relay logic

I think this is right. It was a minor brain sprain, after the banana bread and rum.

formatting link

Imagine designing an entire dial telephone system with relays.

Reply to
jlarkin
Loading thread data ...

Any particular reason you want the four windings in parallel for the 30 volt output? Otherwise it can be easily done with two relays.

Seems to me you would be trusting the transformer consistency a bit too much have four winding in parallel in your solution, won't it get a bit warm if there is any buck/boost from having the windings in parallel?

John :-#)#

Reply to
John Robertson

Anyone remember that guy who used to post on one of the sci.electronics groups several years ago who seemed like he was trying to control a whole house and barn electrical system with relay logic?

Reply to
bitrex

All you need are the two 30V windings and the 90V winding; the rest is fluff.

Take a look at

formatting link

Reply to
Robert Baer

To use all the copper all the time, which keeps the output impedance down and gives max VA on all settings. The 600 watt class-D amp chip is $6, but the huge 120 VA transformer costs $65.

Transformers are wound with exactly the number of designed turns. Windings in parallel are OK.

I also want a "fault" state, namely open circuit. I can drive three relay coils independently, which is 8 states. Maybe I can find or make an "open" state, still with three DPDT relays.

I don't have any systematic way to do relay logic. The only formal structure that I know of is the tree that makes a 1:2^N mux; everything else is just fiddling. Hurts my head.

Reply to
jlarkin

See my comment to JR about using all the expensive copper.

Reply to
jlarkin

Your schematic sketch is slightly off then, relay K2's second contact layout is reversed - the drawing shows one SPDT switch going clockwise and the 2nd SPDP has to mentally go counterclockwise for it to work correctly. You need to redraw the second half of K2 so it looks like the

2nd half of K3. Electrically the same, schematically then correct.

I still work on EM stuff all the time... I had ignored the Class D note, didn't think it was important, my mistake!

John :-#)#

--
(Please post followups or tech inquiries to the USENET newsgroup) 
                      John's Jukes Ltd. 
 Click to see the full signature
Reply to
John Robertson

The convention here is that all relays are shown in their de-energized state. I could flip K2B, but that would be cosmetic.

One somewhat counter-intuitive thing about a class D amp is that its power output and efficiency don't much care about the power supply to output swing ratio. That gives one more freedom when buying power supplies and transformers.

Reply to
jlarkin

Hi, here is a slightly different take on that one.

Schematic:

formatting link

Again, the circuit uses 3 relays, but K1 and K2 can be paralleled, or if a relay with 4 changeover contact pairs ("4-pole 2-throw") is available then K1 and K2 can be combined into that one relay.

The basic principle is that windings form virtual pairs and each pair is controlled by one relay to be either in series or in parallel. A final relay controls the "pair of pairs" in the same way.

It is possible to use independent K1 and K2 too, this allows an extra output selection of 90V in addition to 30V, 60V and 120V. The 90V option has a somewhat reduced current capability, its configuration is that one pair of windings is in parallel and all the others are in series.

However when K1 and K2 are used independently, an invalid state must be avoided: When K1 and K2 have unequal states (one is on, the other is off, no matter in which order) then K3 must never be on at the same time. That would result in a transformer overload (paralleling of unequal windings is like a short circuit), however this is avoidable by disabling K3 through an override whenever "K1 xor K2" is true.

Regards, Dimitrij

Reply to
Dimitrij Klingbeil

That's nice, more orderly than mine. Relay drivers are cheap (we use TPIC6595) so I may as well control the three relays independently and offer 90 volts as a feature. Thanks.

These are nice looking DPDT relays

formatting link

with several sources.

I can kill the class-D amp input when we switch the relays, to avoid transient faults.

Reply to
John Larkin

You don't understand you own circuit. Relays K1 and K3 are driven with the same logic according to your table. So you can combine them into a single relay and retain all three voltages, 60, 90 and 120.

--
  Rick C. 

  - Get 1,000 miles of free Supercharging 
 Click to see the full signature
Reply to
Rick C

torsdag den 12. september 2019 kl. 21.37.21 UTC+2 skrev John Larkin:

they should available everywhere, they fit the standard DIN rail relay sockets

Reply to
Lasse Langwadt Christensen

What would it be a silly idea? You don't need high speed there and relays are insanely immune to surges, especially when they are in a stable state during the surge.

Best regards, Piotr

Reply to
Piotr Wyderski

Relay control of lights and outlets in high end homes was common in the '60 s and '70s. A load could have a local switch, be remote controlled or from a timer and all with low current 24VAC wiring. A single switch could turn o n every light on the property, if needed. That could be connected to the dr y output of an alarm panel for security. I knew one electrical contractor w ho installed and maintained them, when I was teaching people to install ala rm wiring.

Reply to
Michael Terrell

What's wrong with just stacking all the secondaries in series and then picking off the selected voltage at the appropriate tap? Isn't that just two relays? And 30-60-120 is a weird progression.

Reply to
bloggs.fredbloggs.fred

Actually, i am building a 12V/48V/72V/... switching system. The batteries would be in serie and/or parallel, depending on config. However, if one relay fails, there will be a big-bang. So, perhaps triple redundant relay/logic.

Reply to
edward.ming.lee

no problem intrinsically other than that as I recall the potential implementations he showed were horrible and would have likely burned his house down, much less met some kind of like "electrical code"

Reply to
bitrex

as I recall the system this guy was working on was more like a relay computer, switching 120V loads, and the relays themselves were 120V coil types inside of feedback-loops driven off the loads themselves.

Like hairball-logic built with line voltages

Reply to
bitrex

That wastes VAs by not using all the copper all the time.

30-60-90-120 is even better. I'm simulating alternators and there are all sorts of alternators.
Reply to
John Larkin

ote:

cs

'60s and '70s. A load could have a local switch, be remote controlled or f rom a timer and all with low current 24VAC wiring. A single switch could tu rn on every light on the property, if needed. That could be connected to th e dry output of an alarm panel for security. I knew one electrical contract or who installed and maintained them, when I was teaching people to install alarm wiring.

Reminds me of a 1970s commercial scale misting unit, all relay logic run at 240v. Yes, in a soaking wet zone. No inbuilt checks, redundancy, fuses, RC D, anything. And made of asbestos.

NT

Reply to
tabbypurr

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.