site  contact  history  index

MPPT regulator is genuine MPPT

May 22, 2019 — BarryK

I posted about a couple of very cheap Chinese MPPT regulators that I purchased, for use with my 12V battery and photovoltaic solar panel:

As I mentioned in that post, there are many cheap regulators labelled as "MPPT" but are actually only PWM type. There is one particular model, that is very cheap yet does seem to be genuine MPPT, but there is some conflicting information in YouTube videos. So, I wanted to find out for myself.

As reported in that post, I ordered two, that seemed to be the same, and upon arrival I opened them up and they are identical inside. I paid AU$46.71 and AU$48.44, and just now looking at the eBay links, I see that the prices have gone up, now AU$64.99 and AU$49.46 (latter with discount code). Vendors are "intl-trading" and "sunyee-au", respectively, Aussie businesses, with AU stock.

As they looked identical internally, I only tested the one from intl-trading, and I am pleased to report that it is a genuine MPPT regulator. Photo of test setup:


I have not tested conversion efficiency, nor idle current, just wanted to confirm that it is genuine MPPT.

Passing puffy clouds in the sky, which was a very good opportunity to observe peak-power point tracking of the regulator. The digital power-meter on the input side, displays voltage, amperes, and power, and I was very pleased to observe the regulator accurately tracking the peak-power point as the sun faded and came back.

The three state-of-charge LEDs also did their job. I started off with the battery slightly discharged, which I achieved by running a desk lamp via a small 12-240 volt inverter. After plugging the regulator to the battery, solar panel not yet connected, two LEDs lit up. Yes, I also tested before having discharged the battery, and got three LEDs lit up. Very good.

After connecting the solar panel, all three LEDs lit up, which is to be expected. The LEDs would only be expected to give an accurate indication of charge without charging, and the battery voltage stabilized. Note, a stabilized battery voltage is quickly achieved by connecting a load to the battery (without charging connected), then remove the load, and the voltage will rise up, within 3 minutes and stabilize -- that voltage is what you read to determine state-of-charge. There is a bit of variation depending on what type of lead-acid battery, but for most deep-cycle batteries, from memory, it is 11.7V 0% up to 12.7V 100% -- I will check this with my battery, need to recharge it fully first.

Probably when I get setup for a camping trip, will configure the two sides of the folding solar panel in series, giving "24V", as this will result in lower current along the cable and into the regulator, which may be "better" for the regulator. 

Tags: nomad

Measurements for three solar panels

May 18, 2019 — BarryK

Today is Saturday 18th May, 2019, Federal voting day here in Australia. I got back from the voting booth about 9.00am and saw that the sky looks quite blue, just a few wisps of cloud, barely visible, no wind at all, very cool, ambient about 14 degrees C. Looks like a good day for testing my solar panels.

I now have test instruments to perform more accurate measurements, and a variable resistive load to more exactly locate the maximum-power point.

I tested three panels...

100W flexible panel

I purchased this panel in 2017, see blog report:

At that time, got a very low current out of the panel. I think now, that there was something wrong with the regulator. Well, something was wrong, as now getting much better performance.

Although the sky looked blue, the sun was low in the sky so more atmosphere to go through, this being almost mid-winter, and I read 800W/m2 irradiance (note, I have the habit of referring to the sun intensity as "insolation", though technically that is the solar energy over a certain time period, whereas "irradiance" is the instantaneous intensity, so the latter is correct when quoting W/m2). Here is the setup:


...the panel is set as close to directly facing the sun as I could judge. Notice the current-shunt (with crocodile clips attached), used to measure short-circuit current, and the brick holding down the ends of the nichrome wire. Testing at about 9.45am, ambient 14degC, 800W/m2:


...this graph shows output voltage versus current drawn from the panel. The peak-power point is at the knee of the curve, which is 4.96*16.95 -> 84.07W.

This is pretty good, considering that the cell temperature will be above 25 degC, and irradiance is only 800W/m2. The panel is rated at 100W @ 25degC and 1000W/m2.

So, if I perform some magic calculations, attempt to estimate what the power output would be at cell-temp of 25degC and 1000W/m2 irradiance...
Delta-temp: 46 - 25 -> 21 -> +5 for actual cell temp -> 26
Power change, temp: 0.4%/degreeC (0.4/100)*84.072*26 = 8.74W
Power compensation, irradiance: (1000/800)*84.072 -> 105.09W
Extimated power o/p of panel, at cell temp 25 degC @ irradiance 1000W/m2:
105.09 + 8.74 -> 113.85W I think that this estimation is too generous. It is based on the actual cell temperature being 5degC higher than the reading on the back of the panel (46degC), the power changing 0.4%/degC, and the power o/p being linear with irradiance. Also, I am assuming that the sun-power meter is reading accurately.

I definitely have to revisit these assumptions, as it is pretty obvious that I would not be getting power output above what the manufacturer is claiming!

250W Atem-Power folding panel

I had previously tested this, and got a disappointing power output, see this blog post:

Hoping now to get more power out of it! Here is the setup, panel angled to directly face the sun:


Testing one half of the panel only, at 10.30am, ambient 16degC, 850W/m2, temp at back of panel 50degC:


Peak power is 4.85A @ 16.21V -> 78.62W
As this is for half of the panel, multiple by 2: 157.2W

I am not going to apply my estimated corrections, as they are too generous. Instead, I want to do another test when solar irradiance is close to 1000W/m2, and will attempt to get the cell temp close to 25degC (possibly by shading the panel until just before taking a reading).

120W Powertech folding panel

This is a traditional glass-fronted folding panel, and very heavy -- the weight is what motivated me to look for something else. Don't recall where I purchased it, think it was in 2016. This is what it looks like:


Testing both panels, at 11.00am, irradiance 850W/m2, ambient 16degC, back of panel 47degC:


The peak power point is 6.48A @ 16.58V -> 107.4W

Again, will hold off with applying corrections.


I want to hold off with conclusions until I have sorted out the compensations for cell temperature and solar irradiance. For now, just comparing the readings as measured:

100W flexible
250W Atem-power
120W Powertech

The specification plates on all three panels state that the figures are at cell temp of 25degC and irradiance of 1000W/m2. My readings are slightly higher cell temperature and lower irradiance. The third column shows percentage output relative to that claimed.

Note, irradiance was slightly lower, at 800W/m2, when measuring the 100W panel, than when measuring the other two, at 850W/m2, so that 84% really should be a tad higher.

These results are very good, higher than I expected.  Except, the Atem-Power panel is still disappointing. 

Tags: nomad

Thoughts about solar distiller prototype no.1

May 16, 2019 — BarryK

There was some feedback after yesterday's post:

Derek suggested that I use cling film (also known as cling-wrap or glad-wrap) on the glass, so will be able to lift the glass off afterward and peal off the film.

I need to experiment with that option. Selleys 401 sealant is "acidic cure", and it might bond to the cling-film, so I might be able to get the glass off, but have to leave the cling-film behind.

Max expressed concern about aluminium and link with alzheimer's disease. Yes, it is a matter of trying to find the lesser evil. The Carocell panel is made with aluminium and plastic, and both of these may have issues. At least I am using a glass pane rather than plastic film.

However, it seems the jury is still out, linking alzheimer's with aluminium:

Anyway, this is prototype #1, and I will be able to change the aluminium sheet with something else, or coat it. The wicking cloth is another unknown -- I have obtained a black eco-friendly felt, and a quick test it seems to have OK wicking properties -- but that is another thing that may need to be improved. 

One interesting point. The solar distiller could be use to heat water. When camping, that could be very handy when want to have a shower, as there are portable outdoor shower units that have a 12V pump available. I think, if the water input flow rate is increased, so that most goes through without getting evaporated, just warmed, we could then fairly quickly collect enough for a shower. But of course, once collected, it will start cooling. Interesting possibility, definitely needs checking out! 

Tags: nomad

Solar distiller prototype underway

May 14, 2019 — BarryK

I posted recently about the Carocell 1000 solar distiller panel, that was given to me about five years ago:

Then, posted about initial plans to do a DIY panel:

Construction is now underway. Originally, I planned to build a dual-glass-pane panel, like the Carocell, however, ran into various technical difficulties regarding suitable materials.

The dual-pane design gives highest output, as condensation occurs on both front and back panes. There is a cloth membrane that water trickles down, and evaporated water may condense on the front pane. The heated air will rise, and will then flow down behind the cloth membrane, a much cooler environment, and condensation will occur on the back glass pane. This gives rise to a circulating current of air.

I did see an article in "Mother Earth News", back in the 80's, with a similar dual-pane design.

The dual-pane design will require a sheet of glass/metal/plastic, with black cloth laid on top, mounted at an angle of about 25 - 30 degrees. As I wanted to build the first prototype with wood, I would have had to coat the central wood sheet to make it totally waterproof, and to tolerate 100°C. The coating would have to be suitable for potable water and high temperature. Could not find any suitable paint, though there was one epoxy that met the specs -- but the fact remained, enclosing a plywood sheet with waterproof membrane may not be a good idea, as any vapours emitted from the wood will cause the coating to lift off.

Decided that the dual-pane design would have to be entirely aluminium frame, with glass or plastic inner sheet. Could not find a plastic sheet that met the specs -- though I know that there are such plastics, as my electric kettle is made of plastic.

Instead, for the first prototype, decided to build a trickle-down design with just one glass pane, no back pane. Wood construction, "premium ply" sheet. On top of the ply sheet is a 0.5mm thick sheet of aluminium. The water collection runoffs, for the leftover "dirty" water and the distilled water, has been fabricated with 0.3mm aluminium flashing. Selleys 401 silicone sealant is to be used, as this is rated as food-safe, for potable water, and a very high temp rating.

The construction is such that no wood will be exposed to water vapour. The solar collection area is about 600 x 880mm.

Construction is at an advanced state, well, maybe 2/3 done. One thing not yet figured out is how to mount the glass pane. I want a semi-rigid silicone "gasket" around the wood frame, onto which the glass can be lowered, and held in place with some aluminium brackets.

I could just squeeze out the Selleys 401 sealant around the frame, then put the glass on top, however, the glass will then be stuck on. I want to be able to remove the glass pane, so as to be able to make internal changes. This is a prototype, and want to be able to pull it all apart and put together again.

If there was some kind of silicone tape, that would be nice. Otherwise, will attempt to spread the 401 sealant as an even layer around the frame, let it set before putting the glass pane on top. may need to construct some kind of spreader to achieve a perfectly level surface.

Here is info on Selleys 401 sealant:

Photos will be posted when it is completed.

Tags: nomad

Car ball and nut steering

April 30, 2019 — BarryK

A different blog post from usual!

Modern cars have "rack & pinion" steering, which gives tight (minimal sloppiness or deadzone).

Many years ago, cars had "ball & nut", or "recirculating ball", steering. The main problem with this is a deadzone, or sloppiness, which got worse as the linkakes/bushes wore.

I was reminded of this recently, when read about the new 2019 Suzuki Jimny, and posted to this blog:

I was surprised that this Jimny still has ball & nut steering.

Back in the 70's I owned an early Suzuki 4wd, one of the "LJ" series, with a 2-stroke engine. I recall, it was dangerous on wet roads. Especially a wet road with tight bends -- I found that the steering would get out of my control, and I had to slow right down.

In the mid to late 80's, I owned an old Holden, an "EH" model I think, a 3-speed automatic. I was cash-strapped at the time, and bought this secondhand. Don't recall what price I paid, but do remember selling it for AU$300.

It had various problems, such as tending to overheat -- but they made engines out of iron in those days, it could survive repeated overheating.

It had the old recirculating ball steering, don't recall if it was power steering. It was worn, with considerable slack. Near where I lived, there was a dirt road, that turned to mud when it rained. I discovered that when driving home on that road, and the car lurched to the right, the steering wheel snatched violently out of my hands. I pulled the wheel left, and it got snatched out of my hands again as the car lurched to the left.

Thus I zigzagged across the road, until I had reduced speed to a crawl.

The problem was the tyres sinking slightly into the mud. If the front tyres turn very slightly from true forward direction, a small wall of mud then forces them to turn more. Which can happen due to the slackness in the steering linkages. When the wheels have been twisted more than a few degrees, the wheels get violently pushed to the side, and the steering wheel spins out of my control.

So, pretty awful type of steering, hey! However, I am over-dramatizing the risk. My Holden was old and the steering linkages very worn. If I had got it fixed, replaced some bushes, the steering slack would have tightened up considerably, and I would probably have been able to race straight through that muddy road.

Apart from the Jimny, apparently some trucks still have this, I don't know about other 4wd vehicles.

With the Suzuki 4wd cars, a popular after-market addon was a "steering damper", a hydraulic mechanism that will minimise the kind of scenario that I have described above.

It seems that Suzuki added it to the Jimny 4wd series, and it is in the 2019 model. Thank goodness. There is also power steering, so it would seem that the two mechanisms would be fighting each other -- well, apparently, it does cause the steering to have a "dead" feel, with no feedback from the wheels.

Why hasn't Suzuki gone over to rack & pinion steering for the Jimny? I don't know, there must be reasons. I don't know enough about the topic to guess why. More info here:


The recirculating ball mechanism has the advantage of a much greater mechanical advantage, so that it was found on larger, heavier vehicles while the rack and pinion was originally limited to smaller and lighter ones; due to the almost universal adoption of power steering, however, this is no longer an important advantage, leading to the increasing use of rack and pinion on newer cars.

The steering damper does make driving safer, however there is another concern. The deadzone means that the front wheels are not necessarily going in the direction in which you are pointing the steering wheel. This has been observed by testers of the 2019 Jimny on highway driving, where the car tends to wander to left or right, and has to be continually corrected.
This means that the driver has to be always alert, and it does make the driving experience more tedious -- though, the continual correction does become habitual. I also briefly owned a Suzuki 1.3 litre Sierra, the model after the LJ series and before the Jimny series, and I recall this wandering problem, but I adapted to it and found highway driving to be OK, if a bit odd.

Thought that I would post these interesting observations! 

Tags: tech

Instruments for testing photovoltaic panels

April 28, 2019 — BarryK

I posted a few days ago about disappointing power output of my "250W" folding solar panel:

...I only got about 120W out of it.

What I want to do is get setup to do more precise measurements, including taking into account the intensity of the sun and the ambient and cell temperatures. To achieve this, I needed to purchase some measuring instruments, that I have done.

Solar power meter

I hunted online to find out if any local store (in Perth, Western Australia) has one of these, and at a reasonable price. Found this, at a clearance price of AU$99.95:


First impressions, quality construction, works OK.

Infra-red thermometer

Measuring solar cell temperature is difficult, as they are sandwiched between materials both in front and back. I decided to use an infra-red non-contact thermometer, and bought this one, AU$34.95, from Altronics:


First impression, easy to use.

There is going to be a temperature gradient from the cells to back surface of the panel. If the thermometer is used to measure back temperature, then I read somewhere that very roughly the cell temp will be 7°C higher.

I also read somewhere, that if the thermometer is used at the front of the panel, at a steep angle, not perpendicular, it can get an IR reading direct from the cell. That remains to be seen.

There is another approach to testing. Instead of letting the panel warm up in the sun, keep it in shade, so the cells are at ambient temperature. When everything is connected to take a reading. move the panel into the sun, take a reading very quickly, then move the panel back into the shade. I saw this technique being used in a YouTube video. However, I don't know whether this will work, as I imagine the cell junction will get hot very quickly. I intend to play with this method.

Resistive load

One way to do a super-quick evaluation of a panel is to test short-circuit current with a multimeter. With the multimeter on DC current-reading setting, and being sure that the multimeter can handle the expected current, connect it directly to the terminals in the panel junction-box. The panel could be in shade, then bring it into the sun and observe  the current.

To obtain the peak power point, though, needs more equipment. An MPPT regulator connected to a not-fully-charged battery would do it. If the MPPT regulator is doing it's job, it will hunt for and find the peak power point. Some means of measuring voltage and current will be required.

An alternative is to use a variable resistive load. For the previous test on the "250W" panel, I used nichrome wire, just some lengths that I had. However, some kind of continuously-variable adjustment would be better, to accurately locate the peak power point. Also, it would need to be designed to handle the power dissipation.

Such a load, I am planning to build very soon. 

Tags: nomad

Atem Power 250W solar panel is a disappointment

April 24, 2019 — BarryK

EDITED 2019-05-19: Improved testing of 250W panel
EDITED 2019-04-27: Edits in bold red (for legal reasons)

I posted about a 250W folding solar panel that I purchased from XXXXXXXX, via eBay:

Very cheap, only AU$288 including delivery.

It arrived a couple of days ago. The sky was a bit murky yesterday, but today brilliant sunshine, totally blue sky, not even a wisp of cloud, ambient temperature cool, low 20s, location Perth WA, date Wednesday April 24 2019. I set it up about 10.00am, but was only getting about half of the expected current. Hmmm.

I am a retired electronic engineer, and I once worked in solar energy research, so I am competent in this field. Of course, I had to investigate further. This is a folding panel, and each side has a junction box. It is easy to pop off the covers by inserting a screwdriver:


Standing at the back of the panel, this is inside of the right-hand junction box:


...the left-hand panel is connected in parallel to the right-hand panel, and output then goes to the solar regulator:


...what I have done here is removed the wires from the regulator, bypassing the regulator. The solar panel now goes directly to the output cable. At the other end of the cable I used nichrome wire as a resistive load. A 150A current-voltage-power meter is plugged in:


I used a digital multimeter to confirm that the in-line 150A meter was reasonably accurate, yes it is. That is also a recent purchase, so at least that works!

The peak-power point was somewhere between 18.55V @ 5.65A (104.8W) and 16.87V @ 7.08A (119.7W). Yep, only about half what it should be. Obvious next step is to separate the panels. I disconnected both panels:


...with a jumper lead, I can now connect either panel to the load.

The right-hand panel gives about 3.5A at the peak power point. Drawing more, bringing it down to 11.64V get 4.69A.

The left-hand panel also gives about 3.5A at peak-power, and I loaded it some more, down to 10.98V at 4.69A.

As a final check, I also measured the short-circuit current out of each side, directly on the junction-box terminals, and yeah, only about 4.7A from each (measured mid-afternoon, with the sun a bit lower in the sky. Midday it would have been a tad higher).

So, yeah, they are crap panels, giving less than half of that claimed. They have labels on the back of each side, stating peak power at 18.4V @ 6.8A and short-circuit current of 7.09A.

The labels specify sun insolation (intensity of the sun) at 1000W/m2 and cell temperature of 25 degrees C. The sun intensity would have been about that for my measurements. In the sun, panels get hot, so the cell temperature is going to be well above 25°C, and a rough guide is power drop of 0.45% per degree rise. So, it would be reasonable to expect a panel rated at 250W to give about 200W in practical use.

What really bothers me about this, is that most people would not know. They would connect the panel to the battery, using the supplied crocodile clips, and think it is working. Well it is, except they are getting far less power than advertised.

The problem is, this overstating of power output is endemic to many of the cheap panels from China. What to do? Something should be done, I should report this somewhere, but what government dept would be appropriate?
The panel that I purchased was from an Australian business, and they have an obligation not to do false advertising. 

EDIT 2019-05-19:
I have a duty to be as accurate as possible when testing solar panels, and to this end have improved the testing setup, with temperature and solar-insolation measuring instruments, a more gradually-variable resistive load, and attention to details (such as making sure the panel is oriented optimally toward the sun, rather than its default angle). A follow-up test has now been conducted, alongside two other panels for comparison: 

Tags: nomad

Planning a DIY solar water distiller

April 23, 2019 — BarryK

I posted recently about this Carocell 1000 solar water distiller, that was given to me about five years ago:

The problem is the size, 1.1m x 1.1m. I would need a roof rack, and preferably slide it under the rack, so would have to buy a roof rack with 45mm clearance above the roof. Well, there is such a roof rack, but currently that is the only purpose that I have for the roof rack.

I plan to create a shelving system in the back of the car, and can have a shelf into which a solar water panel can slide. But the width of the back door is only about 1m, too narrow for the Carocell.

So, playing with ideas for building my own. It will be a flat panel, with water trickle-down, through a cloth. I have ordered a special eco-friendly cloth to test, and also have corresponded with a chap in the US who has made a similar type of trickle-down panel. He advised a certain cloth that I could try.

I was thinking of making the frame from pine, pure pine, nothing laminated or chipboard. The danger is chemicals evaporating in the hot wet conditions inside the panel. But then, should the wood be left bare, or painted? If painted, same problem with dangerous chemicals.

It would be good to seal the surface of the wood, but it would have to be a "safe" product. I have asked for a recommendation on the "Green tech" thread of the Whirlpool forum:

Don't know if anyone will reply. Might have to take a punt on some product. Or, make the frame from aluminium. 

Tags: nomad