Is your unobtainium power switch worth $5 and an hour or so of your time?

Do the Triac mod so you have a resistive 120 ma contact load and the switch will outlast you and no more arcing. Don't forget to remove the RC filter across the switch contacts.

After removing it from the switch contacts, would I install the RC filter cap across M1 and M2 pins on the triac?
 
The Triac doesn't need it, the Littlefuse Alternistor series I mentioned is designed for high inductive loads and no filter needed. A Triac switches off as the sine wave goes through zero so there is no arc to be suppressed.
It is a surplus part as long as you use an Alternistor series.
No I have no ties with Littlefuse. I just stumbled on a good thing.
My LR-9090 is still cranking fine and I tend to turn it off and on every time I go past my test bench to continue looking for a failure even though I do not expect one.
Does your switch have 1 or 2 contacts and if 2 are they both used?
I need to download the SM for your SX and look at the ckt.
 
The Triac doesn't need it, the Littlefuse Alternistor series I mentioned is designed for high inductive loads and no filter needed. A Triac switches off as the sine wave goes through zero so there is no arc to be suppressed.
It is a surplus part as long as you use an Alternistor series.
No I have no ties with Littlefuse. I just stumbled on a good thing.
My LR-9090 is still cranking fine and I tend to turn it off and on every time I go past my test bench to continue looking for a failure even though I do not expect one.
Does your switch have 1 or 2 contacts and if 2 are they both used?
I need to download the SM for your SX and look at the ckt.

Sorry to drag this up. yes a triac opens when the current goes through the zero crossing. However the voltage and current aren't likely to be in phase. Therefore it is possible that the triac would open with a high voltage across it.
 
Sorry to drag this up. yes a triac opens when the current goes through the zero crossing. However the voltage and current aren't likely to be in phase. Therefore it is possible that the triac would open with a high voltage across it.

That is an internet myth. It has been repeated so often it has taken on an aire of authenticity. Best guess was a bored EE student for a laugh. Sounds like something I or a roommate might have done.

A transformer reflects the load (secondary) impedance to the input (primary). If the secondary load is resistive, then the load as viewed at the input is resistive / in phase.

Amplifier power transformer secondary's are wired to rectifiers. That is DC i.e. resistive. A tiny secondary AC load may be incandescent pilot lights also resistive / no phase shift. Some of the newer units have switching power supplies, but the ones I have looked at start with a rectifier to a DC buss then to the switcher so still DC / resistive as viewed from the input.

Think about it. If that were not the case the power grid would never work. The substation transformers would add so much phase shift any AC motor would burn up from over current, your light bulbs would barely glow. At 90 degrees there is no real power available.

We got to do the calcs to determine the transformer addition to and effect on the secondary as seen from the primary in first or second year EE. The component of load viewed from the primary added by the transformer had best be negligible or the transformer is trash.

Questioning everything is a good. The older we get the less we question what we were taught and learned over the years. That is why most discoveries and inventions are done by newbies that didn't know it couldn't be done.
 
That is an internet myth. It has been repeated so often it has taken on an aire of authenticity. Best guess was a bored EE student for a laugh. Sounds like something I or a roommate might have done.

A transformer reflects the load (secondary) impedance to the input (primary). If the secondary load is resistive, then the load as viewed at the input is resistive / in phase.

Amplifier power transformer secondary's are wired to rectifiers. That is DC i.e. resistive. A tiny secondary AC load may be incandescent pilot lights also resistive / no phase shift. Some of the newer units have switching power supplies, but the ones I have looked at start with a rectifier to a DC buss then to the switcher so still DC / resistive as viewed from the input.

Think about it. If that were not the case the power grid would never work. The substation transformers would add so much phase shift any AC motor would burn up from over current, your light bulbs would barely glow. At 90 degrees there is no real power available.

We got to do the calcs to determine the transformer addition to and effect on the secondary as seen from the primary in first or second year EE. The component of load viewed from the primary added by the transformer had best be negligible or the transformer is trash.

Questioning everything is a good. The older we get the less we question what we were taught and learned over the years. That is why most discoveries and inventions are done by newbies that didn't know it couldn't be done.

In the time you took to impressed us with your EE degree you could have proven your alleged myth by capturing the input current and voltage on a scope to show they were in phase. You might also capture the peak current at turn on, with power supply caps completely discharged so you would know if you exceed the I2T rating of the device. that would tell you if the triac could survive in larger amplifiers with large capacitance.

Years ago I used a phase angle controller chip to softstart a large power supply. I had DiDt troubles with the triac I used and didn't want to use snubbers. I abandon the project. The alternistor devices may solve that problem and some issues with other power switches.
 
  • Like
Reactions: mbz
This TRIAC mod is what I'm presently doing to my amps. It is the latest solution I chose for my perceived problem and it works to my satisfaction till something better comes along. It is shared as a point of departure for those who care. The modder must do their own risk / benefit assessment.

If you feel a similar need and come up with a different solution, I would love to read a thread on it's successful implementation. If it is easier or cheaper I may adopt it.

Power on is not my expected failure mode. Shorted output, filter cap or plugging something with a shorted cord into the switched power outlet (read my own stupidity) is what I expect to save my power switch from. I will happily replace a TRIAC if it blows faster than the amp fuse to save my switch.

Maybe should have given my considerations for development:
ALPS rotary power switch, no ratings, eye ball at contacts and spacing guess 5A at 120 on a good day maybe 7A at 220V
ALPS switch replacement not available so need to protect it.
Tired of installing ice cube relays and MOVs to reduce the switch load to 1 relay coil and remove possibility of amp failure blowing switch contacts.
Transformer fuse 7A
Supply 120V outlet string fed with a 20A breaker.
600 VAC at 25A continuous TRIAC rating designed for inductive loads has plenty of head room for my application.

What would happen if my idea failed?
It shorts a TRIAC and I'm out about 4 bucks and some time. Yawn!
In event of failure: Try something else or go back to relays.

As far as TRIACS on larger supplies. What is Larger? Talking home stuff that is.

I adapted the circuit from May 1987 QST "A Line-Side Regulator for High-Voltage Power Supplies" for my 3KV plate supply. I used a Ho Hum ECG56006 400 V 15A TRIAC with a 250V MOV across it for spike suppression. Line supply is 240 VAC. Soft start / comparator / regulator to protect the diode string on power up and regulate the output voltage is a simple OP amp described in the original article. Picked up another Ho Hum TIC 265D 400V 25A TRIAC to use when the ECG crapped out. I thought I had installed , but found the 15A one still there. Over 25 years and the 25A TRIAC is still a spare.

When I originally tested the HV supply in 1987 or 8 with the amp hooked to a dummy load it easily put out 2KW as measured with a Bird 43 / 2500H slug. That was not enough load on the power supply to require the TRIAC to remain turned on all the time. Transformer is absurdly large for a 2KW supply, but the price was right.

The old HV Transformer some one gave me is 2800VAC out with 110 / 220 in. Iron core is about 5 X 5 X 10 transformer is about 11 1/2 X 10 X 7. Don't think I ever weighed it but it's a back buster to pick up. It's over 10 years since I vacuumed the dust out of the PS and it sure needs it.

Gee that was (wasn't) interesting, whazit got ta do wid da subject at hand

Source impedance: Lower 220 V end run of AWG 10 in present location vs. audio amp 120V AWG 12 jumped though multiple outlets.

Biggest audio amp I have is SX 1050 transformer KVA unknown. HV PS transformer manufactured probably in the 40s KVA last know by whoever applied it allegedly in a radar. Transformer KVA limited by core saturation regardless of winding size. More iron = more KVA. HV PS transformer KVA guessed > 10X biggest audio PS transformer I have.

HV PS voltage regulator is an electronic power switch / proportional control. Flips on at bottom of dead band and off at top of dead band or remains on if full load is reached. No zero crossing detector same as I am doing on my audio amps. Key up, biased off, maybe once a minute. Key down, SSB, totally dependant on voice wave form guessing between 100 and 1000 times a minute. Equivalent switch operations in around 25 years of intermittent service < the number of hamburgers served by McDonalds
 
In the time you took to impressed us with your EE degree you could have proven your alleged myth by capturing the input current and voltage on a scope to show they were in phase. You might also capture the peak current at turn on, with power supply caps completely discharged so you would know if you exceed the I2T rating of the device. that would tell you if the triac could survive in larger amplifiers with large capacitance.

Actually, a while back I did a post on some thread using what ever amp was on my bench at the time that was a little more definitive. Used 2 Fluke 87 for I & E and Simpson 290-2 Watt meter. No surprise what's so ever, the old text books were correct. I X E = W => PF = 1
 
Did this tonight on my mac, the 6100 was demonstrating a shorting power switch from time to time- easy as pie. Cut the switch wired about 2 inches where it exits the tube, mounted the triac to a screw and wired it up- I endorse this mod!
 
Still haven't placed an order for one (but it's sitting in my cart at Digikey). I did think of one thing...most of the AC cords on these pieces aren't polarized, so if the plug is oriented so that the neutral is going to M2, how is this thing going to behave?

Edit: Nevermind...the only quadrant they won't trigger on is QIV (positive gate, negative MT2). Ought to work fine even with the AC plug reversed.
 
Last edited:
Simple AC Power Switch Using a Thyristor

A little better schematic, with a couple of part numbers that I've sourced, even though the only part I've used so far is the smaller TO-220 part. Still, I had a Kenwood KA-8100 here with a fried power switch and after managing to get the switch working well enough to trigger the triac, installed the device as shown in the attached schematic and it works like a champ.

I'm working on a couple of Yamaha B-2's next, and plan on adding this mod in a similar fashion.

Edit: The server is barfing pine needles at the moment, so I can't get the schematic uploaded. I'll edit this post with the scat as soon as the server is officially 'out of the woods'.

Edit 2: Got it...I think...

Edit 3: Got a few Q's on this, so I'll mention this here...the location of the load (at MT1 or MT2) is not important...Littelfuse makes that clear in their application notes. But, triggering should always take place between the gate and MT2 (and even this may not be critical, but I prefer to follow at least one standard).

Edit 4: You can, according to the applications sheet, leave off the cap and 100 ohm resistor running from MT1 to MT2. Theoretically, this can leave the unit susceptible to inadvertent power-up of power-off due to a power spike, but I've done this mod without the cap and resistor on a couple of units and seen no such thing.
 

Attachments

  • Thyristor_Switch_annotated.jpg
    Thyristor_Switch_annotated.jpg
    44.5 KB · Views: 2,428
Last edited:
The easiest, quickest and cheapest way to extend the life and eliminate the replacement of amp and receiver and auxiliary power switches is to use a switchable power strip.

Plug all the equipment of the audio system into the power strip and use its switch to control all. This also takes the load of the ac powered outlet on the rear of some receivers and off of the receivers power switch.

Turn volume control to minimum (or select no speakers connected) and stop tape and cd transports.

The switchable power strip also makes unplugging seldom used systems to eliminate damage from power and lighting strikes very easy.

The switch on the power strip is sacrificial and easily and cheaply replaced in event of failure.

Charlie
 
Last edited:
Well... there are at least a couple of reasons I won't ever go that route. First, the majority of vintage audio gear doesn't suffer from failure-prone power switches like the Setton RS-660, Pioneer SX-450/650 et. al. None of the gear I own (other than the RS-660, which I've already modified with an internal AC relay to prolong the switch life) has that issue and I've serviced so many of the same Tandberg models I know it just isn't an issue I'm ever likely to encounter.

Second, turning on all of the components simultaneously via a power strip is begging for annoying or potentially speaker-damaging transients with some equipment. I will ALWAYS turn my power amp (or receiver or integrated amp) on last, after any other components in the system are booted up and running. That way you don't send any DC transients to your speakers, unless the amp itself has an offset problem or lacks the kind of start-up protection to guard against spikes during power up.

I really think this concern about power switches being the Achilles heel of a lot of vintage gear and needing to be bypassed in this way borders on paranoia, but that's based on very rarely having to deal with failed power switches, YMMV.

I do however generally try and avoid using the rear panel AC outlets if possible - no reason to add the extra load to the power switch, esp. if I'm turning everything on manually anyway.

John
 
I've been doing some research on the Alternistor triacs. For my application I used a Q6030LH5. Note that although it is considered a triac, the data sheet indicates that the switch is actually two reverse paralleled SCRs. From what I recall older triacs did not switch evenly on both the positive and negative cycles. This resulted in a DC offset that would make a transformer run hotter. The alternistor solves this problem.

Some versions of the Phase Linear 700B used a 3 amp switch for the power switch. The switch is failure prone, due I assume to the inrush current, and nearly impossible to find a replacement. The images I've attached were taken to make sure the inrush current did not exceed the rating of the triac.

Using a fluke 87 set for peak min/max, 1 millisecond, and a Fluke 80I-600 current clamp I recorded a peak current of 147 amps. I did approximately 30 turn on tests to verify max current. Each test was done with the power supply caps completely discharged to get max inrush current. The peak 1 cycle surge (Itsm) for the Q6030LH5 is 350 amps. So I should be safe there.

The stock P/L 700B has 2 x10,000uf caps. After taking data on the stock amp I doubled the capacitance to see if the inrush was any higher.

The current waveform is the red trace and the voltage present at the input transformer is the yellow trace. Looking at the voltage waveform, it drops dramatically at turn on which helps to limit the inrush currrent. My bench is about 25 feet from the 200A breaker panel. The house is fed from a pad mounted transformer in the back yard about 75 feet from the house. Line voltage runs about 121 volts. Conclusion: I should have a relatively stiff source so the inrush currents are probably worst case.

The voltage waveform doesn't show any ringing due to the insertion of the triac. About 15 years ago I played with some triacs that would ring each cycle and I could not find a practical snubber that would stop the ringing. As advertised this triac does not require any snubber.

Scope setting for the current waveform is 50A/division. the first waveform is the stock amp, the second is with double the capacitors. The scope is a USB Stingray DS1M12.
 

Attachments

  • 700B in rush current reduced.jpg
    700B in rush current reduced.jpg
    40.7 KB · Views: 299
  • inrush current3 w 40,000uf.jpg
    inrush current3 w 40,000uf.jpg
    43.1 KB · Views: 277
I forgot to mention the schematic and mounting information in the previous post.

I used the schematic posted by EchoWars in post #34. I did not use the snubber he shows.

The triac was mounted in the middle of the chassis using the lower machine screw that holds the driver board. The nylon standoff was replaced with a shorter aluminum standoff.

I loaded the power supply to pull a continuous 8A load through the triac. Temperatures taken with a thermo couple stabilized at 67 degrees C. Although painfully hot to touch that's barely warm for semiconductors. The output transistors were in the mid 80's under these conditions.
 

Attachments

  • Thyristor_Switch_annotated.jpg
    Thyristor_Switch_annotated.jpg
    44.5 KB · Views: 555
Last edited:
This mod works like a champ. But beware of false current readings if you use a kill-a-watt to monitor current draw. I just wasted 2 hours trying to find a ghost 60 watt power draw on a Sansui AU-9500.

This Alternistor must toss some interference back into the power line that throws off the kill-a-watt's internal current sensing chip. A dim-bulb is enough of a filter to block it.
 
Back
Top Bottom