A few small repairs

[kitestrings]

A few of you may remember an album by this name.  It occurred to me that, on his list of repairs, “Sonny” torched the house as I recall.  Okay, getting past that – I really was just working down my list of fixes, and, maintenance before winter.  I greased the turbine bearings, yaw head and lubed the tail hinge.


I also replaced this switch.  Unfortunately the gasket hadn’t sealed well and water got in to it.  It is just a parking brake, but it is handy and adds an element of comfort in really high wind conditions.

I put a new PVC cover on it and seal the plugs with silicone.


Back on the ground I replaced the water heater elements.  They’ve been in since about 2009 IIRC.  One had died earlier, so I’d run for the summer with just the one, until it too failed.


The new ones are stainless, and can be wired for 1,000 or 2,000 watts.  After a bit of observation, I decided to leave them at 1,000w.  At 56V, where we more typically are diverting, it is closer to 22-23A or a little over 1,200w



I also flushed the LPG unit that this preheat tank delivers to:

[Bruce S]

Bummer about the seals not working out, good to see you got to it long before is was needed.

Cheers
Bruce S

[kitestrings]

A couple things I’d meant to mention –

The reason the elements are cut is because it was the easiest way to remove the port bushings.  This type of tank uses a large NPSM (straight thread) element port with flat gaskets, so there are port bushings that adapt from 1” NPT for the elements used (both old and new) to this larger port.

The stock thermostats control the elements via solid state relays mounted on a large heat sink.  My understanding is that diversion at high frequencies can be a challenge for these relays.  The Classic, in “Waste-Not” mode operates in the range of about 500-700 Hz IIRC, so there can be voltages spikes that exceed the rating of the devices.  The engineer at VB Controls was quite helpful some years back.  I had been using fly-back diodes, but he also provided us with a capacitor that further suppresses these spikes.  It seems to have worked well.


An earlier pic (before the fin tubes were reoriented)

[joestue]

yeah you do not generally want to pwm a solid state dc relay at more than i'm guessing.. 10 hertz.

the optically powered mosfet gate drivers can't drive the mosfets very hard.

this is a random 200v rated 40 amp relay, has an on state voltage drop of 2.8 volts at full load, so its not a large mosfet in there..

600us turn on time and 2600uS turn off time.

https://www.mouser.com/datasheet/2/418/5/NG_DS_SSRDC_0218-1488273.pdf

[kitestrings]

Bruce,

The seal that I was talking about is just the cover seal, or gasket for this toggle switch, if that wasn't clear; nothing on the turbine itself.

joestue,

These are operating at much higher frequency, at times.  A typical day might start with a very low pulse rate, just burning off the excess relative to your set-points but rises rapidly as production exceeds need at the batteries.  Then, in our case, the relay might be full ON for a chunk of the late day period, and resume PWM near dusk.

Waste-not is described here:

https://www.midnitesolar.com/faqList.php?faq_ID=137&parent_ID=2&faqCat_ID=51&faqCatName=Classic%20Charge%20Controller&showAnswer=true&subCatName=



These are the relays we've been using:
https://www.power-io.com/products/hdd.htm

[MagnetJuice]

Thanks for explaining why the heating elements are cut, I was puzzled, was wandering how an element can ‘explode’ like that. 

Time to get ready for the winter.

Ed

[OperaHouse]

I've been testing a number of DC to DC solid state relays and they are quite disappointing.  The fastest ones are transformer coupled and the optical ones are quite slow, up to 8.6ms with slow ramping.  Most concerning is the rapid shutdown circuit on both.  If the turn off has an analog component with a slow ramp down on the control voltage, the drive to the FET gate is linear and allows the gate drive to be as low as 1V.  That will burn out the FET.  Pulsing can cause that.  Self powered from PV can be dangerous if drive voltage becomes low.  I wouldn't buy one which is rated less than 60A, that typically gets you an IRFP4332. Two SSR in parallel is the best solution for reliability dropping the heat generated dramatically.

[MattM]

By comparison,how fast are magnetic relays?