Keston C25 h/w demand mod

[fred]

For Ed Sirett and anyone else who's interested, here's the mod I made to the control panel of my Keston Celsius 25 system boiler to create full heat demand for hot water operation.

The circuit is not the simplest one possible but for reliability it avoids having a relay in the high impedance signal path to the boiler demand level. The 4k7 resistor means that there is always a small current flow on the relay contacts which is something I prefer. The worst that can happen if the relay contacts fail is that the mod ceases to function rather than the whole c/h packing up.

The relay is mains powered and must have gold plated contacts (which is rare on mains relays), I used RS model 394-9116.

The mod is straightforward to implement but requires some soldering skill. The 10k resistor needs a cut in the boiler demand level signal which I made by cutting the control pot wiper lead (middle connection) and putting the 10k res in its place. The diode & the 4k7 resistor can then be soldered to the pot terminals on the back of the board and the relay connections wired up.

Full demand for the Celsius 25 is when the demand control level is at

0V; I have not found that the diode drop affects the operation significantly.

Control board mod:

+5V --- | ctrl / 10k modified panel \ demand pot / | level (100k) \ +5V | _|_ --- | 0V | | \4k7 | h/w / | demand >----- \ 1N4001 | 240V~ | |--|---------> boiler 240V RTN demand demand

Against my better judgement, here's a version that doesn't involve a mod to the control board but places a cut in the loom instead. Lots simpler, just a relay and wiring, but with the relay in the signal path so beware of dirty contacts affecting operation in the years to come. Again, relay contacts must be gold plated.

Loom based mod:

Control Board Loom

+5V --- | ctrl / RL1a modified panel \ demand pot / \o----| level (100k) \ o |---------------(#---------|------------->

_|_ 0V

h/w demand >----- 240V~ | --- - | | | o RL1 | | RL1b / coil | | o - | | | --- c/h >---------> boiler 240V RTN demand demand I both cases, be careful to separate the wiring on the two relay contacts as one is low voltage and one is mains, neat soldering is required and best to sleeve the joints or gloop with hot melt glue.

I've posted a couple of pictures of the first mod on alt.binaries.pictures.misc under the same thread name. If anyone is on a non binary server then don't worry, you're not missing much ;-)

HTH

[Ed Sirett]

Thanks a lot. The latter arrangement I had worked out for myself, but I can see your first design is less invasive.

After thinking about this I'm not yet inclined to implement the mod. At present the bathroom rad is also a simple restricted lock-shield bypass, so it would become lethal at 80C during water heating. It would also grossly over heat the 1.8m x 1.6m room.

I might eventually put in a smart bypass and put a TRV on the bathroom rad.

[fred]

In article , Ed Sirett writes

You're welcome.

I often run the flow at 50-60 so just a little low to have a cyl stat set to 60. The following mod would let you set the h/w demand flow to 65 or so to avoid the superheated bathroom rad. There's plenty of room on the panel for an extra pot, perhaps with the spindle cut short and a screwdriver slot cut to avoid accidental movement.

Just another one of many variations that could be applied but if your system is working I understand it is hard to work up the enthusiasm to tear it apart.

Variable h/w demand control board mod:

+5V --- | ctrl / 10k modified panel \ demand pot / | level (100k) \ +5V | _|_ --- | 0V | | 1k \ 1N4001 | h/w h/w /---------> boiler 240V RTN demand demand

Regards,

[Tony Bryer]

The two in our church are working just fine and I think my fellow members would rather they stayed that way . I am interested to know though what the basic control algorithm is. I assume that the knob sets the flow temp, and the control board looks at the return and flow temps to decide how much burner power is needed. But obviously somewhere it's also deciding how fast to run the pump. Any insights would be welcome.

[fred]

No great insights I'm afraid, at least no more than I am sure you have gleaned from just watching the boiler in operation, as you have described above. I've guessed that the pump speed increases in stages to match the power output but that is just based on what would seem sensible to get the increasing heat away.

Andy Gabriel seems further ahead with his control system than I am and I am hoping he will publish new insights as he comes across them, much as he did for his recent slime beast solution ;-). I've picked up a lot from his posts on this boiler and his system, also quite a bit from yours, thanks.

[Andrew Gabriel]

Cheers. Some things I have gleened -- may not all be true...

I think the knob sets the flow temperature -- using a clip-on pipe thermometer (not accurate), I calibrated it from 30C at the lowest setting through to 85C (IIRC) at the highest.

The pump seems to run at full speed until the temperature differential across the flow and return starts getting small, at which point it slows down in a number of steps. This is presumably to achieve a better temperature differential.

The gas valve electronic control seems to be a straight on/off affair. The gas rate seems to be varied by the fan speed, presumably by using the pressure drop in the gas/air mixer to control the gas flow -- there is a small hose from the mixer back to the gas valve which looks like it might be providing pressure feedback for this.

If there are not enough radiators operating to absorb the 7kW minimum boiler output but there is still a call for heat, the boiler seems to let the water temperature creep up by about an extra 5C before cutting off the burner completely. There is then quite a large hysteresis before the burner fires up again.

Several people here have noted the temperature control is not adiquate for ideal heating and hot water in a condensing system. Another thing the Keston is missing is provision for connecting remote frost stats. I believe it has an internal one which will bring the boiler on until the return water is 30C. However, in my case the boiler is in a room where this is unlikely to trigger, but much of the pipework is in an unheated area under a suspended floor where it's more likely to go below freezing. Ideally, I should be able to trigger the boiler's frost protection operation remotely from this location too. I have temperature monitoring in this location, and intend to do this by causing it to run at lowest temperature setting (30C) when this area goes below zero, with no other call for heat.

[Tony Bryer]

Thanks to you and Fred for the info

[Ed Sirett]

That's my understanding too. The figures are about right - the 1 o'clock I have it on is about 65C on the flow.

And to get the boiler into condensing mode if possible.

?? Either direct from the mains to the demand terminal in which case the boiler will heat the bypass circuit, or direct from the mains to the CH zone valve if the whole house needs protecting.

Or have I failed to understand what you want to do?

[John]

Via or in series with a Danfoss AT/F pipe stat set to about 15/20 degrees which will only keep the pipes just warm then shut off

[Andrew Gabriel]

Well, ideally have the pump running continuously when some pipework is in sub-zero areas, and have the boiler firing up occasionally as necessary to maintain the pipework temperature above freezing. A similar mod to fred's which instead forced temperature to lowest setting would do this. It was just rather annoying that the Keston already has that logic internally, but no ability to trigger it from a remotely positioned frost stat. John's solution would work in a similar way but without the continuous circulation, which for lagged pipes is probably good enough most times.