Preamp with LM1036N and TDA7050

[cedric88]

Hi all,

I'm currently building a preamp. The schematic is based on something I've found in an old dusty electronics book in the local library
It uses a LM1036N for vol/bass/treb/bal control and a TDA7050 as a headphone amplifier.

Based on the schematic I have made a PCB with Eagle. Inputs are selected on another pcb and Vsupply (12V and 5V) is also provided by another pcb using 7812 and 7805 regulators. The audio input signals will be connected via solder pens and shielded audio cable.

I've made a DIY amplifier a little while ago with PCB kits, and I've also made already some PCB's (not audio). However, this is the first time I've made an audio PCB and I'm not very sure if the ground plane I've drawn is good enough to avoid hum or noise. I've already read some stuff concerning ground loops.

Could any of you please take a look at the board to see if I haven't made any mistake or screwed some things up ?

Thx

 

[blu_glo]

My tuppence worth:

TDA7050: voltage gain = 27dB according to datasheet. From the headphone amps I've made, I've found such gains to be excessive from line level inputs. Typically I use 6dB and thats it! You do not need much to deafen yoursef even with high-impedance types! Attenuate the input to the headphone stage. Use low value resistors no more than a couple of K-ohms to keep noise to a minimum.

LM1036: Signal level handling - dependant on what you use, this device's max input of 1.6Vrms is a liitle on the low side. Most modern CD players will give out max signal = 2.2V rms. The older (?) LM1035 seems to have higher supply rail capability of 18VDC operating and as such comfortably
takes 3V rms input.

 

[cedric88]

Hi blu_glo, thanks for your reply and your advice !

I've put an attenuator before the potmeter. This should attenuate with -20 dB (please correct me if I'm wrong !)

Apparently I've also made a little mistake with the potmeter, the connections of the left and right channel were reversed

Here is a new improved schematic:
[IMGDEAD]http://www.flickr.com/photos/cedpol/8758965564/sizes/h/in/photostream/[/IMGDEAD]

Considering the signal level handling on the LM1036, would it be a good idea to also put an attenuator before the audio input of the LM1036 ?

Another concern I have is the ground plane. I haven't used a star in the grounding because that would have made the routing a lot more difficult. Is it a good idea to use a ground plane the way I did on a preamp PCB ?

Greetings
Cedric

EDIT: adjustment link to image

 

[blu_glo]

Your 20dB attenuator is about right, leaving you with 7dB of max headphone gain. This means you should at least be able to turn the headphone volume to a sensible position before deafening yourself!

LM1035/36 are interchangeable get your circuit running with the 1036 if you alread have it; then search out a 1035.... I wouldn't bother with the LM103x-attenuator complication, you'll loose signal/noise ratio (noise may get audibly worse)

http://www.datasheetcatalog.com/datasheets_pdf/L/M/1/0/LM1035.shtml

Ground, hereafter referred to as 0V on line preamps is not quite as critical as you may be led to believe.

Input sockets->PCB input 0V [signal wire]
PCB output 0V-> supply 0V [supply wire]
PCB output 0V->ouput RCA 0Vs [signal wire]
Do not worry about using common 0V for left and right; channel seperation is already compromised within the IC and you won't hear the difference anyway.

Connect 0V supply to metal chassis somehere near input with 22-100R resistor and 0.1uF capacitor in parallel both component values non critical!

I'm assuming by "preamp" you do mean such and the power amps are in seperate enclosures elsewhere?

 

[blu_glo]

looking at your PCB if it's not too late I'd put the input L and R input sockets 0V closer together (better still use 1 single wire) and ditto the L and R output ones closer together or same wire. Take power supply 0V input from close to the ouput sockets 0V.

 

[cedric88]

I've made a redesign of my pcb with your comments in mind. Inputs and outputs are now grouped. Output 0V is also close to power supply 0V.

This pcb will be in a separate metal enclosure. Power amplifiers will indeed be in a different metal housing. I will draw the power amps later, these will probably be based on LM3886's.

Concerning the LM1035, I've searched on Farnell, RS Components, Digikey and a couple smaller webstores but I haven't found anything ! Could it be that this LM1035 is out of production ?

Already thanks for your help !

 

[blu_glo]

Could it be that this LM1035 is out of production ?

I beleive it has long since been discintinued however NOS may still be available....

 

[cedric88]

I've found some of the LM1035's on the internet. I will order a couple of these but I will test the board first with the LM1036 since I have these already in my possession.

I will make and test the pcb this weekend, I'll keep you informed !

Regards
Cedric

 

[cedric88]

Allright, so I etched and tested the pcb. At first everything seemed ok, I've tested it with a small power amp board I once bought as a kit (with TDA2050) and the sound is good.

However if I connect a headphone to it I've got a constant "hissss". This "hisss" sound is constant (doesn't change with volume higher). I can hear and manipulate the input sound with the volume pots but the "hisss" remains

Test setup can be seen in attachments.

Anybody got any recommendations what can cause this "hissing" sound ?

Regards
Cedric

 

[blu_glo]

It's most likely the high gain of the headphone amp IC.

Quick questions (on my way to work!)

other than the hiss is the headphone circuit working?
is it quiet hiss in the background?
what position on the volume is loud but comfortable listening?
Is the outside of the POT earthed?
have you tried an earthed metal plate underneath the PCB?

 

[cedric88]

other than the hiss is the headphone circuit working?

Yes, the circuit works as it should
​

is it quiet hiss in the background?

Yep, the hiss remains constant whatever the position of the pot
​

what position on the volume is loud but comfortable listening?

I can open the pot completely without hearing distortion or ruining my ears (sound is not too loud I mean with that)
​

Is the outside of the POT earthed?

The outside of the pot is connected with the ground plane of the pcb. You can see this at the pic of the PCB I've posted earlier. However, my pcb is NOT connected to the ground in the test setup.
​

have you tried an earthed metal plate underneath the PCB?

No, I haven't. Could you explain this a bit more ? Is the earthed metal plate for shielding ?

​

Again, thanks for your help and quick reply

Greetings
Cedric

 

[blu_glo]

other than the hiss is the headphone circuit working?

Yes, the circuit works as it should
Good. So fundamentally it's working!

​

is it quiet hiss in the background?

Yep, the hiss remains constant whatever the position of the pot
So that's going to be noise from the TDA7050 which is a high gain chip. Not necessarily the best option for a headphone amplifier.....

​

what position on the volume is loud but comfortable listening?

I can open the pot completely without hearing distortion or ruining my ears (sound is not too loud I mean with that)
Don't forget the LM1035 volume will also affect the signal and possibly noise passed on to this chip! try with LM1035 VOL pot set to max, then halfway and note differences (if any).

​

Is the outside of the POT earthed?

The outside of the pot is connected with the ground plane of the pcb. You can see this at the pic of the PCB I've posted earlier. However, my pcb is NOT connected to the ground in the test setup.
OK. What value is the POT? Couldn't see it in your schematic. I might suggest 10K and not much more. Also a 100pF from each chip input to earth.

​

have you tried an earthed metal plate underneath the PCB?

No, I haven't. Could you explain this a bit more ? Is the earthed metal plate for shielding ?
Yes, shielding does help a lot especially where high gains are involved.
​

Again, thanks for your help and quick reply
Not a problem glad to be of assistance.

Greetings
Cedric​

​

 

[blu_glo]

Me: - OK. What value is the POT? Couldn't see it in your schematic. I might suggest 10K and not much more. Also a 100pF from each chip input to earth.

Sorry picked it out of the original, I assume you left it 22K which is not bad at all.... use ~100pf caps as previously mentioned on input pins.....

 

[blu_glo]

How did this project go?