[Amps] SSR in amp PS not working as expected

[Ulf Tjerneld]

Hi George,
I think the main recommendation is to use random fire rather than zero-crossing. But I’m new to the subject, so I’m open for arguments.

/Ulf

Sent from Mail for Windows 10

From: GEORGE WALLNER
Sent: den 9 januari 2019 14:25
To: Ulf Tjerneld; amps at contesting.com
Subject: Re: [Amps] SSR in amp PS not working as expected

Ulf,
I have used 100-s of SSR on various ship-board systems. They work great provided they are used correctly and are of the right specifications, and a few precautions are taken. 
Generally speaking, there are two kinds: zero crossing and random switching. For your application you should use zero crossing. Also, you should add 270 VAC varistors (MOV) between "neutral" (non-switched leg) and both the input and output. In an RF environment it also makes sense to add a 10nF capacitor parallel with the control input and both outputs (from ground). Also, for 240 V applications, I always use 480 V SSR-s. Just to have enough headroom: switching inductive loads can create high voltage transients. An SSR witll have about 1.8 V of forward voltage drop when on. At 10 Amps that is 18 W of dissipated heat. You want to mount it either on a large metal surface or a heat-sink.
It is better to find the right SSR for the application, rather than finding an SSR and then trying to use it in an application. You need to know the specifications, as there are many variations.
 
73,
George,
AA7JV/C6AGU
 
 
On Mon, 7 Jan 2019 19:31:21 +0100
Ulf Tjerneld <ulf at sm0nor.com> wrote:
Thanks a lot for all this information. I got some help from the local ham community here too. Most of the feedback I get is to avoid them and stay with traditional relays. I have learned a lot recently 😊.
However, I don’t give up that easily so I will investigate this a bit further. If someone on the reflector has experience of SSR:s that actually do work, please advise! I know of at least one power supply project published in QEX where zero crossing SSR:s are used (W8ZR, QEX sometime back in 2013). At least the Chinese ones he proposes are very cheap, about 10 dollars, so I’m inclined to order a couple of them and try them out. But the delivery time from China is something like 60 days, so I might give because of that ☹.
 
PS… the snubbing network didn’t work. Well, maybe a bit, I actually measured 103 volts on the 105 volt secondary. But still a very loud hum that I don’t like.
 
73’s Ulf
 
Sent from Mail for Windows 10
 
From: Manfred Mornhinweg
Sent: den 6 januari 2019 23:26
To: amps at contesting.com; ulf at sm0nor.com
Subject: Re: SSR in amp PS not working as expected
 
Ulf,
 
there are two fundamentally different kinds of SSRs. One type uses 
thyristors (TRIACs or SCRs), the other uses back-to-back MOSFETs.
 
These two (or three!) types have very different behavior. Not only that 
MOSFET SSRs can handle AC and DC while the others are usable with AC 
only, but also in that their handling of inductive loads is very 
different. Generally TRIACs are poor with inductive loads, antiparallel 
SCRs are much better, and MOSFET SSRs might be best if rated for 
inductive loads, but that needs a special circuit inside them to avoid 
damage from inductive kick.
 
Thyristor SCRs will switch on either the instant you apply a control 
signal, or the next voltage zero crossing after that, depending on their 
type, and will always switch off at the current zero-crossing. With a 
purely inductive load (and the primary of a very lightly loaded 
transformer is very close to that), the current zero-crossing is 90 
degrees out of phase with the voltage zero-crossing. This leads to some 
head-scratching.
 
TRIAC SSRs will often trigger far more easily in one polarity than in 
the other. This is likely the cause for your transformer hum: It's 
getting only a semicycle instead of the full grid waveform, so there is 
a huge DC component present, and the transformer core gets hopelessly 
saturated, with the magnetic flux leaking out of the core.
 
Another possible reason for hum is that the SSR is switching on too much 
after the zero crossing, so that there is a hard voltage step inside 
each half cycle. The harmonics contained in such a step tend to be very 
loud.
 
You will need to look for SSRs rated for highly inductive loads, and 
preferably of the back-to-back MOSFET type. Or else use old-fashioned 
mechanical relays.
 
One last thing: In some cases you can cure such thyristor misbehavior by 
adding snubber networks. You might want to try this first. Simply add a 
100 ohm resistor in series with a 100nF capacitor, of adequate AC 
voltage rating, directly across the output side of each SSR, right at 
the SSR. With some luck that might tame them.
 
Manfred
 
 
========================
Visit my hobby homepage!
http://ludens.cl
========================
 
_______________________________________________
Amps mailing list
Amps at contesting.com
http://lists.contesting.com/mailman/listinfo/amps