Mic Preamp. Adding things and getting ideas. Would love help please.

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Leek

Well-known member
Joined
Jul 11, 2016
Messages
105
So, I have built a working transistor based mic preamp.
Here is the schematic I used, it's identical in every aspect.

http://www.talkingelectronics.com/projects/200TrCcts/images/MicPreAmp.gif


In short, could I use a 10k potentiometer, connected to the input/output of the circuit (and ground) to control the gain? And, if not where could I connect a potentiometer to control the gain of the preamp.
 
The 100 ohm resistor in series with the 10uF capacitor and their associated 10K resistor set the gain. The 'normal' way to vary the gain of this circuit is to replace the 100 ohm resistor with a variable resistor. The gain is given approximately by the 10K divided by the 100 ohms which in this case is 100 times or 40dB. Replacing the 100 ohms with a 10K pot would allow the gain to be varied from 6dB to 40dB. The 10uF is however too small for adequate bass response; its 3dB point with the 100ohms is about 160Hz.

Cheers

Ian

PS. The 100nF input and output capacitors are probably also too small for adequate bass response.
 
ruffrecords said:
The 100 ohm resistor in series with the 10uF capacitor and their associated 10K resistor set the gain. The 'normal' way to vary the gain of this circuit is to replace the 100 ohm resistor with a variable resistor. The gain is given approximately by the 10K divided by the 100 ohms which in this case is 100 times or 40dB. Replacing the 100 ohms with a 10K pot would allow the gain to be varied from 6dB to 40dB. The 10uF is however too small for adequate bass response; its 3dB point with the 100ohms is about 160Hz.

Cheers

Ian

PS. The 100nF input and output capacitors are probably also too small for adequate bass response.


Ian, would replacing the 10uf with a 47uf, 63v electrolytic cap and replacing the 100nf caps with Wima 0.22uf caps help with bass response?
 
Leek said:
So, I have built a working transistor based mic preamp.
Here is the schematic I used, it's identical in every aspect.

http://www.talkingelectronics.com/projects/200TrCcts/images/MicPreAmp.gif
Note that this circuit is way too simple to have good linearity. Meaning the signal will be distorted even at low levels. A Fuzz Face guitar pedal would probably have better linearity. This circuit has no feedback.

There are Neve and Telefunken circuits that are discrete like this but with way better linearity and features. They were considered state-of-the-art at their time and a lot of good music was recorded with them. So discrete circuits like this are still worth building for sure. But "performance" is going to be nothing like an op amp based design.
 
JohnRoberts said:
A short cut back of the envelope math if 100 ohm and 10uF delivers a half power point of 160 Hz, increasing the cap 4.7x will reduce the cut off to 1/4.7  so between 35 and 40 Hz.

.22 uf in series with input should be fine. .22 uF in series with output depends on the load impedance it drives, while that load can't be very low with only a 10k collector load.

JR

This is only a test of components I have laying around. I'm trying to get a decent configuration that will outperform the stock preamps in my DIGI 002. Something that not only sounds better but will be beneficial as a learning process for me.

Ok, got the cap replacements squared away on my breadboard. Next, I plan to add a THAT 1250 (-0dB) line transmitter into a 2nd gain stage of some sort (possibly based around NJR NJM12904AD) with another potentiometer to control output gain. I plan to use a Crimson C4000B output transformer to balance the output and to help with unwanted distortion. Does this sound adequate? If the NJR op amp isn't adequate, how could I use something like a NE5532, OPA2134 or equivalent as a single supply op amp?
 
JohnRoberts said:
unfortunately that can just confuse other people trying to learn with misinformation. yes I have been ignoring that thread...
You guys both need to start with learning about discrete device design (biasing etc). Swiftly becoming a lost art as off the shelf ICs will deliver superior performance to anything you come up with.

Then after you learn some basics, there is literature about console design. Look for the Steve Dove series that ran in studio sound, but I think it has been reprinted in some later book(s).

Sorry if this sounds harsh there are many here who can help you besides those of us who designed consoles for a living. You must walk before you can run.

JR

I don't think it's harsh. Any information I can gather and use to learn from is always welcome!
 
80hinhiding said:
Yeah, it is harsh John, in my opinion.

You may feel it is harsh but it is good honest advice. There a re loads of circuits on the internet and you can probably cobble something together from them. Each one is probably fine in isolation but you need to ensure input impedance and output drive capabilities are compatible. It is clear from some of the circuits you have posted that they will not work well together. That I why I recommended you start with regular audio op amps as they have all the basic characteristics that mean they will work well with each other. That way you can get something running knowing it has a good chance of both working and sounding good. After that is the time to start thinking about discrete circuits.

Cheers

ian
 
ruffrecords said:
You may feel it is harsh but it is good honest advice. There a re loads of circuits on the internet and you can probably cobble something together from them. Each one is probably fine in isolation but you need to ensure input impedance and output drive capabilities are compatible. It is clear from some of the circuits you have posted that they will not work well together. That I why I recommended you start with regular audio op amps as they have all the basic characteristics that mean they will work well with each other. That way you can get something running knowing it has a good chance of both working and sounding good. After that is the time to start thinking about discrete circuits.

Cheers

ian


Could I not just use a output stage based around a STMicroelectronics STX13005 power transistor and a Crimson Audio 4000B line transformer to get rid of any unwanted distortion? It seems quite unjust that I came here for help and tips putting this together and the only answer I get is "scrap everything, get a few ic's and build something".
 
80Hinhinding, I know the feeling. I'm using a solderless breadboard for Christ sake. I however don't want to use IC's. I want to use a discrete signal path if I can. The transistor's should be adequate as far as noise goes. The components are all top notch and with 1%. Caps have been changed to compensate for bass response. Instead of telling us we're not ready for this or that. Perhaps share knowledge and the things you guys know to help make us ready? You can't honestly tell me you read a few books and could build a console. It was through working on the equipment, and experimenting that things were built.

So, I'll try this again. I have my potentiometer set and can now adjust gain. Thanks Ian for that. I have now replaced cap's to compensate for a better bass response. Apparently distortion is going to be a problem with this circuit even at low levels. How can I compensate for that and eliminate it. Would adding a STMicroelectronics STX 13005 power transistor and a Crimson 4000B output transformer reduce the distortion? Would a better 2nd gain stage based around a few more transistors or even an IC help with this? Or would I be better off adding a 2nd gain stage based around a few transistors (or an IC) and the output transformer?
 
JohnRoberts said:
I'd say I'm sorry if i was.. I went back to your thread and remain uninterested. So if you don't listen and take advice you don't get more advice. Taking a wild guess we may have a generation gap here.

You don't seem to appreciate the massive resources you have at your finger tips using modern search engines and the internet. I wish I had that as a youth.

If you demand to be spoon fed an education that takes years/decades to acquire, good luck with that. As I said there are many people here higher up the latter of knowledge that can help you. 

I also have a personal problem with passive mixers in general (no I won't explain why). Your thread appears to have little to do with passive combining anyhow.  Maybe break it down to separate threads with single questions each, not a blog.

You will live without my help, many have.

JR

PS: I really do appreciate people who truly want to learn, good luck.

Can we please not use my thread as a means to sprout uselsss information? If you want to talk to 80hinhinding about his design or this or that, use his thread. You're literally just taking up space in mine wher i'm actually trying to get useful information and use it. Not trying to be rude, but your post was pointless and didn't help me at all.
 
> http://www.talkingelectronics.com/projects/200TrCcts/images/MicPreAmp.gif
> This circuit has no feedback.


10K Q2C to Q1E looks like a feedback. What am I missing?

Assuming I see what I see, open-loop gain is well over 1,000, closed-loop gain is about 100. "Significant" (though not huge) feedback.

The 9V supply limits the possible output. No-NFB the THD will approach 20% at 2Vrms out. With NFB, more like 2% at 2V and 1% near 1V.

That puts input overload at 10mV-20mV. This is lower than many modern microphone applications. A gain of 100 under a 9V supply is pretty crazy. But he was asking how to reduce it. Note that a post-pot will distort and then turn-down, so it still overloads. A pre-pot throws-away S/N. With better understanding of basic resistor and circuit operation a gain-change plan will be clear.
 
> I ... don't follow all advice, so what?

So advice-offers will get tired of being dissed and stop advising.
 
PRR said:
> http://www.talkingelectronics.com/projects/200TrCcts/images/MicPreAmp.gif
> This circuit has no feedback.


10K Q2C to Q1E looks like a feedback. What am I missing?

Assuming I see what I see, open-loop gain is well over 1,000, closed-loop gain is about 100. "Significant" (though not huge) feedback.

The 9V supply limits the possible output. No-NFB the THD will approach 20% at 2Vrms out. With NFB, more like 2% at 2V and 1% near 1V.

That puts input overload at 10mV-20mV. This is lower than many modern microphone applications. A gain of 100 under a 9V supply is pretty crazy. But he was asking how to reduce it. Note that a post-pot will distort and then turn-down, so it still overloads. A pre-pot throws-away S/N. With better understanding of basic resistor and circuit operation a gain-change plan will be clear.

Thanks PRR for the helpful information. I can up the power supply no problem. I was thinking of perhaps using 15v for the added headroom. If I were to use a 15v supply, and incorporate the STMicroelectronics STX 13005 power transistor and the Crimson Audio 4000B output transformer, could I get an adequate signal while reducing distortion? Perhaps use a THAT 1250 IC to unbalance the signal, pass it through another stage based around a NJR NJM3414AD, a STX 13005 transistor and a Crimson 4000B to balance the output again?
 
PRR said:
> http://www.talkingelectronics.com/projects/200TrCcts/images/MicPreAmp.gif
> This circuit has no feedback.


10K Q2C to Q1E looks like a feedback. What am I missing?

Assuming I see what I see, open-loop gain is well over 1,000, closed-loop gain is about 100. "Significant" (though not huge) feedback.

The 9V supply limits the possible output. No-NFB the THD will approach 20% at 2Vrms out. With NFB, more like 2% at 2V and 1% near 1V.

That puts input overload at 10mV-20mV. This is lower than many modern microphone applications. A gain of 100 under a 9V supply is pretty crazy. But he was asking how to reduce it. Note that a post-pot will distort and then turn-down, so it still overloads. A pre-pot throws-away S/N. With better understanding of basic resistor and circuit operation a gain-change plan will be clear.
You right. Of course. I glanced at this and it looked like 2 inverting stages. But I see current supplied by the 10k will serve to reduce the Vbe of the 1st transistor and thus throttle it off. Unlike in a Fuzz which uses an inverting stage with a follower and then mixes the output with the input.
 
Leek said:
So, I'll try this again. I have my potentiometer set and can now adjust gain. Thanks Ian for that. I have now replaced cap's to compensate for a better bass response. Apparently distortion is going to be a problem with this circuit even at low levels. How can I compensate for that and eliminate it. Would adding a STMicroelectronics STX 13005 power transistor and a Crimson 4000B output transformer reduce the distortion? Would a better 2nd gain stage based around a few more transistors or even an IC help with this? Or would I be better off adding a 2nd gain stage based around a few transistors (or an IC) and the output transformer?

As PRR pointed out, there is plenty of gain in this circuit so distortion will not be a problem. It is very similar to a circuit I used in a small mixer for location recording back in the 70s. You now have a basic gain block and you can follow it with a fader and then you need to add an output stage to make it into a complete mic pre.

Cheers

Ian
 
Here's more advice.

Go BUY "The Art of Electronics" 3rd ed by Horowitz & Hill and start reading it.  They'll be lots you don't understand but there will be lots that you will .. including the explanation for simple circuits like yours.

And you will still be referring to it when you are a Guru in your old age.

It should tell you how to add your STM STX 13005 power transistor and the Crimson Audio 4000B output transformer etc.  But don't expect this to reduce distortion etc. to an "acceptable" level without a LOT more understanding.

Also download LTspice and start learning to use it.

I'm going to side with JohnRoberts and PRR on this.  Us oldies have done a lot of design(s) like this and while "good" performance is possible, its not simple.  Ruffrecords is ex-Neve so his idea of "acceptable" is slightly easier.

(Sorry Ian.  Ex-Calrec can't resist a dig at ex-Neve  8) )

If you just want to play with something on a breadboard, play away .. but the breadboard, TAOE and LTspice will teach you a lot more and more quickly than anything we could tell you in a dozen long posts.

But have a look at RIAA preamps in old HiFi amps if you want tested circuits of this sort.  Variable gain which is important for Mike Preamps is harder to do nicely.

I've just looked back at your original post.  Ian's reply is the most appropriate.  It just answers your question.  Curing your circuit's other ills and turning it into a practical mike preamp will take a lot longer.
 
ruffrecords said:
As PRR pointed out, there is plenty of gain in this circuit so distortion will not be a problem. It is very similar to a circuit I used in a small mixer for location recording back in the 70s. You now have a basic gain block and you can follow it with a fader and then you need to add an output stage to make it into a complete mic pre.

Cheers

Ian

Ian, would it be plausible to use (for an output stage) 1k resistors feeding a STMicroelectronics STX13005 power transistor that feeds the Crimson 4000B Transformer into a Amphenol 1/4" connector? If this doesn't seem plausible, could you provide a simple output stage that would work on a single power supply. I can skimp on the transformer if need be. Or, if need be I can spend a little more to get a decent linear power supply so I can use a dual supply op amp to make a balanced line driver. But, I'd like to not do so if I can. I know the 1073 doesn't use linear power. So, I know I can build a decent single supply output stage that connects via a 1/4" jack.
 
Leek said:
Ian, would it be plausible to use (for an output stage) 1k resistors feeding a STMicroelectronics STX13005 power transistor that feeds the Crimson 4000B Transformer into a Amphenol 1/4" connector?

Probably not. The transformer will probably be OK, but the manufacturers specs contain virtually zero design information so it is difficult to be certain. However, if you feed it into a typical 10K input it should be OK but its headroom will be limited.

As for the STX13005, I am not sure why you chose that; it is meant for fast  switching applications and it comes in a TO92 package so it is not really suitable for an audio output stage. I think you would be better off with something like a BD131 on a small heatsink. Even then it does not really have enough current gain on its own to act as an output stage.

Cheers

Ian
 
ruffrecords said:
Probably not. The transformer will probably be OK, but the manufacturers specs contain virtually zero design information so it is difficult to be certain. However, if you feed it into a typical 10K input it should be OK but its headroom will be limited.

As for the STX13005, I am not sure why you chose that; it is meant for fast  switching applications and it comes in a TO92 package so it is not really suitable for an audio output stage. I think you would be better off with something like a BD131 on a small heatsink. Even then it does not really have enough current gain on its own to act as an output stage.

Cheers

Ian

I just had a STX transistor laying around. So, was trying to figure out how I could use it. What would be the most simplistic way I could use a Crimson 4000B transformer (or heck how about a Triad TY-250P as I have a bunch laying around) to make a decent output stage that connects to a 1/4" jack. I plan to feed the output stage into the line input of my 002. If I'm not mistaken it is a 10k input as most modern interfaces. would it be as simple as connecting the the output and ground from the gain block to the transformer, then connect that to the 1/4" jack?

And if using the transformer is too complicated, a transformerless output stage would be fine. As long as it was single supply. I've upgraded my psu to 18v as well for the added head room. So, any ideas Ian? I was thinking maybe I could either use the transformers or use some cheap single supply op amps to create an electronically balanced output stage.
 
Leek said:
I just had a STX transistor laying around. So, was trying to figure out how I could use it. What would be the most simplistic way I could use a Crimson 4000B transformer (or heck how about a Triad TY-250P as I have a bunch laying around) to make a decent output stage that connects to a 1/4" jack. I plan to feed the output stage into the line input of my 002. If I'm not mistaken it is a 10k input as most modern interfaces. would it be as simple as connecting the the output and ground from the gain block to the transformer, then connect that to the 1/4" jack?

And if using the transformer is too complicated, a transformerless output stage would be fine. As long as it was single supply. I've upgraded my psu to 18v as well for the added head room. So, any ideas Ian? I was thinking maybe I could either use the transformers or use some cheap single supply op amps to create an electronically balanced output stage.

Just to get it going and hear it through your 002 all you need is a simple emitter follower to buffer the  existing mic pre. The output of your BC557 is already at about half the supply voltage so get rid of the 100n output cap and connect the base of another NPN to the collector of the BC557. Use the STX if you wish. Connect its collector to the +ve supply, Connect its emitter to ground via a 1K resistor. It should then draw about 9mA. Connect the +ve of a 47uF electrolytic to the STX emitter and connect the other end of the 47uF to 0V via a 10K resistor. Your mic pre output is across the 10K resistor. it is unbalanced, quite low impedance and should drive you 002 line in no problem.

Cheers

Ian
 
Back
Top