Homepage

The Model

Draft projects

Articles published

Technologies

Annexed material

Downloads

 About Bakens Verzet

STICHTING BAKENS VERZET

1018 AM AMSTERDAM, THE NETHERLANDS

Director,

T.E.(Terry) Manning,

Schoener 50,

1771 ED Wieringerwerf,

The Netherlands.

Tel: 0031-227-604128

Homepage: http://www.flowman.nl

E-mail: (nameatendofline)@xs4all.nl : bakensverzet

 


MODEL FOR SUSTAINABLE SELF-FINANCING INTEGRATED RURAL AND POOR URBAN DEVELOPMENT FOR THE WORLD'S POOR

Incorporating innovative social, financial, economic, local administrative and productive structures, numerous renewable energy applications, with an important role for women in poverty alleviation in rural and poor urban environments.

 


 

"Money is not the key that opens the gates of the market but the bolt that bars them"

 

Gesell, Silvio The Natural Economic Order

Revised English edition, Peter Owen, London 1958, page 228

 


 

Edition 9: 07 December,2006

 


SOLAR SPRING : MAINTENANCE

How to install the supplementary I/b remote control card to transform a MkI Sunprimer into a Mk I/b unit.

Please consult the accompanying electric diagrams.
Unscrew the two small screws from the Sunprimer box wall and, without moving any wires from the Sunprimer box holes, lift the circuit card from the bottom of the Sunprimer.
Connect the new 6 ampere light blue coloured D11 diode by interrupting the copper track A-B by scraping away the small length shown in green on the drawing, using an already assembled MKI/b unit as an example. You may find the D11 diode in the MKI/b units sent is connected at different places, but the best place is on the A-B track as shown. The D11 diode is there to protect the BDX88C Q2 transistor against inverse voltage. The D11 diode is a Darlington type resistor with a built-in diode to protect against inverse voltage, though it has proved too weak in one or two situations.
Next, solder off and discard the 8200 ohms resistor marked in green in the drawing which is no longer necessary, and use its two holes to connect two small new supplementary circuit wires. It is not difficult to connect and solder other wires and make new holes in the Sunprimer box (use sample MkI/b units sent for comparison). Be sure to use a good silicon sealant or other product to seal the screws and holes, against humidity. For the same reason it is better to connect the aluminium cooling flanges of the transistors dry to the box walls and then seal with silicon, than to use special heat conducting grease with a risk of not getting a perfect protection against humidity. We also suggest the wires be sealed at the point they enter the Sunprimer box by using a drop of kyanoacrilic fast drying glue or similar to prevent the wires from sliding through the holes of the Sunprimer box when pulled.
The remote switch, when in a CLOSED position, cuts out Q2 and Q4 polarisation and discharges the 22000 uF capacitor.

Use of controller self-diagnostics

See Use of controller diagnostics
See Checking LED diagnostics

Detection of controller failures.

First, use the controller diagnostics as above indicated.
Second phase, check on controller. while CONNECTED.

Step 1:

Disconnect all cables and free the controller. Switch remote switch off, if there is one. Do not open the controller box.

Step 2:

Measure resistance from brown input to brown output. The correct resistance is INFINITE. If resistance is not infinite, the main transistor Q2 has burned out, and the only the D11 diode resistance or that of the burned out transistor remain. If only the Q2 transistor has burned out, the resistance measured will probably be about 800000 ohms on controllers MkI/a and Mk I/b or just a few ohms on the Mk I units, or alternatively just a few ohms with the Mk I/a and Mk I/b units where both the main transistor Q2 and the diode D11 have burned out. The correct resistance from brown output to brown input is INFINITE with the Mk I/a and MK I/b units and about 800000 ohms with the Mk I. However, this measurement is not so significant because if the D11 diode is good, transistor Q2 could be burned out without affecting this resistance value. This last situation is VERY DANGEROUS FOR PUMP SAFETY and probably represents 99,9% of controller failures.

Step 3:

Measure resistance from blue input to black input, and vice versa. The correct resistance in each case is about 27000 ohms. If one of the readings is just a few ohms, then the main transistor Q1 has burned out. This is NOT DANGEROUS for pump safety, but the controller remains in series mode and the pump will either not start at all or will only start with good insolation. Replace failed transistors and/or diodes and repeat the above check. Otherwise proceed to make a more detailed check as follows :

Maintenance of the Sunprimer Controller : Detection of Controller Failures

To carry out a more detailed check, the following equipment is needed:
- an amp gauge and a volt gauge placed on the output towards the pump
- a volt gauge from blue input to black input
- an oscilloscope from blue input to black input
- a volt gauge between the poles of the large 22000 uF capacitor
- an oscilliscope between the legs (poles) the small 0.22 uF capacitor
- a tap or valve on the pump output
- a pressure gauge on the pump placed for reasons of safety BEFORE the tap or valve.

The checks are easier to carry out:
a) if the PV panels are positioned at about half sunlight so as to get 1.5 - 2 amps when shorted,
b) if a 2.6 cam pump is used (but this is not essential)
c) if the pump is in a small tank nearby.

Step 1:

Open the controller box and check with volt gauge to see if the large 22000 uF capacitor is completely flat. If it is not, wait for it to go flat or short it. BE CAREFUL! WARNING!!! The sparks are nasty!

Step 2:

Connect the instruments and set the switch inside the box on 4 panel mode, connect the pump, and OPEN THE TAP OR VALVE COMPLETELY.

Step 3:

FINALLY ....... connect to the panels.

Step 4:

To connect the legs (poles) of the small 0.22 uF capacitor use special flexible long armed WELL INSULATED tweezers. Oscilloscopes are needed to view the PUT Pulses of the pulse generator oscillator, but small, cheap, piezoelectric buzzers without built-in oscillator but with small loudspeakers joined to external oscillators can also be used. Since they are very high impedence they sound only when voltage suddenly changes, so pulsation noises can be heard with the ear near to them.

Step 5:

We recommend step 4 first be done with good controllers to check instruments and equipment and get used to good controllers before checking failed ones. Connect FIRST the pump and ONLY THEN, SECONDLY the panels. When the panels are connected, output current is initially zero or just a few milliamps and the large 22000 uF capacitor voltage goes from zero to about 60V over a period of about 2 minutes. The capacitor then releases its surge, the pump starts, and the output current becomes the current needed to drive the pump. If 22000 uF capacitor voltage does not go beyond 34V and does not give surge to the pump, one or more panel connections are not good. If the 22000 uF capacitor gives surge to the pump but the pump does not start, this means that input power is too low or output pressure is too high. If the pumps starts and output voltage is more than 40V, then the controller is in series mode. Blue/black input voltage will be 1-2V. The small 0,22 uF capacitor will produce oscilloscope or piezoelectric loudspeaker voltage pulsations. Blue/black input will produce either no oscilloscope or piezoelectric pulsations or very minute ones. If the pump starts and output voltage is less than 34V, then input power is too low. In this case put the panels in a better position, or wait for more sunlight, or use more powerful panels. If output voltage is still below 34V, then the controller is in parallel mode and maintains this position. Blue/black input voltage is about the same as output voltage. The small 0.22 uFcapacitor produces oscilloscope or piezoelectric loudspeaker pulsations. Blue/black input will also produce oscilloscope or piezoelectric pulsations. In this situation, the controller is in order and the fact it does not switch from parallel to series, even with high input power, means that the pairs of panels are not equal or are not receiving an equal amount of sunlight.

When we initially connect the electric cables, output current is zero or a few milliamps and the 22000 uF capacitor voltage goes up from zero to about 60V in 2 minutes, and the controller is in series mode. If blue/black input voltage is observed during this phase, it will be 1-2V. The 0.22 uF capacitor will produce oscilloscope or piezoelectric loudspeaker pulsations. The blue/black input will produce no, or very small, oscilloscope or piezoelectric loudspeaker pulsations. Therefore, with a good controller, the PUT pulsation generator oscillator will always be running (each pulsation being between 0.2 - 0.8 seconds) in both the series and the parallel modes. In parallel mode, there is voltage between blue and black input with pulsations; in series mode there is no (or a very low) voltage nor are there any pulsations (or just very very small ones) between the blue and the black input. If, with the pump running, the PV panels are very slowly turned away from the sun, thus decreasing insolation, output voltage lowers very slowly, and when it goes below 22V the controller stops the pump. Before this, between 32V and 24V the controller passes from series mode to parallel mode. When the panels are in parallel, the pump noise is different (it is less "round") and the blue/black input voltage is not near zero but is about the voltage of the panel pair, and oscilloscope or piezoelectric buzzers will reveal input voltage pulsations from blue to black. If the PV panels are turned very slowly back into the sun, thus increasing insolation, output voltage increases very slowly, and the controller passes from parallel to series mode, blue/black voltage goes down, and the input voltage pulsations between blue and black either disappear completely or become very very weak. If the panels are turned away from the sun again, the controller passes from series to parallel mode again, and so on. If a controller passes from series to parallel but not from parallel to series, either the panel pairs are not equal, or they are not receiving equal sunlight. If we slowly turn the panels out of the sunlight, when output voltage is about 24V, the controller stops the pump, output current is zero or just a few milliamps, and the whole cycle starts again.
If, by way of experiment, various resistors are in turn now added (never more than one in all!) to the cable between the controller and the pump, we will find there is a threshold. If resistor value is below that threshold, measurements will be as described above. If the resistor value is higher than the threshold, the pump stops when output voltage is low, but the controller DOES NOT STOP CURRENT and the output current is always that of the panels, the cycle does not start, the capacitor does not charge, and so on. Nothing works any more. This is a situation which is DANGEROUS FOR PUMP SAFETY. To start the system off again we must switch off either the panels or the pump. The resistance threshold mentioned is NOT FIXED. It is lower with higher panel power and high pump current load, and vice versa. So a present limitation of the Sunprimer Mk I controller is that it does not work well if cable resistance is too high, or if current is too high, because switching off is currently based on an output voltage threshold of 20-21 V. Thus, if:
(cable resistance + motor resistance) x (output current) > 20V
The Sunprimer MK I controller does not switch off. This is why it is so IMPORTANT to use LOW RESISTANCE cables, with 1.5 ohms as a good safety limit, to avoid heads which are TOO GREAT FOR THE CAM SIZE used (load must be less than 4 amps), to AVOID CLOSURE OF VALVES and taps on water pipes. Since the motor resistance is 1.8 ohms, the system is inherently safe if the prescribed precautions are taken. If we very slowly close a valve or tap instead of turning the panels into and out of the sun, as we close the valve, output voltage lowers gradually and when it is between 32V and 24V the controller passes from series to parallel mode. If we open the valve slowly again, the controller passes from parallel to series mode. If we continue closing the valve or tap to the point where voltage is about 24V, the pump stops (KEEP YOUR EYE ON THE PRESSURE GAUGE!!!!! DO NOT ALLOW PIPE PRESSURE TO GO TOO HIGH!!!!!). Once the pump the valve has been closed and the pump has stopped, the controller will switch from off to on according to the value of the product of:
(cable resistance + motor resistance) x (last output current)
If the controller stops the current, a new cycle can start, and if, purely by way of experiment, the valve or tap remains closed, the pump will try to start again every two minutes. If the controller does not stop the current, output current is always the current of the panels, a new cycle does not start, the capacitor does not charge, and nothing works any more. THIS SITUATION IS DANGEROUS FOR PUMP SAFETY. To start the cycle again, either the panels or the pump have to be switched off.

Step 6:

With the pump running, short the remote switch wires if there are any. If all is in order, Q2 and Q4 polarisation is cut out, the 22000 uF capacitor is discharged, the controller is switched off and will remain switched off for so long as the wires remain shorted.

Step 7:

We can now proceed to check faulty controllers. The PUT 2N6028 oscillator should be in operation and we can check this by looking at the oscillator or piezoelectric loudspeaker connected to the 0.22 uF capacitor. If PUT 2N6028 is not working, the controller remains in parallel mode. This situation is NOT dangerous to the pump, causes only 0.01% of failures, and is due to the fact that the PUT 2N6028 unit is mechanically rather fragile. If PUT 2N6028 is not working, first replace the 10000000 ohms resistor using a 1800000 1W resistor, which is mechanically more resistant than the 10000000 ohms resistor and has more or less the same function. Then replace PUT 2N6028 with either another 2N6028 or with a 2N6027 unit.

Step 8:

To check the repair operation, when the controller is reconnected, if current is NOT either zero or just a few milliamps but equal to panel current and if capacitor voltage does NOT gradually go up to 60V, or if current remains just a few milliamps lower than the current from the panels and the capacitor voltage remains low,
- disconnect pump and panels
- connect pump and panels and try again.

Step 9 : diagnosis

If starting current is still not zero or just a few milliamps and capacitor voltage still refuses to go up to 60V, then damage has occurred either to the main transistor Q2, to transistor Q4, to thyristor BTA16, or to diac DB3. This situation is DANGEROUS TO PUMP SAFETY. If starting current is a few milliamps and lower than the current from the panels, and capacitor voltage remains low, then either thyristor BTA16 or diac DB3 has burnt out. THIS SITUATION COULD BE DANGEROUS TO PUMP SAFETY. If all appears to be in order, but the 22000 uF capacitor releases its surge and starts the pump at voltage 30V, then the 27V zener diode has burned out.

Step 10 : further diagnosis

If all appears to be in order, with initial output current zero or just a few milliamps, but capacitor voltage goes up to maximum 40V and then stops without starting the pump, first check PV panels and connections, then check whether the controller is in series mode or parallel mode by viewing blue/black input voltage and pulsations and at the 0.22 uF capacitor pulsations. If blue/black voltage is low, the controller is correctly in series and the problems lie with the panels. If blue/black voltage is high and there are NO blue/black pulsations, then check whether the 0.22 uF capacitor pulses. Of it does not, then the oscillator has failed. If it does, the problems are in the nearby solderings.

Step 11: further diagnosis

The 22000 uF capacitor reaches 60V, releases it surge and discharges. If the pump still does not start, input power is too low or output pressure is too high, as the probability that the Q2 or the Q4 resistor fail when the system is switched off is so remote it can be disregarded.

Step 12:

If the pump starts, either close the valve or tap very slowly, or slowly turn the panels out of the sun, to see whether the controller passes from series to parallel mode, and then open the valve or tap slowly or turn the panels slowly back into the sun to see whether it switches back from parallel into series mode. If the controller remains in series amd refuses to pass into parallel mode then either transistor Q1 or transistor Q3 has burnt out. This is NOT dangerous to pump safety, but the pump starts only with a good insolation or does not start at all. If the controller passes from series to parallel, but refuses to go back into series, then the pairs of panels are not equal or are not getting equal sunlight. If the panels are turned out of the sun to the point where output voltage is less than 24V, the controller should stop the pump, and output current should be zero or just a few milliamps. If the pump stops when output voltage is low, but the controller does NOT stop the current and current output is the same as that of the panels, either the main transistor Q2 or transistor Q4 or thyristor BTA16 or diac DB3 has burnt out, which is VERY DANGEROUS TO PUMP SAFETY or wire resistance is too high. If the pump stops when output voltage is low, but the controller does not stop current, and output current is lower than that of the panels (just a few milliamps) and capacitor voltage remains low, either the thyristor BTA16 or diac DB3 has burned out. This COULD BE DANGEROUS TO PUMP SAFETY.

Step 13:

Q2 transitor failure in an electrically DISCONNECTED controller can be detected by measuring voltage from brown input and the brown output wires.
Q1 transistor failure in an electrically DISCONNECTED controller can be detected by measuring voltage from blue input and black input wires.

Step 14:

Series/parallel oscillator failures can be detected in an electrically CONNECTED controller by measuring 0.22 uF capacitor pulsations.Failure to pass from series mode to parallel mode : failure of transistor Q1 or Q2 Failure to switch off and to disconnect pump from panels with panel current always present :failure of transistor Q2, transistor Q4, thyristor BTA16, or diac DB3
Failure to switch off, failure to disconnect pump from panels, with current a few millamps and lower than the current from the panels : failure of thyristor BTA16 or diac DB3 Capacitor gives surge at much less than 60V when the switch is in the 4 panels position : failure of the 22000 uF capacitor.
Failures of other components and solderings can only be found by checking each component and soldering.

THE ONLY SITUATION DANGEROUS FOR PUMP SAFETY IS WHEN THE CONTROLLER FAILS TO CUT OFF CURRENT FROM THE PANELS, that is when either transistor Q2 or transistor Q4 has burnt out. Since it is highly unlikely that transistor Q4 burn out, the fault will nearly always lie with transistor Q2.

Step 15:

For passive lightning protection refer to the Solar Spring installation manual. Active lightning protection is limited.

Instructions for the substitution of thyristor BTA16 and DIAC DB3

. Sunprimer controllers are built using rugged parts manufactured by leading american electronics manufacturers. Despite careful screening at the factory, on very rare occasions it may occur that either the thyristor BTA16 or the DIAC DB3 not be produced according to their manufacturers' published figures. Qualified dealers may themselves rectify such defects as follows:
1) The Sunprimer circuit soldering has been covered by a protective paint. To make sure the replacement work is well done, the paint must be carefully scraped away at the work points before de-soldering and re-soldering is carried out.
2) Use a soldering iron suitable for electronics. Work carefully. Please do NOT try to use electricians' or metal-workers' equipment.
3) In case of difficulty sourcing a BTA16, any other thyristor (or any SCR) will do the job provided it is 16 amps nominal.
4) In case of difficulty sourcing a DB3, any other DIAC is satisfactory provided it is 28-32 volts nominal.


Menu for the maintenance of Solar Spring pumps.


Main menu Solar Spring pumps.


Some recommended technologies,


List of attachments to the Model.


Complete index of the Model.


Model Homepage.


Typical list of graphs and drawings.
List of abbreviations used.
List of key words.
Documents for funding applications.


Homepage Bakens Verzet