OPEN SOURCE DIY Mic Project - ORS 87 - Stripped Down u87

[AdamFrisen]

More neophyte questions.

If I'm following correctly, I can power up the mic (Capsule disconnected) with My Focusrite Scarlet interface and feed the signal in REW on my computer. Simple enough...

The part I don't fully get is step 4: injecting a signal, what am I using to connect and the source of the signal? Am I using just 3.5mm audio output from my Mac and connecting wires directly into the circuit. I'm sure I have a couple sacrificial cables around I can appropriate for the effort.

Would one of you be so kind as to post a photo of the setup? My REW experience is primarily with a uMic and speaker testing so this is quite different.

eso

I learned how to use REW thru excellent videos I linked here:

https://groupdiy.com/threads/measuring-distortion-with-a-soundcard-and-software.88188/#post-1161365

 

[eso]

I learned how to use REW thru excellent videos I linked here:

https://groupdiy.com/threads/measuring-distortion-with-a-soundcard-and-software.88188/#post-1161365

Thanks Adam, that's a great resource for calibrating and basic measurements. I've been measuring loudspeakers with REW for a few years so I am familiar with some of the basic functions.

But this still doesn't tell (or show) the means of injecting the signal into the mic. For loudspeaker measurements I use audio out into an amp to power the drivers.

From Khron earlier in this thread: "Just cobble together a cable from a (volume-adjustable) line out... " This is what I want to clarify.

If I take a typical 3.5mm audio cable in the line out from my Mac, snip an end off and connect the + signal to the sgnl on the mic and the - to ground does this give me the connection I need for the measurements.

 

[Khron]

If I take a typical 3.5mm audio cable in the line out from my Mac, snip an end off and connect the + signal to the sgnl on the mic and the - to ground does this give me the connection I need for the measurements.

Do you not already have / use an external audio interface (where I assume the tested microphone plugs into)? What's wrong with the line outputs of that?

 

[cello4u]

I have the KSA board, and BOM. Trying to attribute each part to its place. No value for C3, C7 on schematic?

 

[Khron]

No value for C3, C7 on schematic?

At least in theory, C7 can be the same 22uF as the source bypass capacitor, or 1uF as the DC-blocking capacitor coming off the JFET's drain. C3's value is still to-be-determined, for a given high-pass frequency. Could try 100nF for starters, and go up or down, depending what you're after.

Sadly I'll be busy moving house during august, so I won't be able to tinker much with my projects, until we're done with that...

 

[eso]

Do you not already have / use an external audio interface (where I assume the tested microphone plugs into)? What's wrong with the line outputs of that?

Until I set up the O2r i'm using a Focusrite Scarlett 2i2, so yes that has 1/4" outputs. So I'd power the mic on one input and use the other output to inject the test signal?

And then again, referring to you "Cobble together a cable..." comment, I'm looking at connector in the output with wires directly to the signal connection in the mic and to ground in the mic.?

eso

 

[Khron]

So I'd power the mic on one input and use the other output to inject the test signal?

Yup - each if those is selectable in REW.

And then again, referring to you "Cobble together a cable..." comment, I'm looking at connector in the output with wires directly to the signal connection in the mic and to ground in the mic.?

Yep again

 

[AdamFrisen]

Thanks Adam, that's a great resource for calibrating and basic measurements. I've been measuring loudspeakers with REW for a few years so I am familiar with some of the basic functions.

But this still doesn't tell (or show) the means of injecting the signal into the mic. For loudspeaker measurements I use audio out into an amp to power the drivers.

From Khron earlier in this thread: "Just cobble together a cable from a (volume-adjustable) line out... " This is what I want to clarify.

If I take a typical 3.5mm audio cable in the line out from my Mac, snip an end off and connect the + signal to the sgnl on the mic and the - to ground does this give me the connection I need for the measurements.

I think this is minimum you need as long as you just want to get the drain bias voltage right using audio interface:

0. Calibrate your audio interface and get some bearing on the actual output voltage generating 1KHz sine from REW, measuring with voltmeter from one of it's line outputs you plan to use.

If your audio interface THD is measured to be <= -100dB, you can measure as DC voltage if you can't measure AC. About 1.1V DC should be 775mV AC (0dBu).

Use the Calibrate level-button in Generate function to set the actual voltage (in RMS). Test by changing the voltage to 100mV in REW and confirm your voltmeter shows 100mV RMS (about 141mV DC).

1. Connect transformer to the output so you get low voltage enough for the audio interface. If you complete the output to XLR then even easier as you can use an ordinary XLR cable.

2. Connect voltmeter to measure DC between drain and ground. It can be tricky clearing line of sight for trim pot if the trimpot has the screw is on the side. Next BOM I will order Bourns with the trim on the top (even it is slightly more expensive).

3a Make sure the phantom power is off on your audio interface
3b. Connect XLR cable from your mic build to microphone input of your audio interface
3c. Turn on the phantom power, the voltmeter should show the current drain voltage (like 12-14 V)
4. Turn the trimpot screw and confirm the change in the drain voltage

5 Turn off phantom power

6a. Connect line out you calibrated in step 0 to mic input (FD, BP)

6b. Turn on the phantom power. Wait until drain voltage gets stable.

7. Bring up Generate function dialog in REW. Make sure to set the voltage to about 100mV-200mV in the text box.
8. Start RTA function and click on record button in REW
9. Click on play in Generate function dialog

You should see a spike in RTA window at 1 KHz. If you see it you can continue.

10. Choose to see the distortion in RTA

11. Click on the top of the highest distortion harmony spike in RTA graph so you get a ref. line

12. turn the trimpot until you find lowest level on the 2nd harmony.

13. Check with little higher input voltage, say 200mV-300mV the THD is lowest even there. Otherwise adjust at this higher level and keep the lowest THD from this reading.

Or you can also use the Scope function in REW and increase the input voltage until the sine wave starts to clip top or bottom, then lower little and minimize the distortion.

 

[eso]

And circling back to Adam's links to the setup videos I need a cable I can loop for the output to the input to make a calibration file for the Scarlett for more accurate measurements too.

eso

 

[AdamFrisen]

And circling back to Adam's links to the setup videos I need a cable I can loop for the output to the input to make a calibration file for the Scarlett for more accurate measurements too.

eso

I think calibrating the audio interface is a good investment.

 

[eso]

Adam & Khron, tanks for the help!

eso

 

[AdamFrisen]

This ghetto-style measurement of microphone looks fun and maybe good enough if you wanna compare one mic to another.

 

[eso]

Rather than cobbling a cable together for setting bias this is an option:

https://www.amazon.com/dp/B08BHTNWNZ?psc=1&ref=ppx_yo2ov_dt_b_product_details

 

[untamedfrontier]

For my last test while I'm waiting for another capsule to arrive, I just used an iPad to output the 1khz tone and used alligator clips off of a 1/8" cable. Not sure if there are any downsides there, but seemed to work ok for me

 

[oongbac]

I've made pair of ORS-87. Used Homero's PCB and with 36v zener and LSK189. Thank you all who provided their great knowledges for this fun project.

According to @kingkorg 's advice, I attached 560R between R9(10k)and ground to inject test signal.

I tried to calibrate THD with REW, the lowest is 0.025% with 100mv @1kHz. And I've got 0.43% with 407mv @1khz.

Used 28dollars K67 style capsule from ebay. If Arienne Audio will release cardioid only K87 I will change with that.

 

[kingkorg]

I've made pair of ORS-87. Used Homero's PCB and with 36v zener and LSK189. Thank you all who provided their great knowledges for this fun project.

According to @kingkorg 's advice, I attached 560R between R9(10k)and ground to inject test signal.

I tried to calibrate THD with REW, the lowest is 0.025% with 100mv @1kHz. And I've got 0.43% with 407mv @1khz.

Used 28dollars K67 style capsule from ebay. If Arienne Audio will release cardioid only K87 I will change with that.

View attachment 133549View attachment 133550View attachment 133547View attachment 133548

Thanks for confirming this, i was wondering if i was right about that 0.5% ballpark. I still can't remember how i got to that number. What value did the trimmer end up on?

 

[oongbac]

Thanks for confirming this, i was wondering if i was right about that 0.5% ballpark. I still can't remember how i got to that number. What value did the trimmer end up on?

One is 1.78kohms, another is 2kohms. Just in case I'm saying this, I used UAD Apollo X6 for THD calibration. And it's minimum gain for microphone input is 10dB.

 

[AdamFrisen]

Thanks for confirming this, i was wondering if i was right about that 0.5% ballpark. I still can't remember how i got to that number. What value did the trimmer end up on?

I didn't dare to put in more than 400mV RMS when I tested with capsule still soldered on, but THD is then about 0.42% at -16dBu in @ 1KHz. The signal in should be fairly calibrated as I used a true RMS measuring voltmeter. Used HL-95 fully assembled body with Vamisound K87 capsule.




Using K47 flat capsule in a smaller body with modified circuit (no C6 and no C9) displays interesting difference in the 2nd harmony:



Peeking at FET847 and KM84 circuit for LDC, I wonder if I should put a 4 pF between gate and rain and if that would make any difference in THD (and gain)?

To my old ears, K47 flat and modified circuit sounds great. Little more bass than the U87 I built in the bigger body I built. Small but audible differences.

I followed that ghetto-style capsule measurement I posted video about earlier and used the K47 flat microphone as a reference and got this graph:


I need to get my test setup better but maybe showing U87 circuit with C6 as 150 pF, has more attenuation in the lower and higher areas compared to the flat circuit + k47 flat capsule.

PS.

I remember an audio course at university where a scientifically conducted comparison of hearing impairments between elderly men and women was presented. Men's hearing of higher frequencies deteriorates much faster than women's.

So if you want to get an opinion on how something you build sounds and you only have grandma and grandpa nearby to ask, choose grandma.

I should be able to find the papers somewhere. Maybe there is some new research with different findings.

DS.

 

[AdamFrisen]

I found the ref. microphone for Tact RCS 2.0 with calibration file (don't know if it still working) and compared to that, the graphs looks like this. I used psychoacoustic filter in REW as smoothing.

 

[Voyager10]

I didn't dare to put in more than 400mV RMS when I tested with capsule still soldered on,

The capsule will have 40V or so across it in normal operation, and will generally tolerate 60V before you need to worry, so a few volts here isn't a problem.

Peeking at FET847 and KM84 circuit for LDC, I wonder if I should put a 4 pF between gate and rain and if that would make any difference in THD (and gain)?

The original U87 schematic (this post) has a 10pF cap C3 hidden away, which isn't in the ORS87 circuit.

The capacitor will reduce gain - it adds negative feedback - but also reduces THD. (The FET"s internal drain-to-gate capacitance varies with voltage, which is a non-linearity which causes distortion. Adding a fixed capacitor in parallel reduces the relative effect of this).

The actual value is 'select on test', as there will be a few pF of parasitic capacitance there already, particularly if stripboard is involved.