Swappin' Mech Power Relay for Solid State

[Pure_Brew]

Howdy folks,
I'm going to try another stab at replacing the Magnecraft W199AX-9 relay in my PS Audio 200c, for a new Solid State relay (SSR). The reason for this is due to coil noise in the mechanical relay, which is audible and somewhat annoying listening to music at low levels.

Relay Item # W199AX-9, 199 Open Style Power Relays / DPST-NO, 40 Amp Rating (AC) on Schneider Electric US - Magnecraft Relays

My theory is that the open relay in this amp is also responsible for some distortion of the AC at the zero cross point, enough to also make this transformer hum. That would also mean that a simple DC filter, that will not allow DC to pass or "DC like" component to pass on an AC line ( like the Humbuster, also PS Audio), would actually not be able to do anything to help the situation, since the AC would be bad after the relay. I actually own one, it works on other amps but not this one.

I could be way off on this, however, on 2 separate occasions, 2 different transformers were tested independently and serviced by 2 different companies (one being the OEM Paraline). Both transformers were found to be actually quiet in operation UT. One of the members of this forum re-installed the transformer, still had Hum even after vacuum dip/service. However, after installing a Teledyne SSR, everything went nearly silent.

So, I figure, why not try it. Plus it's fun to learn

I have been fortunate to run across a little info that points to this solution as a workable one, and that the open style relay of this type can be successfully replaced with an SSR to reduce or eliminate noise.
However, I have not been fortunate in being able to speak to the individual(s) directly on exactly what was used and how it was wired. So, I'm wondering if I could get some help from the community here.

1): First of, based on the drawings below, does this look like I have this planned out correctly?

keep in mind that the W199 is being jumpered on both ends, tying the poles together electrically. In turn, this also serves as a pathway for the control voltage, traveling from AC (+) to over the relay, to the heat sensor and then to the control connection. The other end of the control of course goes to the switch then back to AC(-).

The plan for the SRR is nearly identical to the W199, with the exception of having to jump 2 poles. So in this case the job would seem to be perfectly fine for a single pole SSR, with the control voltage being fed at the same point as the open (+) side of the relay, just as in the original layout.

I'm not sure why a 2pole was originally used, other then for connection flexibility. Furthermore this appears to be very simple with no soft starting, which may be the reason why the manufacturer of this amp suggested that it be left on at ALL times.

2): Should any resistance be added on the control line? If so where?

The coil resistance in the W199 is 290ohms. Also it would be very easy to connect a resistor and the switch at the unused portion of the barrier strip, then connect the other end of the resistor to the AC (-) end.

I would suppose that as long as I'm within the operating range of the control switch, that would be fine for the relay. However, the question comes up for the actual power switch on the amp, and the heat sensor. Otherwise I guess I could try to match the resistance-overall?

3): What would be the best SSR choice?

Matching specifications based on voltages/amps doesn't look too difficult. However, there are several different internal configurations to choose from:

switch type: (random/zero cross?)(SCR/SCR(2)/Triac?)
contact configuration: (SPST-NC/SPST-NO/DPST-NO) w199=DPST-NO. Looks like I'll be using a SPST-NO (Single Pole Single Throw - Normally Open)

4): Heat...

The heatsink recommended for a 40A SSR is pretty huge. Here is one example

Item # SSR-HS-1, Heat Sink on Schneider Electric US - Magnecraft Relays

In the end, it seems like 40A is such a huge number, and I can't image that that this is going to run that high. However, the transformer has been recently Hipot tested @ 2.5k, the resistance across the mains is 1.6ohms. I know that the amp can output up to 1000w x 2 @ 1.5ohms- But normally 200 x 2 @ 8ohms. Any help here with this would be great. I can't fit an almost 5x6 heatsink in the amp....

I guess that's about it...

Thank You

[Mooly]

Power control solid state relays I have seen use a logic level to switch, usually accepting 3.5 to perhaps 30 volts as a control signal. So you need to check that for definite on any relay you intend using.
I can't for a moment imagine heat is a major consideration for an application like this. Easy to calculate roughly, just look at the quoted on resistance and look at the current you will be pulling through it.

[Pure_Brew]

Originally Posted by Mooly

Power control solid state relays I have seen use a logic level to switch, usually accepting 3.5 to perhaps 30 volts as a control signal. So you need to check that for definite on any relay you intend using.

Yes. 120VAC is used to supply the original relay. While many SSR's use a small control voltage, the "drop-in" replacement will need to handle the same as the original (unless I modify).

This SSR from Magnecraft seems to fit the bill, although the voltage control is wider(90-280):Item # 6240AXXSZS-AC90, Class 6 Solid State Relay / SPST-NO, 40 Amp Rating on Schneider Electric US - Magnecraft Relays

Originally Posted by Mooly

I can't for a moment imagine heat is a major consideration for an application like this. Easy to calculate roughly, just look at the quoted on resistance and look at the current you will be pulling through it.

While I couldn't disagree, I did find a few smaller foot-print heatsinks that would fit anyway. I think it would be prudent not to attach it directly to the bottom panel, since the main caps are laying on the panel, right in front of the relay position. Any potential heat would transfer to the surface of those capacitors.

I'm not sure what resistance I should be looking at. Do you mean the resistance of the mains of the transformer?, or the amplifiers rated output to the speaker load? I believe you mean the latter. The tranformer is rated for 2500w, being fed 120v, then the current draw would be at most 20.83Amps-no? I thought that the mains resistance was like 1.6ohms, but I've misplaced the data sheet...

*********************************

Also, on a different note: I did read that back to back SCR's (SCR(2)), seem to be the way to go for low distortion AC control. (which the SRR above has).

[kevinkr]

Originally Posted by Pure_Brew

Howdy folks,
I'm going to try another stab at replacing the Magnecraft W199AX-9 relay in my PS Audio 200c, for a new Solid State relay (SSR). The reason for this is due to coil noise in the mechanical relay, which is audible and somewhat annoying listening to music at low levels.

Relay Item # W199AX-9, 199 Open Style Power Relays / DPST-NO, 40 Amp Rating (AC) on Schneider Electric US - Magnecraft Relays

My theory is that the open relay in this amp is also responsible for some distortion of the AC at the zero cross point, enough to also make this transformer hum. That would also mean that a simple DC filter, that will not allow DC to pass or "DC like" component to pass on an AC line ( like the Humbuster, also PS Audio), would actually not be able to do anything to help the situation, since the AC would be bad after the relay. I actually own one, it works on other amps but not this one.

I could be way off on this, however, on 2 separate occasions, 2 different transformers were tested independently and serviced by 2 different companies (one being the OEM Paraline). Both transformers were found to be actually quiet in operation UT. One of the members of this forum re-installed the transformer, still had Hum even after vacuum dip/service. However, after installing a Teledyne SSR, everything went nearly silent.

So, I figure, why not try it. Plus it's fun to learn

I have been fortunate to run across a little info that points to this solution as a workable one, and that the open style relay of this type can be successfully replaced with an SSR to reduce or eliminate noise.
However, I have not been fortunate in being able to speak to the individual(s) directly on exactly what was used and how it was wired. So, I'm wondering if I could get some help from the community here.

1): First of, based on the drawings below, does this look like I have this planned out correctly?

keep in mind that the W199 is being jumpered on both ends, tying the poles together electrically. In turn, this also serves as a pathway for the control voltage, traveling from AC (+) to over the relay, to the heat sensor and then to the control connection. The other end of the control of course goes to the switch then back to AC(-).

The plan for the SRR is nearly identical to the W199, with the exception of having to jump 2 poles. So in this case the job would seem to be perfectly fine for a single pole SSR, with the control voltage being fed at the same point as the open (+) side of the relay, just as in the original layout.

I'm not sure why a 2pole was originally used, other then for connection flexibility. Furthermore this appears to be very simple with no soft starting, which may be the reason why the manufacturer of this amp suggested that it be left on at ALL times.

2): Should any resistance be added on the control line? If so where?

The coil resistance in the W199 is 290ohms. Also it would be very easy to connect a resistor and the switch at the unused portion of the barrier strip, then connect the other end of the resistor to the AC (-) end.

I would suppose that as long as I'm within the operating range of the control switch, that would be fine for the relay. However, the question comes up for the actual power switch on the amp, and the heat sensor. Otherwise I guess I could try to match the resistance-overall?

3): What would be the best SSR choice?

Matching specifications based on voltages/amps doesn't look too difficult. However, there are several different internal configurations to choose from:

switch type: (random/zero cross?)(SCR/SCR(2)/Triac?)
contact configuration: (SPST-NC/SPST-NO/DPST-NO) w199=DPST-NO. Looks like I'll be using a SPST-NO (Single Pole Single Throw - Normally Open)

4): Heat...

The heatsink recommended for a 40A SSR is pretty huge. Here is one example

Item # SSR-HS-1, Heat Sink on Schneider Electric US - Magnecraft Relays

In the end, it seems like 40A is such a huge number, and I can't image that that this is going to run that high. However, the transformer has been recently Hipot tested @ 2.5k, the resistance across the mains is 1.6ohms. I know that the amp can output up to 1000w x 2 @ 1.5ohms- But normally 200 x 2 @ 8ohms. Any help here with this would be great. I can't fit an almost 5x6 heatsink in the amp....

I guess that's about it...

Thank You

One thing to consider about an SSR is the switching element is either a TRIAC or back to back SCRs and these devices turn off slightly before the zero crossing point so if you were concerned about distortion at the 0 crossing point before you should be more so with an SSR. (The dead zone can be several volts above and below the zero crossing.)

Buzzing relay coils usually have to do with the winding being energized with AC or very ripply DC - either of which can be fixed.

Make certain that the SSR can handle the worst case current on power application which may be a great deal more than 40A for a couple of cycles.

Personally I would not use an SSR in this application, but would find a quiet mechanical relay instead. I've also had some problems with really big transformers and SSRs and getting them to turn on and off at the right time due to the phase angle of the load currents. (power factor issue)

[event horizon]

I think Mooly is refering to the on resistance of the solid state relay. Seeing as you know the maximum current (20A+) that the transformer could draw then working out an appropriate heatsink for the solid state relay is made pretty simple considering any voltage drop.

Bests.

[Pure_Brew]

Thanks people

Originally Posted by kevinkr

One thing to consider about an SSR is the switching element is either a TRIAC or back to back SCRs and these devices turn off slightly before the zero crossing point so if you were concerned about distortion at the 0 crossing point before you should be more so with an SSR. (The dead zone can be several volts above and below the zero crossing.)

I did actually get the point after some study, that the Back to Back SCR's were going to be a good choice for AC control. However, based on what your saying about the zero cross point and some images I looked at, what your saying seems logical. The truth is, I don't know exactly what I need, but I do want to get rid of the mechanical noise in the end

To me, it would make sense that a solid physical contact should provide less distortion in the AC. But if I can elminate the mechanical noise, and it worked for someone else in the same amp, I feel compelled to try it.

Originally Posted by kevinkr

Buzzing relay coils usually have to do with the winding being energized with AC or very ripply DC - either of which can be fixed.

I thought by replacing the original mech relay, which was most likely in service for 25 yrs, with the identical replacement might solve this - I was wrong. I also thought that a DCfilter/rectification of the incoming AC would also solve this issue-wrong again. (although it reduces transformer hum in other amplifiers that I own). What would you suggest? I don't think I've ever seen an amp that uses a big open relay like this that flashes bright blue electricity like this one does at turn on.

Originally Posted by kevinkr

Make certain that the SSR can handle the worst case current on power application which may be a great deal more than 40A for a couple of cycles.

I actually this I'm good on this. The peak amperage handling on the SSR's I've been looking at are quite huge, way above 40A.

Originally Posted by kevinkr

Personally I would not use an SSR in this application, but would find a quiet mechanical relay instead. I've also had some problems with really big transformers and SSRs and getting them to turn on and off at the right time due to the phase angle of the load currents. (power factor issue)

I really don't know, but are mechanical relays typically used in high power audio amplifiers? I would use the best of what's out there if only I knew what that would be... As far as I know, some SPST-NO SSR's have adjustable phase angle. But I think that would be used in 3-phase AC with 3 of those relays for motor control.

I guess I'm also still wondering why a double-pole relay was used in the original application. They are simply tied together. More physical contact area?

[Mooly]

Originally Posted by event horizon

I think Mooly is refering to the on resistance of the solid state relay. Seeing as you know the maximum current (20A+) that the transformer could draw then working out an appropriate heatsink for the solid state relay is made pretty simple considering any voltage drop.

Bests.

I was yes

Pure_Brew
You can not measure the primary resistance of a transformer as an indicator of how much current will flow. At AC it's completely different due to impedance rather than resistance.

Maximum surge currents occur (which goes against normal thinking) when you apply power and the AC waveform is at a zero cross point. The amount of magnetism remaing in the core from the last switch off also modifies this.

Perhaps that could explain why kevinkr has had problems with SSR.

Nelson Pass has a simple SS switch on the A75 amp... maybe something along those lines might be better.

[Pure_Brew]

Originally Posted by Mooly

I was yes

Pure_Brew
You can not measure the primary resistance of a transformer as an indicator of how much current will flow. At AC it's completely different due to impedance rather than resistance.

Ah, ok thanks. But deriving the current from the rating of the transformer(2500w)divided by the voltage (120v) is correct yes?
I=P/E ......20.83A

Originally Posted by Mooly

Maximum surge currents occur (which goes against normal thinking) when you apply power and the AC waveform is at a zero cross point. The amount of magnetism remaing in the core from the last switch off also modifies this.

Perhaps that could explain why kevinkr has had problems with SSR.

So, wait, would that mean that whenever you switch on the amp with an SSR designed to switch at zero cross, rather then random, your gonna get switched on at the highest possible surge?..lol?

Originally Posted by Mooly

Nelson Pass has a simple SS switch on the A75 amp... maybe something along those lines might be better.

Found it thanks:
http://www.passdiy.com/pdf/a75p2.pdf

Well, I see he used a triac in that design, as well as a bunch of other parts/protection in that circuit. If you look at the layout on the 200c, it has absolutely none of this. That original design is probably NOT a "best" practice. (well, I guess you never want to shut it off).

[Mooly]

The power/current calculations are near enough to work with... remember it only holds true for resistive loads with a power factor of 1. It's possible to have loads of current and voltage in a circuit that dissipates no "power". A capacitor connected across the mains draws current, and has line voltage across it, but it doesn't even get warm... there is no power dissipated.

Same for an inductor. A tranny with a low (2 ohm or less) draws a very small current when not connected to anything (The secondaries)... just enough to maintain the core magnetising conditions.

Switching at the zero cross point ? It's a hugely complex subject and goes against everything that might seem the obvious.

Found this today for you, specifically on SSR and switching inductive loads,
http://relays.tycoelectronics.com/ap...fs/13c3206.pdf

[Pure_Brew]

Originally Posted by Mooly

The power/current calculations are near enough to work with... remember it only holds true for resistive loads with a power factor of 1. It's possible to have loads of current and voltage in a circuit that dissipates no "power". A capacitor connected across the mains draws current, and has line voltage across it, but it doesn't even get warm... there is no power dissipated.

Same for an inductor. A tranny with a low (2 ohm or less) draws a very small current when not connected to anything (The secondaries)... just enough to maintain the core magnetising conditions.

Switching at the zero cross point ? It's a hugely complex subject and goes against everything that might seem the obvious.

Found this today for you, specifically on SSR and switching inductive loads,
http://relays.tycoelectronics.com/ap...fs/13c3206.pdf

OMG Many thanks on that link!

Obviously a potentially DANGEROUS situation. 500KVA transformer getting 400a turn-on peaks at zero crossing!?!? -wow

I see that the "type" of load must be defined: Resistive/capacitive/inductive, and at some point most likely has some component of each.

Although I'm trying to turn on a transformer, isn't there a huge capacitive load?

For example: Quoted from the "Son of Zen" pdf. http://www.passdiy.com/pdf/sonofzen.pdf

The do-it-yourselfers I know are usually generous with the
power supply capacitance, resulting in high turn-on currents from the AC
line.

(and suggests suppressors)

Due to the design of the uncovered, open style relay, I believe this is exactly what I am seeing, since the contacts arc and flash at turn on
However, this is only after the amp has been off for awhile, and those capacitors have discharged.

Once the amp has been on for a bit, you can switch it off and on without this flash reoccurring.

Also I see, the transformer itself would be inductive, however, this confuses me a little:
http://www.tycoelectronics.com/docum...age_Relays.pdf

Quoting this document:

Inductive load – interrupting dc inductive loads is more difficult than resistive loads.The stored energy (1/2 LI2) in the inductance induces a voltage (-L [di/dt]) that tends tomaintain the current. This continues until the energy of the inductance dissipates. Unless
special quick-opening contacts or other means are used to interrupt it, the arc’s persistence depends on the load’s time constant (L/R), a direct relationship. AC inductive loads do not create the same problem as dc loads because polarity reversal at the end of each half-cycle forces the current to zero. Also, the current is out of phase with the voltage and, during the last part of the current half-cycle, the supply voltage is opposing the voltage of self-induction.

So anyway.... once the caps are full I don't see any big arc as before. I would imagine that, according to your link, If I ever had a 400amp+ inrush as a result of transformer issues, the contacts would have probably exploded long ago.

But then again, we are talking about zero-cross voltage. How could one ever know at what point of the wave form is reaching when the open relay strikes down? It would have to be completely random, and, even then, couldn't it couple at the zero-cross at some point anyway.....?

Thanks