Dear Sirs and Madans!
I am working on my first preamp project and I have finished designing input selector board. Would you, dear forum members with more knowledge and experience than me, take a look at sch and pcb and tell me your opinion (what is missing, what must be better, which are strong points, etc ...)?
My first opinion is I have to place bandpass filters (20Hz-20kHz) at every analogue input to filter unwanted frequencies, am I right? What else is missing? Input selection process will be handled digitally (by CPU) via CS,A0,A1,A2 pins of analog mux and board will be powered with +/-15V via connector J2 from power supply pcb.
Here is schematics:
and here is pcb:
I am working on my first preamp project and I have finished designing input selector board. Would you, dear forum members with more knowledge and experience than me, take a look at sch and pcb and tell me your opinion (what is missing, what must be better, which are strong points, etc ...)?
My first opinion is I have to place bandpass filters (20Hz-20kHz) at every analogue input to filter unwanted frequencies, am I right? What else is missing? Input selection process will be handled digitally (by CPU) via CS,A0,A1,A2 pins of analog mux and board will be powered with +/-15V via connector J2 from power supply pcb.
Here is schematics:
An externally hosted image should be here but it was not working when we last tested it.
and here is pcb:
An externally hosted image should be here but it was not working when we last tested it.
Hi Marko,
I have a question, have you checked if the MPC508 multiplexers are suitability for audio signals? As I can see the datasheet does not even mention the THD+N and the crosstalk figure looks not so spectacular too. It also looks like the part is very sensitive to the input and load characteristics.
Regards,
Oleg
I have a question, have you checked if the MPC508 multiplexers are suitability for audio signals? As I can see the datasheet does not even mention the THD+N and the crosstalk figure looks not so spectacular too. It also looks like the part is very sensitive to the input and load characteristics.
Regards,
Oleg
Hi, Oleg!
You exposed good point, thank you, I've sent email to Texas Instruments, waiting for answer, are there any other design flaws (possible or existing)? And which multiplexer IC would you choose?
Sincerely,
Oleg
your input filtering can be either passive or active.
I recommend passive single pole filtering. It's simple and cheap and works.
But you would need to set the F-3dB turn over frequencies to roughly a decade beyond the wanted passing frequencies, i.e. 2Hz to 200kHz.
If you go to active filtering you can increase the poles/slopes of the filters and that would allow F-3dB frequencies to be much closer to your wanted pass band.
Ok, will redesign the board so it will be using relays, however which relays to use, I need relay with two contacts for audio signal (left, right) and working contact on +12V, because I will drive them using ULN2003 IC, can someone give me a hint which relay to use, becase there are like zillion different relays on the market?
Sincerely,
Marko
Ok, I've redesigned the input selector board, here is schematic:
An externally hosted image should be here but it was not working when we last tested it.
and here is pcb:
An externally hosted image should be here but it was not working when we last tested it.
Now, by yours opinion, dear forum members, should I put bandpass filters (10Hz-30kHz) on every input or not? And is PCB better?
Sincerely,
Marko
I did not fully inspect the design but some things I already noticed.
1. Separate the audio ground and control ground. Make a provision to connect them together at some point should there be a need.
2. Do not short unused ULN2003 outs to GND or you will fry the chip since the default state of the outs is HIGH. Just let them floating.
3. Move ULN2003 closer to the relays they control or add back EMF suppression diodes across relay coils.
4. Don't use individual traces to the output connectors from each input relay. Make a bus for each output and connect relay outs to the bus.
5. Your mount holes look like 1mm diameter. If you really need them then make them bigger, let say 3.3mm dia.
6. Make provision for RC low pass filters for all inputs.
Regards,
Oleg
1. Separate the audio ground and control ground. Make a provision to connect them together at some point should there be a need.
2. Do not short unused ULN2003 outs to GND or you will fry the chip since the default state of the outs is HIGH. Just let them floating.
3. Move ULN2003 closer to the relays they control or add back EMF suppression diodes across relay coils.
4. Don't use individual traces to the output connectors from each input relay. Make a bus for each output and connect relay outs to the bus.
5. Your mount holes look like 1mm diameter. If you really need them then make them bigger, let say 3.3mm dia.
6. Make provision for RC low pass filters for all inputs.
Regards,
Oleg
1) Audio ground (AGND) and Control Ground (CGND) are now separated. However, do I need to connect them together in some point, for instance, with jumper?
2) ULN2003 outputs now left floating.
3) Ok, however, as far as I understand Texas Instruments ULN2003A datasheet, especialiy its chapter 8.4:
this has already been achieved since my COMMON pin on obath ULN2003A is connected to +12V power supply pcb polygon or am I wrong? And I've also moved ULN2003s closer to relays.
4) I've merged outputs into bus.
5) TODO
6) TODO
Here is upgraded pcb:
An externally hosted image should be here but it was not working when we last tested it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.