FET mic designs and germanium?

[mateus]

Heyas!

I've been building a few stompboxes to improve the number of tools in my studio while learning and practing the whole DIY process before trying my luck with new projects, all been going great, having fun/hate and learning a LOT, increasing complexity and introducting new components step-by-step.

The last few weeks I've been playing around with building saturation effects, mostly for guitars, bass and keys, trying different saturation devices and combos, testing and testing. I came to a conclusion that seems to be "kind" of usual, when it comes to saturation, and amplifying something 'till it clips, I found that I like it better germanium transistors to silicon transistors, at least with all the different models I've tested.
Yeah, my testing conditions ain't no lab test, it consists in reamping stuff to the stompbox and micing it back. I know that all my chain will be doing their thing in the resulting audio but one thing remained the same everytime, I like how germanium transistors sound.

Short story, is it possible/worth the effort designing a FET mic around germanium transistor?

Sorry if this sounds dumb or redundant.

Cheers!

 

[Sorr]

A jfet is used because of its high input impedance it doesn't load down the capsule. If I was going down that path. I would use a jfet for the front end and directly DC connect a germanium from the fet drain ( which normally goes via capacitor to the output Tx ) as a germanium emitter follower.
However due to the nature of germanium transistors (and due to scarcity, limited choices) you might run into noise problems and base current leakage.
For noise you just have to go through them and find the least noisy one. May not be suitable on vocals but you could get away with it with loud sources such as electric guitars. You might want to try changing from a DC coupled emitter follower to a capacitor coupled one.
I would built just the prototype amplifier on vero board, with sockets for the transistors. So you plug them in. And have a trimpot on the fet source. This needs to be adjusted for the correct drain voltage as these fets used in Mics vary a lot. And on the germanium's emitter. So can adjust values easier.
After you have optimized it which involves listening to the noise. and you are happy with it. Then install it into a mic.

 

[Bo Deadly]

Germanium transistors are generally not suitable for low noise, low distortion, full bandwidth circuits. It sounds like it would be a cool thing to put into a circuit but in practice they're noisy, highly sensitive to temperature, leaky, have limited bandwidth and highly variable gain between parts.

They are potentially good for deliberately adding distortion to a circuit. But in that case you need to create a high gain, temperature compensated circuit and pick a transistor that works well in the circuit. You want the Germanium transistor to be the part of the circuit that clips and thus creates the distortion. And in practice it's probably not going to sound a whole lot different from a Silicon transistor unless you find one that actually has limited bandwidth at high gain. Then it will clip relatively softly compared to a Silicon transistor.

So I would design the circuit around the specific transistor. Specifically, I might use a high impedance current source (or sink since Germanium are usually PNP) as the load to get as much gain out of the thing as possible (but try transistors with different gain - a medium gain part would probably sound better). Then I would use a small coupling cap and series "grid stopper" to deliberately caused Vbe rectification to charge up the coupling cap and shift the bias when it's overdriven. These sort of things might create subtle tonal effects. You might actually hear the effect of it being Germanium. Meaning do things that push the characteristics specific to Germanium transistors. Otherwise there's no point in using a Germanium transistor and you might as well just use a conventional Silicon circuit.

 

[Gus]

I would be worth building to find out if it sounds "good" or is a good special effect sound.

To repeat somethings and add to the above posts

The question is what type of circuit to build

You will need a high input resistance follower or gain stage using a JFET or other devices.

Do you want follower/gain to Ge buffer(emitter follower (this might be the most temp stable))

Or follower/gain to a circuit like the big muff distortion stages with Ge diodes instead of the Si

OR

replace the PNP Si emitter follower in the circuits like in the MXL 2001 v67 and other microphones You will need to change a resistor for the biasing or direct couple or add another resistor for voltage divider bias

Or

one of the early AKG 414 eb etc circuits

There are other circuits to try

Now part of the Ge sound is limited bandwidth you can try a med power Si transistor

FWIW I post as Gus at diystompboxes and I like to use Si instead of Ge in distortion fuzz effects.

I have built treble boosters and FF like circuits with Si med power transistors and people seem to like the "sound"
the link is to a Si version(note the operating points of the transistors) of an older 3 Ge transistor fuzz. IMO Si can be fun to work with. Ben also posted a video of a build in the thread
https://www.diystompboxes.com/smfforum/index.php?topic=114210.0

 

[mateus]

Wow! Thanks a lot guys! For now this is far away for my super limited knowledge about electronics but it's trully inspiring and intriguing, again thanks a true lot! I'll keep studying and trying to learn as much as possible 'till I can understand and try something like this.

Now part of the Ge sound is limited bandwidth you can try a med power Si transistor

FWIW I post as Gus at diystompboxes and I like to use Si instead of Ge in distortion fuzz effects.

I have built treble boosters and FF like circuits with Si med power transistors and people seem to like the "sound"
the link is to a Si version(note the operating points of the transistors) of an older 3 Ge transistor fuzz. IMO Si can be fun to work with. Ben also posted a video of a build in the thread

Heya Gus, yup, I've seen some of your posts talking about this and have them printed =)
I just recently started building fuzzes and other "kind of" fuzz effects, so far so good, got one Muff built, working and sounding great, super silent, a Fuzz Face, also working and sounding great, also super silent, and a Rangemaster, this one not sounding so great but I think I have cooked the germanium transistor, or that one is somehow faulty =/

I just got some old power transistors here, like this one:



And some cool others, along a bunch of other components, from an old destroyed radio transmitter from the 70's.

 

[Whoops]

HI came to a conclusion that seems to be "kind" of usual, when it comes to saturation, and amplifying something 'till it clips, I found that I like it better germanium transistors to silicon transistors


Short story, is it possible/worth the effort designing a FET mic around germanium transistor?

Depends on what you expect from the Mic.
You like how Germanium transistors distort compared to Silicon, I'm also a big fan.
But do you want your mic to saturate? Do you want the sound of the mic to sound like a Fuzz pedal?

If that's the goal, maybe do a more dirty mic then you already know that you like Germanium so go for it.

But if you want a mic to sound clean, have good headroom, be stable at different temperatures then the FET mic circuits already available do that very well.

 

[mateus]

Depends on what you expect from the Mic.
You like how Germanium transistors distort compared to Silicon, I'm also a big fan.
But do you want your mic to saturate? Do you want the sound of the mic to sound like a Fuzz pedal?

If that's the goal, maybe do a more dirty mic then you already know that you like Germanium so go for it.

But if you want a mic to sound clean, have good headroom, be stable at different temperatures then the FET mic circuits already available do that very well.

Heya!

Yup, lots and lots of times I've seen myself adding all kinds of dirty nasty things to mic'ed tracks, from drums to vocals. This kind of pattern behaviour have made me always look at all sorts of dirty things in a very sexy way
Drums? Yeah, distorted snares and kicks all the time.
Rooms? Squash and grit!
Guitars? Balls!
Bass? Destroy it!
Vocals... Guess what...

That's why that idea got me thinking.

Cheers!

 

[mateus]

Guys, what kind of impedance a germanium transistor like receiving?

 

[Bo Deadly]

It depends. The impedance of a bipolar transistor is not linear. If it is biased low, the impedance can be very high. If the bias is high, the impedance can be very low. The impedance can vary as the signal swings higher and lower depending on the impedance at the emitter, the gain, the temperature, etc.

But very generally, the optimal source impedance (OSI) of a bipolar resistor is going to be in the 1K to 10K range.

 

[mateus]

I was thinking about trying my luck with a ribbon transducer and a small transformer for testing. I learned how to make corrugation and work with 1.8-2.5u ribbons and I can work around some ferrite trafos, I know I know, but I have been doing some sweet sounding ribbons with ferrite cores, nice color and saturation, so maybe a ribbon + trafo (step-up to around 1k) then germanium?

 

[Bo Deadly]

Sure. Whatever floats your boat. But again, Germanium are pretty noisy. A ribbon requires a LOT of gain. Possibly 70dB+. So that's not an ideal combination. But when you're just getting started in DIY EE, what you're building matters less than what you learn from trying.

 

[Whoops]

Like Bo Deadly said a Ribbon mic requires a lot of gain, that gain is normally from the output transformer and the Mic Preamp circuit that follows it.

If you want to use Germanium transistors why don't yo do a Germanium Mic Preamp?
That would be a good implementation for the Germanium transistors, there's some nice Neve circuits using Germanium as the early Neve modules were Germanium based, you also have the first version of the Helios mic amp circuit that was Germanium based.

check this thread for the Neve Germanium circuits, there's schematics over there:
https://groupdiy.com/threads/diy-dual-germanium-mic-pre-eq-based-on-neve-early-designs.18356/
Also this memeber had pcb's for sale:
https://groupdiy.com/threads/feeler-neve-germanium-pcbs.63680/

 

[mateus]

Cool! This is a great idea!! I like it a lot!
Too difficult to get the trafos for these?

Maybe I could etch my own boards?

Cheers!

 

[Whoops]

Maybe I could etch my own boards?

If there's cards already made by Salomonander no need to etch it, just ask him if he still has some spares.
But if you need to etch them it will be easy to do.

As for the transformers I don't remember now what might work, but I'm pretty sure you will find transformers from Carnhill, Sowter, Jensen or Cinemag that will do the job

 

[mateus]

If there's cards already made by Salomonander no need to etch it, just ask him if he still has some spares.

Cool! Sent him a message yesterday but since the thread was from couple years ago wasn't sure if he was still around.

As for the transformers I don't remember now what might work, but I'm pretty sure you will find transformers from Carnhill, Sowter, Jensen or Cinemag that will do the job

The folks at Cinemag have always been very kind and helpful to me, gonna check wich ones they can recommend me.

Thanks again man!

Cheers!

 

[Whoops]

read the thread I posted there's some good info there, you can ask about possible transformers over there also

 

[mateus]

read the thread I posted there's some good info there, you can ask about possible transformers over there also

Yup, already did and downloaded all that I could.