TROUBLESHOOT: White noise in D-251 build

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

SparkleBear

Well-known member
GDIY Supporter
Joined
Jun 21, 2016
Messages
308
Location
Santa Cruz, CA
Greetings All,

I've been on and off trying to track down the white noise in my D-251 build for about a year now. I have noise tested the 6072a and its super quiet, Replaced the CT12 capsule, de-soldered the whole thing, cleaned it, re-soldered it up again, changed the output caps, twisted the audio wires in the power supply and kept them from the filter caps, tried filtering the B+.... all to no avail. The white noise makes it pretty un-usable for any voiceover/vocal or acoustic guitar application as the hiss is too noticeable.  I am going to post a few sample recordings, perhaps anyone has any ideas. Listen to the lead in before the guitar comes in... I am posting a recording with my D251 and next with my MK Audio U67. The 67 is dead quiet in the lead in...

251 link: https://www.dropbox.com/s/doga2sonunml4u6/Will%20She%20Be%20Gone%20%28251%29.mp3?dl=0

u67 link: https://www.dropbox.com/s/1dmqcloash2avqx/Will%20She%20Be%20Gone%20%28u67%20no%20filters%29.mp3?dl=0

Ideas Anyone?

I am out of them... Perhaps there is a qualified technician who may be able to troubleshoot this on their bench for a fee? Anyone out there servicing DIY microphones? Any help is so appreciated.

Thanks a ton!
 
I posted in the other thread for you...I know what its like hoping someone will answer when everyone is busy  so I wanted to let you know. Again, I don't know if it'll help. but thought I would share
 
When you said you "cleaned the thing" are you referring to the capsule or the board?

The high impedance section is very susceptible to oils from handling as well as flux from soldering.  If you haven't given your board(s) a good scrub with Isopropyl Alcohol you may want to do that.  Be careful if you have any Styroflex capacitors on the board as they are sensitive to the alcohol (may want to remove them before cleaning and re-install after being very careful to keep things clean.

I've built a D-251 and mine is dead quiet, so it can be done (more luck than skill on my part I'd guess).

 
Can you post some pictures? Take a snapshot of the high impedance section if you can.

Listen to the white noise on headphones and shut off the power supply. Do you hear the white noise as the sound is dying?
 
I could repair it for you, but the distance is probably too long...
So lets try figure out something here ;)
Good pics of your build is a must if you want to find any answers.

What output transformer did you used? Haufe? ami?
Are you sure that you connected it properly?

What are the values of high impedance resistors?

Is the CT12 backplates are disconnected  from eachother? You can check with "beeper" - continuity tester- in your multimeter.

If the backplates are disconnected from eachother, then desolder back diaphragm and related to it half of backplate - set mike for cardioid and check is it still noisy.

 
For what it's worth, i had a potentially similar problem the other day. Granted, with chinese CK12-type capsules, but it might be relevant.

These ones don't have the backplates exposed on the sides, and the screw(s) for the backplate wire(s) need to be long enough to reach all the way to the bottom of the threaded holes, past a good 3-4mm or so of plastic insulator on the sides.

Initially, i used some shorter screws that did hold the capsule tight enough in the saddle, but didn't reach deep enough to make contact with the backplate. Tapping the mic body did produce some signal, but no reaction to sound, and all in a wash of white noise. After employing some long-enough screws, noise was gone and sound did get picked up as expected.

No idea if this might be the case, or useful, but i just thought i'd put it out there :)
 
For what it's worth, I have had the original capsule returned to me and replaced it with another.  It is possible for debris between the backplates to cause a similar noise in a 251 circuit. I am quite aware of how problematic this can be and try to be as concientious as possible concerning cleaning and handling of my capsules resulting in a very low failure rate in this regard.

The screws I supply for the backplate connections are 6mm long and so extend about 1.5mm into the brass backplate.

What about the transformer?
 
I had a similar problem.
The backplate connection is very close to the brass tensioning ring.
This means that any contamination or high humidity (condensation) between these points will cause a very small leak current.
After good cleaning and applying conformal coating around the backplate connection, this problem never occurred agian.
 
Hi,
it is a 251 clone from poctop?
From all advice and experience given before it seems related either from the out transformer either from a backplate connection problem...
Have you considered the relay (if it is a D251) as being the culprit? It does disconnect the backplate of one half of the capsule and if this is not the poctop specified part or a faulty one it could be a problem being part of the high Z circuitry.
 
i listened the sample  and i i suspect also the HiZ section (everything around the capsule , capsule included)

for schematics : http://www.filmsoundsweden.se/backspegel/akg-c12vr.html

1 : clean with isopropyl alcohol diluted (20%) the 8Mohms and the  30Mohms resistors
2 : clean the capsule  sides only, with  slightly moisturized q-tips and distilled water or with d.w.+ 5% of isopropyl if it's very dirty, including the mount
3: clean the tube around the pins with isopropyl diluted
4: clean the tube socket with isopropyl
5 : even clean the cables in the highz section
6: clean the 4,8nF and 10nF polystyrene with distilled water
7: clean every trace of soldering flux with isopropyl
8: relay can only be cleaned if you can access to the inside , clean it with a brush and isopropyl

important : use gloves

i personally wouldn't use some coating or varnish as it could infiltrate the capsule or dirty parts and install the problem forever...

if you still have noise issue , here're some ideas:

-check tube bias ( should be  around 1,3 v max 1,6 ) rebias if needed
-test another 6072a  , tube can only be tested inside another equivalent and working tube mic
-check transformer symetry : input  1khz 1v sinus and measure AC voltage at both hot and cold outputs : should be equal otherwise change it
-check also how the heater voltage is stiff , with a scope , and then add RC filters if needed
-check also the B+, same procedure
-clean all the connectors
-check the mic cable: max 5m, touch it and move it when the mic is powered on, if you're hearing  noise , change cable for mogami
-remove the earth ground and check the difference
-find hypothetical ground loops and remove them

regards
Fred

 
Hey all,
Thanks for all the responses and ideas...

To answer a few of the questions:

I cleaned the board not the capsule. I cleaned the board and the high z sections being careful of the styroflex caps. I have a noise tested 6072 that has been tested multiple times by multiple sources and is definitely not the issue. I have taken a ton of photos of the mic and high z section. I also made a recording of turning the microphone on, letting it idle then turning it off too (@Delta Sigma)...

@ln76d: I have used the AMI transformer and I do believe it is connected correctly.  I used all the resistor values given to me by the D-Elam251 poctop BOM. The backplate connections do not beep continuity from the meter. I don't understand the rest of your test with the "If the backplates are disconnected from eachother, then desolder back diaphragm and related to it half of backplate - set mike for cardioid and check is it still noisy"

@granger.frederic:
The capsule is brand new from Tim Cambell. He had tested it right before sending it over. I will try to follow your procedure in the next few days here. I don't have a scope in my possession though. The circuit for this microphone is self biasing so i cant change the bias. 25' Mic cable is mogami and does not show microphonic behavior.

Ideas: If After another full cleaning of the board and high z sections the noise is still persistent; Perhaps I should purchase another set of relays and take apart the whole board, desoldered cleaned and start from there to ensure all parts are cleaned along the way right after soldered and also to do this with  gloves...


Here is a dropbox link to a folder with all the photos and the recording of the mic on/off procedure.

https://www.dropbox.com/sh/h9acbzhptb3cowj/AACHdoODRU73UoZeat_efALGa?dl=0

Thanks a ton all.

-Eric

 
hi eric

self bias doesn't mean you can't adjust the bias voltage
change the cathode resistor (from 1,8k to 2,7k) value and experiment

the tube can be noisy in a mic (very high impedance) and noiseless in a preamp or another standard impedance stage

be careful when you clean a capsule : only slightly moisturized q-tip or coton , you don't want to infiltrate  the solvant inside

you must also check the electrodes insulation resistance of the relays in the datasheet if you can (should be around 1000Gohms at 100v), it should have some hermetically sealed contacts, magnetic shielding and if possible an electrostatic shield ...

you damaged the relays with the soldering iron, i would change all for better ones

check that the hiz insulated wires have the minimum length and that they don't touch anything (in the higZ section and around the capsule also)

i wouldn't twist that much the capsule wires, you create an extra capacitance and increase the noise

Fred
 
HarmonyUnited said:
@ln76d: I have used the AMI transformer and I do believe it is connected correctly.  I used all the resistor values given to me by the D-Elam251 poctop BOM. The backplate connections do not beep continuity from the meter. I don't understand the rest of your test with the "If the backplates are disconnected from eachother, then desolder back diaphragm and related to it half of backplate - set mike for cardioid and check is it still noisy"

Eric - you have properly connected capsule as you wrote, then for test you could  disconnect half of the capsule (back diaphragm and one backplate) from the circuit - just to use it for cardiod (despite that it is disconnected by the relay).
Anyway if the capsule is 100% properly working, then i would try to change styroflex caps for other type - for test.
Unfortunately ksf caps are fragile.
 
granger.frederic said:
hi eric

self bias doesn't mean you can't adjust the bias voltage
change the cathode resistor (from 1,8k to 2,7k) value and experiment

the tube can be noisy in a mic (very high impedance) and noiseless in a preamp or another standard impedance stage

be careful when you clean a capsule : only slightly moisturized q-tip or coton , you don't want to infiltrate  the solvant inside

you must also check the electrodes insulation resistance of the relays in the datasheet if you can (should be around 1000Gohms at 100v), it should have some hermetically sealed contacts, magnetic shielding and if possible an electrostatic shield ...

you damaged the relays with the soldering iron, i would change all for better ones

check that the hiz insulated wires have the minimum length and that they don't touch anything (in the higZ section and around the capsule also)

i wouldn't twist that much the capsule wires, you create an extra capacitance and increase the noise

Fred

Hey Fred,

I am going to purchase the resistor and capacitor components for the mic. I have only found relays with 100GOhm insulation resistance at 500v. Can you look at this data sheet and tell me what you think? Its the relay that Dany put in his BOM. http://www.mouser.com/ds/2/315/mech_eng_ds2y-1076009.pdf
Do you have any ideas about a better part?

Also, i am going to get some regular caps to replace the polystyrene caps just for testing. What voltage rating should I get for those? The PS were rated at 630v and anything else will be huge at that rating. What voltage do those caps see at max?

The tube was noise tested in a microphone by Brent Jesse at AudioTubes.com

I am going to replace all those parts, shorten and un-coil the capsule wires, clean the circuit and see where we are from there!

Thanks!

 
One wise (or at least interesting) thing to do would be to re-test the mic after every component replacement. That way, at least you'd find out which was the actual culprit - that is, assuming you're even curious about that :)
 
By any chance, did you try powering the mic using batteries?  Four D cells in series will light the 6072 heater in parallel or single section, 8 cells if the heater sections are in series.

I went to Radio Shack a good while back and got a dozen or so of those 9 volt battery terminal clips (the ones with the red and white wires coming off them) and connected 12 inexpensive 9 v batteries in series to provide HT for a AC701 tester.  (The A voltage was done with 3 D cells in series through a resistor with the voltages metered..)  This takes power supply AC ripple, etc. out of the equation when testing mic tubes.  For tubes like the EF86 and 6072, more 9 volt batteries can be connected in series to get the required HT.  BE CAREFUL... separate the batteries with fiber board and avoid shorts or accidental contact... for test purposes the grid was terminated with a 150 meg resistor and 20 pf silver mica cap in parallel...

One post was very correct about some tubes working well at normal input such as mic preamps, but there is little documented test data about the same tube operated with input at 150 meg ohms or more...
 
that kind of reed relay (or equivalent)  :  http://www.mouser.com/ds/2/89/2200_series_reed_relay_datasheet-594609.pdf
but be careful at the pins placement.

voltage rating  must be over 160V for styroflex capacitors

change first the high value resistors (Gohms) as you have maybe heated it too much and damaged the coating, then, after cleaning the solvant has perhaps infiltrated  ...
 
Back
Top