I've finally gotten my preamp to a semi-assembled state and am having noise problems. I am using pcb's from Rod Elliot.
I have traced down a rather loud hiss to the balanced output of the preamp. I believe this because when I disconnect power from the SSM2142 balanced output line drivers, the noise goes away. Additionally, when I disconnect signal from the SSM2142's the noise is unchanged, indicating it is not coming from the preamp circuit.
I have used the suggested circuit for the balanced output as presented in the chip's datasheet, utilising 0.1uF ceramic caps as decoupling. The datasheet boasts highly of the chip (link here), thus I do not believe this to be simply the best it can do. This is such a simple circuit, what could go wrong?
The hiss is quite loud and always present, no matter what gain state the preamp circuit is in.
Any help? Where do I start tracing this problem? I am still fairly new to DIY and unfortunately do not possess a scope.
Many thanks; pictures to follow shortly
I have traced down a rather loud hiss to the balanced output of the preamp. I believe this because when I disconnect power from the SSM2142 balanced output line drivers, the noise goes away. Additionally, when I disconnect signal from the SSM2142's the noise is unchanged, indicating it is not coming from the preamp circuit.
I have used the suggested circuit for the balanced output as presented in the chip's datasheet, utilising 0.1uF ceramic caps as decoupling. The datasheet boasts highly of the chip (link here), thus I do not believe this to be simply the best it can do. This is such a simple circuit, what could go wrong?
The hiss is quite loud and always present, no matter what gain state the preamp circuit is in.
Any help? Where do I start tracing this problem? I am still fairly new to DIY and unfortunately do not possess a scope.
Many thanks; pictures to follow shortly
the whole preamp:
the balanced outputs:
schematic from datasheet:
@scott: I have used the circuit above as detailed in the datasheet for the chip... it's very simple!
@Minion: I am using P88 from ESP, details on this web page. As for grounding scheme, I have the power supply (P05a from the same site) grounded to the chassis at one point, and the preamp circuit and the balanced output grounded to the chassis at a separate point. I have played with a few different grounding points and have found no difference.
As for the power supply, details are here , and my big electrolytics are 2200uF (thats what you meant right?)
thanks for the quick replies guys!
An externally hosted image should be here but it was not working when we last tested it.
the balanced outputs:
An externally hosted image should be here but it was not working when we last tested it.
schematic from datasheet:
An externally hosted image should be here but it was not working when we last tested it.
@scott: I have used the circuit above as detailed in the datasheet for the chip... it's very simple!
@Minion: I am using P88 from ESP, details on this web page. As for grounding scheme, I have the power supply (P05a from the same site) grounded to the chassis at one point, and the preamp circuit and the balanced output grounded to the chassis at a separate point. I have played with a few different grounding points and have found no difference.
As for the power supply, details are here , and my big electrolytics are 2200uF (thats what you meant right?)
thanks for the quick replies guys!
You should have all ground wires comeing to a common point and then bolting that to the Chassis not 2 seperate points....
also you are P2P ing the ballanced line Driver stage , because of this you could be picking up interferance because of the Extra long leads and crossed wires, Pluss the extremely long distance of the ballanceing stage from the PSU....
Since the Curcuit is so simple and should work I believe it is probably the fact that your are Point 2 point wireing it....
Have you thought about maybe getting a Piece of vero board or even just makeing a PCB and putting your Ballanced line driver stage on that....?? If you do this I would add a couple bigger Power caps as close to the IC pins as possible in the 10uf to 100uF range.....
If you want I could Quickly design a PCB for you if you wanted to etch it at home or for a couple bucks I could etch it for you....
something like this maybe??
Cheers
also you are P2P ing the ballanced line Driver stage , because of this you could be picking up interferance because of the Extra long leads and crossed wires, Pluss the extremely long distance of the ballanceing stage from the PSU....
Since the Curcuit is so simple and should work I believe it is probably the fact that your are Point 2 point wireing it....
Have you thought about maybe getting a Piece of vero board or even just makeing a PCB and putting your Ballanced line driver stage on that....?? If you do this I would add a couple bigger Power caps as close to the IC pins as possible in the 10uf to 100uF range.....
If you want I could Quickly design a PCB for you if you wanted to etch it at home or for a couple bucks I could etch it for you....
something like this maybe??
Cheers
Attachments
I have tried grounding everything to a common point with no noticeable change in noise floor. I have thought of using a small project board for the output, but I really can't see the very very small amount of wire affecting things much at all, can you really? There is 1" of wire at most extending from each leg of the chips.
Would it help if I got a vero board and put the balanced circuit nearer to the psu? I think I will try that...
Would it help if I got a vero board and put the balanced circuit nearer to the psu? I think I will try that...
There seems to be quite a bit of wire going from the PSU to the SSM Chip , with that much of a distance from the PSU you should try adding a 10Uf cap as close to the IC Pins in Paralell with the 0.1uf ceramic caps and see if that helps with the Noise....
also have you tried it with the Top of your Chassis closed?? this should shield the curcuit from interfernace to some degree....
also have you tried it with the Top of your Chassis closed?? this should shield the curcuit from interfernace to some degree....
Well the 22uF electro's didn't do anything ... I'm trying the vero board now...
by the way Minion I believe you may have pins 1,2 and 7,8 mixed up on your pcb layout. According to the datasheet, pins 1 and 2 are joined together to one phase, and 7 and 8 are joined to the other phase. On your layout, I think you have 1 and 8 together, and 2 and 7 ... ? Did I see that right?
by the way Minion I believe you may have pins 1,2 and 7,8 mixed up on your pcb layout. According to the datasheet, pins 1 and 2 are joined together to one phase, and 7 and 8 are joined to the other phase. On your layout, I think you have 1 and 8 together, and 2 and 7 ... ? Did I see that right?
With electrolytic caps used as PSU bypass caps you allways put the Positive side of the cap at the power rail and the negitive to Ground with + supplies ,and with negitive supplies you put the negitive end of the cap at the Power rail and the Positive to ground....
yes 22uF should work fine....
Cheers
yes 22uF should work fine....
Cheers
first of all nice job on the construction of your preamp! very well done. it looks fly just sitting there with the top off even if it does throw a 'hissy' fit when you turn it on
anyways, i too have found op amps can be tricky with noise problems. here are a few things i have used in the past with some success:
- try and find out for sure if its coming from power supply injection. try hooking the chip up to a different PS if you have one, or even better a pure DC source such as battery(s). if it still hisses on pure DC then you know the source is NOT the PS. also check the snubbing caps parallel to the rectifier diodes ... if one of those caps is defective or leaky it will cause all kinds of noise issues.
- bypass caps on the rails as close to the chip as possible. i know you already tried that but maybe try a few different values for the caps, such as 10uf, 22uf, 47uF etc in parallel with 0.1uF, 0.05uF, 0.01uF etc. i have found that certain combinations of values work better with certain types of noise and with certain chips.
- shield all signal wires as much as possible. use shielded cable or even coax with the outer braid grounded if necessary. check all grounds. try using a bigger wire (lower awg) for your grounding runs to reduce impedance in the ground system.
- when you tried the chip when it was disconnected from the preamp, did you ground the inputs? if not, try it and see it the noise disappears.
- test the unit in a different room of the house. it may sound silly, but i once chased my tail for days trying to get rid of mains hum in a CB circuit only to find out that the noise was coming from the florescent tube ballast transformer right above my head in my workshop.
- is this PA two channel? if so, does it hiss on both channels or just one? if just one try switching the chip from the good channel. that will at least tell you if the problem lies with the chip or not.
- if all else fails, you can quickly, cheaply and easily make a preamp from discrete components by simply implementing just the input and VA stage of your favorite power amplifier.
good luck! i am curious to see what the cause of the hiss finally turns out to be. noise issues can be so frustrating and time consuming to diagnose, but often the cause turns out to be something simple.
anyways, i too have found op amps can be tricky with noise problems. here are a few things i have used in the past with some success:
- try and find out for sure if its coming from power supply injection. try hooking the chip up to a different PS if you have one, or even better a pure DC source such as battery(s). if it still hisses on pure DC then you know the source is NOT the PS. also check the snubbing caps parallel to the rectifier diodes ... if one of those caps is defective or leaky it will cause all kinds of noise issues.
- bypass caps on the rails as close to the chip as possible. i know you already tried that but maybe try a few different values for the caps, such as 10uf, 22uf, 47uF etc in parallel with 0.1uF, 0.05uF, 0.01uF etc. i have found that certain combinations of values work better with certain types of noise and with certain chips.
- shield all signal wires as much as possible. use shielded cable or even coax with the outer braid grounded if necessary. check all grounds. try using a bigger wire (lower awg) for your grounding runs to reduce impedance in the ground system.
- when you tried the chip when it was disconnected from the preamp, did you ground the inputs? if not, try it and see it the noise disappears.
- test the unit in a different room of the house. it may sound silly, but i once chased my tail for days trying to get rid of mains hum in a CB circuit only to find out that the noise was coming from the florescent tube ballast transformer right above my head in my workshop.
- is this PA two channel? if so, does it hiss on both channels or just one? if just one try switching the chip from the good channel. that will at least tell you if the problem lies with the chip or not.
- if all else fails, you can quickly, cheaply and easily make a preamp from discrete components by simply implementing just the input and VA stage of your favorite power amplifier.
good luck! i am curious to see what the cause of the hiss finally turns out to be. noise issues can be so frustrating and time consuming to diagnose, but often the cause turns out to be something simple.
@gain: Thank you so much for your response and your compliments! I am building this preamp to mate with my Dad's new McIntosh 2102 (a beautiful 100W all tube 2ch power amp) and so can not tolerate ANY noise.
I currently have it connected to said amplifier but have bypassed the balanced circuit completely. I know it's not absolutely necessary to have a balanced signal in this application, but it's slightly luxurious. Anyways, it is running completely clean with NO noise at all. I am SURE it is the balanced circuit doing something, but I just don't know what.
I've built a small project board as suggested earlier but had exactly the same results as before. I really don't see what that did beyond shortening the psu run to the circuit.
So I am back to square one. I will look over your suggestions Gain tonight and see what I come up with. Until then, any further help is much appreciated.
I will check back in with any news.
I currently have it connected to said amplifier but have bypassed the balanced circuit completely. I know it's not absolutely necessary to have a balanced signal in this application, but it's slightly luxurious. Anyways, it is running completely clean with NO noise at all. I am SURE it is the balanced circuit doing something, but I just don't know what.
I've built a small project board as suggested earlier but had exactly the same results as before. I really don't see what that did beyond shortening the psu run to the circuit.
So I am back to square one. I will look over your suggestions Gain tonight and see what I come up with. Until then, any further help is much appreciated.
I will check back in with any news.
You know there is a ballanceing method called "Impedance Ballanceing" that simply uses a couple resistors and a capacitor and a couple ferrite beads(optional) this type of ballanceing is used a lot in mic preamps and such...
Look at the output on this mic preamp schematic for Ideas on how to impedance ballance....
http://www.jensen-transformers.com/as/as017.pdf
Cheers
Look at the output on this mic preamp schematic for Ideas on how to impedance ballance....
http://www.jensen-transformers.com/as/as017.pdf
Cheers
yes, Minion is absolutely correct. in fact, i believe you can run a balanced line using nothing but passive components. it is done in RF circuits all of the time, using a device called a balun (BALanced, UNbalanced). basically all it is is a transmission line transformer with the balanced line connected to one coil and the unblananced line connected to the other. one of the most common applications of baluns was in tv's between a balanced 300 ohm 'twin lead' and a 75 or 50 ohm unbalanced coax (the twin lead was a popular antenna feeder for tv's like before we were even born, hence the need to convert it to coax). twin lead its still widely used in amateur radio but thats a subject for another forum!
btw just curious ... how long is the signal cable run between the preamp and the power amp?
i ask because you may not even need to implement balanced lines at all if your run is short (like less than 12-24 feet). in fact, you can actually introduce more noise junk into the signal feeding it through two active chips (balancer and unbalancer, tied to two noisy power supplies) than you would have got from simply just running the unbalanced line level, especially if the run is short.
an obvious exception to this is if you live in an (electrically) noisy area.
i ask because you may not even need to implement balanced lines at all if your run is short (like less than 12-24 feet). in fact, you can actually introduce more noise junk into the signal feeding it through two active chips (balancer and unbalancer, tied to two noisy power supplies) than you would have got from simply just running the unbalanced line level, especially if the run is short.
an obvious exception to this is if you live in an (electrically) noisy area.
after all of this junk I'm thinking it is worth it to scrap the balanced output. I thought it would be easy to do with just one chip and I read good words about the performance of the chip and thought it was worth a shot... less circuitry is better, right?
Thanks everyone for helping so much, I sure love this place!
If what you hear is really hiss, like what you hear between stations on a radio, than few of the things mentioned are likely the cause. If you've bypassed the thing, it shouldn't be PS related unless there's actually something wrong with the PS. A reversed tantalum will do it, but you probably don't have tants- if you do, check 'em. You either have oscillation or RF pickup that's causing something to operate out of it's normal range, or you need to go back to the chip data sheet and calculate the noise level expected given the gain of the amp following the driver chip. I've used the 2142, and it's very good, but I can't remember the noise level. Follow it with enough gain, and you'll hear hiss.
Meaghers, the SSM-2142 never was particularly famous for its noise performance. The spec' sheet quotes it as -109 dB re 10V rms. If we express this as dB's re 1V (o dBu) it is -89 dB. Compare this to the noise of the THAT Corp 1646 which is -101 dBu, making it 12 dB quieter. I think I read that Texas are no longer supplying the 2142 which would be understandable as the 1646 is a better product particularly where behaviour driving unbalanced loads is concerned. Read about it at www.thatcorp.com
Keith
Keith
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.