (4 diodes all up) should be put into a metal enclosure, that would act as a heat sink, and after pressing the diodes against the side of the metal case, use heat transfer paste to get the conductivity for the heat. Also advised to use 2 of the 15 amp diodes for each panel in parallel to reduce heat and voltage drop. he did ask where we may be going and I replied all over Australia. To day I went there again to get more stuff, and the sales person on duty assured me that that person would really knew his stuff. OK that job is done.
The third panel (the one I am working on now) has a plastic box mounted on the roof of the van, that allows me to connect the solar panel wire to the house cable, with room enough to mount the diodes, OK I'll use two as I have all up 6 (thanks Terry and Tony), So here is my question, How much heat can these diode generate? is a plastic box about 75 x 45 x 45mm big enough to handle the heat, can I just join two diodes between the solar output and the house positive, or do I need to have a terminal (tag strip) and hard mount them? and or in a metal box, or one with more air space.
I need advise as soon as possible.
The closer to the max. current rating of the diode the higher the voltage drop, it isn't linear but for simplicity we will assume it is. This means, a standard 15 amp diode with 15 amps being passed through it will have a 0.7v drop between the in and the out. The Schottky diode that I sent you has half that voltage drop, so at 15 amps current the voltage drop would be 0.35v.
OK, now to put the numbers into a formula, the 160w panel you have can produce roughly 160w x 70% divided by 12v = 9.33 amps. Let's round that up to 10 amps. 10 amps is roughly 2/3 of 15 amps or 66%, 0.35v x 66% = 0.23v. We know the figure for amps is 10 amps in this case and we now know the figure for the voltage is 0.23v. 10a x 0.23v = 2.3watts and this is the max energy available for heating. With a standard diode that figure would double to 4.6w and over time that could build up some heat, but with access to air cooling the 2.3w of heat will dissipate quickly so no heat issues would occur. If you were to double up the diodes, or two diodes in parallel on the positive cable, then the heat generation would be halved across each diode, or 1.15w from each diode and that would not be enough energy to heat anything because the heat dissipation would be greater than that.
Next time you are in Altronics, run the numbers past your helper there and see what he thinks.
T1 Terry
LTC4359 - Ideal Diode Controller with Reverse Input Protection:
produktinfo.conrad.com/datenblaetter/1200000-1299999/001255992-da-01-en-IC_OR_CTRLR_N_LTC4359IMS8_PBF_MSOP_8_LTC.pdf
LTC4357 - Positive High Voltage Ideal Diode Controller:
www.farnell.com/datasheets/85443.pdf
LTC4358 - 5A Ideal Diode - Analog Devices:
www.analog.com/media/en/technical-documentation/data-sheets/4358fa.pdf
I have also been working on the Daughters home irrigation system, so was too busy to answer the post yesterday.
Hi Ian
Sorry that I did not see this post earlier
I have read that some silicon types diodes get hot, but that Schottky diodes stay cool
After also hearing that diodes get hot, I measured mine with a thermal heat gun
With 10 amps going through the 15SQ045 HV Schottky diodes, with an ambient temperature of 21°C, I measured 38°C at one diode, and just over ambient temperature at the second diode
My understanding of reading the Internet, (which may or may not be true), is that the diodes must be in perfect length/temperature/resistance
If not in perfect harmony, I am led to believe that the current will flow through the diode with the least resistance, when they are in parallel
The feedback, through the diode at night time, when the solar panel was not producing anything, was 160 millivolt, 0.16 volts, for the diodes we received from Terry
I measured this just to make sure that, I had not damaged a diode while fitting them, or if I have damaged one, it is in the open position, with no current flowing
I have no way of measuring the amps at such low voltage
The DC/DC regulator I have also stops any feedback to the panel/s
I had to connect a positive wire from the battery to the outlet of the diodes, and measure from the inlet of the diodes to the negative of the battery
Below is a picture showing my diodes in parallel, and not in perfect harmony, or perfect soldering, I had to pull them upside down to take a picture
They will live in a plastic Jiffy box
Do not let the different coloured shrink wraps confuse the issue, I was running short of the red ones
Also the bent wire is from my handling, and not from any type of heat
As has already been stated, a parallel pair of diodes will divert the majority of the current load to the one with least resistance. This mismatch in the load sharing is exacerbated if the diodes are thermally isolated. This is because, as a diode becomes hotter, its Vf drops, thereby diverting even more of the current to itself, resulting in even more heating ...
There are some I-V characteristic curves in the following document.
Derating of Schottky Diodes:
cdn2.hubspot.net/hubfs/1871852/Content_for_Degreed.com/Previous_Content/Derating-of-Schottky-Diodes1.pdf
ISTM that if one insists on paralleling two diodes, then they should come from the same batch and be mounted on the same heatsink. A better solution would be a dual diode in a single package (3-terminal) that can be screwed down.