diy solar

diy solar

What ever happened to the Maxim Chip?

svetz

Works in theory! Practice? That's something else
Joined
Sep 20, 2019
Messages
7,201
Location
Key Largo
TL;DR: A 3 diode panel is 3 subpanels in parallel. If a part of a subpanel is in shadow, rather than other subpanels lose power the bypass diodes kick in. The maxim chip replaces the diodes and can even pull available power from the shaded subpanel (so more power and no overheating).

A few years back this article came out: Has Maxim Just Killed The Microinverter & DC Optimiser?
Although this article probably does a better job explaining it: https://www.solarquotes.com.au/blog/solar-panel-optimisation/

It was big news at the time. Jinko “Eagle MX”, Hanwa "Q.Peak-G4.1/MAX", Trina " TrinaPeak", Suntech and ET Solar supposedly all adopted it, but there were reports the chips also caused TV reception issues.

NREL reported in 2014 the Maxim chips work: https://www.nrel.gov/docs/fy14osti/62024.pdf and the information is still on the maxim site https://www.maximintegrated.com/con...ing-maxim-solar-cell-optimizers-in-pvsyst.pdf so it seems like a real thing.
 
Last edited:
Available on Digikey... $6.37, datasheet

Voltage is too low for my panels... pity, really want to see how they work and I don't even have shade! :unsure:

Update: They do have some with voltages upto 37.2: ref
 
Last edited:
So the panels need to come with these chips already built in?

Or can you put them in? Do they replace the diodes? Have any real world pictures of panels with them?
 

Attachments

  • Untitled.jpg
    Untitled.jpg
    72.3 KB · Views: 14
As far as I know, you can just replace the existing diodes with them. Maybe one of the EE types can say if they can be put in series with an existing diode?

It was a big news a while back, then all the news fizzled out and I don't see panel manufacturers saying they're using them (possibly the Jinko Eagles). Maxim has a pretty good rep so I suspect they work.

But, I also suspect they were "oversold", sure they can probably get power out of the shaded bit of the subpanel, but it's probably not as much power as they hoped. Perhaps that's changed with higher powered panels? For example,, a 3 diode 300W panel is three 100W subpanels; so one at 50% power is still 50W.
 
As far as I know, you can just replace the existing diodes with them. Maybe one of the EE types can say if they can be put in series with an existing diode?

It was a big news a while back, then all the news fizzled out and I don't see panel manufacturers saying they're using them (possibly the Jinko Eagles). Maxim has a pretty good rep so I suspect they work.

But, I also suspect they were "oversold", sure they can probably get power out of the shaded bit of the subpanel, but it's probably not as much power as they hoped. Perhaps that's changed with higher powered panels? For example,, a 3 diode 300W panel is three 100W subpanels; so one at 50% power is still 50W.


I would like to know more since soon I'll have 16 panels installed and some will be in shade at times during the day.
 
Me too! The NREL report is encouraging, but why aren't these all over the marketplace?

I suspect if you've a lot of shade then microinverters (e.g., Enphase, AP Systems (Magnum resells an AP)) or DC optimizers (e.g., SolarEdge) are the way to go. I think the chips just optimize the available power out of the panel, so if the whole panel is in shade the voltage will be still be down - so if on a string inverter that one panel's lower voltage could still bring the string's overall power down.

Although, even with a microinverter, since shadows crawl across the surface, you should still see some benefit.
 
Ran across an a useful image in an article that illustrates the concept. The docs in the OP are still better, but at least is shows you can buy them.

The idea is that every panel is made up of "strings" and if even one cell is impacted (shade, bird poop) the entire string cut out by the diode having an overall negative impact. The maxim chips act like microinverters and take whatever power the substrings make to combine it into the total output power. Theoretically, any panel can be retrofitted by replacing the diodes in the Jbox with the chips.

Capture.PNG
 
So, this basically bumps up the shaded string's voltage, so that it can continue to contribute whatever power it is able to convert.
Doesn't that then force current through the inactive cell(s)? If so, is that necessarily good for the cells. I know that forcing current backward through cells as can happen in a parallel array setup without blocking diodes heats up those reverse biased panels.
 
The Maxim chip is more hype then reality. It has some benefit for a bird poop spot but not much more.

Any sub-string section current is going to be limited to the current produced by weakest output cell. If any single cell is fully shaded then very little current will be produced by that sub-string.

The Maxim chip is just a buck DC-DC converter. If a sub-string has a partially shaded single cell and its current output drops, the buck converter will buck down the voltage output of the sub-string so current output matches the current produced by other illuminated series strings. This means same current output as unshaded sub-strings but at a lower sub-string voltage contribution to stack by the partially shaded sub-string.

If a single cell is totally shaded there is little current current to work with and the chip just goes into bypass similar to a bypass diode. A MOSFET switch bypass in the Maxim chip is a bit better then a bypass diode as the diode, at panel current, will have 0.8v to 1.2v diode forward bias voltage drop with about 6 to 8 watts of heating to diode which often stresses them to point of destruction. A MOSFET bypass switch can reduce this voltage drop to a couple tenths of volt with associated watts of heating to less then couple of watts.

Cost of the total Maxim circuit at commerical volume with chip, PCB, buck inductor, and capacitors will be about $3-$4 verses less then $0.50 for bypass diode. The cost-benefit is just not there, especially with the tough panel market price competition.
 
Most charge controllers provide the blocking diode function, but if you want to waste some not insignificant amount of power forward biasing a diode all day, be my guest.
If you want to replace a bad Bypass diode in a PV panels' J-Box, you can use an "Active Diode" which is a FET switch purpose made for solar.
Several companies make them , and they dissipate less heat than a Schottky diode.
https://www.digikey.ca/en/articles/...mprove-solar-panel-efficiency-and-performance
digikey.ca/en/articles/active-bypass-diodes-improve-solar-panel-efficiency-and-performance
 
Was looking at the shade starting to fall on my panels this morning as we head into winter and wondering if there had been any advancements in this technology. I didn't see anything new other than the voltage drop is now down to about 25% of a conventional diode (a gain of about ~+0.3V). But, in reading through the posts had some additional thoughts:

So, this basically bumps up the shaded string's voltage...
No, not a string of panels. It works on the innards of the panel and bumps up the panel's overall voltage. But yes, internally it bumps up the voltage of a sub-array within a panel by replacing that subarray's diode built into the panel. Which means it should work for inverter or microinverter systems (although probably not as much benefit with micros since the output from a string is very affected by the weakest panel in the string).

...Doesn't that then force current through the inactive cell(s)? If so, is that necessarily good for the cells.
Given a diode affects all the cells in the sub-array I suspect the maxim chip would actually be better for them. But, not an EE, so...

...It has some benefit for a bird poop spot but not much more...
Snowy areas where the bottom can be covered would probably see a big increase too. They might not be very useful for leafy shade, depending on how the shadows traverse the panel through the day and subpanel orientation.

In NREL's test case, it was somewhat guaranteed to be a success as they used a solar farm setup at their Colorado facility with no shade other than "row" shading during the winter months:
1697547997577.png1697548020760.png
... will be about $3-$4 verses less then $0.50 for bypass diode. The cost-benefit is just not there, especially with the tough panel market price competition.
The chips are almost $5 now, but I'm not sure how you'd calculate the return as it would really depend on your shade characteristics. In NREL's study, the shade always moved across the same subpanel and didn't affect the others.

My neighbor's trees have grown a bit since installation and there are about 3 to 4 panels at the bottom of the line that get affected from November to March. So, here it would only be about $20 for four chips. Let's see... +20% from a panel at half power would be about about +30 watts. So, 120W for four panels. Winter insolation of 4, so 480W/day. 90 days at $0.14/kWh would be $6 value per year. Over 25 years that would be a net gain of $130. So not much value honestly. Also, my shade moves across all the subpanels so I doubt it works as well as the NREL study. It would be nice to see a "leaf" study.
 
Instead of an optimizer for entire panel, it is an optimizer for one 20 to 24 cell (normally diode bypassed) section of a panel. Suitable for a 60 to 72 cell panel with 3 diode-bypassed sections.

When a cell is half shaded, regular optimizer can either limit entire panel to that current leaving 1/2 power, or bypass that cell's section of the panel, leaving typically 2/3 of the power.

The Maxim chip does same for a single section, so a cell half shaded would result in panel delivering 1/3 + 1/3 + 1/2 x 1/3 = 5/6ths of rated power.

For rooftop, you still need RSD and you may want monitoring. Include those features for the price of the existing chip and they might have something for that market.
 
Back
Top