How quiet are Crydom CKR SSRs?

[Mike]

As SSRs I'd expect the Crydom CKR range to be quiet, however I can't spot any claims to them being 'silent' on the Sensata website, nor anyone having mentioned it here. It's perhaps more of a consideration for my project as they'll be sitting in a cabinet in the hallway, rather than a plant room...

 

I know that a few of you have used them - well, @TerryE at least - so I'd be interested in your experience.

 

 

[TerryE]

Like the grave 😊  Triacs don't make any noise, as they do the switch as the AC crosses 0V.  Remember that they only have an efficiency of ~99-99.5% and that ½-1% ends up as heat, say 5W @ 12A, enough that you may need have some sort of thermal sink.  

 

 

Edited March 3, 2022 by TerryE

[Nickfromwales]

14 minutes ago, TerryE said:

Like the grave 😊  Triacs don't make any noise, as they do the switch as the AC crosses 0V.  Remember that they only have an efficiency of ~99-99.5% and that ½-1% ends up as heat, say 5W @ 12A, enough that you may need have some sort of thermal sink.  

 

 

Can you ( or is there any advantage to ) use 2x SSR's in parallel to boost reliability / shed excess heat from a singe device? Or would I just double the CCC of the SSR to allow longevity / minimise generated heat  I am looking to get away from the Finder relays ( not contactors ) for switching Wills and immersions on / off and need to order stuff up for one client this week ideally.

[TerryE]

Nick, I did initially think of having an SSR and contactor in parallel, and using my control system to do a sequence of:

• <SSR on><Contactor on><SSR off>   ....
• <SSR on><Contactor off><SSR off>

And that way the SSR would only be on for a second or, but the Contactor wouldn't be under load during the switch, but I discarded this as being too clever.

 

IMO the weakness with the CKR range is that they rely on convective airflow for cooling:

 

IMO in retrospect you'd be far better off with the surface mount style directly mounted with thermal paste on the steel equipment enclosure, and here you would be using metal-on-metal thermal conduction to convert the entire enclosure (say ½m² area) into a thermal dump and the ~5W waste heat would only raise the case by a couple of °C.

 

[dpmiller]

in the kit I work with, a SSR generally has a series contactor for failsafe isolation

[Mike]

5 hours ago, TerryE said:

Like the grave 😊  Triacs don't make any noise, as they do the switch as the AC crosses 0V.  Remember that they only have an efficiency of ~99-99.5% and that ½-1% ends up as heat, say 5W @ 12A, enough that you may need have some sort of thermal sink.

 

Thanks - that's what I'd hoped!

 

The fins were a clue that there may be a heat issue. The installation guide mentions that 'to achieve maximum ratings, there must be a minimum spacing of 0.8 in (20 mm) between the devices (horizontally) and of 3.15 in (80 mm) measured between DIN clip tabs (vertically)', so I was going to follow that. I'm planning this one in France so don't want to frighten the electrician (or the inspector who has to sign off the installation) with anything non-DIN rail, if I can avoid it

Edited March 4, 2022 by Mike

[Nickfromwales]

9 hours ago, dpmiller said:

in the kit I work with, a SSR generally has a series contactor for failsafe isolation

Thats the reason why I thought to double ( at least ) the CCC of the SSR, eg to hugely improve reliability.

 

The whole point of moving to SSR(s) will be to get away from the noise of operation which can be hugely problematic in a new TF dwelling. More so when there are multiple loads to switch at different times / intervals / other sporadic loads etc.

 

In fairness, if these are switching end equipment which has its own internal thermostat, if the output of the SSR were to latch permanently on then the worst case would be say a DHW cylinder heating to the immersion max temp setting outside of the desired time-slot or a Willis primary ‘loop’ heating similarly, both with no ill effect to the dwelling or end user ( see next ).

 

This comes full circle to reinforce my comments about always having a TMV on an UFH manifold, as a doomsday device, and similarly with the UVC’s where I install a TMV on the hot outlet to cap the max output water temp to ~50oC. With these types of downstream failsafe disciplines it matters not one jot about delivery heat temps running away and removes totals reliance on the load switching device.

 

IMHO it is a contractor which os more likely to fail than an SSR. @TerryE have you had to replace any yet? Reliability / degradation of any concern? Duly noted on affixing to a metal chassis vs convection cooled. Many thanks. 

[TerryE]

@Nickfromwales Nick after you gave me so much help with my build, I am glad to return the favour in a small way. 

 

Yes, I've had 2 SSRs fail; one failed safe on and one tripped its MCB before failing.   I never bothered to implement the contactor approach.

 

We discussed the root cause on another topic: Because these Crydoms run hot at ~55-60°C, I recommend avoiding wiring any multicore leads (chrimped or not) directly into them.  I have a separate connector strip in my box and I run a 4mm single core link from the SSR to this and wire the (chrimped) flex power lead into this.  This way both leads into the L1/T1 connections are 4mm single core. 

 

The connector strip runs cool (at only a few °C above ambient) so I haven't had any issues since.  

 

[Mike]

On the question of durability, and looking at the thermal de-rate graph (below), while the 20A model has enough headroom to switch 3kW, I guess it may be worth paying the extra £10 to upgrade to the 30A version?

Presumably it either handles the heat better, or generates less, as all models seem to have the same finned case.

[TerryE]

@Mike, I did the same.  I use the 20A versions and my loads all pull just under 12A so I am at the limit of the CKRX2420 spec at 60°C.  I did just think of putting a 5V 100mm quiet fan below the SSRs to give a bit of forced are circulation over the fins to see if this made any significant difference.

 

PS: I do find the nominal ratings of these devices very misleading because at 20A, say, the device needs to dump about 7W and there's no way that this would stay under 60 let alone 25°C without having forced airflow over the finned areas.

Edited March 4, 2022 by TerryE

[Mike]

Well after a few hours investigation, I'm inclined to specify a Siemens model instead - probably the 30A 3RF2330-1AA22 in my case - for a few reasons:

• It has a significantly bigger heat-sink
• At the planned loading it doesn't require additional space around it (though the 30A module is 45mm wide, thanks to the sink)
• The datasheets are clearer on the acceptable currents : 22A @ 40°C according for resistive loads (category AC-51 of IEC 60947-4-3) for the 30A version; 13.2A for the 20A version*
• Prices not too dissimilar to Crydom
• The range is certified for domestic use to IEC/EN 60335-1 (fairly certain I spotted only a reference to industrial applications for the Crydom models).

*Siemens have an article discussing current specifications & the need for additional precautions in higher ambient temperatures for this range here.

 

 

[Mike]

17 hours ago, dpmiller said:

in the kit I work with, a SSR generally has a series contactor for failsafe isolation

What's the reason for not just using a contactor?

[dpmiller]

for fine temperature control. This is in steam sterilisers, we're switching heating loads of up to 50kW and a failed-short would be seriously bad. But control has to be good enough to not overshoot too...

So we have one contactor as a safety device on an independent control, and then actual heater switching is carried out by SSRs or another contactor, depending on brand/model

[TerryE]

@Mike, you need to be careful in selecting SSRs both in terms of input and output specs.  The Cyrdom's take a TTL compatible input and so can be driven directly by form Arduino micros or through a 3.3V to 5.V level shifter. I drive mine through MCP23008 multiport I/O expanders which can run at 5V and can drive the SSR inputs but interface to the microprocessor using 3.3V I²C.

 

The Crydom range seems to be one of the few which match these requirement, but you do have more choice in the Panel Mount format such as the Cyrdom 1 Series (datasheet) which can also be directly driven at 3.3V and they are cheaper too.

 

The Siemens models you reference are really optimised for industrial use and take 40V AC control inputs which means you will need a 6:1 step-down transformer and a separate way of switching the 40V AC input signals getting a USB or RPi SSR board capable of driving 40V AC is a PITA as most AC SRR boards quote 100-250V operating range.  However you have a rich choice of relay boards which will happily drive at the 40V few mA that the Siemens SSRs take.

[Mike]

4 hours ago, TerryE said:

The Siemens models you reference are really optimised for industrial use and take 40V AC control inputs

Thanks for checking, and for the warning!

 

Having read the Siemens datasheet more closely, the section on the control inputs does seem unclear to me - I've posted it below.

 

There's a line for "control supply voltage 1 at AC" of 110 ... 230 V ; I had read that as meaning a voltage in that range switched it on - and a lower voltage off, which is as far as I'd got.

There's also a line for "control supply voltage at AC... at 50 Hz full-scale value for signal<0> recognition" of 40V - meaning below 40V switches it off?

And another "control supply voltage... at AC initial value for signal <1> detection" of 90 V - meaning over 90V switches it on?

 

Are you able to add any illumination? I'm not used to analysing electronics specs in depth, as you may have spotted

 

As for my intended voltages, I was planning to indirectly switch the SSRs at mains voltage. That is, I'm planning to use a Raspberry Pi + MQTT + NodeRed for control, but to switch a Shelly Pro 4PM, which outputs 230V to drive the SSRs.

 

The Shelly's in there for a couple of reasons. Partly because it could be used to control the SSRs directly, either via the manual control buttons or the Shelly app, or its onboard access point (which also allows the setting up of weekly programmed schedules) - so a good fallback if there's ever a problem with the Pi. And also because it doesn't rely on a 'DIY solution', so I hope will not perturb the electrician or the inspector; being France all new installations have to be signed off by a national electrical inspectorate before they'll turn on the supply!

 

In fact, it's slightly more complicated that the above. The Shelly will actually switch the SSRs via a load-shedder, which can temporarily shut down the SSRs & which also operates at 230V; in France your standing charge depends on the mains supply current that you subscribe to, so shedding the load from low-priority circuits at times of peak household demand has a short payback period, and flattens peak demand on the national grid.

 

BTW, although I need 230V control, other Siemens models, such as the 3RF2330-1AA44, can apparently control using 4 to 30V DC (perhaps switching on at 4V and off at 1V, subject to the correct interpretation of their datasheet).

 

 

 

 

Edited March 5, 2022 by Mike

[TerryE]

First I misread  the Siemens datasheet.  The input will accept 240V at a current draw of around 15mA so yes, this could be switched by the Pro SM4.  The advantage of a Wifi based switch is that it removes the RPi from the scope of the inspectors review, though you should have the CE DoC to hand.

The disadvantage of Wifi is that it doesn't work too well in a steel equipment enclosure. 🙃  Though you could try putting the RPi inside the enclosure as well

 

In the UK, the rule are different. Your electrics have to be signed off by a Part P certified electrician who certifies the house registration and does this in an online registry so the Binsp (and insurance companies) can look this up.  Once my electrician read the datasheet for the Crydoms he was happy to ignore the 5V control signals as these are outside the scope of the certification. 

[Mike]

@TerryE thanks again for your feedback - it's good to have someone looking over my shoulder and pointing out potential problems.

 

Enclosures are still plastic on this side of the channel, so WiFi will probably work, however the Pro range of Shellys also have Ethernet ports, so I can hook up to that if it doesn't. The Pi and network switch will be in the (mandatory) Communications Cabinet directly below the consumer unit, so it would only need a couple of patch cables. Keeping the DoC to hand is certainly a good idea; I'd like the electrician and inspector to be as equally disinterested as yours was!

 

The French system is stricter than Part P, in that no electrician can sign off their own work. On the other hand - if you live in a detached house and don't sell it for 10 years after certification (to sidestep other legal issues) - then it's still possible to do a complete DIY installation over here, provided it meets the standards and is signed off by the inspector.

 

Edited March 5, 2022 by Mike

[TerryE]

2 hours ago, Mike said:

Enclosures are still plastic on this side of the channel

My electrician used a plastic enclosure.  I would have been a lot more comfortable with a steel one. ☹️

[Mike]

Well it looks like I may be choosing Crydom after all.

 

After checking the wiring information for the Siemens units, only selected models are suitable for use with MCBs, and they can only be driven by 0-24V DC, or 4-30V DC (models ending -1DA06 & -1DA44; 3RF2330-1DA06 for example). Their other models require fuses, which aren't permitted in domestic installations...