Beesneez Lulu Fet - KM84 SMD mod

[GraemeWoller]

Part One: The Project

I have a pair of beautiful Beesneez Lulu FET small diaphragm condensers.

They sound fantastic on All Of The Things. This thing? Fantastic. That thing? Fantastic. All Of The Things sound fantastic with them. I can't say the same for a KM184 that I put them up against when I had the chance… The poor Neumann couldn't keep up. The Lulus laughed it out of the room.

I'm guessing it's that great capsule Ben made, the Cinemag, and the "relatively" simple KM84 circuit powering it all. I suppose there's a whole bunch more tone in that than the transformerless design that's going on in the KM184.

I'd not really consider myself a proper DIYer, if I'm honest with myself, but I love to learn.

The only other thing I've put together since I was an early teenager has been an Audio Sector chip amp kit back in 2010. That was fun, but I didn't really have to design anything. I just had to solder things together and poke it in a case. (I'd recommend a chip amp to anyone, btw. Bloody nice sound, right there.)

I started wondering to myself what it would be like to change out a few components and tinker away, hopefully without buggering up the great work Ben and his crew had done when he made them. What effect would that have on the tone of the mic?

To cut a long, probably boring, story short, I decided that if I was going to do "something" then I might as well go hard and learn a few new skills to add to my quiver and see what I come up with.

So I present, over a couple of posts, the story of a custom KM84 based SMD PCB replacement that I tinkered up. (I'll upload the Eagle files when I get a chance if anyone is interested).

I'm super-interested to hear what everyone thinks, especially considering the polarising discussions of through-hole vs. SMD!!

- Edited to avoid confusion about why I'm modding already amazing mics -

 

[micaddict]

Ben might chime in, too. 

 

[GraemeWoller]

Ben might chime in, too. 

That would be lovely. I didn't even think to ask what his thoughts were about it, he would've been the first best sounding board given that he is the guru!! He might think I'm an idiot using SMDs!

 

[GraemeWoller]

Part Two: The Circuit

Finding out the Lulu FET was based on the KM84 gave me a great starting point. I didn't have a clue initially… I just thought they sounded really nice, so I wasn't all that concerned about whether they were based on something else or not.

I can read a circuit diagram just enough to get myself in trouble but looking at a board and drawing up the schematic would've taken yonks to sort out by myself and matching an official KM84 circuit diagram from Google to the Lulu PCB made life SOOOO much easier when I was starting.

Like I said in the first post, I've never done this before. I made a circuit board when I was about 13 with a resist pen and some etching stuff from Dick Smith Electronics and that looked sh*t, so I didn't do any after that. The only other thing was the chip amp build, and that's it.

This project?

MILES away from that for me. Lightyears.

I got myself the free version of Eagle (https://www.autodesk.com/products/eagle/free-download) and launched into putting the schematic into digital form.

I STILL don't really understand Eagle…

After using it for this project I STILL don't get why there isn't a universal tool for some of the things. I couldn't get ANY of the keystrokes to work, ever. That slowed me down so much. I got there, obviously, but man I found Eagle a pain to learn anything.

I watched a bunch of tutorials and learnt about the combo move of Device, Package, and Symbol. Really handy when you're planning to make a PCB later on.

I learnt how to add a device, connect it with a "net", replace the device with the right one, replace the device with the other right one…

I noticed that while most of the components were identical the Lulu had a couple of value changes with capacitors. Only a couple, but I decided to stick to Ben's choices instead of the KM84 choices. (I actually don't really know what difference it should make. That's the next thing I need to learn.)

Then when I had it all in there and I started my first attempt at the PCB layout I thought…

"This would be MUCH easier to layout with SMDs rather than through-hole".

"Can that even be done at home?!"

Yes, it can. With a hacked toaster oven or a frying pan.

 

[GraemeWoller]

My Eagle diagram.

 

[micaddict]

Without getting too technical, I'll add that LDCs have lower noise than SDCs.

Also, the modern. trannyless KM184 has lower noise than the older KM84.

Finally, if the output is relatively low, you need a quality preamp that doesn't get noisy at higher gain settings.

 

[GraemeWoller]

Without getting too technical, I'll add that LDCs have lower noise than SDCs.

Also, the modern. trannyless KM184 has lower noise than the older KM84.

Finally, if the output is relatively low, you need a quality preamp that doesn't get noisy at higher gain settings.

That's interesting, feel free to get technical. Is it due to the smaller diaphragm naturally having a lower output and the amp section having to make up for that?

 

[GraemeWoller]

Part Three: The Eagle and the Mouser

When I discovered that people make SMD based boards at home… both etching and "reflow" (that was my new word of the day) it gave me a chance to reassess what I was going to do with the tinker. I didn't know very much about through-hole components really, but it was exciting to shift focus to SMDs and try something I knew basically NOTHING about.

I did a bit of reading up about SMDs in audio and came across some interesting polarised discussion about the pros and cons of the big bits verses the small bits. There were good arguments for both, "well argued arguments" may be a better way of putting it, but I thought at least I can give it a pretty good shot and if I don't like it I can easily swap back to the original boards and/or try a different version. I'm certainly not losing anything by giving it a burl.

I settled on a reasonably large package size, 3216, because I thought I'd probably have a bit more success handling something a bit bigger for my first project, and I tried to only get components that were those dimensions.

The other advantage about going tiny was that I could place the bias trimmer on the top of the board with everything else and keep it all tidy.

Although I felt like getting to grips with Eagle was a bit tedious I did appreciate the huge amount of package options when it came to turning my schematic into a PCB layout. There are heaps! They're often hard to find, (they need a cleaner, smarter, search system in my opinion), but there's a lot there when you start to become acclimatised to the libraries.

At this point I began going back and forth between Eagle and a parts list I was building up on Mouser.com. I swapped out components heaps, and for all sorts of reasons. Most of the time it was just that I didn't have a clue what I was doing. When you don't have any experience you find yourself doing quite repetitive things on a regular basis.

I reasoned that while I didn't know all that much about components then "following the instructions", as such, should get me close to a reasonable result. If a through-hole tantalum is 4.7µf and 25v then looking for the same specs in a SMD should get me most of the way there as long as I chose a respectable manufacturer. As it turned out most of my resistors and capacitors came either from Kemet or Vishay except for the 1gΩ resistors which were from Ohmite.

At this point I think it's worth noting that I'd love any feedback you guys and girls have about EVERY aspect of this project. Eg. "Vishay is fine, but Kemet sucks because the quality is patchy and the CEO eats babies" would be good to know.

The next challenge was to take the packages and try to sensibly lay them out on the PCB. I worked out to make it the right size… I worked out how to put cutouts in the board (hoping that the manufacturer would do that for me ). I aimed to get it "readable" as much as I could. I wanted to have simple lines to follow and I knew I'd need it with basic skills like mine. I think I managed to get a pretty tidy layout really. Sensible, I hope, and easy for anyone to pickup and understand. There are only a couple of vias going through to the otherside where I couldn't avoid it and I reckon both sides look pretty reasonable.

When I had it ready, and checked as well as I thought I could, I sent it off to https://www.seeedstudio.com to get a bunch done with their $5 deal. I couldn't make them for that, even with shipping included, and I would've buggered up a whole bunch working out how to etch them. I made a couple of mistakes in V1, as it turns out, but I'll talk about that next.

Here's a photo of the finished V1 board, followed by the .sch and .brd files.

 

[GraemeWoller]

The schematic:

 

[GraemeWoller]

And the board:

 

[micaddict]

Is it due to the smaller diaphragm naturally having a lower output

Yup.

What preamp are you using?

And, if I may say so, Ben is a very experienced mic builder and you say you're pretty new to this. I personally would think twice before I'd make alterations. These can also change tone BTW. And, as you've found out, tone is where it's at, especially with these, but regardless.

 

[GraemeWoller]

Yup.

What preamp are you using?

And, if I may say so, Ben is a very experienced mic builder and you say you're pretty new to this. I would think twice before you make alterations. These can also change tone BTW. And, as you've found out, tone is where it's at, especially with these, but regardless.

I'm using a Focusrite ISA Two.

Oh, I completely agree. Ben is REALLY good at what he does, and I think I'm going to have to clarify in my first post incase I offend someone.

I'm not just changing things because I think Ben's design is lacking, far from it, I own an Isobel as well and that sounds brilliant too. Ben is The Guru, from my perspective.

The noise floor didn't make me think it was shitty, it made me think, "I want to know more about how this all works, so I might as well have a fiddle and see what I can learn, and see where I end up".

I'd much rather not work in a vacuum on things, so I thought I might share what I've been up to. Those who know more can make suggestions, those who know less might be able to learn from my mistakes and findings.

DIY is supposed to be fun, so I'm trying to have fun learning by standing on the shoulders of giants. And with the Lulu to work with, the shoulders are indeed giant.

Also, I'm pretty tidy and careful.

 

[micaddict]

Man, you're fast. 
I edited my post within a minute after posting.

 

[GraemeWoller]

Man, you're fast. 
I edited my post within a minute after posting.

He he, "Mr. Speedy". That's what all the girls call me!

 

[micaddict]

Poor girls. 

 

[GraemeWoller]

Poor girls. 

Ha! Yeah, best to get good with one's hands.

 

[micaddict]

Oops, let get this back on track.

 

[Khron]

I've been thinking about KM84-ifying my t.Bone SC140's for a while. I've shoehorned in a slightly modded Oktava MK012 circuit into one of them a while ago - one noticeable thing was the way lower output.

***

The "de-railing" reminded me of this AvE fellow's videos over on YouTube. "Connoisseurs" will know what i mean ;D

 

[GraemeWoller]

I've been thinking about KM84-ifying my t.Bone SC140's for a while. I've shoehorned in a slightly modded Oktava MK012 circuit into one of them a while ago - one noticeable thing was the way lower output.

***

The "de-railing" reminded me of this AvE fellow's videos over on YouTube. "Connoisseurs" will know what i mean ;D

Hmm, interesting. I've noticed a lower output too, I've thought it might be the fet bias?

 

[Khron]

In the case of the KM84 circuit, it may well be - the gain of the FET is determined by (and wiser minds, correct me if i'm wrong) the ratio between the drain and source resistors.

With the fixed drain resistor, the variable source resistor will affect that, in addition to altering the DC bias points. That's actually how i implemented a ~10dB pad, in my modified-Chinese-Schoeps design for LDCs (BM800 / sE2200A).