Mic Preamp. Adding things and getting ideas. Would love help please.
- Thread starter Leek
- Start date
Help Support GroupDIY Audio Forum:
Ian, I am having some trouble wrapping my mind around connecting a few things. I was wondering, would it be ok to connect R11 of the line receiver to the - input of the preamp stage and then the - output of the preamp stage to R2 of the line driver? So that I have the line receiver feeding both inputs of the mic preamp and both outputs of the preamp feeding the single input of the line driver? I have provided a picture of my schematic and my small build. Perhaps you can help a little more!
https://s12.postimg.org/gb64e1zbh/Mic_Preamp.png
https://s12.postimg.org/gb64e1zbh/Mic_Preamp.png
metalb00b00
Well-known member
- Joined
- Jul 8, 2012
- Messages
- 294
The correct connection is:
Balanced receiver output to mic input +, and mic input - to ground.
Mic output + to balanced driver input, and mic output - to ground.
Balanced receiver output to mic input +, and mic input - to ground.
Mic output + to balanced driver input, and mic output - to ground.
ruffrecords
Well-known member
Leek said:
Basically yes. You mic pre at the moment is unbalanced in and unbalanced out. For testing you should connect pin1 and pin3 of the input XLR to 0V and pin2 to the mic pre input. The output of the mic pre goes to the tip of the TRS and the ring and screen should be connected to 0V.
Cheers
Ian
OK. Unlike John Roberts, I've succumbed to the Dark Side. Sometimes, the temptation to do something really dirty is irresistible. 
In da old days, this type of design would have been done with a Radford LDO & Distortion Analyser. This is just an exercise in LTspice .. which you can download.
It's nearly early 1970s SOTA ... but not quite. That would require an extra Emitter Follower. But if you are making something like this, why try to make it sound as clean as a 1980s 5534 design? 8)
I was first inspired to do this stuff when Angus McKenzie published a Studio Sound article on improving a Revox A77. Its mike preamps weren't its best feature.
It's loosely based on LNprimer.doc in my Yahoo MicBuilders Files directory which IS SOTA for what it is supposed to do. You have to join.
Read it to understand some of the Low Noise design process ... and "The Art of Electronics".
In da old days, this type of design would have been done with a Radford LDO & Distortion Analyser. This is just an exercise in LTspice .. which you can download.
It's nearly early 1970s SOTA ... but not quite. That would require an extra Emitter Follower. But if you are making something like this, why try to make it sound as clean as a 1980s 5534 design? 8)
I was first inspired to do this stuff when Angus McKenzie published a Studio Sound article on improving a Revox A77. Its mike preamps weren't its best feature.
It's loosely based on LNprimer.doc in my Yahoo MicBuilders Files directory which IS SOTA for what it is supposed to do. You have to join.
Read it to understand some of the Low Noise design process ... and "The Art of Electronics".
Attachments
If you are just feeding the Line I/P of your Digi 002 (simulated by R9), you can use somewhat less than 30V for the supply. Adjust R2 for symmetrical clipping if you do this. Presently, it will clip just before 16dBV. The bias chain R1 & R2 is decoupled by C5 to get rid of noise from the power supply line before it reaches the sensitive input.
It’s unclear what input is required for FS on the Digi 002 line inputs. The most relevant number seems to be 18dBu which is 6.15V or 15.8dBV This needs the 30V supply
Gain is set by switching R6 to give 4 gains between 42dB(sorta) & 12.1dB. You could use a Reverse Log pot but that would be yucky and unacceptable to the BBC
R10, 11 & C4 implement an Impedance Balance O/P stage. Output resistance, Ro, will vary with the gain setting so R10 might need tweaking for a good CMRR compromise.
If the preamp is next to your Digi 002, you can probably use this as drawn. If it is more than a couple of metres, replace R10, 11 & C4 with your Crimson 4000B 1:1 transformer ... giving better balance and adding its distortion to the vintage THD profile characteristic of the 2 transistor amp.
http://www.crimsonaudiotransformers.com//FileStream.aspx?FileID=2
The Crimson spec looks good BUT there is no design impedance or winding resistance specified. Both frequency response & distortion are VERY dependent on this. If it was designed for a low impedance, using a higher impedance at both input & output will degrade LF response & THD. I’m going to assume it is designed for 600R or less on at least one side. Our worst case is Ro = 188 (when R6 = 82) so we should get better than the specified +/- 0.5dB frequency response 20Hz – 20kHz provided the DC resistance of the windings is small enough not to take the total ‘source’ resistance above the design resistance.
'Professional' gear would have a Sowter microphone input transformer and make this 'acceptable' to the BBC in 1970. I’ve shown Lundahl & Jensen equivalents. They add another 14dB gain and this would actually have very good noise performance even today. Rest of document doesn’t include input transformer gain/THD
These transformers are 1:5 so their impedance transformation will be (1:5)^2 = 25. A 150R dynamic mike will present 3K8 to the amplifier (simulated by R12) .. which gives near optimum noise performance for the input transistor Q1, BC548C running at 94uA.
The load seen by the mike is R7 transformed by the transformer; 47k / 25 = 1k88
If your dynamic mike is within a couple of metres, you could probably feed it unbalanced into the preamp without the expensive input transformer. But P48V would require it … noise would be better … and it gives 14dB of clean gain. See the Jensen application notes for details.
It’s unclear what input is required for FS on the Digi 002 line inputs. The most relevant number seems to be 18dBu which is 6.15V or 15.8dBV This needs the 30V supply
Gain is set by switching R6 to give 4 gains between 42dB(sorta) & 12.1dB. You could use a Reverse Log pot but that would be yucky and unacceptable to the BBC
R10, 11 & C4 implement an Impedance Balance O/P stage. Output resistance, Ro, will vary with the gain setting so R10 might need tweaking for a good CMRR compromise.
If the preamp is next to your Digi 002, you can probably use this as drawn. If it is more than a couple of metres, replace R10, 11 & C4 with your Crimson 4000B 1:1 transformer ... giving better balance and adding its distortion to the vintage THD profile characteristic of the 2 transistor amp.
http://www.crimsonaudiotransformers.com//FileStream.aspx?FileID=2
The Crimson spec looks good BUT there is no design impedance or winding resistance specified. Both frequency response & distortion are VERY dependent on this. If it was designed for a low impedance, using a higher impedance at both input & output will degrade LF response & THD. I’m going to assume it is designed for 600R or less on at least one side. Our worst case is Ro = 188 (when R6 = 82) so we should get better than the specified +/- 0.5dB frequency response 20Hz – 20kHz provided the DC resistance of the windings is small enough not to take the total ‘source’ resistance above the design resistance.
'Professional' gear would have a Sowter microphone input transformer and make this 'acceptable' to the BBC in 1970. I’ve shown Lundahl & Jensen equivalents. They add another 14dB gain and this would actually have very good noise performance even today. Rest of document doesn’t include input transformer gain/THD
These transformers are 1:5 so their impedance transformation will be (1:5)^2 = 25. A 150R dynamic mike will present 3K8 to the amplifier (simulated by R12) .. which gives near optimum noise performance for the input transistor Q1, BC548C running at 94uA.
The load seen by the mike is R7 transformed by the transformer; 47k / 25 = 1k88
If your dynamic mike is within a couple of metres, you could probably feed it unbalanced into the preamp without the expensive input transformer. But P48V would require it … noise would be better … and it gives 14dB of clean gain. See the Jensen application notes for details.
Hey Ricardo, I've actually changed things and went with an IC front end followed by a transistor emitter follower output stage. It seemed a lot simpler and I've been debugging it. Here is the schematic. I have also provided two different BC 547 emitter follower output stages. Though, I'm leaning towards the first/top one.
https://postimg.org/image/9h3b2wznt/
https://postimg.org/image/9h3b2wznt/
The only sensible reason for building something like this today is that the vintage circuit provides some euphonic (nice sounding) distortion.
Worst distortion case with gain set to 42dB. THD is 1% at 5V output and mainly 2nd (h2). Clips at about 5.7V But the onset of clipping is quite soft. You could push it a bit harder .. it would distort and still not sound too harsh.
This depends on the music. Good small unaccompanied choirs or piano need really clean preamps with huge headroom not to sound nasty. Other stuff with big dynamic range, like drums, actually don’t sound nasty or even different if clipped.
Input frequency is 1kHz. THD doesn’t change much with frequency at any level below clipping.
Worst distortion case with gain set to 42dB. THD is 1% at 5V output and mainly 2nd (h2). Clips at about 5.7V But the onset of clipping is quite soft. You could push it a bit harder .. it would distort and still not sound too harsh.
This depends on the music. Good small unaccompanied choirs or piano need really clean preamps with huge headroom not to sound nasty. Other stuff with big dynamic range, like drums, actually don’t sound nasty or even different if clipped.
Input frequency is 1kHz. THD doesn’t change much with frequency at any level below clipping.
Attachments
What PRR was trying to tell you in the other thread is that all the 5532 circuits you have posted have terrible noise and would be unacceptable as mike preamps.Leek said:
If you want to use an IC, THAT 1510/12 are probably best. They have excellent datasheets & application notes. If you build them correctly to their spec, you would have something that sounded like your Digi 002 preamps.
What I've posted is something a top 1970's mixer designer at eg N**e would have done. If you use the input transformers, it's noise performance would be difficult to fault. It will have the 'sound' of the famous name mixing desks of that period.
So, what you're telling me is this preamp build would be ok for things like drums, electric guitars and louder sources where distortion and things like would be ok? I'm not looking for a huge super clean preamp. I'm just looking to build a working preamp. Upon learning how to get this circuit going, I can add Jensen/Lundahl transformers later to increase headroom and noise floor. Upon looking over my schematic ricardo, how does it look? Will it work?
Something like my circuit would have been used for everything by the top studios in the early 1970s.Leek said:
If you are recording certain types of high dynamic range stuff like drums, it would still do an excellent job. For unaccompanied choir and piano, good 'modern' IC stuff would give better results. You need to distinguish between stuff with high dynamic range (soft bits & loud bits) and stuff which is just loud. The former is MUCH more difficult to record.
If you want a working preamp, choose one of the simpler circuits like ESP or the THAT datasheets and build it EXACTLY as shown.
If you want to understand how to design good LN amps, you will learn more building my circuit and working out why it is better than your earlier efforts.
I've recommended you build it exactly as shown and add transformers later. I've also told you what you gain when you add the output Crimson and the input Sowter.
To get the full advantage, you need to do your homework ... reading "The Art of Electronics" or my LNprimer from Yahoo MicBuilders.
Ricardo, with my schematic that I've provided and Ian helped me add a pot and unbalanced output too, could I not just add a 10468 style input and 11148 output transformer to balance the INS and OUTS? It seems redundant to now throw my components away and just buy new passives to build your styled preamp.. Ian said it has plenty of gain so distortion (noise) wouldn't be a problem..
If you'll look at it, its a BC547 input, feeding a BC557 that then feeds another BC547 output stage. I could use the 10468 style transformer at input to balance the signal into the BC547. I could then use the last BC547 in a emitter follower stage to feed the 11148 output transformer, correct? Is the 2N3055 more suitable for the output stage with the 11148 style transformer?
If you'll look at it, its a BC547 input, feeding a BC557 that then feeds another BC547 output stage. I could use the 10468 style transformer at input to balance the signal into the BC547. I could then use the last BC547 in a emitter follower stage to feed the 11148 output transformer, correct? Is the 2N3055 more suitable for the output stage with the 11148 style transformer?
How about you post links to 10468 & 11148? You make it difficult for people to understand you .. poor pics on some other website etc. ... the less likely you will get good answersLeek said:
New passives about $0.02 / each for the resistors.
Why don't you build it as "Ian suggested" (I don't think you understand Ian fully) and then build my version.
The most expensive part is the 30V supply but this is needed to drive your Digi 002 to the spec I show. I'm not sure Ian has worked out if your version can drive the Digi 002.
You haven't shown anything for us to look at.
Ricardo, the transformers I have are located at (http://m.ebay.com/itm/201486707839)
They are the same transformers used in N*** clones. I'm pretty sure with the transformers,transistors and etc I can build something acceptable? I've ordered a few OPA134, THAT 1646 line driver and more passives based on each data sheet. I then got a 16v-0v-16v transformer to use with some 1N4002 diodes to build a bipolar power supply.
Between the input transformer, transistors, preamp IC's, line driver or output transformer I should be able to get something going..
I'm certain a transformer balanced input, OPA134 preamp stage and THAT 1646 output stage would be easier to build then my guess work?
They are the same transformers used in N*** clones. I'm pretty sure with the transformers,transistors and etc I can build something acceptable? I've ordered a few OPA134, THAT 1646 line driver and more passives based on each data sheet. I then got a 16v-0v-16v transformer to use with some 1N4002 diodes to build a bipolar power supply.
Between the input transformer, transistors, preamp IC's, line driver or output transformer I should be able to get something going..
I'm certain a transformer balanced input, OPA134 preamp stage and THAT 1646 output stage would be easier to build then my guess work?
If you build this right, you'll have something that sounds just like .. and measure nearly as good .. as your Digi 002 mike preampsLeek said:
Even easier and probably better all round is to use THAT 1510/12 and an impedance balanced output stage.
http://sound.westhost.com/articles/balanced-2.htm
This is an excellent summary of Bill Whitlocks encyclopediac articles. Originals are on the Jensen website.
Concise & accurate argument for this El Cheapo Balanced O/P
ricardo said:If you build this right, you'll have something that sounds just like .. and measure nearly as good .. as your Digi 002 mike preamps
Even easier and probably better all round is to use THAT 1510/12 and an impedance balanced output stage.
http://sound.westhost.com/articles/balanced-2.htm
This is an excellent summary of Bill Whitlocks encyclopediac articles. Originals are on the Jensen website.Concise & accurate argument for this El Cheapo Balanced O/P
I'll see if I can't get a schematic together. But, I'm a little let down upon building something that "sounds nearly as good as my 002 preamps". I'm hoping to get asomething that has more color and is different from the "cleanish" preamps in my 002. That being the case, I'd be better off going with transistor based preamp stage and a transformer balanced output stage, no? If the input transformer has the least effect on the signal using an IC like the THAT1250 balanced line receiver seems a more practical approach. I do have a few different op amps, so if I could use those instead of buying a THAT 1512 it'd be great. They include OPA134, NE5532, LME49710, Analog Design OP27 and TL075's. I do have a build based around the THAT1250 balanced input stage, Transistors and LME49710 preamp stage and the output transformer that I recently bought. But, I'm still trying to decide what to do with everything.
I have
14 TL075
2 LME49710
2 OPA134
2 NE5532
1 AD OP27
2 BC547
3 PN2907
1 BC557
1 (N*** styled input transformer)
1 (N***) styled output transformer)
1 THAT 1250 line receiver
1 THAT 1646 line driver
and about 75$ I can spend to get something "fat and colorful" going. I'm against "clean" as the pre's in my 002 are pretty clean so something with a good mid range, color which I can get from the n*** style output transformer. Which adds color and is known to be one of the biggest parts of the n*** sound. The only issue is trying to figure out what would be the best preamp stage that can drive the output transformer. I'm leaning towards making the front end and preamp stage all IC based for the simplicity and then have the output transformer provide the color. Though, IC's like the OPA134 and etc are known for the BB sound which is sort of laid back and has a good "bitey" mid-range. Any tips in which way to take this Ricardo?
The input transformer has the BIGGEST effect in these vintage circuits. Without it, the noise performance will be dire.Leek said:
THAT 1250 is a line receiver. It is a terrible mike preamp .. like your early 5532 efforts.
If you'd bothered to read my LNprimer in Yahoo MicBuilders, you'd see you FIRST DEFINE WHAT YOU WANT TO DO. Then you work out how to do it with what you have available.
This is BS
This is pure liquid BS.
Kingston had an excellent thread on what is REALLY important.
opamps and local decoupling of rails, some questions
http://www.groupdiy.com/index.php?topic=37307.80
Many true gurus chime in. It proves how OPA rolling takes a VERY poor second place to correct earthing, layout & decoupling. It’s a long thread but read the whole thing from #41 to find pearls of wisdom.
You still haven't specified what you want to do except some vague stuff about 'colour'.
And I can't find any specs on your Neve transformers.
But if you want a Mike Preamp with a vintage distortion profile to work with your Digi 002, I'd use my circuit with the Neve input transformer. Should be in the ball-park ... if not quite as good as my Sowter/Lundahl/Jensen recommendations.
I'm more wary of the Neve output transformer. I'd use your Crimson as this has a 'known' distortion profile which fits in with the circuit, the Digi 002 levels .. what you want. 'Some' spec is better than the Neve 'no spec.'
Or just leave the output unbalanced. If the preamp is within a metre or so of your Digi 002, it should be OK. The transformer(s) add(s) mainly 3rd harmonic while my circuit is mainly 2nd.
PN2907 can probably replace BC557C in this circuit.
There's loads of other stuff that's important for a good Mike Preamp ... like Layout, Earthing, Decoupling. Some of it is in Kingston's thread and there's more in my MicBuilders LNprimer.
Andy Peters
Well-known member
Leek said:
"color" is an utterly meaningless term when you don't qualify your needs.
