Active ribbon-mic

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
f0m3 said:
Ok, thank you.
So what about the 4 jfet thing.  Any suggests?

Not my circuit so anything's just conjecture.
 
1/Doubling up on another pair of the same type J-Fet in parallel would lower noise by up to 3dB.

2/Another possibility would be a pair of suitable J-Fets to cascode the K170's.  Not so much for the decrease in capacitance and extended top end, nor for any elimination of distortion at high frequencies, because the source z is low here. 
But because the K170 is best limited to 15V or lower on its drain to prevent against any possible leakage on its gate. 

Edit:
3/ Add a pair of source followers to the output to lower the impedance and provide a better source to the following mic pre.

4/ Anything's your guess...  ;)




 
Winston O'Boogie said:
1/Doubling up on another pair of the same type J-Fet in parallel would lower noise by up to 3dB.
Provided they run each at the same current than the singles. Simulations lead me to think it is feasible within the constraints of P48.

2/Another possibility would be a pair of suitable J-Fets to cascode the K170's.  Not so much for the decrease in capacitance and extended top end, nor for any elimination of distortion at high frequencies, because the source z is low here. 
But because the K170 is best limited to 15V or lower on its drain to prevent against any possible leakage on its gate. 
And that's about it. IMO there are other (simpler) ways for restricting Vds. For the rest, Miller capacitance is not a real issue and the output impedance, already excessive in the original, is even worse.

[/quote] 3/ Add a pair of source followers to the output to lower the impedance and provide a better source to the following mic pre. [/quote] From the beginning I felt concerned with the fact that the preamp sees an unduly high source Z.
I've simmed a version with emitter-followers added. The raw output Z is less than 5 ohms. Needs build-up resistors for properly driving xfmr-based preamps. I know some use low-power opamps. I believe that would be noisier than a simple emitter-follower.

I've layed out a PCB. Placed the order today. Size 65x28mm, designed to fit into a Neutrik NA-housing. We'll see...
 
abbey road d enfer said:
Provided they run each at the same current than the singles. Simulations lead me to think it is feasible within the constraints of P48.
And that's about it. IMO there are other (simpler) ways for restricting Vds. For the rest, Miller capacitance is not a real issue and the output impedance, already excessive in the original, is even worse.

3/ Add a pair of source followers to the output to lower the impedance and provide a better source to the following mic pre.  From the beginning I felt concerned with the fact that the preamp sees an unduly high source Z.
I've simmed a version with emitter-followers added. The raw output Z is less than 5 ohms. Needs build-up resistors for properly driving xfmr-based preamps. I know some use low-power opamps. I believe that would be noisier than a simple emitter-follower.

I've layed out a PCB. Placed the order today. Size 65x28mm, designed to fit into a Neutrik NA-housing. We'll see...

You're the man. 
Regarding doubling up on input pairs,  I was thinking they'd each have separate ballast resistors for current sharing.  Besides the output impedance issue which you've addressed, the source impedance the J-Fets see is not  the best for noise so another 2 might help.  Anyway...
If you're using K170's, do you need a few?  I have some on hand I can send. 
 
Winston O'Boogie said:
You're the man. 
Regarding doubling up on input pairs,  I was thinking they'd each have separate ballast resistors for current sharing.  Besides the output impedance issue which you've addressed, the source impedance the J-Fets see is not  the best for noise so another 2 might help.  Anyway...
If you're using K170's, do you need a few?  I have some on hand I can send.
Thanks for your generous offer. The board is layed out with LSK389's, which I find incredibly difficult to get. The only source I found doesn't want to ship to France, so I had them shipped to NY where a friend will post them back to me... :mad:
Meantime I have ordered 2SK369's; I'll have to twist leads...
I've laid the board with separate source resistors
 
abbey road d enfer said:
  The board is layed out with LSK389's, which I find incredibly difficult to get. The only source I found doesn't want to ship to France, so I had them shipped to NY where a friend will post them back to me... :mad:
Meantime I have ordered 2SK369's; I'll have to twist leads...
I've laid the board with separate source resistors

Tell me about it!  I've yet to even bother trying the LSK389 because of hassle getting it here too.    I traded most of my original 2SK389's to Fred Forssell about 15 years ago as I needed some "V" grade J74's and K170's. 
I did find a little baggie of a handful of 2SJ109's recently and the price those things now sell for was too tempting so, off they went...     
 
I made progress! Received the boards, populated them in various manners.
First with 4xLSK170's (unknown grade), dun wuk, because too much current, the FET's were like short-circuit.
Then 2xLSK170's, it wuks! But draws about 9mA, which is too close to the P48 limits IMO.
After I receive the LSK389's, I put one on a board. It wuks too! What I means, it makes some sound, but also a lot of noise. Change the LSK and then it works fine.
Finally I put two LSK'sand it works superbly. I made noise measurements; loaded with a SPL EQ Magic preamp of unknown input Z*, the booster provides 27dB of gain I measure an EIN of  -129dBu lin 22-22k with a 200r dummy source. It's too good to be true. It is somewhat consistent with the claimed 1nV/sqrtHz of LSK389. Using a quad makes the noise contribution the same as one. That's all the accuracy I can vouch due to my lack of access to the AP and doing that on the kitchen table.
Now I need to check headroom and THD.

*Actually it doesn't matter much, since the outputs are via PNP emitter-followers à la Schoeps. 2x47r resistors constitute the bulk of the output Z.
 
Very cool.
So just to clarify:  Are you saying that, after swapping out a possibly faulty LSK389,  you ended up with 2 X LSK389 (2 pairs in parallel) ?  I'm guessing you mean that a 2nd pair cancelled out the circa 3dB noise increase of the diff pair to get you back to the 1nV/rtHz of a single transistor yes?

 
Winston O'Boogie said:
Very cool.
So just to clarify:  Are you saying that, after swapping out a possibly faulty LSK389,  you ended up with 2 X LSK389 (2 pairs in parallel) ?
  That was the original plan from the start.

  I'm guessing you mean that a 2nd pair cancelled out the circa 3dB noise increase of the diff pair to get you back to the 1nV/rtHz of a single transistor yes?
Correct. Actually there's a small noise contribution from the source resistors (68r in each) so the actual source Z seen by the FET's is 268r. That's why I say the measurements are too good to be true.
 
Checked headroom and THD.
Max input at onset of clipping is -11dBu. It should withstand the kick drum mic (hopefully).
THD is not terrific with about 0.1 to 0.2% for output level between +4 and +15dBu. Goes down with level but hard to measure cause noise interferes.
After all it's not so bad for a circuit with so little NFB. Doing what I can to comfort myself...  :D
 
abbey road d enfer said:
Checked headroom and THD...

...After all it's not so bad for a circuit with so little NFB. Doing what I can to comfort myself...  :D

Those specs are very good, there's no need for comfort and you know it  ;)
I expect  clipping is a very graceful affair and limited to  quite benign low orders, and you could also argue that if output levels are anywhere near +4 to +15dBu then there's not much need for an additional gain stage anyway. 
Thanks for the details  :)

 
Hey I built this! I previously had designed a very simple passive microphone splitter,  and included this schematic on the transformer isolated output to compensate for the signal loss when splitting. Works great! Maybe a little too good! My signals are now VERY hot going to my preamps when I do live recordings.
Here's a link to the passive splitter I built a while ago.
https://groupdiy.com/index.php?topic=68449.msg870866#msg870866

Please ignore my bad soldering =)
In the picture, the top board is my old splitter design, and the bottom one is the new one with the small amplifier in it.
 

Attachments

  • 2020-01-26 10.58.43.jpg
    2020-01-26 10.58.43.jpg
    2 MB · Views: 129
Whoops said:
66eJRsY.jpg

Yikes. That's a bit disturbing for a product that costs $100. ($69 when this thread was new-ish.) Watching the Mike Delgado video (Booth Junkie on YouTube) where he compares the FetHead to the CloudLifter, you can hear more noise from the Fethead, through YouTube, even without headphones. (!)

I'm thinking Artur Fisher's Bumblebee Ribbon Booster might be a smarter place to direct my funds.

I was on the cusp of buying  a FetHead, so thanks for the internal pics.
 
kato said:
I was on the cusp of buying  a FetHead, so thanks for the internal pics.

All the circuits will be simple, even if some more than others

kato said:
I'm thinking Artur Fisher's Bumblebee Ribbon Booster might be a smarter place to direct my funds.

€149.00 excl. VAT, seems pretty expensive.

Why don't you use this thread and others that exist here in GroupDiy and build a few units for yourself for almost nothing?
 
kato said:
Yikes. That's a bit disturbing for a product that costs $100. 
That's the reality of business. A manufacturer typically operates on a mark-up of 4; that means that the product is sold to distributors at 4 times the price of raw components. The Fethead is sold ex-works at £35 ($45). $11 seems to be correct for component cost.
Then the distributor sells to a dealer, taking about 30% margin, which means multiplying by 1.43. So the dealer pays $64, and he sells to the public at $99.99, making about 35% margin.

  he compares the FetHead to the CloudLifter, you can hear more noise from the Fethead, through YouTube, even without headphones. (!)
I have a Fethead, I didn't find it particularly noisy, although I found that the noise dependance on source impedance was strange. IMO the main issue with Fethead is that its performance varies too much with the impedance of the preamp it is connected to.

 
Whoops said:
All the circuits will be simple, even if some more than others

Agree. Four fets might seem like there's more to the circuit. But it's more the build quality that's offputting, as well as the noise comparisons on YouTube. I've read two accounts of ones that have gone dead. My perception is these haven't been engineered not to fail.

Whoops said:
€149.00 excl. VAT, seems pretty expensive.

Yeah. I'm that kind of consumer. I'd prefer not to buy anything at all if possible.
If I feel like I need an item, I don't want it to be landfill fodder. I'll pay to support the DIY community as well. I'm leaning toward not buying one because I only have one mic with low sensitivity. And my preamps are pretty decent.

Whoops said:
Why don't you use this thread and others that exist here in GroupDiy and build a few units for yourself for almost nothing?

Now you're speaking my language.  ;D
That C-lifter clone in the other thread looks pretty simple. And I probably have all the parts on-hand already.  :) :) :)
Uh, stay tuned. 

I'd love to see a PCB layout if anyone's designed one. Otherwise, I can perfboard it.
 
abbey road d enfer said:
That's the reality of business....

I think you're explaining why we DIY.  ;)  Or perhaps why business is not my calling.

What do you think of the KlarkTeknik MIC BOOSTER CT 1. At $29, it's undercutting the spork out of Cloudlifter, FetHead, Dynamite, Launcher,  SS-1, Radial McBoost, Cathedral Pipes, etc.

What do you suppose is inside? 16¢ worth of SMD components? Or the PRR circuit from 2005.  :p

I dig the inline form factor of the Fethead and the CT1.

 

Attachments

  • KT.jpg
    KT.jpg
    5 KB · Views: 34
kato said:
What do you think of the KlarkTeknik MIC BOOSTER CT 1. At $29, it's undercutting the spork out of Cloudlifter, FetHead, Dynamite, Launcher,  SS-1, Radial McBoost, Cathedral Pipes, etc.

What do you suppose is inside? 16¢ worth of SMD components? Or the PRR circuit from 2005.  :p

I dig the inline form factor of the Fethead and the CT1.

Well it's a Behringer product, probably a clone of something around, released for an unbeatable price.
Might be good

Behringer stuff works,  there will be smd for sure, they want to sell it for cheap but are not dumb, they still make it profitable so yes SMD
 
Whoops said:
Well it's a Behringer product, probably a clone of something around, released for an unbeatable price.
Might be good

Behringer stuff works,  there will be smd for sure, they want to sell it for cheap but are not dumb, they still make it profitable so yes SMD

Whoa! That's a Behringer product!!?!  Thanks for the tip. I was unaware of the acquisition.

That makes up my mind. I have a "No Behringer equipment" policy in my studio.    ;)
 

Latest posts

Back
Top