My Ribbon Microphone
Here is a repost of my article from gearslutz
PART 1 – CONSTRUCTION
Ok so i had a MXL 770 i stopped using ages ago, a Neve BA283AM module clone lying around, Sowter 8074 transformer and a very nice sounding NADY ribbon motor. A friend was telling me about the royer 122 with its impedance matching circuitry and suddenly it hit me.
Why not build a ribbon mic with its very own built in Neve preamplifier and see what happens…
This is subtantially different from the Royer design – He uses a transformer to bring up the voltage levels and then an impeadance converter to send it to teh preamp – My design is the NEVE 1073 preamp engine inside the microphone body using standard ribbon technology
This means than Royer’s design is designed to deliver the ribbon to the preamp with as little coloration as possible. In this case I love the Neve 1073 sound and I am looking for the coloration.
Anyway I built it last night and I have pictures and sound clips…
Step 1 Modify the neve amp.
Originally designed to be operated at 24V the quiecent current is something like 0.5ma which works well with phantom power. I also tested it and works well at 12V (output is limited at +/- 6V which is extreme headroom for mic applications )
here is the module after the changes:
Description of changes:
1. How to get a balanced output from an unbalanced amp? Look at Rod Elliots’s article Balanced Transmitter and Receiver II. The output is buffered with 150 ohms on both the hot and cold lines. Simulated in pspice the impeadance imbalace chanages with frequency but shouldnt be too high and CMRR rejection shoud remain at good levels (at least 60-80db). (After hooking up the rig without connecting the transformer the CMRR is good)
Added a resistor network to power the rig (see 6)
2. I removed the R20 resistor and reversed the output cap (and the impeadnace network once since it would be connected to 48V
3. Added the recommended feedback resistor based on the NEVE 1073 design
4. Replaced the 120K resistor that is there to load the cap to a lower value to load the transformer
5. replace the default gain resistor to 100ohm (more or less 30db)
6. Cap and zener to keep the voltage at 12V. originally I did 24V but there isnt enough current available the quiecent voltage with no regulation is about 18V so i decided to go for 12V – ample headroom to record an airplane. The final cap (underneath) is additional 470u. This gives enough power for 3-4 secodns if phantom is disconnected – enough for strong transients.
The transistors are 2n5089 with ultra high beta low noise (beta is 1200) which translates to extremely low noise floor.
This is the NADY motor – very nice motor. It seems NADY was the source for all the cheap ribbons out there – and the design was stolen in china from them. They really sound nice. Please read this A stirring story: EQ Mag’s “Chinese Connection” by Yours Truly – 3dB I feel sorry for them
The green wire is my attempt in adding humbucking to the motor. I removed it as it has no effect. Does anyone know how to improve the hum rejection of the motor? I have to direct it to find place where there is no hum pickup. I know it is the motor since I moved each component seperately and also tried with amp only, tranformer without ribbon and finally whole setup.
Ok now to next picture:
Hooked it up and worked first attempt!
Ok now to closing it up, note the sponge used to damp the cavity resonances – i stuffed more stuff in there after the pic :
And this is it ready to go!
When all said and done
1) MXL – $60 for the case
2) NEVE – u need to build one now… or buy one on ebay and saw it in half
3) NADY Motor – $70
4) Sowter Transformer – $80
not cheap but hey very solid design…
Test run in parallel into identical preamp – regular 5532 based pres. The mics are side by side but some level differences might occur dues to
- Ribbon proximity effect
- Pointing the instruments at one mic
- Different physics of the mics
Bass trumpet (valve trombone)
I am not a trumpet player but can dabble in trumpet if i need a few licks here and there. I wanted to test LOUD and also soft trumpet. trumpet can get very very lound I decide to do both soft playing and extreme loud to see if it can hold up.
Dumbek has all the frequencies in the world and places both exciting trebbles and bass note. It also has very spiked transients and I wanted to see how they differ. If u examine the waveforms you will see incredibally different fave forms! At the beginning I was sute the phases were reversed, but i think some of the difference has to do with figure 8 pickup reflections from the walls. I dabble in dumbek – I cant keep rhthm well on it but for this test we can get a sense of the sound…
The recorder has a shrill sound full of high overtones but need to remain sweet. Here is a unforgettable (ok full of mistakes) performance:
Here is my son on his small classical guitar with metal strings… I hate that instrument but its what we have around here…. Tried to tune it as much as I could. I had to record far (3-4 feet) from mic because of the pronounced proximity effect of the ribbon.
Bottom line – The ribbon has some real nice qualities. It sounds really good on lead trumpet players that can blow their head off and still sound great.
I can now take the Mic to any studio we record with and not worry if they don’t have good enough preamps.
The ribbon + NEVE exhibits interesting qualities I will test drive it over the next few weeks.
The following spectrum analysis of the signal says it call:
1) Ribbons have a flat response whereas condensers have an enhanced high range that give that “sparkle”
This a “Max” type FFT of guitar player the whole track. It simulates the actual curve of the mic when presented with a “pink noise” type signal. (I should really record a pink noise signal from a speaker )
You can see the high end boost in the condenser starting at 9-10K (414 Condenser is the Red line, Ribbon is blue line)
Same type of graph can be seen here with a single note analyze a 1kh B natural (Si) (Taken from the recorder flute track above).
Notice the 6th and 7th harmonic are more pronounced in the red graph (414) because of the added HF boost. This tends to make certain instruments harsh sounding.