AC Reverse Polarity Protection

[chromesphere]

Probably overkill and ALOT more expensive then the familar 1n400x diode reverse polarity protection, i was reading this page about reverse polarity protection which was quite interesting, and noted the last example.  Maybe its an ad for the LTC4365?   

I was going to leave a reminder for me to order one to play around with but alas, you cant even find the LTC4365 even if you wanted one.  (probably on mouser though).  Under voltage protection is obviously useless for effects. But AC. overvoltage and reverse polarity checks alot of boxes.

Anyway heres the page, was an interesting read all the same:

http://digital-diy.com/general-electronics/419-power-protection-circuits.html

Paul

[armdnrdy]

On all newer 120 volt wall receptacles and cord ends the reverse polarity protection is a larger blade size for the neutral. The cord end only fits in the wall receptacle one way.

[chromesphere]

I was thinking more for things like....my damage control timeline uses 12v AC (believe it or not).  Fortunately it has a very distinct connector and i think its 2.5mm as well.

Paul

[greaser_au]

In a perfect world, a (nearly) "free voltage"  input is the way to go... Figure on, say, a 6.5V minimum (and some arbitrary maximum like 20V - even AC) input. Hang a bridge rectifier and capacitor off the  (isolated) DC socket, & limit the output to about 5V with *something*. Use this 5V output  - with some sort of isolated "advanced switchmode supply" (say a Valor device such as PM6078-004*;  yes, IT IS nearly the same size as a PP3 battery)  to produce a 9V isolated supply for the pedal circuit. Ground reference this any way you want; positive or negative ground- whatever- NO SHORTS!!!...  Happy for anyone to contact me for my concept licensing arrangements  (along with my idea for lefty pedals!!!!)

david

*http://cgi.ebay.com.au/ws/eBayISAPI.dll?ViewItem&item=271156805344

[greaser_au]

oops... double post

[R.G.]

Yep, that is right. The pedal should be able to work with AC or DC sources of power from a minimum to a maximum level.

I messed with this some years back, trying to come up with something simple(-ish) that would increase a pedal's tolerance for AC or DC power. I failed on one or the other counts because (1) the simple stuff doesn't work, as you cannot retrofit all pedals, and older pedals always connect ground to one or another of the incoming power leads, and (2) the high-immunity circuits are much more complex than most pedals.

Not that it can't be done, it's just reasonably costly, as witness the $4 number for the IC mentioned in the link. When I posted the idea for the MOSFET protector, people groused because the MOSFET then cost $0.50 to $0.75 each.

Guitarists insist on using half-century old pedals and value them more for being unmodified - "original" - so you can't insist that all pedals change. And these same old pedals will amplify any ground offsets or noise to a level up with which a guitarist will not put.

Indeed, in a perfect world, a pedal should be tolerant of various power sources. However, that won't happen, because of human psychology. The thing is, the original ancestor pedals so worshipped assumed they would have a cheap battery, and can't be modified. It has fallen to the power supply to match a battery or else. The onus to conform is on the power supply. Changing the paradigm to making the pedal include circuitry as or more complex than the pedal circuits themselves to be power supply tolerant amounts to the pedal designer deciding to spend money to make life easier on the power supply makers.  Won't happen.

[greaser_au]

Indeed, in a perfect world, a pedal should be tolerant of various power sources. However, that won't happen, because of human psychology. The thing is, the original ancestor pedals so worshipped assumed they would have a cheap battery, and can't be modified. It has fallen to the power supply to match a battery or else. The onus to conform is on the power supply. Changing the paradigm to making the pedal include circuitry as or more complex than the pedal circuits themselves to be power supply tolerant amounts to the pedal designer deciding to spend money to make life easier on the power supply makers.  Won't happen.

Hi R.G.

This is where your West Texas wisdom #37 seems most appropriate  That said, given a few experiments to simulate a PP9 battery on the output side, it might be useful to some of us here for clones. I imagine that someone with a lot more insight than me could carry this idea to it's logical conclusion.

david


Note: I bought a quantity  of these  DC-DC converters (for a couple of non-effects related projects) and landed them for somthing under AU$1 each - I'm not affiliated with the seller further than that.

[R.G.]

There are a number of commercial DC-DC converters that work. They're expensive unless you get them surplus, at least here. You got a good deal.

I keep coming back to a design along those lines for a battery faker that's been playing in my head for a few years. If you had the right high frequency transformer, you could take most anything from about 5V to 24V, AC or DC, and make 9Vdc, isolated and regulated, out of it. I believe there is a commercial product that does something similar.

Today's power converter control chips make this easy enough for the electronics. One 8-pin IC, possibly an external MOSFET, and the normal slew of Rs and Cs can do everything needed. The trick is that you need a fairly well-designed, if small, custom transformer to do the isolation. Easy enough to do, but then it pulls in the need to source ferrite cores and bobbins, wind well for high frequency switching, and so on.

The concept is easy enough. The practicalities of doing it cheaply, reliably, and not making it spew spurious RF emissions always push it just over the edge of doing it for me.

[Paul Marossy]

I keep coming back to a design along those lines for a battery faker that's been playing in my head for a few years. If you had the right high frequency transformer, you could take most anything from about 5V to 24V, AC or DC, and make 9Vdc, isolated and regulated, out of it. I believe there is a commercial product that does something similar.

Nice idea. But could it have a whine like those charge pump IC chips?

[PRR]

> you cant even find the LTC4365 even if you wanted one.  (probably on mouser though).

Not Mouser, but DigiKey has 'em.

Surface-mount only. $4 a pop. Plus several resistors and a pair/dual MOSFET.

http://www.digikey.com/product-search/en/integrated-circuits-ics/pmic-power-management-specialized/2556708?k=LTC4365

Or: http://www.digikey.com/ and Search for LTC4365, then check "in stock".

You can order direct from Linear Semiconductor. 2 bucks, but too many choices.
http://www.linear.com/purchase/LTC4365


> 120 volt wall receptacles ...larger blade size for the neutral

Yes, but that is not what he is talking about.

(Anyway, inside our work we should never care or rely-on any L-N distinction.)

The LTC4365 will only turn-on a load if the source is the right direction and within a specified voltage window. It works to 34V either way, doesn't die until 60V. So this goes on the low-volt side of wall-warts, not the raw 120V line.

So you could set up to reject all negative connections, reject 0V to 7V, pass 7V to 12V, and reject anything higher.

On AC.... seems to me it would turn-on for part of each cycle. If the pedal has a massive rail-cap, it might even play. More likely it would buzz real bad. There is a turn-on delay, so maybe it just won;t turn-on with normal AC line frequency.

It also appears to drop voltage some. The MOSFET Gate is only driven to the highest available voltage. MOSFETs typically drop 1V-3V Gate to Source.

[greaser_au]

There are a number of commercial DC-DC converters that work. They're expensive unless you get them surplus, at least here. You got a good deal.

I keep coming back to a design along those lines for a battery faker that's been playing in my head for a few years. If you had the right high frequency transformer, you could take most anything from about 5V to 24V, AC or DC, and make 9Vdc, isolated and regulated, out of it. I believe there is a commercial product that does something similar.

Today's power converter control chips make this easy enough for the electronics. One 8-pin IC, possibly an external MOSFET, and the normal slew of Rs and Cs can do everything needed. The trick is that you need a fairly well-designed, if small, custom transformer to do the isolation. Easy enough to do, but then it pulls in the need to source ferrite cores and bobbins, wind well for high frequency switching, and so on.

The concept is easy enough. The practicalities of doing it cheaply, reliably, and not making it spew spurious RF emissions always push it just over the edge of doing it for me.

Cheaply, reliably, quietly;  and the other hard one: efficiently (though up to 100mA or so it probably doesn't matter that much)...    

I wasn't suggesting this so much as a 'faked battery' add-on (I guess it is, anyway), but as an alternative easy-diy suggestion in the vein of Paul's original point - an almost bullet-proof power socket.  In this case the bridge/capacitor/regulator are all pretty much standard junkbox items, there is only one 'hard item' -  the brick itself - and that isn't too bad while they're still available on the surplus market.  No messing around with many unusual and specialised parts - and the usual pain of developing skillz, e.g. winding.

The PP3 reference was intended to show that the things are physically HUGE in the scheme of things- no easy 1590A builds with these!!! With any useful lead length, they are probably thicker than a PP3  

david

PS: I should point out that the "licensing arrangments" comment is one of my feeble attempts at ironic humour, I've used it few times on various fora.  I reserve the right to keep using it, even where COMPLETELY inappropriate!!!