Active ribbon-mic

[abbey road d enfer]

Awesome, thanks. Now I see how I had the biasing all messed up. I played around with this in LT Spice today and these values give 25dB gain and ok distortion (<0.1%THD) at -45dBu input.

I drew up a little PCB: https://oshpark.com/shared_projects/w3wEWLH9

Again, I haven't built this yet so use at your own risk.

There's only 1.5V Vce. I would increase the base resistors.

 

[waltzingbear]

took a look at your gerbers, you're shorting the drain of one of the FETs to ground at the neg of the 100uF cap.
looks like you placed the run from FET drain to base of PNP and then placed the cap, which fell on top of it on the back. 

Cheers
Alan

 

[Meathands]

Thanks abbey and Alan. Here are new design files with your suggestions and a couple other changes:

1. Changed input Z to 10k and scaled C1 and C2 accordingly
2. Removed 47R resistors on output. SPICE calculates output Z at 30 Ohms, close enough to a mic output for me to eliminate two resistors.
3. Changed T2 and T3 to BC327. Just bc I have them around.
4. Corrected output polarity
5. Spread out LSK389 footprint for less treacherous soldering.

By the way, I got a couple of the "leading" mic boosters to do some real world tests against this circuit, and the results are impressive. This circuit is better than the two I bought in terms of noise, distortion, and stability of gain. To be honest, it's surprising to me that people continue to use just the two JFETS or cascode circuit without a followers at the output. The output Z of those circuits is ~6.8k, so it's not surprising their effective gain changes drastically based on the preamp input Z.

Here are the new gerbers: OSH Park ~
New schematic PDF is attached.

 

[MidnightArrakis]

Thanks abbey and Alan. Here are new design files with your suggestions and a couple other changes:

1. Changed input Z to 10k and scaled C1 and C2 accordingly
2. Removed 47R resistors on output. SPICE calculates output Z at 30 Ohms, close enough to a mic output for me to eliminate two resistors.
3. Changed T2 and T3 to BC327. Just bc I have them around.
4. Corrected output polarity
5. Spread out LSK389 footprint for less treacherous soldering.

By the way, I got a couple of the "leading" mic boosters to do some real world tests against this circuit, and the results are impressive. This circuit is better than the two I bought in terms of noise, distortion, and stability of gain. To be honest, it's surprising to me that people continue to use just the two JFETS or cascode circuit without a followers at the output. The output Z of those circuits is ~6.8k, so it's not surprising their effective gain changes drastically based on the preamp input Z.

Here are the new gerbers: OSH Park ~
New schematic PDF is attached.

Meathands:

While I found your schematic circuit of the "Inline Mic Preamp" to be interesting, there seems to me to be something missing in this schematic diagram. However, since I am - NOT - a circuit designer myself, then perhaps there is something going on here that is escaping my layperson's knowledge. In any case, perhaps you will be able to clarify a simple point for me, OK??? THANKS!!!

What I fail to see in this schematic is any "Power Source" to operate this circuit and it does not seem to be "Phantom-Powered". From what little I know about schematics, I believe there needs to be a " V+ " or a "9VDC" or "a something" symbol connected to "C5 +" in order to show where a power-supply needs to be connected and by what voltage. Or, is this circuit "self-powered"? I am guessing since C5 is a 63VDC capacitor, that V+ is probably 48VDC. If so, then a " 48V " symbol connected to " C5+ " should be shown.

In a distantly related sense, with this being a low-noise mic-preamp circuit primarily for ribbon microphones, it would be most useful to anyone attempting to build this circuit to know what - types - of capacitors should be used for both C1 and C2 and C3 and C4. Are ceramics OK? Or, Mylar caps? Or, better still, Polystyrene types? A simple piece of text next to each capacitor or maybe a general " NOTE: " shown within the schematic would eliminate any confusion and/or doubt. Just my 2-cents worth.

I would be very interested in hearing back from you about this, primarily to know if my observations are correct or not. THANKS!!!

Regards,

Jerry B. Williams
Midnight Blue Designs

/

 

[MidnightArrakis]

Thanks abbey and Alan. Here are new design files with your suggestions and a couple other changes:

1. Changed input Z to 10k and scaled C1 and C2 accordingly
2. Removed 47R resistors on output. SPICE calculates output Z at 30 Ohms, close enough to a mic output for me to eliminate two resistors.
3. Changed T2 and T3 to BC327. Just bc I have them around.
4. Corrected output polarity
5. Spread out LSK389 footprint for less treacherous soldering.

By the way, I got a couple of the "leading" mic boosters to do some real world tests against this circuit, and the results are impressive. This circuit is better than the two I bought in terms of noise, distortion, and stability of gain. To be honest, it's surprising to me that people continue to use just the two JFETS or cascode circuit without a followers at the output. The output Z of those circuits is ~6.8k, so it's not surprising their effective gain changes drastically based on the preamp input Z.

Here are the new gerbers: OSH Park ~
New schematic PDF is attached.

Meathands:

I believe that you need to change the T2 and T3 type BC327 transistors as I just received this message from Digi-Key about the BC327:

"This product is no longer manufactured and will no longer be stocked once stock is depleted".

In addition, the pin-out of the Emitter and Collector of the BC327 shown in your schematic is reversed as to what is shown in the BC327 datasheet, which I have attached for your review. Is there a similar or better type of transistor available that is more common and more readily stocked?

The BC327 is also available in what is called a "kinked" configuration. The "kinked" version also has a readily available "spread out" footprint for downloading.

On your LSK389 schematic symbol, I believe it would be rather useful for you to also show where the "Source" and "Drain" pin locations are on your schematic symbol since the "Gate" location is obvious. In addition, while these transistors are available from Digi-Key, there is a 5-piece minimum purchase and at $10 each, that is $50 just for 5-transistors!!! So, everyone should be prepared to build at least 5 of these "Inline Mic Preamp" circuits!!!

Regards,

Jerry B. Williams
Midnight Blue Designs

/

 

[waltzingbear]

what do you intend to use for C3 & 4 ? COG gets pretty expensive at that value. and hard to find. Mouser has one cap, a Kermet 0.47uF/50V but it is 6.35mm lead spacing, $3.88 ea and 9 in stock

 

[Meathands]

https://octopart.com/r82dc3470aa60k-kemet-20075732?r=sp
But you could also go down to .22u or even .1u without creating an audible HPF. I just used .47u because we stock them for a different product,

 

[abbey road d enfer]

What I fail to see in this schematic is any "Power Source" to operate this circuit and it does not seem to be "Phantom-Powered".

From the beginning it is clear that this circuit is intended to operate in conjunction with phantom powering, which involves a 48V source and distribution resistors (6.8k) connected to the two legs of the microphone connection, the shield being the 0V reference. I suggest you google phantom power.

From what little I know about schematics, I believe there needs to be a " V+ " or a "9VDC" or "a something" symbol connected to "C5 +" in order to show where a power-supply needs to be connected and by what voltage. Or, is this circuit "self-powered"? I am guessing since C5 is a 63VDC capacitor, that V+ is probably 48VDC. If so, then a " 48V " symbol connected to " C5+ " should be shown.

Since the 48V source is in the mixer/preamp, and not in the preamp, it is not necessary to represent it in the circuit. Do you represent the power plant in the schemo of a mains-powered apparatus?

In a distantly related sense, with this being a low-noise mic-preamp circuit primarily for ribbon microphones, it would be most useful to anyone attempting to build this circuit to know what - types - of capacitors should be used for both C1 and C2 and C3 and C4. Are ceramics

Capacitor technology has very little to do with noise in this circuit.
There may be other considerations, though. For example C1 & C2 should be cearmics because they are the less inductive, which is paramount in filtering EMI/RFI, although, to be perfect, they should be wired directly on the connector, not on the PCB. However, we often don't need perfection, so having them on the PCB, with adequate layout is acceptable.
More important is that they are properly matched, 2% being a desirable target. Which orients the search towards COG/NP0 types.
Other types have too much tolerance and variation. They also have pronounced non-linearities, but considering teh usually low levels there it's not a major issue.

OK? Or, Mylar caps? Or, better still, Polystyrene types?

Most ceramic capacitors of this value would have significant distortion issues, so the only alternative is film or 'lytic.
Purists would frown at 'lytics, so there's not much choice there, although I bet some would advocate tantalum types.
For my MicBooster I use 0.47uF polyester types. The residual AC voltage across the capacitor is so low the resultant distortion is negligible.

A simple piece of text next to each capacitor or maybe a general " NOTE: " shown within the schematic would eliminate any confusion and/or doubt. Just my 2-cents worth.

I would be very interested in hearing back from you about this, primarily to know if my observations are correct or not. THANKS!!!

Regards,

Jerry B. Williams
Midnight Blue Designs

/

 

[Meathands]

Meathands:

I believe that you need to change the T2 and T3 type BC327 transistors as I just received this message from Digi-Key about the BC327:

"This product is no longer manufactured and will no longer be stocked once stock is depleted".

Thanks for taking a look at the circuit. There are lots of BC327s still made by several manufacturers: https://www.mouser.com/Search/Refine?Keyword=BC327

In addition, the pin-out of the Emitter and Collector of the BC327 shown in your schematic is reversed as to what is shown in the BC327 datasheet, which I have attached for your review. Is there a similar or better type of transistor available that is more common and more readily stocked?

Right you are! Will correct that, thanks.

On your LSK389 schematic symbol, I believe it would be rather useful for you to also show where the "Source" and "Drain" pin locations are on your schematic symbol since the "Gate" location is obvious. In addition, while these transistors are available from Digi-Key, there is a 5-piece minimum purchase and at $10 each, that is $50 just for 5-transistors!!! So, everyone should be prepared to build at least 5 of these "Inline Mic Preamp" circuits!!!

LSK389 are available from NAC semi here: LSK389A-TO-71 | LINEAR SYSTEMS | NAC Semi

Here's a google Drive folder with updated schematic, BOM, and a more condensed PCB layout without the PC-mount XLR jacks: GroupDIY Active Ribbon-mic Files - Google Drive

 

[MidnightArrakis]

Thanks for taking a look at the circuit. There are lots of BC327s still made by several manufacturers: https://www.mouser.com/Search/Refine?Keyword=BC327

Right you are! Will correct that, thanks.

LSK389 are available from NAC semi here: LSK389A-TO-71 | LINEAR SYSTEMS | NAC Semi

Here's a google Drive folder with updated schematic, BOM, and a more condensed PCB layout without the PC-mount XLR jacks: GroupDIY Active Ribbon-mic Files - Google Drive

I have captured your schematic into my CAD-software and included 2 resistors for what I believe would be the phantom-power for this circuit. I was criticized for previously asking if this circuit is phantom-powered, but in my opinion, just sticking the C5 capacitor out there without some type of V+ symbol, really is not all that obviously clear to me. It is probably a failure on my part but my experience and exposure to schematics is that they are clearly drawn out, detailed and documented so they can be easily understood. You don't have to assume anything nor do you have to go back through 20-pages of previous information just to find something out. Let me know what I have done incorrectly and I will update my version accordingly.

While I do know that BC327s are still available, I was merely alerting you to the fact that they have been discontinued and someone in the future who may be reading this forum thread and wish to also build this circuit might end up being out-of-luck. That's why I had asked you if there were any similar, but more current transistors, that would also perform as equally well or better than the BC327. I have also seen that the original TOSHIBA 2SK389 had been discontinued, but then LINEAR Systems re-introduced it as the LSK389. Lucky us, huh???

Are there any additional tweaks to this circuit that you can think of which would be good to have? Are there any component values that could be better optimized? THANKS!!!

/

 

[abbey road d enfer]

I have captured your schematic into my CAD-software and included 2 resistors for what I believe would be the phantom-power for this circuit. I was criticized for previously asking if this circuit is phantom-powered, but in my opinion, just sticking the C5 capacitor out there without some type of V+ symbol, really is not all that obviously clear to me. It is probably a failure on my part but my experience and exposure to schematics is that they are clearly drawn out, detailed and documented so they can be easily understood. You don't have to assume anything nor do you have to go back through 20-pages of previous information just to find something out. Let me know what I have done incorrectly and I will update my version accordingly.

As I wrote earlier the 6.8k resistors are not on this board; they are in the mic preamp, with the 48V source.
You may not have noticed that, but the conduction of the transistors is all it takes to power the circuit.
And 48V is not applied where you think.
You must consider the system, not just this part of the circuit.

 

[MidnightArrakis]

As I wrote earlier the 6.8k resistors are not on this board; they are in the mic preamp, with the 48V source.
You may not have noticed that, but the conduction of the transistors is all it takes to power the circuit.
And 48V is not applied where you think.
You must consider the system, not just this part of the circuit.

 

[MidnightArrakis]

THANK YOU!!! -- I just took a "stab in the dark" when placing the 2 6K8 resistors to power the circuit because earlier, I couldn't get an answer to a question I had asked about it. Right now, all that I am looking for is for someone in this thread to proclaim that all of the refinements and tweaking are done and this circuit is as good as it is going to get. But, since this thread is now over 14-years old, it seems as though every few months or so someone comes along and points out if - these - resistor values were changed, the gain improves or the noise is lowered or by using - these - transistors instead, some other parameter is greatly improved or whatever. Just when can all of the tweaking end and this circuit is safe to build? I'm a "Circuit Builder" and - not - an "Endless Circuit Design Tweaker"!!!

When a company hires me to design one of their new product design projects, they hand me all of their - completed and finished - schematics and not a folder containing all of their engineering efforts during the past 6-months or more containing all of the circuit changes, improvements, updates and general engineering discussions and efforts. These completed schematics are clearly drawn out, detailed and well documented. My job is to then take these finished schematics and perform all of the mechanical and PCB designs, along with anything else that is necessary, for me to create a "first-article-build" prototype. The equipment shown within this link are a few examples of the equipment I have designed only from completed schematics:

https://app.box.com/s/ppsp38kb5rijeadt6i73q4u99nwurqqv
[the conduction of the transistors is all it takes to power the circuit] -- Since I am not a "Circuit Designer", that aspect is not apparent to me. While I certainly know what "Phantom Power" is (I've been using it since probably before you were born), I am simply not aware of powering a circuit in this manner.

[48V is not applied where you think] -- OK!!! Could you please be so kind and relate to me as to where it is applied? THANKS!!!

When and "IF" this circuit is ever considered completed and finalized, I am planning on designing 2 different versions of a PCB, one for Surface-Mount parts and the other for Thru-Hole parts. But, I am right now hesitant to buy any parts because next week someone in this thread will proclaim, "WOW!!! Instead of using a BC327, a 2NXXXX transistor works THIS MUCH BETTER because of its "blah, blah, blah" parameters"!!! I would just like to hear a general consensus of all of the Circuit Designers in this thread to say, "Yeah, well.....I cannot see or imagine how any additional improvements can be made to this circuit" and call it quits. Is that possible here???

I have a ROYER SF-12 stereo ribbon microphone, as well as a stereo-pair of CASCADE "Fat-Head" ribbon microphones and they both could use a little bit of "OOMPH" prior to my DigiMax mic preamps, especially the SF-12. This is why I am keenly interested in building this circuit.

I "THANK YOU!!" for any and all of your assistance.

JBWilliams

/

 

[abbey road d enfer]

So who produced the schemo?
I feel somewhat cheated because I have published my own version about a year back, that looks strangely similar (but not identical). Now I see someone is taking advantage and produces a commercial version. I can see a couple of points that could improve performance/practicality/reliability, but I'll keep it to myself for now.

 

[Meathands]

I'm sorry you feel cheated and if I've contributed to that. I revived this thread earlier this year with an attempt to synthesize all the best ideas in this thread into one schematic. The most important of which (to use the Schoep's output buffer instead of cascode) was yours originally I believe from a few years ago. I wasn't aware you already had a completed schematic until you posted it on the previous page of the thread, but when you I asked you about why you did certain things and incorporated them into my schematic.

Anyway that's all just to say its no strange coincidence that my schematic above looks like yours—I compiled it from this thread then incorporated suggestions you and a couple others kindly gave. By sharing the schematic and PCB layouts, I was attempting to give back something to this thread. I have tweaked some values but I am certainly not attempting to pass the design off as mine and I apologize if it came across that way.

 

[MidnightArrakis]

So who produced the schemo?
I feel somewhat cheated because I have published my own version about a year back, that looks strangely similar (but not identical). Now I see someone is taking advantage and produces a commercial version. I can see a couple of points that could improve performance/practicality/reliability, but I'll keep it to myself for now.

[I can see a couple of points that could improve performance/practicality/reliability, but I'll keep it to myself for now] -- OK!!! I guess your answer settles my question of whether or not to build this circuit!!! My now knowing that this circuit - could - be improved upon in different ways, but that such information is being withheld for whatever reasoning, I will just abandon the idea of building this circuit and instead go buy a "Cloud-Lifter" or something.

I had always thought that one of the underlying principles of this forum was to work together on a project and achieve some level or type of "Utopia" that we all could share. But, now.....learning that a certain circuit design "Utopia" is being withheld from this thread shatters my faith in what I thought this forum was all about!!! And.....that's sad.

[it's no strange coincidence that my schematic above looks like yours] -- Since this is an open and public forum, I would then believe that there is no "ownership" of - anything - that is posted here!!! Everything that is shared on here is for "The Greater Good" and - THAT - is not owned by anyone!!! So, why would anybody claim that a particular schematic is "theirs" and then also withhold any contributing improvements, when said schematic is being posted on an open and public forum to begin with??? What am I missing here?

As I have mentioned previously, I am - not - a "Circuit Designer", but a "Circuit Builder". Depending upon the level of technology used, I am able to take a schematic and package it into a 1U rack-enclosure or have it shrunk down to the size of "the head of a pin" and package it into a watch enclosure!!! With that said, my presence here is to also assist any of those who may be interested, in taking any of this forum's schematics and designing any PCB's and/or performing any sheet-metal design work which would help someone who is not skilled in these areas complete a project that they are interested in building. And, I certainly would - NOT - lay any "ownership" claim to any of my PCB or sheet-metal designs as being "mine"!!! They're for EVERYBODY!!!

/

 

[ricardo]

MidnightArrakis wrote : Right now, all that I am looking for is for someone in this thread to proclaim that all of the refinements and tweaking are done and this circuit is as good as it is going to get.

There is no circuit that cannot be improved. I look at my own designs from 40+ yrs ago, some of which have been 'improved' by others, and I still see possible improvements (and faults which I've made.)

But a good circuit gives good performance even if it isn't 'perfect' and many of the circuits in this thread fall into this category.

The reason for this 'continual improvement' is that we have many gurus (and pseudo gurus like me) who are happy to share their knowledge & skill ... and also educate others with their insight. They don't have to do this.

We need to do our utmost to keep these gurus happy. Some truly SOTA designs have emerged from their contributions to our pages.

Some ways to really piss them off ...
- turning their contributions into commercial products without permission and attribution
- spurning their efforts at education. As you haven't bothered to look at all 20 pgs, you've probably missed Abbey taking Meathands very patiently through his design in the last few pages.
- modifying their designs without due thought. If you had built Abbey's circuit EXACTLY (with perhaps a few clarifications on capacitor types), it would have worked and performed well. ie better than most of the commercial designs

It is a LOT of time & trouble to explain circuits fully .. especially to those who can't be bothered to do their homework. Again they don't have to do this. This isn't a commercial contract. The 20 pgs are what you see & get.

I won't go on but your posts are unlikely to encourage Abbey to share his thoughts. You might like to take your concerns to the Brewery where they will get the attention they deserve ..

 

[clintrubber]

Time flies.... started this thread back in 2005... (back then triggered by the active Royer ribbon)

Just happened to see that meanwhile quite some commercial products have emerged.

Can't tell for sure which ones took some inspiration from this thread (but Ok, wasn't the printing press also invented in several places at the same time? ).

Seeing that at least one manufacturer is already offering a +Germanium version (really? ) I expect/fear we haven't seen the latest commercial addition yet... so it might be milked to the last drop... let's see, tube emulation... hand-rolled foil-capacitors... gold-plated XLR-contacts... premium enclosures...

But possibly the #1 thing to regret is that these additions might/will be used where they're often not needed at all. Much confusion about it out there, misinformed people talking about increased headroom etc. People thinking it all sounds better because this-or-that gizmo is inserted for additional mojo.

OK, enough rambling for now

 

[kato]

…
Seeing that at least one manufacturer is already offering a +Germanium version (really? ) I expect/fear we haven't seen the latest commercial addition yet... so it might be milked to the last drop... let's see, tube emulation... hand-rolled foil-capacitors... gold-plated XLR-contacts... premium enclosures...

The germanium ones are LOL funny. I wouldn't be surprised to find the same 2 silicon transistors they use in their non-germanium model, but with different color paint on the outside.

…
But possibly the #1 thing to regret is that these additions might/will be used where they're often not needed at all. Much confusion about it out there, misinformed people talking about increased headroom etc. People thinking it all sounds better because this-or-that gizmo is inserted for additional mojo.

I've been telling everyone who asks about cloudlifter type devices, "If you have a mic pre that will do 65dB or 70dB of gain you don't need one."

Everyone seems convinced this is a must-own device. But YouTube has blown up with folks using SM7bs with Focusrite 2i2's, so I guess they've created a market for such devices.

 

[abbey road d enfer]

Seeing that at least one manufacturer is already offering a +Germanium version (really? ) I expect/fear we haven't seen the latest commercial addition yet...

What about a phantom-powered tube version?

But possibly the #1 thing to regret is that these additions might/will be used where they're often not needed at all. Much confusion about it out there, misinformed people talking about increased headroom etc. People thinking it all sounds better because this-or-that gizmo is inserted for additional mojo.

Agreed, but...see next post