PSU dedicated temperature control channel?

EagleB3

The temperature control circuits built into the PSU are far from ideal.

They rarely have more than one fan speed switching limit (only two set values are supported - the minimum rotation speed and the maximum rotation speed when the boundary temperature is exceeded).
Even the minimum speed is set so that the fan is guaranteed to start, and even at these speeds the noise from the fan can be quite strong.
When replacing the fan, these parameters are difficult to correct without replacing parts in the PSU.

I myself was ready to formulate my question as follows: I have a Duet 3 MB6HC (with firmware version 3.3, DATE: 2021-06-15 21:45:56 stable), which has several free channels for measuring temperature and outputs suitable for controlling a fan. I would like to add a temperature sensor (B3950 or PT1000) to my PSU and control the fan speed in my PSU. At the same time (in order to reduce the priority of requirements):

It is imperative to implement at least a minimum control characteristic (fixed fan speed in three sections, determined by two boundary temperature values) - otherwise the idea generally loses its meaning.
Optional: It is highly desirable to be able to turn on 100% fan speed for some time after turning on from the "off" position (analogous to parameter "B" in command M106; for example, M106 B0.5).
Optional: It is highly desirable that not a single output channel for heater control, which is not in the PSU, is occupied. PSU heats itself up.

It is almost not necessary: It would be ideal if there was a way to realize a proportional increase in the fan speed in some areas, for example, from a temperature of 45 to a temperature of 65:

Tell me, is there a way to implement what I want in the current RRF implementation? If it is described somewhere, please give a link to the description!

engikeneer

@eagleb3 I believe M106 already does your option 4 by using a proportional thermostatic mode (in fact the example uses the same temperature numbers your are!):
https://duet3d.dozuki.com/Wiki/M106

EagleB3

@engikeneer ,
As far as I understand it, the M106 command does not make it possible to create a thermostatic channel without assigning some kind of heater to this channel.

But there is no heater in PSU. Besides:

I do not want to take a single release channel with a heater even without real heater connectivity to this output;
I do not have the confidence that the thermostat channel will be flicked correctly without adjusting the PWM control (M303) of the heater used in it. And how to perform it, if there is no real heater?

Of course, perhaps I understand something wrong to apply the command M106 ...

That is, I understand perfectly, how can you create a temperature measurement channel with the M308 command. But I absolutely do not see the way to create a thermostatic channel for this sensor:

without heater
with a three-stage control characteristic that I need.

If I create a fan S3 command M108:
M308 S3 P"io.whichever" A"PSU_temp" Y"thermistor" T100000 B3950 ; define PSU temperature sensor

I will get if using the M106 command
M106 P3 L0.40 X1.00 T40:65 B1.0

Here is a rule of control:

I would not like to burn my power supply or DUET when experiments ...

engikeneer

@eagleb3 the H parameter is the sensor number (i.e. your thermistor), not the heater. I think you can only do one stage of ramp, but you can do the left picture for your option 4

What I think you want:

M308 S4 P"io.whichever" A"PSU Temp"  Y"thermistor" B3950; Create a temperature sensor (number 4) on io whichever, using a thermistor
M950 F2 C"io.whicheverfan"; create fan 2 on pin io.whicheverfan
M106 P2 H4 L0.4 T45:65 ; set fan 2 to be thermostatically controlled by temp sensor 4. Fan will be off below 45C. Fan will ramp from 40% to 100% between 45 & 65C, and be on 100% above 65C.

Obviously, you'll need to get your pin names sorted etc, and might want to change the sensor/fan number

EagleB3

@engikeneer , THANKS A LOT! How do you think it is necessary to use the "X" parameter in the M106 command? For example,
M106 P3 L0.40 X1.00 T40:65 B1.0

engikeneer

@eagleb3 I think it defaults to 100% (I've never specified the x parameter before) but can't hurt to define it like that

EagleB3

@engikeneer said in PSU dedicated temperature control channel?:

I think it defaults to 100% (I've never specified the x parameter before) but can't hurt to define it like that

Many thanks! I think even such a configuration will be enough for most applications.
In any case, it is much (MUCH!) better than what is offered by the manufacturer in PSU industrial standards!