Hi community,
In a small portable amp I would like to use a dual coil latching dpdt relay as an input selector (latching because the amp will sometimes be battery driven). From my research I have found that the TQ2-L2-5V from Panasonic Electric Works should work fine. Because the amp will probably be driven sometimes by a 12 V battery, sometimes with a 21 V wall wart, I plan to use a voltage regulator to power the relay.
Now, even if with a dual coil relay the wiring is supposed to be simpler, I’m not sure, if I got it right. The schematic in the attachment shows what I have come up with so far (it shows the schematic of the relay, as well as my plan to wire it).
I’d be happy if someone could help me shed some light on the following:
Cheers
In a small portable amp I would like to use a dual coil latching dpdt relay as an input selector (latching because the amp will sometimes be battery driven). From my research I have found that the TQ2-L2-5V from Panasonic Electric Works should work fine. Because the amp will probably be driven sometimes by a 12 V battery, sometimes with a 21 V wall wart, I plan to use a voltage regulator to power the relay.
Now, even if with a dual coil relay the wiring is supposed to be simpler, I’m not sure, if I got it right. The schematic in the attachment shows what I have come up with so far (it shows the schematic of the relay, as well as my plan to wire it).
I’d be happy if someone could help me shed some light on the following:
- Is the wiring correct?
- Can I power the relay with a voltage regulator?
- Do I need flyback diodes? Do I need one for each coil? If yes, are they at the right place?
Cheers
Attachments
If you use latching relay I would suggest to use two push buttons to select inputs. The switch as it is put now will make one relay coil always energized while it is not necessary. Instead of 20V Zener diodes you can use reverse biased 1N4000 diodes otherwise you do not really suppress back EMF of 12VDC coil with 20V Zener.
Also, standard signal relays with 12VDC coils can have around 10mA current rating. This current is only necessary to switch the relay and later can be reduced to the "hold" current which can be two-three times smaller. This way you can use your switch with distinct position showing which input is ON. If you build a power amp, then 5mA of constant current to hold the relay will not drain battery too much compared to the amp itself. Note that the relay will only be energized if one of the two inputs is selected. The "default" input setting will not drain any current.
Regards,
Oleg
Also, standard signal relays with 12VDC coils can have around 10mA current rating. This current is only necessary to switch the relay and later can be reduced to the "hold" current which can be two-three times smaller. This way you can use your switch with distinct position showing which input is ON. If you build a power amp, then 5mA of constant current to hold the relay will not drain battery too much compared to the amp itself. Note that the relay will only be energized if one of the two inputs is selected. The "default" input setting will not drain any current.
Regards,
Oleg
Second try...
Oleg,
Thanks a lot for your reflections and suggestions!
I didn't realize that with a maintained switch one of the relay coils will always be energized. But it totally makes sense.
For the simplicity of the whole design I will nevertheless stick with the maintained switch. As you suggest, the impact on battery life should be bearable.
Although, if this way one relay coil always stays energized anyway, it doesn't make sense to use a latching relay, if I understand it right. Using a non latching relay instead, will make the circuit even simpler.
And to try to keep the power consumption low, I will try to follow some of the suggestions I found on electronics.stackexchange.com. I will use a high sensitive 12V relay, and power it with a 10V linear voltage regulator. According to the datasheet of the relay 10V should be enough to get the relay to switch.
I would be happy to hear from you, if this all makes sense... I attached a new version of the schematic to show, how I understood your inputs.
Best,
Simon
Oleg,
Thanks a lot for your reflections and suggestions!
I didn't realize that with a maintained switch one of the relay coils will always be energized. But it totally makes sense.
For the simplicity of the whole design I will nevertheless stick with the maintained switch. As you suggest, the impact on battery life should be bearable.
Although, if this way one relay coil always stays energized anyway, it doesn't make sense to use a latching relay, if I understand it right. Using a non latching relay instead, will make the circuit even simpler.
And to try to keep the power consumption low, I will try to follow some of the suggestions I found on electronics.stackexchange.com. I will use a high sensitive 12V relay, and power it with a 10V linear voltage regulator. According to the datasheet of the relay 10V should be enough to get the relay to switch.
I would be happy to hear from you, if this all makes sense... I attached a new version of the schematic to show, how I understood your inputs.
Best,
Simon
Attachments
Hi Simon,
The way you have it now looks good to me. I usually use Omron G6K miniature relays for signal switching. They have around 100 mW coil power rating which is really low. I even made a PCB for such relay selector switch as you need except for the regulated power supply. My selector switches both, signal and its return.
Regards,
Oleg
The way you have it now looks good to me. I usually use Omron G6K miniature relays for signal switching. They have around 100 mW coil power rating which is really low. I even made a PCB for such relay selector switch as you need except for the regulated power supply. My selector switches both, signal and its return.
Regards,
Oleg
Hi Oleg,
Thank you for your feedback. Much appreciated! I will definitely use the Omron G6K.
I have a question concerning your PCB you linked to. I don't know if I can formulate this understandably:
You used two DPDT relays. So, you have 4 input pins per channel, but would in my understanding only need 3 if you switch both signal and ground. How did you use the 4 pins per channel (NC-A 1+2 and NC-B 1+2; NO-A 1+2 and NO-B 1+2)?
Cheers,
Simon
Thank you for your feedback. Much appreciated! I will definitely use the Omron G6K.
I have a question concerning your PCB you linked to. I don't know if I can formulate this understandably:
You used two DPDT relays. So, you have 4 input pins per channel, but would in my understanding only need 3 if you switch both signal and ground. How did you use the 4 pins per channel (NC-A 1+2 and NC-B 1+2; NO-A 1+2 and NO-B 1+2)?
Cheers,
Simon
On my PCB the returns of the left (named "A") and the right (named B) channels are not connected. I meant the PCB for dual mono configuration for maximum channel separation. If needed the GNDs can still be connected together by jumpering the corresponding bottom side relay pins and skipping that relay.
GNDs of individual inputs of the same channel are also separate which means you always need to switch two wires for each channel.
Grounds can be terminated at the input board.
if it is differential then it must use 2 switches per channel
if it is differential then it must use 2 switches per channel
That was my idea when I designed my PCB. If necessary GNDs can just be tied together on the underside of the PCB or switched using the second relay depending on the needs. This allows using the PCB for single ended and differential signal switching due to close coupling of signal traces. They just follow each other on the opposite sides of the PCB which makes the smallest loop area between the two traces.
Will my PCB work?
Hi,
My project advances step by step. Now I’m in the process of designing the little PCB with the input selector on it.
Besides the input selector, there will also be a Stereo-2-Mono-Switch on the same PCB as described by myself in the thread «Stereo to mono with switch: Do my circuits work?», as well as an LED. The schematic for both elements combined is attached to this post. I tested it on a breadboard with the parts I plan to use, and it works.
I will not solder the LED and the SPST switches to control the relays directly on the PCB therefore in the schematic you see symbols of JST connectors at their place.
Also attached to this post is the PCB I designed, in a size that would fit into my amp case. The fat traces are 40 mil, the thiner ones 25 mil.
Now, I’m wondering if this is supposed to work, and I’d be happy for suggestions on what I could improve.
Cheers,
Simon
Hi,
My project advances step by step. Now I’m in the process of designing the little PCB with the input selector on it.
Besides the input selector, there will also be a Stereo-2-Mono-Switch on the same PCB as described by myself in the thread «Stereo to mono with switch: Do my circuits work?», as well as an LED. The schematic for both elements combined is attached to this post. I tested it on a breadboard with the parts I plan to use, and it works.
I will not solder the LED and the SPST switches to control the relays directly on the PCB therefore in the schematic you see symbols of JST connectors at their place.
Also attached to this post is the PCB I designed, in a size that would fit into my amp case. The fat traces are 40 mil, the thiner ones 25 mil.
Now, I’m wondering if this is supposed to work, and I’d be happy for suggestions on what I could improve.
Cheers,
Simon
Attachments
The PCB layout is a bit chaotic to follow but the schematic is fine. Just one question, why did you use 10k resistors in series with the signal?
They're normally shorted out, unless mono is selected, when they mix the channels together.
Last edited:
But this adds a lot to the output impedance of the circuit in mono mode which seems to be the default state. Such high value can interfere with any low pass (EMI) filter downstream. Wouldn't 1k resistors be sufficient for the job?
It's usually best to keep impedance levels as low as possible, without loading the sources excessively.
Tube preamps have often used 10k mixing resistors. Solid state, usually smaller.
Last edited:
Thanks OlegSh and rayma for your comments so far.
@OlegSh: I like how you put this: «The PCB layout is a bit chaotic to follow but the schematic is fine.» 🙂 So, the PCB should work, but design wise there is room for improvement. Do you think the reason for the chaos is primarily due to the placement of the components?
Concerning the resistors: Rayma is right about their function. Concerning the value I'm actually following someones suggestion. As far as I've understood the value has to be smaller than the impedance of the source connected to the mixer. In my case this would be a bluetooth module, an iPhone, Laptop, CD-Player etc. So, I assume that I'm on the safe side with 10k. But maybe it's a bit high.
@OlegSh: I like how you put this: «The PCB layout is a bit chaotic to follow but the schematic is fine.» 🙂 So, the PCB should work, but design wise there is room for improvement. Do you think the reason for the chaos is primarily due to the placement of the components?
Concerning the resistors: Rayma is right about their function. Concerning the value I'm actually following someones suggestion. As far as I've understood the value has to be smaller than the impedance of the source connected to the mixer. In my case this would be a bluetooth module, an iPhone, Laptop, CD-Player etc. So, I assume that I'm on the safe side with 10k. But maybe it's a bit high.
Yes🙂 Also keep in mind that the ground trace should follow signal as close as possible. On your PCB design ground and signal go different ways creating a loop susceptible to interference.
Yes, probably 5k would also be fine. Any lower, check the specs for your sources to be sure.
Certainly 1k would be pushing it for many sources with op amp outputs.
Thanks OlegSh, and rayma.
I really appreciate your inputs!
I'll definitively give another try concerning the layout, and will lower the resistors to 5k.
I really appreciate your inputs!
I'll definitively give another try concerning the layout, and will lower the resistors to 5k.
Try something like this arrangement. The ground line(s) can go up the middle.
Consider using a wall wart DC power supply instead.
https://www.digikey.com/schemeit/project/selector-1UAID58303HG/
Last edited:
don't use a 10V regulator for a 12V relay.
use the full 12V to trigger the relay.
look at adding a current saver circuit to reduce the operating current after the relay has successfully triggered.
using 12v for a 12V relay means the current saver is not very effective, but still worth using.
Don't use lower voltage relays, they use too much current.
When you use the high voltage supply, then use a regulator to bring down the voltage to what the amp circuit requires.
Use a separate regulator to power the relay/s. This time you can use a higher voltage, I suggest somewhere from 15V to 18V for a 12V relay. And again use a current saver circuit, this time the higher supply voltage makes the current saver very effective.
The back emf diode can be a variety of devices.
A single reverse bias signal diode 1n4148/914 does the job well. But it slows down the relay release time a lot.
A combination of signal diode and resistor helps speed up the release slightly. Even faster is a diode + Zener. The Zener voltage should be selected to be near the relay voltage, anything from 9V to 15V should do. ESP has a good webpage detailing this slow/fast release of relays.
use the full 12V to trigger the relay.
look at adding a current saver circuit to reduce the operating current after the relay has successfully triggered.
using 12v for a 12V relay means the current saver is not very effective, but still worth using.
Don't use lower voltage relays, they use too much current.
When you use the high voltage supply, then use a regulator to bring down the voltage to what the amp circuit requires.
Use a separate regulator to power the relay/s. This time you can use a higher voltage, I suggest somewhere from 15V to 18V for a 12V relay. And again use a current saver circuit, this time the higher supply voltage makes the current saver very effective.
The back emf diode can be a variety of devices.
A single reverse bias signal diode 1n4148/914 does the job well. But it slows down the relay release time a lot.
A combination of signal diode and resistor helps speed up the release slightly. Even faster is a diode + Zener. The Zener voltage should be selected to be near the relay voltage, anything from 9V to 15V should do. ESP has a good webpage detailing this slow/fast release of relays.
Last edited:
- Status
- Not open for further replies.