I want to customize the existing preamp using signal relays and the existing IR remote control hardware. Unused functions on the remote will be converted to another purpose. This is a one off custom DIY project.
I have four remote keys that I want to each control one relay in the preamp. We are talking small high quality DC coil DPDT pcb signal relays.
I have the schematic for the remote receiver and decoder, and have located a signal pulse from the 5V remote decoder IC that goes low for 150ms on each key press. Each of the four keys has a dedicated control line for this purpose. I was hoping to find a control line that latches high or low but I can only find a pulse signal. I want to use that pulse to toggle a relay, the first time the remote key is pressed the relay energizes, with the second press it de-energizes. I need a simple circuit that converts the 150ms pulse into a latched high or low drive signal that I can use to drive a transistor which in turn drives the relay. The part that I can't figure is how to derive the toggled latched output from the pulse input. A simple thru hole TTL or CMOS cct is what I have in mind, something I can make on perf board. What do you recommend? The IR remote receiver circuit uses the 5V TC9134 chip, Toshiba I think, 1990s vintage, the control lines I am thinking of using are the outputs type SP.
Ted in NY
I have four remote keys that I want to each control one relay in the preamp. We are talking small high quality DC coil DPDT pcb signal relays.
I have the schematic for the remote receiver and decoder, and have located a signal pulse from the 5V remote decoder IC that goes low for 150ms on each key press. Each of the four keys has a dedicated control line for this purpose. I was hoping to find a control line that latches high or low but I can only find a pulse signal. I want to use that pulse to toggle a relay, the first time the remote key is pressed the relay energizes, with the second press it de-energizes. I need a simple circuit that converts the 150ms pulse into a latched high or low drive signal that I can use to drive a transistor which in turn drives the relay. The part that I can't figure is how to derive the toggled latched output from the pulse input. A simple thru hole TTL or CMOS cct is what I have in mind, something I can make on perf board. What do you recommend? The IR remote receiver circuit uses the 5V TC9134 chip, Toshiba I think, 1990s vintage, the control lines I am thinking of using are the outputs type SP.
Ted in NY
How about a D-type flip-flop? When the clock line sees a low-to-high transition, it copies the state of the inputs to the outputs. I guess you would need to reverse the polarity. Then you can use the data as the clock, just include a cap/resistor to introduce a small delay on the clock line. Connect each input to the clock pin via a diode.
So (again, assuming polarity is reversed), when you press a key the corresponding input goes high and the clock pin sees a l2h transition. The outputs are switched to reflect the state of the inputs, so the output line corresponding to the pressed key is high and the others are low. Then you don't care about the previous state.
This is what I use in my preamp input selector, but it is driven by mechanical keyswitches.
There is probably something much better, or that correctly handles the normally-high outputs of your decoder, but this works and a quad flipflop is pennies.
See my gallery for a schematic (6 poles instead of 4).
So (again, assuming polarity is reversed), when you press a key the corresponding input goes high and the clock pin sees a l2h transition. The outputs are switched to reflect the state of the inputs, so the output line corresponding to the pressed key is high and the others are low. Then you don't care about the previous state.
This is what I use in my preamp input selector, but it is driven by mechanical keyswitches.
There is probably something much better, or that correctly handles the normally-high outputs of your decoder, but this works and a quad flipflop is pennies.
See my gallery for a schematic (6 poles instead of 4).
Last edited:
I took a look at your circuit, that's helpful, thank you, it is just the sort of thing I am looking for, a simple cct that I can build with a few ICs and discrete parts. It's almost exactly the application I am interested in, however my relays will be doing a stereo-width switching function rather than source selection.
I think I understand the polarity inversion requirement, in your circuit the switches pull the D inputs high whereas mine go low.
I have some questions. I'm not sure how the relay cancelling requirement is met. In your circuit when S1 is pressed and released the LED and relay (off page) turn on. What happens when S2 is pressed? Does relay S1 get released? That's what I need, to be able to toggle the relay by sequential presses of the remote key. Hope I've explained it well enough.
PS another way to think of it is this, I want only one of the four relays energized at any moment, selecting #2 must cancel #4 etc.
I think I understand the polarity inversion requirement, in your circuit the switches pull the D inputs high whereas mine go low.
I have some questions. I'm not sure how the relay cancelling requirement is met. In your circuit when S1 is pressed and released the LED and relay (off page) turn on. What happens when S2 is pressed? Does relay S1 get released? That's what I need, to be able to toggle the relay by sequential presses of the remote key. Hope I've explained it well enough.
PS another way to think of it is this, I want only one of the four relays energized at any moment, selecting #2 must cancel #4 etc.
Last edited:
PPS another question. What is the policy with unused 40174 pins, should they be tied to ground, or supply, or output?
I guess I did not understand your requirement when I read your first post. This circuit won't toggle, that is, putting another pulse on input1 will not set output1 low. Assuming relay1 is on, then when you put a pulse on another input (say #2), that sets input2 high while the rest remain low, the clock goes high, copies input state to output state, and now only relay2 is engaged.
I had understood you to mean that the decoder would put a pulse on one of 4 pins.
Been a while since I read the data sheet (that preamp was a one-time project 4 years ago!), I think they cannot be floating, so I would tie them to ground, but I would do it through a resistor so they can't interact. I don't know if that is needed but you could likely find out via google in a few minutes.
Maybe I did understand you. If the decoder has a dedicated line corresponding to each key, and you just need to use the 150ms pulse to trigger the corresponding relay (and only that relay), then the above circuit will work.
Funnily enough, when I built it I was concerned about the possibility of triggering 2 relays at the same time (hence selecting 2 simultaneous input sources) by accidentally hitting 2 buttons at the same time. I tried for quite a while but could not make it do that, the timing is just too tight for me to get both buttons closed at the exact time, even though the time constant I used on the clock pin was way bigger than needed (I just used what I had in spare parts). There needs to be a short delay ("settling time") between setting an input high and triggering the clock, but it's only something like 20ns. In 4 years there was exactly one occasion when it didn't work right, it turned on in a bad state and pressing buttons did nothing, but a power cycle fixed it and I could never replicate it to troubleshoot. When I prototyped it I found I needed to have a decoupling cap right on the supply pin, but that wasn't exactly surprising.
I used a 40174 because I had 6 relays to switch. You can get D-type flip-flops in dual, quad, hex, octal, etc. For your needs a quad flip-flop is what you need (40175 I think), then you don't have to worry about the extra inputs.
Last edited:
See my gallery. It is the circuit i usee in my Mezmerise, that you helped me with so much.
Unused inputs should be tied directly to wherever is most convenient for board layout, supply, ground or even an adjacent output pin from an active part of the device.
Leave unused outputs floating.
Thank you Nezbleu. Yes, now we are getting into the subtleties of what exactly it means to select one from multiple relays. In the case of source select, your example, we want one relay always selected, and selecting one should cancel the others. I believe this is what your circuit does. How does your system initialize? Which source relay is selected on power up?
There are four keys, four control lines and four relays. On power up no relays are selected. After I have used the keys (in some random manner) then I need to be able to get back to this "no relay selected" condition somehow.
Mooly, thank you for the advice.
There are four keys, four control lines and four relays. On power up no relays are selected. After I have used the keys (in some random manner) then I need to be able to get back to this "no relay selected" condition somehow.
Mooly, thank you for the advice.
Last edited:
Hi Nezbleu
Another question about logic stuff. The decoder is powered from 5V. I want to use 12V for the relays with a driver transistor for each relay. Should the CMOS flip-flop between the decoder and the transistor be powered from the decoder supply (5V) or the relay supply (12V)? I want to avoid accidentally blowing the decoder IC!
About interfacing the decoder and the flip-flop. I attached the data sheet spec for the control pins, marked SP22 - SP1. They are active low. Do I simply connect the SP pin to the D pin? Can I rely on the decoder outputs to pull the D inputs high? Or do I need pull up resistors?
PS what is the drawing package you use for the schematic? I need one, is it free? 🙂
Another question about logic stuff. The decoder is powered from 5V. I want to use 12V for the relays with a driver transistor for each relay. Should the CMOS flip-flop between the decoder and the transistor be powered from the decoder supply (5V) or the relay supply (12V)? I want to avoid accidentally blowing the decoder IC!
About interfacing the decoder and the flip-flop. I attached the data sheet spec for the control pins, marked SP22 - SP1. They are active low. Do I simply connect the SP pin to the D pin? Can I rely on the decoder outputs to pull the D inputs high? Or do I need pull up resistors?
PS what is the drawing package you use for the schematic? I need one, is it free? 🙂
Last edited:
First: think of the flip-flop as a pair of data registers (4bits wide in your case). When the clock pin goes from low to high the chip copies the state of the input register to the output register. This could be one line "on" as in my case, or any combination. The point is that whatever input pins are high or low, the corresponding outputs will be set to that state.
There is also a "clear" pin which sets all outputs low when it goes low. In my case i just tied it to gnd via a resistor, and fed it from V+ via a small RC network so it stays low for a few ms at power-on, so it always starts with nothing selected. You could add something to pull it low when you need to clear everything.
I would power the flip-flop from the 5V supply, but the transistors will be in series with the relay coils. It just needs to cause the transistors to switch on, play with the base resistor values to get them switching and not dissipating much power. Any small signal TO92 bjt should work. AndrewT helped me immensely with understanding that part, you can find our posts in the Mezmerize build thread from a few years ago.
There is also a "clear" pin which sets all outputs low when it goes low. In my case i just tied it to gnd via a resistor, and fed it from V+ via a small RC network so it stays low for a few ms at power-on, so it always starts with nothing selected. You could add something to pull it low when you need to clear everything.
I would power the flip-flop from the 5V supply, but the transistors will be in series with the relay coils. It just needs to cause the transistors to switch on, play with the base resistor values to get them switching and not dissipating much power. Any small signal TO92 bjt should work. AndrewT helped me immensely with understanding that part, you can find our posts in the Mezmerize build thread from a few years ago.
Last edited:
Drawing pkg was gEDA on my old linux system (which died). It is a powerful and flexible electronic design package and I do not recommend it. I expect there are many newer and easier packages. It was free though 🙂
I guess I missed or ignored this bit. The problem is that the outputs are active low, which is the opposite sense we expect on the flip-flop. We want the selected output to go high, and the clock pin expects a low-to-high transition. I expect there are NOT gates you could use between decoder and flipflop for this purpose.
I expect the decoder outputs can do all the work, but the data sheet should tell you. I am really no expert with this stuff. I just like this little circuit because I figured it out, based on something I had seen a few decades previous, and because I think it is elegant (data is clock!).
Thank you. I have studied the schematic of the decoder some more and things are getting more complicated unfortunately. The decoder output ports I described, the ones that go low for 150ms, are already driving PNP transistors in a key matrix, which is biased to -22V, this is to interface with the vacuum fluorescent display driver that receives the control signals from all the ports. I'm beginning to lose confidence that by adding another load to the output port I may screw up the existing display functions, I think I may be wise to pause and think this over, this is getting deeper than I expected! It has been helpful to have this discussion, I appreciate your help Nezbleu, thank you.
Thinking is good, as is study of the relevant data sheets. However, I would not worry too much about loading the decoder outputs, esp since you need to have an inverting buffer between decoder and flipflop.
Best of luck, ping me if you have more questions.
Best of luck, ping me if you have more questions.
Hi,
here´s a 4-in-1-out switch with an additional optional output relay.
It uses a 4-in BCD-to-decimal decoder CD4028.
From this decoder it uses only the single-bit outputs 1,2,4 and 8.
Diodes D1-D4 feed the output condition back to the associated input to achieve a latched, self-holding condition.
The 4 outputs feed into a NPN transistor-Array that drive the relays and LEDs.
The optional C1 realizes a preset condition and decides which input is chosen after power-up ... in this case Input No.1.
If another one of the input tactile switches is closed, the decoder momentarily sees two inputs active and switches from the former output to the new output (which is anything different from 1,2,4 or 8).
As now there´s no feedback from output to input any more, only the (still) pressed switch is seen as new input.
The decoder switches to the corresponding output (1,2,4 or 8) which is then latched again.
The switching action is "break before make".
D5-D10 collect the unused outputs, of which one is momentarily active when a new input switch is pressed.
This pulse is used to trigger one half of a Dual Monostable Multivibrator CD4098.
Every pulse at its input triggers the CD4098 output for a certain delay time, that depends on R5/C2.
The "Out"-relay falls off for a period spanning the delay time.
After the delay has passed the "Out"-relay releases the output again.
The amount of parts depends on wether or not the Out-option is used.
Without the option IC1, 4 Diodes and 4 NPN-transistors suffice.
jauu
Calvin
here´s a 4-in-1-out switch with an additional optional output relay.
It uses a 4-in BCD-to-decimal decoder CD4028.
From this decoder it uses only the single-bit outputs 1,2,4 and 8.
Diodes D1-D4 feed the output condition back to the associated input to achieve a latched, self-holding condition.
The 4 outputs feed into a NPN transistor-Array that drive the relays and LEDs.
The optional C1 realizes a preset condition and decides which input is chosen after power-up ... in this case Input No.1.
If another one of the input tactile switches is closed, the decoder momentarily sees two inputs active and switches from the former output to the new output (which is anything different from 1,2,4 or 8).
As now there´s no feedback from output to input any more, only the (still) pressed switch is seen as new input.
The decoder switches to the corresponding output (1,2,4 or 8) which is then latched again.
The switching action is "break before make".
D5-D10 collect the unused outputs, of which one is momentarily active when a new input switch is pressed.
This pulse is used to trigger one half of a Dual Monostable Multivibrator CD4098.
Every pulse at its input triggers the CD4098 output for a certain delay time, that depends on R5/C2.
The "Out"-relay falls off for a period spanning the delay time.
After the delay has passed the "Out"-relay releases the output again.
The amount of parts depends on wether or not the Out-option is used.
Without the option IC1, 4 Diodes and 4 NPN-transistors suffice.
jauu
Calvin
Attachments
Last edited: