Preamp noisy - 100K pot?Relays?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi there. I've built a pre using Rod Elliott power supplies - all good. I'm using a pcb from Ebay to switch channels - it has 4 relays and a 100K volume pot.This is powered from the aux output of the supply (which is unregulated). Problem is it's noisy at low levels and on full - fairly quiet in the middle range - where am i going wrong?:(
 
Administrator
Joined 2004
Paid Member
Hi Tex,
Two things.
1. Can you post a schematic and take a picture of them to post also?
2. **** Power supply ****

Your power supply is very important from a noise standpoint. It must be quiet electrically. A "wall wart" unregulated supply is the worst thing I can imagine to run a preamp off. Maybe a bicycle generator with rectifiers might be worse.

Your grounding is important of course. Did you ground the case of the volume control (this would normally make it noisy in mid travel).

-Chris
 
Layout

Just learning.....
 

Attachments

  • pre 2.jpg
    pre 2.jpg
    27.1 KB · Views: 504
Hi Tex
Buzz caused usually from power supply. Hiss caused usually from the filters builded around operational amplifiers or from the operational amplifiers alone. As i understand from that you wrote, the ground of power supply it is also connected in chassis through the volume pot. If it is so, try to seperate the ground of power supply from the chassis because this is electrically earthed to the mains electric ground through the mains IEC socket. As it appears in the photo the IEC socket it is filtered type. Try to feed the power supply directly from the mains by bypassing the IEC socket. The most important it is the seperation of mains earth from the signal ground which is connected to the power supply ground rail. The type of the operational amplifiers that you use it is good (NE5532). The input sellection relays when activated, maybe cause a little noise due to electromagnetic field generated from the coils that includes. I preffer a passive input selection rotary switch, because it is a little difficult to find out a noisyless type of relays. The impedance of inputs is very high (a normal case) and they are very sensitive to any electromagnetic field source. Be carefull with the inputs RCA sockets; their metal sleeve may don't touch in the chassis because this cause again connection between the electric earth and the power supply ground. Or in the worst case a ground loop between inputs and outputs. Generally you may check first that no one part cause a connection between the electric earth and the power supply ground. With the aid of a ohm meter (connect one of the test leads in the electric earth of the IEC input socket and the other in the power supply ground permanently) check anything that can cause a short between the chassis and the power supply ground. Make this first of all and if the problem remains inform us for further assistance. Good luck.
Fotios
 
Improving.....

Thanks very much for that info. I have tried a straight through IEC and it made very little difference. I then got rid of the whole relay switching PCB and just went straight into a 20k pot I had lying around. This cured the whole varying hum and hiss issue, but I still have a slight residual noise at all volumes - not bad, but enough to be annoying at very low levels. I am using DIP8 connecting blocks for all the IC's and i read somewhere that this can cause problems - I just wanted to be able to try different IC's. Any other theories?
Thanks again!
 
Administrator
Joined 2004
Paid Member
Hi Tex,
If you had a 'scope and THD meter you could get more information on this. I don't like sockets for audio applications. With the extra capacitance and connections, your op amps may be a little unstable. Your decoupling on the supply pins is not nearly as effective as it is with a soldered IC.

-Chris
 
very close now....

I found the main fault was that i was using two power supplies - one for each side of the pre. When I just paralleled the single supply to both sides most issues disappeared. I have re-instated the relay switch and vol and my only issue now is a hiss a low levels - this pretty much disappears with the pot cranked up - problem is I listen at low levels quite often, so I would like to sort this. The pre is designed for the AD797 at the output stage and i am currently using NE5534's. Should this be an issue? Should I change the value of the volume pot to 50k or 20k? Any more ideas welcomed thanks!
 
Hi again Tex
I don't see a response that you maked a check in the RCA in/out sockets as i proposed in my previous reply. Are the sleeves of RCA insulated from the metal plate of chassis? Please make it. It is mandatory to not touch on the case of your preamp. Some little details make the big difference many times. Don't be confussed with the electronic part of the whole construction. This is well checked (usually) from the seller of the kit. The operational amplifiers are very good and characterized from their very low noise figure. The DIL sockets - according to my opinion - they are not bad; but you can make an improvement by changing these from DIL8 sockets with turned pins which have usually gold plated holes. As for the schematic, i don't understand the existence of these 20K trimers in the input pin of each IC. Evidently the designer of circuit make an effort to eliminate a possible d.c. component from entering in ICs. Try to remove the 20K trimmers and observe if there is any improvement in the noise. Also, as i see in the schematic, the output of preamp it is balanced but you use it as single ended (only the positive out node) as i can see in the photo from the existence of RCA output sockets. A possible improvement it is that you can remove the inverting part of output, simply by dessoldering R128 and extracting U103 from its socket. The thought it is verry simple; the less the number of ICs, the less the noise added on the total noise of circuit. I have to make one more remark; the distance of preamplifier pcb it is some big from the input sockets. Make an effort to reduce the length of cables between the inputs and the pcb as more as it is possible. Change the place of pcbs near the in/out sockets and move the pcb with relays and volume pot more front. You have enough space in your case to do this. Change the IEC mains inlet with an unfiltered and ensure a good contact of your metal case with the electric earth; but avoid the contact of power supply ground rail with the case. Check it with an ohm meter.
In pending of results
 
Re: very close now....

Tex Houston said:
I found the main fault was that i was using two power supplies - one for each side of the pre. When I just paralleled the single supply to both sides most issues disappeared. I have re-instated the relay switch and vol and my only issue now is a hiss a low levels - this pretty much disappears with the pot cranked up - problem is I listen at low levels quite often, so I would like to sort this. The pre is designed for the AD797 at the output stage and i am currently using NE5534's. Should this be an issue? Should I change the value of the volume pot to 50k or 20k? Any more ideas welcomed thanks!

Tex
The use of individual supply for each channel it is a very good thing. You can use the two supplies, but the ground node of each one may DO NOT TOUCH with the other! From these of you write, i understand that you have a nice ground loop between the two supplies and thus between the two preamp modules! Check again the sleeves of in/out RCA. This is your major problem. Keep the ground nodes of the two channels seperatelly from input to output. I don't know if you can understand that i say. Maybe Chris (Anatech) can to explain you with better words because my english are not so good. As for the type of operational amplifiers, indeed the AD797 it is better from NE5534. But the change of NE5534 presupposes also the change of filters value. I propose to leave you the operational amplifiers as is. The profit it is negligible in relation with your major problem with the cabling, the interconections and thus the ground loops.
 
Thanks Fotios

Thanks again for your help. I have checked all RCA connections and they have no connection to chassis ground. There is however a connection between the shield of all inputs and the power supply ground - I figure this is happening at the relay board where the shields are tied at each stereo input (3 connection points) and the output. The 20k trimmers are to adjust DC offset and I have this down to zero or less than 1mv. Thanks for the tip with the balanced outs - I don't plan on using them, so I may as well lose that part of the circuit.
As I said the whole thing is pretty quiet now apart from the slightly annoying hiss at low volume - what about the pot value question?
Is it worth perservering with the dual supplies? i would need to lose the relay switcher right?
Have changed IEC connector thanks!
 
Re: Thanks Fotios

Tex Houston said:
Thanks again for your help. I have checked all RCA connections and they have no connection to chassis ground. There is however a connection between the shield of all inputs and the power supply ground - I figure this is happening at the relay board where the shields are tied at each stereo input (3 connection points) and the output. The 20k trimmers are to adjust DC offset and I have this down to zero or less than 1mv. Thanks for the tip with the balanced outs - I don't plan on using them, so I may as well lose that part of the circuit.
As I said the whole thing is pretty quiet now apart from the slightly annoying hiss at low volume - what about the pot value question?
Is it worth perservering with the dual supplies? i would need to lose the relay switcher right?
Have changed IEC connector thanks!

Why not 2 power supplies? The only you have to make it is the absolute seperation of the 2 grounds. That is all. Check it with an Ohm meter. I thing the contact of the 2 supplies ground it take place possibly in the relay board as i understand from that you wrote. From the other hand, don't worry so much with the use of one supply for the two channels because the power consumption it is small. I think one supply it is in position to sustain the two channels simultaneously. Tell me something - because it is not appears in the schematic pdf - the relays are placed after the input buffers? And the volume pot it is placed after the output of U105? If it is so, the value of 100KÙ it is proper because the small loading of U105 (AD797) output it is indispensable. The smallest the value of potentiometer the heaviest the loading of output of U105 thus more thermal noise. Now for the hiss heared, this is caused probably from the filters around the input buffers and not from the ICs itself. This is the famous Johnson or Nyquist or thermal noise. The use of metal film resistors as you know solve this problem by some way. The rest it is the tolerance of value of capacitors which form the filters around the operational amplifiers or before these and the inputs. If you hear buzz, this is owed to the mains transformer. I know verry well this; because we are using toroidal transformers we are thinking that the strong electromagnetic field it is eliminated; but no one thinked that if the transformer core (usually formed from cylindrical turned plate of soft iron) it is not tighten causes a constant buzz of 50Hz (in Greece so much is the frequency of mains electric). You can hear the noise of your transformer with this trick: place the tip of a long screwdriver on the metal cover of the transformer and place your ear on the handle; by pressing your ear a little on the handle you can hear the buzz from the core nicely reproduced from your speakers. It is an enervating thing! To solve this, when i order a toroid transformer from the workshop i pay extra euro for dipping the core into epoxy resin first, and after the winding and before the insulating with the mylar tape again a dipping in epoxy resin. I wait for your response
Fotios
 
Thank you again - I may now stick with the single supply as i can use the other one on a new project. The relays are after the input buffers yes but the volume pot is on the board directly after the relays and before the U105 IC - will this make a difference to the ideal pot value? There seems to be no noise directly after the input buffers, so I guess it's happening after that.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.