Or seal with some coating, Røde style?
AKG Perception 420 Mod
- Thread starter Gbout2
- Start date
Help Support GroupDIY Audio Forum:
Yes, Clearly any unwanted elements at the jFET gate must be avoided due to the enormous input impedance
My puzzlement is how Rode manages to put the SMD jFET directly on the PCB and still get 4...6dB noise! because in that area I noticed traces, flow residues, dirt, etc.,above the protective layer. It was a nt1000 a few years ago, contaminated, made a noise like an old steam locomotive. I insisted on cleaning with isopropyl alcohol many times, until I realized that there were fibers stuck to the pcb, very fine filaments of medical cotton wool, they simply behaved like electric wires. I changed it to thin, soft cloth and here it is! I did it. It still works perfectly today...
Last edited:
That is interesting, Neumann TLM103 manages to get very low self noise also, but I read they use ceramic boards and some sort of conformal coating to counter leakages? IMSAI guy has a great video on it.
Maybe using a small gauge magnet wire, 30 gauge to jumper from the gate of the FET? Idk...im still a noob.
What kind of conformal coating counters leakage? I've never used conformal coating before, but I read it's good to protect against moisture (which I assume is bad also @ high impedance parts).
Maybe using a small gauge magnet wire, 30 gauge to jumper from the gate of the FET? Idk...im still a noob.
What kind of conformal coating counters leakage? I've never used conformal coating before, but I read it's good to protect against moisture (which I assume is bad also @ high impedance parts).
Nothing inherently wrong with soldering gate directly to the board, it becomes issue when dirt, and humidity is present. Coating should prevent this from happening, and i've never seen this issue with Røde. Coating should be also appropriate type, as it can start conducting after a while.
Acrylic or Urethane conformal coatings are fine. Clean the board with appropriate cleaning agents, suitable for the type of flux you have to remove and suitable for the components on the PCB. If you don't do this, the coating will not adhere properly to the surface and dendrides will still grow underneath the coating (talking from experience here). Coating only makes sense on PCB material with low moist absorbtion, such as FR-4. Do not use cheap XPC, FR-1, FR-2, CEM-1 or CEM-3.
Other tips 'n tricks when you'd rather not use the "flying leads" method: apply guard rings (see KM184), use ENIG or OSP finishing and not HASL, leave high-Z tracks free from soldermask (!) and place slots between high-Z nets and other tracks (see TLM102 PCB). Or apply any combination of these methods.
I found TLM102 PCBA pictures on the web, but that is just FR-4. Maybe it was another microphone? Anyway, it would puzzle me if ceramic substrates were used because, AFAIK, their Surface Insulation Resistance is higher than FR-4. And it's more expensive. But maybe I'm wrong, because those GOhm resistors in cheap mics are on ceramic substrates. Anyway, it contradicts information from literature I've read on this subject.
Jan
Oops, mixing up TLM102 and TLM103. TLM103 PCBA looks like it's ceramic, indeed. So yes, I'm puzzled now... 
Jan
Jan
Hi here is the video of imsai guy talking about the 103.
Thanks for sharing the link. Interesting circuit.
Just noticed a typo in my first post: what I intended to say was that, to the best of my knowledge, the SIR of ceramics was lower, not higher than FR-4. Couldn't find the article back which stated this, but it might have been wrong. Anyway, ceramic substrates are not DIY friendly, so whatever the truth is, this knowledge is of little practical use to us. Only nice to know and to satisfy our hunger for knowledge.
Jan
Just noticed a typo in my first post: what I intended to say was that, to the best of my knowledge, the SIR of ceramics was lower, not higher than FR-4. Couldn't find the article back which stated this, but it might have been wrong. Anyway, ceramic substrates are not DIY friendly, so whatever the truth is, this knowledge is of little practical use to us. Only nice to know and to satisfy our hunger for knowledge.
Jan
Finally got the 2SK30A in from China via Aliexpress. Here are some pictures. And yes, that is my brand new FNRSI tester saying it is a resistor. Interesting...
Tested a 2n3819 and it came up as JFET.
Testing a different K30A, came up as another resistor.
.....
Tested all 10....all resistors!
Ive been bamboozled.
Pictures attached.
#1 A look at the k30a...
#2 A resistor it says...
#3 2n3819 JFET testing as JFET.
Tested a 2n3819 and it came up as JFET.
Testing a different K30A, came up as another resistor.
.....
Tested all 10....all resistors!
Ive been bamboozled.
Pictures attached.
#1 A look at the k30a...
#2 A resistor it says...
#3 2n3819 JFET testing as JFET.
Attachments
Hi audio freq!
all my perfectly working k30a's show on the same FNIRSI multitester as some resistances!
It's a problem with correct detection that I only encountered on the k30a.Attachments
In a few days I have to receive from China 60 pieces of k30a, 300 of a1015 pnp low noise bjt and others. Of course I will measure and test them in the circuit.
My advice is to test them to see if they work like jFETs.
Attachments
Is there a breadboard circuit I can create to test them?
Equipment I have:
Oscilloscope
Signal Generator
Power Supply
Multimeter
Breadboard
One of the methods of testing and measuring Idss that I have been using with good results for several years. The resistance also can be 1ohm or 10ohm.
Attachments
https://sound-au.com/articles/jfet-design.htm#s2
Might wanna be careful with the breadboard, wonky contacts can have you running around your tail forever.
I received k30a faster than I expected.
As I said before, although on the multimeter it measures as a resistance, in reality the k30a are real, excellent, according to the specifications and even with low noise. I tested compared to other well-known and quality brand.
I am excited. Being in the group of Idss=2.6...6.5 I will use them even instead of 2n3819 in u87, I have done it before with very good results.
I bought 60 pieces for $10.38
I don't care who and where manufactured them.
Attachments
Hi audio freq!
all my perfectly working k30a's show on the same FNIRSI multitester as some resistances!
Which would be your exact set up? 12v power, and 1 ohm resistor? I would like to replicate your exact testing measures, so I can compare notes with you on the quality of the ones I received.
Attachments
Set the multimeter to the hundreds of mA scale for a more accurate reading.
0.030V = 30mV
30mV : 10 ohm = 3mA
Idss=3mA
