Upgrading a channel strip

[godzofguitar]

Hi all !
today is my first post in the group and I’m looking for some feedback on a mic input circuit design for upgrades or some basic clean up to drop the noise floor any help is much appreciated…

 

[JohnRoberts]

You could brute force it some with lower noise components but that preamp input stage is running open loop so will have higher distortion than preamps with negative feedback returned to the input device emitters. Most typical preamps require one or two more op amps to do that.

JR

 

[godzofguitar]

Thank you John for the input on the circuit design proposed, any suggestions for specific low noise devices based on the circuit? it is using the old work horse ne5532 except for the channel led clip indicator that is a rc4558, believe it or not it is pretty stable with a decent sound it is an aggressive forward sounding channel strip, I have a very high quality power supply powering the board thinking that could be helping the board a good bit with the noise floor although it could use some harmonic love, as far as negative feedback closed loop network it would take some rework to make that happen might have to make some circuit boards for that for a physically stable and professional build, thanks again for your time and input.

 

[JohnRoberts]

The mic preamp noise with that topology is dominated by the input bipolar transistors (PN4355). I am not familiar with those parts and expect there are lower noise modern PNPs. Back in the day I used 2sb737s extremely low noise PNPs but they went obsolete around the turn of the century. To optimize the mic pre for different devices may involve several tweaks, and you still lack the benefit of wrapping NF around those input devices to reduce distortion.

Using a 4558 for the clip indicator is ok, it is not in the audio path.

JR

 

[ruffrecords]

Have you actually measured the noise floor? Many of these designs are within a small number of dB of the theoretical maximum so there are no big gains in noise performance to be had.

Cheers

Ian

 

[gyraf]

I don't see anything here that would account for unneeded noise - as JR and Ian mentions, your improvement would most probably be maximum 1-3dB lower noise, and this only with genuine low-impedance microphones, say 50 Ohms or 150 Ohms perhaps.

That said, I find it worrying if 5532's are used in the EQ circuits pictured there: The potentiometers are not isolated from amplifier input-bias-current so these will go unnecessarily scratchy over time. For these positions, A3-4-5-6, I would recommend using low-input-current opamps i.e. JFET types like TL072.. Noise contribution would be minimal, that part is dominated entirely by the input transistor pair..

/Jakob E.

 

[Matt Syson]

Getting a good EIN figure for medium gain is pretty difficult and in days gone by only EIN at MAX gain was specified thus overlooking the gain settings that are used in practice. as it is a mixer are you sure you are measuring the most important aspects as presumably you want it as noise free as possible when recording ONE channel at a time. As Ian poiunted out the noise figures obtained were rarely more than a few dB different unless a piece of gear was very badly designed.

 

[Newmarket]

potentiometers are not isolated from amplifier input-bias-current so these will go unnecessarily scratchy over time. For these positions, A3-4-5-6, I would recommend using low-input-current opamps i.e. JFET types like TL072.. Noise contribution would be minimal, that part is dominated entirely by the input transistor pair..

/Jakob E.

Just a note that if the OP wants optimum fidelity an low noise then there are now better options than TL07x opamps. Although there are well regarded desks full of them so not a big deal in my book. Also OP wants some "harmonic love" so not sure where modifications go with that. But to note that most preamps don't really show their "character" until pushed.

 

[Newmarket]

, believe it or not it is pretty stable with a decent sound it is an aggressive forward sounding channel strip, I have a very high quality power supply powering the board thinking that could be helping the board a good bit with the noise floor although it could use some harmonic love, as far as negative feedback closed loop network it would take some rework to make that happen might have to make some circuit boards for that for a physically stable and professional build,

Not understanding the "believe it or not" tbh. Unless I'm misunderstanding what you mean by stability.
What desk / kit is it ?

 

[godzofguitar]

Thank you all for the comments I really do appreciate you guys taking the time to comment on my post, the console use to belong to Greg Allman when he was doing the adverts for Carvin it’s the MX2488 came out of a rehearsal spot, I’m doing a complete rebuild on it and thought I would do the most I can to bring it up to a more respectable sounding unit while I’m digging in and recapping the thing, I realize I’m attempting to polish what some consider a turd but it’s really just to have some fun and see what can be done to bring it up to another level…

 

[godzofguitar]

Not understanding the "believe it or not" tbh. Unless I'm misunderstanding what you mean by stability.
What desk / kit is it ?

Referring to the pre amp stage running open loop the console is a Carvin mx2488.

Not understanding the "believe it or not" tbh. Unless I'm misunderstanding what you mean by stability.
What desk / kit is it ?

 

[godzofguitar]

I don't see anything here that would account for unneeded noise - as JR and Ian mentions, your improvement would most probably be maximum 1-3dB lower noise, and this only with genuine low-impedance microphones, say 50 Ohms or 150 Ohms perhaps.

That said, I find it worrying if 5532's are used in the EQ circuits pictured there: The potentiometers are not isolated from amplifier input-bias-current so these will go unnecessarily scratchy over time. For these positions, A3-4-5-6, I would recommend using low-input-current opamps i.e. JFET types like TL072.. Noise contribution would be minimal, that part is dominated entirely by the input transistor pair..

/Jakob E.

Yeah I think they must’ve been getting a crazy deal on them so instead of adding another parts purchasing column to the list they just used 5532’s every where, I have bunch TLE2072’s on hand pretty fast ic compared to the 5532 so not sure if that will work or not without some rework, but I agree with your comment about the 5532 in the EQ circuit thanks for your input I really appreciate it…

 

[godzofguitar]

Getting a good EIN figure for medium gain is pretty difficult and in days gone by only EIN at MAX gain was specified thus overlooking the gain settings that are used in practice. as it is a mixer are you sure you are measuring the most important aspects as presumably you want it as noise free as possible when recording ONE channel at a time. As Ian poiunted out the noise figures obtained were rarely more than a few dB different unless a piece of gear was very badly designed.

I’ll be using the console for tracking and mixing drums and guitars and basic back up vocals along with keyboards, unless I need something esoteric for a delicate vocal or some string arrangements with quite passages I already have that covered so prolly will not use the console for that anyway, thank you for the comment I really appreciate it.

 

[godzofguitar]

Have you actually measured the noise floor? Many of these designs are within a small number of dB of the theoretical maximum so there are no big gains in noise performance to be had.

Cheers

Ian

No I haven’t taken any measurements as of yet, I’m building a test bed for the channel strips to do just that so I can try more modern transistors and opals to optimize the performance of circuit, thanks for the question I appreciate that.

 

[godzofguitar]

The mic preamp noise with that topology is dominated by the input bipolar transistors (PN4355). I am not familiar with those parts and expect there are lower noise modern PNPs. Back in the day I used 2sb737s extremely low noise PNPs but they went obsolete around the turn of the century. To optimize the mic pre for different devices may involve several tweaks, and you still lack the benefit of wrapping NF around those input devices to reduce distortion.

Using a 4558 for the clip indicator is ok, it is not in the audio path.

JR

Thank you John! I found some ROHM 2sb737 I haven’t looked at the specs just yet to see how far off from the 4355 they are pretty sure they won’t just be drop in ready, the console is 48 channels and they have just enough to do the console, I just need to work out the details for the NF network for the transistors…thanks again for your help!