A work-in-progress, but running tolerably for a few years. Cye badgered me into posting following the purchase of one of his pumps...
It's my approximation of the Solartwin concept.
Collector is about 5' square, and is 10mm twinwall roofing front and back, screwed through into battens, and hung on galvi brackets hooked under the tiles. Actual collector panel is DPC membrane w/ a snake of 10mm black vinyl pushfit hose.
Since this photo, I've doubled the piping with a second staggered snake. Twenty-odd metres of pipe in the collector now.
Flow connects to the DHW flow/expansion pipe, inlet to the DHW supply.
Circulation is via Cye's pump. After a previous controller like Norm's (same velleman kit but modified for reverse-action and greater hysteresis, sensing flow from the collector) I invested in one of Solartwin's new 3-sensor differential controllers. I power this via a 7.5v unregulated DC supply with a cheapo 10W PV ontop for speed control.
So how does it do? Not bad. Currently I'm getting a full tank of water above 40C.The Solartwin controller works well. Panel exit temps can be around 55C so I'm working to reduce the flow a touch. This is tricky as the pump likes to catch entrained air and airlock if flows drop too much.
David, thanks for sharing details of your solar heating system. That's a really nice panel you've made.
I am not familiar with the SolarTwin controller - What does the solar twin controller do that another general purpose differential temperature controller wouldn't?
And is the 7.5v PSU topped up with PV providing power directly to the pump and if so how have you wired these together? What is the controller powered from?
The jacksons camping site is a good read and the solar pods are an interesting design and look really nice. Hard to compare 'panel size' of pods with traditional two dimensional glazed flat panels, but what does strike me is that the solar pod plus, with an effective area of perhaps 0.6-0.7 sq metres, are probably on the dear side at £230 given that a 2 square meter commercial/accredited flat panel is available in NI for around the £300 mark. On the other hand the 3-dimensional aspect of the pods would however be more tolerant of poor roof inclination and orientation, and being smaller, would be easier to fit?
Cye, the ST controller is *now* nothing special. But when it first became available it was the only controller that could be powered by PV., that I could find. Everything else was mains-driven, for a mains pump. What does it do? simultaneous display of all three temps- panel, plus tank top & bottom, adjustable on and off differential as well as run-on; plus overtemp. It's powered by the PV and automatically switches sampling time between day/night based on V-in.
Best case the PV is large enough to fully control the pump; mine is a touch small to provide sufficient start-up current when the pump is first called for. So the best answer for me- remembering that the controller wants to see 0v overnight- is the PV and DC supplies in parallel, the PV protected by a series diode. The wallwart is plugged into a cheap timeswitch so that supply is on between 0800 and 1800 or so. I'm south-facing but a stand of conifers kill the morning sun for a while at this time of the year.
Who's selling panels for £300 out of interest?
The pods intrigue me. Having not previously considered a H/E or dual-coil tank, freeze-tolerant was always my focus and plastic is more stretchy than copper. I've also looked at the flexible-mat pool collectors, but they tend to big to build a housing for.
I like the idea of the two power sources in parallel. Is the PSU switch mode and/or how does the psu draw less power when the PV is working at peak output in perfect lighting conditions?
Re the jacksons site, i had also been eye-ing the flexible mat solar water heaters. If these could be glazed cheaply and easily, they'd be very good value. I wonder can the mats be cut up and made into smaller units? If they could be adhered to the dark side of a pv panel they would not only increase the efficiency of the pv, but they would also produce useful heat too.
The guy that was selling the flat panels is Vinny Hamilton in enniskillen. I've inspected one and they're very well put together. He's the guy that makes his own biodiesel - see the pic of the panels he's selling at the tail end of the following thread
the psu is an ordinary transformer one, a cheapo switchable. So consumption is minimal but doesn't vary much on load. Ideally you want flow to match energy, IMHO, to keep flow temps up. But motor starting current, flow restriction, pump cavitation, all affect the range you can work within. There's a reason ST use that positive-displacement pump...
re. the pool heaters, I've looked at moulded panels as well as the El-Nino tube variety. Pretty sure the tube type could be cut down to suit a sensibly-sized enclosure but I'd really love to see one come unassembled.
That's a very useful technique, using PV in parallel with a psu to vary pump speed. Do you also have a diode on the smpsu output as well as on the PV output?
I think that varying pump speed to match available solar energy is more of an issue for direct flow systems such as your's and Colin Lloyd's and shouldn't be a great concern for indirect systems.
With pumped direct flow arrangements, too high a pump speed and you encounter the problem of destratification in the hot water tank. (Destratification = Lots of luke warm water all through the tank and no hot water at the top of the tank). Too low a pump speed and your panel boils/stagnates. A fine balance is required.
In terms of pure collector efficiency, i.e. how much of the solar energy is converted to heat energy, it has been shown that energy capture efficiency on any type of panel is maximised when the temperature difference between the panel and ambient air is minimised. This means that ideally a panel shouldn't be allowed to get very much hotter than ambient air if one is aiming to maximise panel efficiency, which in turn suggests that minimising flow rates may not always be ideal. For flat panels the energy capture efficiency of the panel drops much more significantly at high panel temperatures than with vac tubes.
So I guess that minimising pump speed appears to be more a necessary compromise for direct flow systems, is likely to result in lower collector efficiency, and should be far less of a concern for those with those indirect flow systems where destratification is not influenced by pump speed (twin coil tank, solasyphon, plate heat exchanger, etc).
I'm aware that there may be a few readers who are possibly unfamiliar with terms used here such as 'direct flow' , the SolarTwin (ST) type of design, etc & I'll try and explain these here. David, perhaps you can provide additional info where I've missed or misunderstood anything?
David's system (described here) is a homebuilt variation of a solar water heating system design commercially available from a GB company called SolarTwin. There are quite a few of these homebuilt systems made by replicating the basic features of the SolarTwin system, and these are lovely simple low-cost systems which are very suitable for people on a low budget and/or keen to make as much of their solar system themselves. Whilst SolarTwin sell their commercial system for thousands, this type of system can be cloned at home for probably under £500.
The main features of the ST design are:
 Direct Flow of hot water into the domestic hot water (DHW) tank. This means that, unlike most other solar water heating systems, the water that is heated in the panels actually is the same water that ends up coming out of your hot tap or shower.
No additional heat exhanger is needed in your hot water tank for the solar hot water and so in many cases your existing hot water tank can be retained. This is one reason why the 'direct flow' method of getting your solar-heated water into your tank appeals to people on a budget and DIYers alike. The plumbing required in your hotpress is limited and you generally don't need to rip your existing tank out!
The advantages of direct flow are it's simplicity and low cost. The disadvantages include the fact that you cannot use antifreeze, that direct flow is unsuitable for hard water areas, and that you have to be very careful with the speed the water is circulated from the panels to the tank and back.
 Freeze tolerant pipework and freeze tolerant solar panel. Because antifreeze cannot be used, the external pipework to/from the panel and the panel itself would be damaged in winter if the pipework was made of the traditional copper. One solution here for the direct flow method is to drain down the panel in winter - This requires effort and is fiddly. Another solution is to have the pump controller circulate warm water from your tank out to the panel when it is in danger of freezing to prevent bursts - This is called 'freeze protection' and, whilst slightly wasteful on a vacuum tube panel, is very wasteful on a flat panel setup and you'll just be burning oil or gas to keep the panel warm. The ST type method is very zen and rather clever, and uses silicon pipework which can freeze without bursting. David's panel is a variation on this theme and employs less expensive PVC pipe which is also freeze tolerant. All in all a clever solution to working with a flat panel connected to a hot tank without a heat exchanger.
David - can you add any further background info on the ST design, e.g., on their rationale for using the little PV panel?
All good, Cye. The PV? Two good reasons. One is that there's no necessity for mains power (and along with this the fact that small DC motors are more efficient than the AC brutes derived from C/H pumps), the other is the ability to alter pump speed and therefore flow dependant on the available solar energy. Then, their (expensive) diaphragm pump is happy to pump low flowrates into a high head- ST only move a couple of litres per minute but can do this against the 3m head from hot tank to roof.