D-EF47 Tribute To Oliver Archut U47 Build Thread.

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Eagletone PSU
EAGLETONE+CM90-2.JPG
 
Anyone experiment with other R3/R4 values?

I read somewhere - posibly on this forum - that part of the U47 character (-ve FB?) is the 29 ohm kathode load.  Oliver's EF-80x design puts R3//R4 for an effective audio kathode load of 22.5ohms.

To keep the same 29:129 ratio for bias, and have an R3//R4 value of 29, I calculated R3=37.41 and R4=129.

I plan on trying different values in the build I'm working on but my results (as most things microphone) will be very subjective.
 
Hey All,

Just posing my finish OA Tribute. It's the first DIY mic I have done but with all the info here on the forum I found it easy enough to get right the first time!

D-7 Capsule
T47 Transformer
Telefunken EF800

Best,
Will J
 

Attachments

  • OA Tribute Complete.jpg
    OA Tribute Complete.jpg
    1.7 MB · Views: 101
Built this circuit into one of Dany's bodies so that the pattern is variable when in the omni setting. R8 and R7 are both 3M and the B+ is 126V.  I started with R6=68k because if I had the same anode current as Dany, I'd hit the voltage drops I was looking for. Using Dany's measured voltages from his schematic as a guide I knew I was way off with the first of my two EF802s. I swapped tubes then began playing with R6 values. R6=47k gave me the best numbers and the best sound. I dropped R3 to 22 from 27 (to raise the anode current) and it sounded even better.

Since Oliver's design gives an AC kathode load of 22 ohms instead of the U47's 29 ohms I thought I'd try a different R4/R3 combination. I upped R4 to 160 then played with R3. Unfortunately, I didn't think of the smaller load drawing less H+ current and the voltage rising so two measurements have an unknown H+. Once I realized, I dropped H+ as far as it would go then added another droppping resistor to get back to 5.05.

The best sounding combination when R4 was 160, was R3=36 which coincindentally (or not) gives the same anode voltage and current as the best sounding combination when R4 was 100 (R3=22).

Tube 1
R6=68k, R4=100, R3=27, V@R5/R6=96.0V, Va=57.6V, Ia=0.384mA, H+=5.05

Tube 2
R6=47k, R4=100, R3=27, V@R5/R6=99.5V, Va=52.0V, Ia=0.475mA, H+=5.05
R6=56k, R4=100, R3=27, V@R5/R6=96.0V, Va=51.2V, Ia=0.448mA, H+=5.05
R6=68k, R4=100, R3=27, V@R5/R6=94.1V, Va=50.9V, Ia=0.432mA, H+=5.05
R6=82k, R4=100, R3=27, V@R5/R6=91.0V, Va=50.5V, Ia=0.405mA, H+=5.05
R6=47k, R4=100, R3=22, V@R5/R6=99.0V, Va=49.0V, Ia=0.500mA, H+=5.05
R6=56k, R4=100, R3=22, V@R5/R6=96.0V, Va=48.6V, Ia=0.474mA, H+=5.05
R6=47k, R4=160, R3=36, V@R5/R6=99.0V, Va=49.5V, Ia=0.495mA, H+=? ??
R6=47k, R4=160, R3=30, V@R5/R6=98.5V, Va=48.0V, Ia=0.505mA, H+=? ??
R6=47k, R4=160, R3=27, V@R5/R6=98.5V, Va=47.5V, Ia=0.510mA, H+=5.45
R6=47k, R4=160, R3=27, V@R5/R6=98.2V, Va=47.5V, Ia=0.507mA, H+=5.25
R6=47k, R4=160, R3=27, V@R5/R6=98.8V, Va=47.6V, Ia=0.512mA, H+=5.05
R6=47k, R4=160, R3=36, V@R5/R6=99.2V, Va=49.2V, Ia=0.500mA, H+=5.05


Now since I'm doing this with only my voice and my ears, all of my opinions are extremely subjective (especially since it's sinus cold season here in Canada). I am mostly posting this info to show that there is much to be gained by experimenting with values and bias in this circuit. Especially when the two tubes I received from Christian W had very different performances, whatever tube you select may require different components to acheive similar results. With that in mind here are my subjective thoughts:

The R6=47k, R3=22 & R4=100 combo sounded very clean with a nice saturation in high end frequencies. The R6=47k, R3=36 & R4=160 combo sounded less clean and was more mid forward with a nice high-mid range saturation. For those looking for a clean amp, I'd suggest starting with R4=100. Some have criticized the EF800 design for being too "hi-fi" sounding; I think that R4=160 might provide a much less "hi-fi" sound.

I always pick whatever component provides the niced colour or saturation. I also worried that I might have a little confirmation bias since I was curious about bringing the kathode load up. With that in mind I struggled to decide between the two amp component combinations but ended up going with R4=160, R3=36 for now.

A note about capsules: I built it with Maiku K47 and it sounds great; a very accurate sounding capsule with a little high end sparkle. I bought a Beesneez K7 and put it in; the K7 was 4dB quieter (not sure why; 85pF vs Maiku's 77pF) but sounded a little flatter across all frequencies and had some of that "3D" sound that many desire. Overall I prefer the Beesneez K7 in this mic but wonder why the output is lower.

My plan for the Maiku is a U47 FET and hearing the capsule in this mic makes me that much more excited to build the 47FET.
 
You have changed the operating point of the tube. Your chart doesn't list cathode voltage or grid - cathode diff. You will obviously get grittier sound if the difference is too small. Also the EF80X is no VF14, they don't exactly operate similarly at the same operating points. Btw. I use diode instead of the R3 and the R4 in my version is connected to anode supply using much larger resistor, but the operating point is very close to this one.
 
mhelin said:
You have changed the operating point of the tube. Your chart doesn't list cathode voltage or grid - cathode diff. You will obviously get grittier sound if the difference is too small. Also the EF80X is no VF14, they don't exactly operate similarly at the same operating points. Btw. I use diode instead of the R3 and the R4 in my version is connected to anode supply using much larger resistor, but the operating point is very close to this one.

Grid should be 0V. I can't measure because my DMM will load the grid resistor so I just assume kathode to ground = kathode to grid. I should've included the kathode voltage but didn't because I messed up my notes and don't have kathode voltages for some. Here's what I have:

Tube 1
R6=68k, R4=100, R3=27, V@R5/R6=96.0V, Va=57.6V, Ia=0.384mA, H+=5.05, Vk=1.11V

Tube 2
R6=47k, R4=100, R3=27, V@R5/R6=99.5V, Va=52.0V, Ia=0.475mA, H+=5.05, Vk=1.11V
R6=56k, R4=100, R3=27, V@R5/R6=96.0V, Va=51.2V, Ia=0.448mA, H+=5.05, Vk=1.11V
R6=68k, R4=100, R3=27, V@R5/R6=94.1V, Va=50.9V, Ia=0.432mA, H+=5.05, Vk=1.11V
R6=82k, R4=100, R3=27, V@R5/R6=91.0V, Va=50.5V, Ia=0.405mA, H+=5.05, Vk=1.11V
R6=47k, R4=100, R3=22, V@R5/R6=99.0V, Va=49.0V, Ia=0.500mA, H+=5.05, Vk=0.904V
R6=56k, R4=100, R3=22, V@R5/R6=96.0V, Va=48.6V, Ia=0.474mA, H+=5.05, Vk=0.906V
R6=47k, R4=160, R3=36, V@R5/R6=99.0V, Va=49.5V, Ia=0.495mA, H+=? ??, Vk=1.003V
R6=47k, R4=160, R3=30, V@R5/R6=98.5V, Va=48.0V, Ia=0.505mA, H+=? ??, Vk=0.869V
R6=47k, R4=160, R3=27, V@R5/R6=98.5V, Va=47.5V, Ia=0.510mA, H+=5.45, Vk=0.805V
R6=47k, R4=160, R3=27, V@R5/R6=98.2V, Va=47.5V, Ia=0.507mA, H+=5.25
R6=47k, R4=160, R3=27, V@R5/R6=98.8V, Va=47.6V, Ia=0.512mA, H+=5.05
R6=47k, R4=160, R3=36, V@R5/R6=99.2V, Va=49.2V, Ia=0.500mA, H+=5.05


Changing the ratio of R3/R4 will change the operating point but changing the total value won't. When I kept R4 as 100 and dropped R3 to 22, it changed the operating point but it didn't make it sound more "gritty", it made it sound cleaner to me. What made the mic sound warmer was keeping the same R3/R4 ratio (and operating point) as R4=100 and R3=22 but upping the values. I didn't write down the kathode voltage for R3=36 and R4=160 but it should be very close to 0.904.

I haven't tried using a diode yet but that's certainly an option.
 
In U47 there is about 40mA current going through the 29 ohms cathode resistor, so a 1mA (or any practical) change in Ia moves the bias point only little. Now the 100 + 22 ohms connected to 5V heater supply makes about 41mA current which is about on par with the U47. If you increase the R3 and R4 to 36 and 160 ohms you will actually have only 26 mA current flowing through R3 in addition to cathode (anode) current, so with the input voltage the bias point  will not be as fixed anymore though still quite. R3 and R4 just don't make simple voltage divider, you have to count in the Ia through R3.  Anyway, you are free to experiments. In practical U47 or other mics the bias point may be (at least now as these mics are really old) well off and they may still sound good or even better than when new. It would be nice to have easily adjustable bias point in mics, sometimes a little bit saturation directly from mic would be a nice to have feature.
 
Thanks Mikko. Gives me a lot to think about. I'm sure liking the way this thing sounds, especially now that I have the Beesneez K7 in there. I may play with values now that I have a different capsule though I expect I'd favour the same values for the amp regardless of the capsule.
 
With much help from David (TLRT) i get my grounding issue sorted with my EF47 (and M49b too).
I want to share it to everyone, if you have problems with noise from grounding problems.
Connect the pins as in the schematic, the things I changed with help from David I attached in the picture.
If someone has further explanations, please share your opinion.
If you want to use Dan's D47 body, connect CRD with COM on PSU.

 
TillM said:
With much help from David (TLRT) i get my grounding issue sorted with my EF47 (and M49b too).
I want to share it to everyone, if you have problems with noise from grounding problems.
Connect the pins as in the schematic, the things I changed with help from David I attached in the picture.
If someone has further explanations, please share your opinion.
If you want to use Dan's D47 body, connect CRD with COM on PSU.


For grounding Dany's D47 body, I grounded the cable shield to the shield pin (sorry, can't remember which pin) then grounded that shield pin immediately to the mic body using one of the tapped holes for the heat sunk resistor. The 0V, I bonded to the mic body where Dany's drawing indicates.
 
I forgot to follow up on my B+=126.

I realized the simplest way to drop the voltage above R5 was to put a load in parallel with C3. This draws more current through R6 to create a bigger drop. I think I kept R6 as 30k and used ~150k in parallel with C3. This gave me roughly the same values as Dany with the largest difference being the tubes we selected.

I still intend to play with the bias when I place my next mouser order. For now, the mic sounds great and putting the switch in omni mode, I can adjust polar patterns from the power supply.
 
Hi everybody,
I have finished my version of the U47.
It works very fine, no noise, no shuffle, perfect!
I'm a little bit confused with the PSU settings.
To adjust  the H+ and B+ i connect the mic, i adjust the H+ to 5.05v and B+ to 105V, great, no problem.
Before closing the case, I say, what's happend if I switch from Cardio to Omni ?, I did it and control the voltage,  oups !! there is only 85v on B+ ???
What's wrong ?

Thank for your help, best regards from France
 
jcdignan said:
Hi everybody,
I have finished my version of the U47.
It works very fine, no noise, no shuffle, perfect!
I'm a little bit confused with the PSU settings.
To adjust  the H+ and B+ i connect the mic, i adjust the H+ to 5.05v and B+ to 105V, great, no problem.
Before closing the case, I say, what's happend if I switch from Cardio to Omni ?, I did it and control the voltage,  oups !! there is only 85v on B+ ???
What's wrong ?

Thank for your help, best regards from France

Maybe your relay has higher power consumtion , also check in PSU R9 and R5. (decrease value of R9)The relay is powered from B+ line,and this is the biggest drawback of the design - if your relay ever goes bad the omni switch will put approx 200V to your mic. Even with 2:3 divider that is 120V to your capsule.
 
Hi,
I have finished this project and the mic sounds great. BUT, there is this strange thing :
- when I turn on the psu, Then connecté the cable to the mic, it works.

- when I turn on the Psu with the mic already connected, it doesn’t work, even after few minutes. I need to remove the cable form the mic and to put it back again, and then it works.

Does anybody know what happens ?
 
Hi all,

First post here—I’ve loved reading this board as a lurker and am thrilled to be able to contribute. I’m about halfway through a D-EF47 build, and I have what is maybe a weird or naive question. I’m considering putting in a Neumann TLM-49 cardioid-only capsule, and using only the cardioid parts of the circuit. Does anyone know if this is feasible, and if so, what sort of adjustments I might need to make to the PSU and microphone circuits in order for this to work (beyond the single connection to the capsule)?
 
Oh, and here's a follow-up question: does anyone know if the TLM-49 capsule holds a candle to an actual K-47? Do you think that the one-sided version of the K-47 capsule is even worth putting in a build like this in comparison to all of the cool capsule options out there (Dany's D7, Thiersch, Beesneez, etc)?
 
wstratton said:
Hi all,

First post here—I’ve loved reading this board as a lurker and am thrilled to be able to contribute. I’m about halfway through a D-EF47 build, and I have what is maybe a weird or naive question. I’m considering putting in a Neumann TLM-49 cardioid-only capsule, and using only the cardioid parts of the circuit. Does anyone know if this is feasible, and if so, what sort of adjustments I might need to make to the PSU and microphone circuits in order for this to work (beyond the single connection to the capsule)?

Make a wire jumper on the PSU pcb between "COM" and "CRD" then you are in fixed cardioid mode, leave out the lorlin switch.
 

Latest posts

Back
Top