Repairing an u87

[aiyiadam]

I have difficulty repairing an old u87 and really need some help. I don’t have much experience with microphones.

There have been previous repair attempts here and I have also done a couple of things.. I checked the components, and the connections, changed the electrolitic caps, corrected a few connections and soldered a new 2n3819. There still are problems.

My soundcard can receive the signal from the microphone, but there is no amplification.. The signal levels on the drain and the gate are similar. Instead of the resistor I have put a trimpot on the source, which doesn’t effect anything. I have checked the caps and they seem to be okay. I might be missing something obvious. Any ideas?

The voltages also seem to be low (without the capsule and an input signal). It is possible that my soundcard is unable to supply enough current.. I don't think that is the reason, but I will also check that.

 

[Khron]

The first commandment of troubleshooting is "thou shalt measure voltages". This looks to be an early U87i (as in, not the later U87Ai). Googling for "u87 schematic" will easily provide you / us with a starting point.

Compare the DC voltage readings at the points of the circuit where voltages are indicated, and we can work up from that.

Where did you get your "new 2N3819"? They have been discontinued years ago, and i've read there are a lot of fakes floating around. Since that's the case, it's not entirely impossible your 3819 is not a 3819 at all, or the pinout is messed up, etc.

 

[aiyiadam]

The first commandment of troubleshooting is "thou shalt measure voltages". This looks to be an early U87i (as in, not the later U87Ai). Googling for "u87 schematic" will easily provide you / us with a starting point.

Compare the DC voltage readings at the points of the circuit where voltages are indicated, and we can work up from that.

Thank you.. I must admit that I haven't checked all of the voltages on the schematic. I will measure and share the voltages, but I really think that there is another problem.

 

[Khron]

but I really think that there is another problem.

... Such as..?

 

[aiyiadam]

... Such as..?

I wish I knew.. I have attached the voltages. Those are measured without the head.

 

[McIrish]

I agree on the 2N3819. The ones you can get now, like from Central Semi, have nothing in common with the originals. I wouldn't use them. the bias points are way off.

 

[Khron]

The voltages in the capsule area WILL be lower than you might expect, since your voltmeter will have a finite impedance (usually 1Mohm), which forms a divider with whatever resistances are "upstream" of it. That's likely why you're reading 46V at the R18-R19 common point, as you should, while after R2 you're only seeing 34V.

But this question is still very much unanswered:

Where did you get your "new 2N3819"? They have been discontinued years ago, and i've read there are a lot of fakes floating around. Since that's the case, it's not entirely impossible your 3819 is not a 3819 at all, or the pinout is messed up, etc.

Case in point...

I agree on the 2N3819. The ones you can get now, like from Central Semi, have nothing in common with the originals. I wouldn't use them. the bias points are way off.

And i'll definitely never ever comprehend what's with worshipping certain "magical" components... I kinda wonder just how noisy a J113 actually is, in this sort of circuit - it's readily available from Mouser at least, and most of the parameters are pretty close to the 2N3819 (keeping in mind the pretty-humongous variance of JFET parameters in general).

 

[Tim Campbell]

Do you have true 48v from your soundcard? The U87 will not operate properly with less.

 

[Khron]

Do you have true 48v from your soundcard? The U87 will not operate properly with less.

There seems to be a 46v reading inside the mic (R18-R19 common point), which sounds pretty right to me..?

 

[aiyiadam]

Do you have true 48v from your soundcard? The U87 will not operate properly with less.

46V.. I am not sure if the soundcard is able to supply enough current.

Where did you get your "new 2N3819"? They have been discontinued years ago, and i've read there are a lot of fakes floating around. Since that's the case, it's not entirely impossible your 3819 is not a 3819 at all, or the pinout is messed up, etc.

The 2N3819 are from a reliable lokal dealer, but it is possible that they are fake.. However, the pinout is right..

I can try other ones.. I have J201s(real ones), BF245A, BF256C and a few more.

 

[Ricardus]

The 2N3819 are from a reliable lokal dealer, but it is possible that they are fake.. However, the pinout is right..

I can try other ones.. I have J201s(real ones), BF245A, BF256C and a few more.

Did you put the transistor on a tester to see if it was actually a FET? I did get fake ones in the past, but managed to find some NOS after scouring the internet.

 

[aiyiadam]

Did you put the transistor on a tester to see if it was actually a FET? I did get fake ones in the past, but managed to find some NOS after scouring the internet.

I did.. It obviously showed N-BJT and I didn't realize/care.. It is too hot here in Turkey now.. I just can't think right.

I just tried a BF245A. I am not sure, if it is a perfect choice, but the mic seems to work now. I will try to set the bias and be done with this..

@Khron , @McIrish , @Ricardus

Thank you.

 

[Asiandude]

Are you live the US? Maybe you need to send to a repair shop to get it fix. I sent my to Tom Onofrio and very happy to get it back working on my 1988 model.

 

[Khron]

Are you live the US?

It is too hot here in Turkey now.

I'm gonna guess that's a "no"...

 

[aiyiadam]

Are you live the US? Maybe you need to send to a repair shop to get it fix. I sent my to Tom Onofrio and very happy to get it back working on my 1988 model.

I am in Istanbul.. The mic doesn't belong to me. I was trying to repair it for a friend / customer..

 

[McIrish]

The voltages in the capsule area WILL be lower than you might expect, since your voltmeter will have a finite impedance (usually 1Mohm), which forms a divider with whatever resistances are "upstream" of it. That's likely why you're reading 46V at the R18-R19 common point, as you should, while after R2 you're only seeing 34V.

But this question is still very much unanswered:


Case in point...


And i'll definitely never ever comprehend what's with worshipping certain "magical" components... I kinda wonder just how noisy a J113 actually is, in this sort of circuit - it's readily available from Mouser at least, and most of the parameters are pretty close to the 2N3819 (keeping in mind the pretty-humongous variance of JFET parameters in general).

In my KM84 clones, I used an LS846 fet. Worked very well and bias resistor value was similar to an original 2N3819.

 

[wave]

A 2N5457 would work as well although the pinout is different... Plus side they are usually under 2mA IDSS

 

[tomas.borgstrom]

What’s the voltage at the source?

Are you sure the capsule is working?

Since repairing attempts have been made before I assume the microphone has been used with other preamps and therefore we can conclude that the phantom powering is not the problem.

In my experience it's quite easy to find 2N3819 from reliable dealers. Some random seller on ebay is of course not one of them.

 

[aiyiadam]

What’s the voltage at the source?

Are you sure the capsule is working?

Since repairing attempts have been made before I assume the microphone has been used with other preamps and therefore we can conclude that the phantom powering is not the problem.

In my experience it's quite easy to find 2N3819 from reliable dealers. Some random seller on ebay is of course not one of them.

It wasn't a voltage or capsule problem. I had fake 2N3819s.. The microphone seems to work with a BF245 now.. BF245 is probably not a very good choice though..

Since I shouldn't comment on its frequency response, noise level etc., I am giving the microphone back to its owner.. We can later try other FETs. The capsule might also need cleaning or reskinning..

 

[living sounds]

Try a BF256B. That's what Andreas Grosser put in my U87ai. I had success with those in clone circuits as well. They're cheap, buy a bunch, test them in the circuit and find a good one.