SunAmp PV spares and maintenance issue

[TerryE]

The back story here is that I was one of the SunAmp early adopter buying and self-installing 2×SunAmp PVs rigged in parallel and heated by mains E7 electricity.  I have been very happy with the product and I did not see the need to change to the UniQ range as I expected at least a 10-year operating life. 

 

A couple of day ago, Jan noticed that the HW isn't running too hot at the mo. Checking through my HA data logs, one of the SunAmps failed about 10 days ago, and cooled down over a couple of days.  So we were mixing roughly 50:50 60°C:20°C into our HW manifold -- which is at about the temp we shower at in our warm climate, hence our not noticing.

 

So I opened up the unit and did some diagnostics.  The line through the inline heater and thermal cut out is reading 18Ω which is nominal for 240V 3 kW heater, so this isn't a heater failure.  The two thermistors register the same resistance; I don't know what the part no is but since they are within 1% of each other, I suspect that they are fine.  Power is arriving correctly to the control board, but and it seems to be sequencing the pump fine.  It just that the 12A 240V heater O/P isn't turning on.  So I suspect that the power relay has died and since this a soldered-on component then it is time for a controller board change, IMO.

 

So Q for @Nickfromwales, @JeremyR etc. do you have a SunAmp contact that I can discuss getting a now controller board from. 

 

Overall I am impressed by the internals of the unit apart from a few quibbles and one more serious concern.  The quibbles are about the layout of the pipe runs in the head -- for example replacing the inline heating element would be a pig to do -- but this is the sort of issue that could be expected from early iterations of the product.  My greater concern is the controller board design.  There seems to be what I would regard as inadequate isolation between the 240V and low voltage areas on the board. Not good.  They've also seemed to have use the same budget relay for switching the pump and the heater.  Even though many of these units are nominally rated to switch up to 12A at 240V, this is at the limit of their operating range.  Switching AC at the power is really hard on the relay contacts and so these tend not to have a good operating life. Whilst this might be acceptable in a low cost easily replaceable unit, it is not in the critical control board of an expensive piece of kit. 

 

Jeremy and I have discussed off-the-self vs. self-built trade-offs but my concerns here are: What is the realistic life of such a board? How long will they be obtainable given that the product is past end-of-life?  Will SA require some price prohibitive replacement cost / installation cost?   I am loath to go to the hassle of replacing what seems to be excellent units because of design flaws in a controller board.  My fallback would be to swap out the controller entirely and do my own control implementation but using a CE certified unit for the 240V switching, and that way I can get my local electrician to check over the 240V installation.  Not my preferred route, but easily doable.

[Jeremy Harris]

@TerryEHave you checked the thermal switch on the heater block?  This was a common failure point on early Sunamp PVs, mine failed and I had to replace it.  The usual symptom is that the thing trips (can be reset by depressing the button in the centre of it, with the power off).  If it keeps tripping then it probably needs replacing.

[TerryE]

6 minutes ago, JSHarris said:

Have you checked the thermal switch on the heater block? 

 

Yup I thought about that which is why I measured the resistance across the switch and the heater -- which gave 18Ω rather than open circuit, which would have been the case if the switch had tripped.

Edited August 3, 2019 by TerryE

[Jeremy Harris]

6 minutes ago, TerryE said:

 

Yup I thought about that which is why I measured the resistance across the switch and the heater -- which gave 18Ω rather than open circuit, which would have been the case if the switch had tripped.

 

Bugger, thought that might have been a quick win.

 

The relay on the circuit board always seemed to arc a bit as it switched, so, as you suggest, that could be the failure.  There's a diagnostic mode available on the circuit board, that sequences the four LEDs and does a self-test on the logic side, plus tests things like the ultrasonic flow sensor output.   I'll try and see if I still have a copy of the instructions for entering the self-test mode.

 

IIRC, the power relay is a standard part, so replacing it should be fairly straightforward.  I have a feeling that the relays are all soldered to the control PCB though, rather than in sockets.

[TerryE]

1 minute ago, JSHarris said:

I have a feeling that the relays are all soldered to the control PCB though, rather than in sockets.

which why I prefer socketed or Din-mounted contactors.  Line replaceable parts.

 

As far as BITE goes, I see that the board has a couple of UART jacks as well as the SMS interface, so something should be possible, though I am not sure how it could self diagnose a relay failure.

 

.  Even having a test point to allow you to probe the relay demand inputs would be nice.

[TerryE]

I am not really comfortable poking around a PCB blind when it has 240V tracks on it

Edited August 3, 2019 by TerryE

[Jeremy Harris]

17 minutes ago, TerryE said:

I am not really comfortable poking around a PCB blind when it has 240V tracks on it

 

I'd be inclined to just buy and fit a couple of new relays and see if that fixes it.  Perhaps worth buying several spares, given that the postage is probably going to be a fairly significant part of the total price.  FWIW, I tend to stick to well-known brand names when it comes to relays.  Omron are generally pretty good.

[vivienz]

I can't add anything to the discussion of the circuitry itself but I asked @Nickfromwales to acquire me at least 1, preferably 2 spare control units for my SA that will be installed very shortly. Given the projected lifespan of the PCM obsolescence of the unit is far more likely to occur through failure of the electronics than other matters. Given the direction that SA as a business seems to be going in, there's no guarantee that spares will be available in a few years' time, so I would far rather pay for a couple of spares now and mothball them than find myself with no way of replacing them in the fullness of time.

 

[Nickfromwales]

Future proofing is never a bad shout. 

@TerryE Wait until Monday if you can, and I’ll make some calls.   

When exactly did you buy the units? 

[Jeremy Harris]

With luck they may still have our old Sunamp PV PCB, as when I sent it back I'm pretty sure they only wanted to examine the internals of the heat cells.  Our Sunamp PV was still working perfectly when I sent it back, although I had replaced the thermal cut-out.

[MikeSharp01]

Is there scope to create an aftermarket control board or is the software to control the thermal cycles a bit special.

[Jeremy Harris]

Just now, MikeSharp01 said:

Is there scope to create an aftermarket control board or is the software to control the thermal cycles a bit special.

 

Shouldn't be too hard for the Sunamp PV.  Just needs a control loop to adjust the pump speed, using the flow sensor and temperature sensors.  The control's pretty simple, it just aims to keep water flowing around the charge hydraulic circuit at ~70°C if it can, shutting down if there isn't enough exported PV to maintain the required charge temperature.  There's a few added bells and whistles, like an over-run timer that keeps the pump on after the power to the heater drops, to prevent/reduce heat soak, plus a sensor that detects when excess PV is available on the main power input and uses that to turn the control system on and off, but that's pretty much it.   Unlike the Sunamp Uniq range, there's no sensors down in the PCM, so the control system is, in essence, a lot simpler (although electromechanically it's more complex).

[dpmiller]

you could socket the pcb for replaceable relays.

[TerryE]

1 hour ago, Nickfromwales said:

When exactly did you buy the units? 

 

Nick, they were installed an 17 Jan 2017, but apart from sharing my good experiences on the forum, I've not had anything to do with the company itself, so perhaps you might have a bit more leverage so thanks very much for your kind offer of doing the contacting.  Having a replacement  / spare controller would avoid me doing one of the options below

 

1 hour ago, vivienz said:

Given the projected lifespan of the PCM obsolescence of the unit is far more likely to occur through failure of the electronics than other matters.

 

The cell is themselves as a simple as can be: about the size of a 20 ltr Jerrycan, with water in & out at the top and a safety vent on the other side.  The heater and mechanicals should have a 10-year life at least.  The weakness is the control board and especially those these budget mechanical relays. Other than the relays, the main design weakness on the boards are the couple of chunky electrolytic caps on board which (depending on quality) tend blow or leak in that 10-year time frame.

 

I've been brooding about this, ILS and reliability.  Sorry ILS - Integrated Logistics Systems and probably JSH is the only other sad fart on the forum that I could had this conversation with.  The military were into this stuff 40 years ago and probably still are: how you repair and maintain kit in the field; when you repair in place, when you junk and replace, when you sap out and send back to central maintenance for repair, etc.  The current wave of consumer devices don't even think about such stuff any more: if its broken then get junk it and get the latest model.  But for me as an end-user, I don't want to be out of water or heating for weeks just because some manufacturer has declare a model end-of-life.

 

Parallel failure modes are good; they give you a step improvement in MTBFs. Serial failure modes just make things worst: example a single engine plane and the engine fails -- you are going to come out of the sky and you might get out of with you life. A twin rotor Chinook is far worse -  either rotor fails and you are going to die, but with a twin engine plane when one engine fails -- you will need to land at a nearby airfield (though the latest twin jets such as the 777 have can an ETOPS rating of over 300 which means that "nearest" can be 5 hours flying away.)

 

So having twin SunAmps means that I can risk taking a month or so working out the best approach. (I also have one working system to instrument and to work out the operating characteristics.)  By way of another example with our UFH: having an ASHP + Willis means that neither system is critical.

 

Back to the SunAmp, my preference would be switch the 240V AC with a couple of panel mounted Crydom SSRs which can be driven by a 50mA 5V TTL input.  Field replaceable on failure.  These and everything else could be driven by my favourite ESP8266 or ESP32 and could configured as MQTT clients driven directly by  my HA system, and this would also simplify my central panel as these would now be always-on devices from a power distribution PoV with no active control functionality.

 

[Jeremy Harris]

Probably no reason why the PCB mounted relays in the Sunamp PV control board couldn't be replaced with remotely mounted SSRs.  Most SSRs accept a wide range of input voltages, so there's a pretty good chance that they would switch OK with the signal that normally drives the electromechanical relays on the PCB.

[ToughButterCup]

In common with, I suspect,  a good few of us on BH who are both building and watching any post on Sunamp closely, I'm one heartbeat away from saying Stuff-It-SunAmp.   

 

I understand about 40% of the content of posts about SA. Why has hot water - any hot water -  been made so damn hard?

Quote

there's no sensors down in the PCM, so the control system is, in essence, a lot simpler (although electromechanically it's more complex).

 

[vivienz]

It wasn't so much a great nervousness about SA that prompted me to request an additional unit so much as our experience at work. OH and I run a small manufacturing business and our machines are all varying degrees of old. The mechanicals themselves are robust and have stood the test of time very well but the thing that leaves us most exposed is potential motherboard failure on one of our bought in brush making machines as the electronics go back to the late 80s and will no longer have manufacturer support in a few years time. I was pondering this whilst discussing the SA with Nick and it struck me as an obvious and easy to achieve way of time proofing the SA unit as one can never be sure how long manufacturer support will be available, particularly with a relatively young company who's owners may want to sell it one once established.

[Ed Davies]

31 minutes ago, AnonymousBosch said:

Why has hot water - any hot water -  been made so damn hard?

 

The Sunamp PV is pretty simple, simpler than most oil or gas boilers.

 

It's just a question of getting support for something which was made in the hundreds, not the thousands. What I can't figure out from this thread is with @TerryE has had any problems getting said support or hasn't tried yet.

[ToughButterCup]

1 minute ago, Ed Davies said:

The Sunamp PV is pretty simple, simpler than most oil or gas boilers.

[...]

 

Thanks Ed. Simpler than PV Direct hot water heating?

 

7 minutes ago, vivienz said:

It wasn't so much a great nervousness about SA that prompted me to request an additional unit so much as our experience at work.

[...] 

a way of time proofing the SA unit as one can never be sure how long manufacturer support will be available, particularly with a relatively young company who's owners may want to sell it one once established.

 

Thanks for your directness, @vivienz.

That's it - if someone as sensible, thoughtful and cautious as you  is buying two off to ensure redundancy (as did @TerryE) then it isn't  for us for us any more. We just don't have the budget. Or the rather we don't have the spare financial and emotional budget to cope with the roller coaster associated with early adoption.

 

Bugger.  - PV Direct, then.

[joth]

10 hours ago, TerryE said:

Parallel failure modes are good; they give you a step improvement in MTBFs. Serial failure modes just make things worst: example a single engine plane and the engine fails

 

This is nice analysis, and got me thinking about the equivalent issues on the mixergy tank we're planning to use. (Posted over on the Mixergy topic)

 

I'm curious though that with your focus on avoiding serialised failure modes, you have an HA + mqtt setup you are confident enough for this. That means IP network needs to be operational to get heating on? I've seen too many cascading failures with IP networks to trust myself to run a mission critical one in my own home. 

My own policy is lighting and heating must work without any IP. Do you have a write up of your HA setup somewhere? Always interested to see how others have partitioned things. 

 

One other question: is this SSR relay controlling the immersion? With your proposed ESP32 controller would you modulate it at all, to maximize free excess from PV, or just binary switch it on/off?

I'm contemplating making my own PV redirect using solid state contactor with 0-10v input, but again interested in what others have done.

 

Thanks!

 

Edited August 4, 2019 by joth

[Jeremy Harris]

2 minutes ago, joth said:

One other question: is this SSR relay controlling the immersion? Do you modulate it at all, to maximize free excess from PV, or just binary switch it on/off?

I'm contemplating making my own PV redirect using solid state contactor with 0-10v input, but again interested in what others have done.

 

Thanks!

 

 

The excess PV diverter that I made several years ago still works flawlessly.  I used an electricity meter chip to do all the heavy lifting and spit out true power, then added some pretty simple code to take advantage of the 1 Wh dead band that's built in to electricity meters.  This means that you can shuffle less than 1 Wh across the grid without incurring charges.  I have my system set so that when the energy bucket reaches 0.75 Wh from PV export the 3 kW heater is turned on until the energy bucket depletes to 0.25 Wh.  There's a small leak in the bucket to ensure that cumulative errors don't build up over time.  I wrote the thing up and posted it here: https://picaxeforum.co.uk/threads/photo-voltaic-immersion-heater-power-diverter-safety-warning.24286/

 

I've since changed the design so that the unit in the meter box transmits a coded signal via an 868 MHz FSK link to a remotely mounted switch box in the house, that just has a receiver, decoder and 25 A SSR.

[ProDave]

My own home built solar PV diverter works by burst firing the SSR in a half second time frame.  So for each half second it can be on for anything from 0% of the time to 100% of the time.

 

I can report it works perfectly and is well withing the 1Wh dead band, which with my immersion heater being about 2.8Kw, equates to about 0.77 seconds.

[joth]

Great I didn't know about the 1Wh meter slack, but figured there must be something like that. We're on a smart meter so I'll have to experiment to see if I can see where it switches over.

At 3kW that's a about 1s grace, so half second sampling sounds minimum.

I was looking at this project that suggests the proportional SSR, which presumably just offloads the need to keep up with driving PWM while in a steady state. Still need to sample at about the same frequency to detect household load or generation changes. 

I like that it is din rail mounted so could all be installed by my electrician and all I have to deal with is sending the low voltage signal into it to set the target heating level.

 

Anyway, I'm going way off topic here! Thanks both for the excellent pointers, as ever. 

[Onoff]

2 hours ago, AnonymousBosch said:

 

I understand about 40% of the content of posts about SA.

 

About four times the understanding I have then! All the home brew controller stuff is where I get lost.

[vivienz]

2 hours ago, AnonymousBosch said:

 

Thanks Ed. Simpler than PV Direct hot water heating?

 

 

Thanks for your directness, @vivienz.

That's it - if someone as sensible, thoughtful and cautious as you  is buying two off to ensure redundancy (as did @TerryE) then it isn't  for us for us any more. We just don't have the budget. Or the rather we don't have the spare financial and emotional budget to cope with the roller coaster associated with early adoption.

 

Bugger.  - PV Direct, then.

 

I think I may not have been clear here, @AnonymousBosch , as it's not an additional SA in its complete form that I have requested but just the electronic controller unit.