CM67 Optimum Start algorithm

[IMM]

No open loop. You are on about simulated proportional feedback, but it was obvious you knew that.

[IMM]

Er no. You take away the power and it says at the position it was when power was taken away (de-energised)

That is true, but they are accurate enough once settled. That is why there is a 0-10v (simulated proportional feedback) version, so the controller knows exactly where the valve is.

It appears to have two coming out of it in some diagrams, but on read some more in depth data it will not modulate a valve. That is why L&S stuff is much superior.

A waste of time now, now the cm67z is here.

[IMM]

The cm67 can't modulate, tat is why it is best to go for a good L&S model that can and have finite temp control.

All snakes should be killed, especially the poisonous ones.

[Andy Hall]

It is open loop between the controller and the actuator because there is no direct feedback of the valve position to the controller. The only feedback is the outer control loop which is the temperature of the controlled environment. That has a much longer time constant than the operating time of the actuator

.andy

To email, substitute .nospam with .gl

[Andy Hall]

It is pretty obvious that the damn thing will stop when the power is removed because it has only three wires and no separate source of power.

Nope. The (temperature) controller does not know where the valve is directly at all. There is separate motor control electronics in the valve head, and that has feedback locally about the motor position. The effect of this is that the actuator position is more accurately repeatable from an applied voltage on the control input but there is still no direct feedback regarding the motor position to the temperature controller. However the arrangement is much more accurate than one with simple on/off/reverse control which relies on the assuming speed of movement of the actuator

Which is what I said in the first place.

Not at all. This is but one componen. I'm already doing what this system does and more.

Each room has temperature sensing leading to a central control system and then there is individual control to each radiator valve.

The CM67z system is interesting, but each controller only manages two zones so it is going to get expensive to implement. Since each TRV requires 2xAA batteries (not a lot of power available) it is going to keep Duracell sales quite buoyant....

.andy

To email, substitute .nospam with .gl

[Andy Hall]

Which is what I said in the first place.

Now you're just getting silly and emotional. :-)

.andy

To email, substitute .nospam with .gl

[Andy Hall]

The Hometronic system itself isn't new and has been around in Germany for a number of years, where it is quite suited to apartment installations. The earlier version of the radiator valve was RF controlled as well but AFAICR did not have transmission capability to a boiler interlock device.

The CF67z is new, although one point to realise is that it will only cover two zones to a set point. Each HR80 can be set in itself but then that is not under CM67z control. The data sheet leaves that bit out.

Also, since each HR80 uses 2 x AAA batteries, it is going to make Duracell shareholders very happy. :-)

Some interesting ideas though

.andy

To email, substitute .nospam with .gl

[IMM]

A 0-10v simulated proportional feedback conbtroller has to guess to some degree. The word simulated gives it away.

[IMM]

A controller acting on a 3 position actuator would pulse it one way or the other (a pulse to not 6 minutes as you mistakingly believe). In short, it nudges the valve one way or the other to reach setpoint. Some controllers can be used with any 3 position actuator of any make. It doesn't need to know the speed of the actuators travel. Simulated proportional feedback are more suited to PID controllers.

No, throw it all out and get the cm67z

[IMM]

You never mentioned L&S. You never knew of their controllers.

No. Poisonous snakes are no use whatsoever and kill people.

[Andy Hall]

I didn't believe that at all, you did.

The L&S one certainly does, and any controller working in this way will need to know the speed of travel at least approximately, or as a loop there would be far too much or far too little gain. The data sheet for the L&S controller certainly refers to a range of speeds of actuator

I don't think so. I am doing some far better things and have some original ideas that I'm working on. I won't discuss them here though..

.andy

To email, substitute .nospam with .gl

[Andy Hall]

Of course, but because there is local position detection at the valve, the result is more accurate and the system loop behaviour much more predictable.

.andy

To email, substitute .nospam with .gl

[Andy Hall]

L&S was never the subject of the thread. You introduced it because you had painted yourself into a corner with respect to the CM67.

Who said anything about poisonous? The vast majority are not, and even those that are are seldom aggressive towards humans.

Of course if you're a rodent, then that's a different matter, whether you can tango or not...... cha cha cha doesn't help either.. ;-)

.andy

To email, substitute .nospam with .gl

[IMM]

A decent "well adjusted" controller on a 3 position valve gives the same accuracy.

[IMM]

You said a pulse can be minutes long. Duh!

controllers

I'll let you into a secret. These controllers have adjustments on them.

No. You will F**k it up. Throw it all out and get a ready made solution. Stop trying to be a controls engineer, when you clearly have missed the basics.

[IMM]

The thread had moved to finer temp control via modulation. Better choice is a heat bank, using a cheap simple condensing boiler and a 3-way mixing valve from a L&S controller for the heating off the heat bank.

Do you feed this reptile live mammals?

[Neil Jones]

I looked at Hometronic last year but I think it's far too expensive for what it is.

I'm vaguely interested in the new solution, but only if it's comparable in price with a normal CM67 + TRV solution. If it's a lot more, I'd be more interested in getting a solution with a bit more intelligence which could be integrated into a whole house system.

Neil

[Andy Hall]

That's not possible. There is no position detection, apart from driving the thing periodically to the end stop and timing from there.

DC controlled actuators have servo electronics in the head with position detection based on a stepper motor. That is inherently a lot more accurate than something run end to end with on/off control.

.andy

To email, substitute .nospam with .gl

[IMM]

Doesn't need it.

It doesn't time it. The proportional side is in control until it is inside the proportional band and when it settles inside the band the Integral pulses it to set point. I have had "highly accurate " control with simple controllers and actuators.

Not necessarily, but generally I prefer this method.

[IMM]

A real cheap solutions is to create a manifold and have all the rads flows off this. Each rad has a 2-port zone valve (cheap at £12 at BES) and each rad hasn electronic wall stat.