2N3819 versus 2SK170 in MXL V50M

[doorunrun]

Since becoming a member I've discovered I have a large quantity of old 2N3819 JFETs. They were in a parts bin I purchased at a silent key ham estate sale several years ago.

I got interested in donor mic mods and did a BM-8000 mod following the YT video of Sounds Sleuth using JLI components. While getting more familiar with mic mod techniques I noticed the popularity of the '3819 and then the "penny dropped" looking through all the old ICs and transistors in that bin looking for a LM301a replacement.

Most of the ones are National Semiconductor, but there are some from what appears to be Teledyne from looking at the logo.


I own a couple of older MXL V50Ms and I substituted its 2SK170 JFET for one of the NS.
Initially, the only thing I notice is what seems to be a lower output level. At this point I haven't done any real comparison between the one I modded and its mate.

I've read over some of the posts on the two components and would like to know more about which circuits the 2N3819 works better or worse.

Thanks!

 

[McIrish]

Did you re-bias the circuit after swapping the FET? That's going to make a huge difference.

 

[Ricardus]

2SK170 is as highly regarded as 2N3819 in these applications. At least from what I have seen hanging around this forum for a few years, so there was probably really no need to make the swap.

 

[kingkorg]

Since becoming a member I've discovered I have a large quantity of old 2N3819 JFETs. They were in a parts bin I purchased at a silent key ham estate sale several years ago.

I got interested in donor mic mods and did a BM-8000 mod following the YT video of Sounds Sleuth using JLI components. While getting more familiar with mic mod techniques I noticed the popularity of the '3819 and then the "penny dropped" looking through all the old ICs and transistors in that bin looking for a LM301a replacement.

Most of the ones are National Semiconductor, but there are some from what appears to be Teledyne from looking at the logo.
View attachment 105594

I own a couple of older MXL V50Ms and I substituted its 2SK170 JFET for one of the NS.
Initially, the only thing I notice is what seems to be a lower output level. At this point I haven't done any real comparison between the one I modded and its mate.

I've read over some of the posts on the two components and would like to know more about which circuits the 2N3819 works better or worse.

Thanks!

The only reason it seems 2N3819 is so popular is because it was used in classic Neumann models, and then got discontinued. But then, those used by Neumann were selected for specific properties. Many discussions about this all over the web. As McIrish said you would need to re-bias.

Not familiar with that mic, but many MXL mics have different bias arrangement than say u87. MXL uses one that allows more tolerance for fet with two source resistors instead of one.

 

[doorunrun]

Did you re-bias the circuit after swapping the FET? That's going to make a huge difference.

...As McIrish said you would need to re-bias.

Not familiar with that mic, but many MXL mics have different bias arrangement than say u87. MXL uses one that allows more tolerance for fet with two source resistors instead of one.

Thanks for the suggestions.
No, I did not try re-biasing. I haven't found a good schematic for this board yet. I understand it's a Schoeps variant and probably does a good job with the 2SK107 in place.

This is what my my board looks like. It's taken from a thread on V57m mods: (https://gearspace.com/board/geekzone/104906-mod-mxl-v57m.html)

There's also a bias board (photo of it in that thread), but I'm not sure how to change the bias.

BTW, I did come across a mod for this board and changed the ceramic caps, next to the large electrolytics, to WIMA films and replaced the gate cap, also ceramic, to polystyrene.

Cheers!

 

[kingkorg]

Thanks for the suggestions.
No, I did not try re-biasing. I haven't found a good schematic for this board yet. I understand it's a Schoeps variant and probably does a good job with the 2SK107 in place.

This is what my my board looks like. It's taken from a thread on V57m mods: (https://gearspace.com/board/geekzone/104906-mod-mxl-v57m.html)
View attachment 105600
There's also a bias board (photo of it in that thread), but I'm not sure how to change the bias.

BTW, I did come across a mod for this board and changed the ceramic caps, next to the large electrolytics, to WIMA films and replaced the gate cap, also ceramic, to polystyrene.

Cheers!

Yes, seems like Schoeps. Forget what i said then about u87 style biasing. Yet it will still need re bias. There are many topics on this, you shouldn't really need a schematic to pull it of.

 

[Khron]

Thanks for the suggestions.
No, I did not try re-biasing. I haven't found a good schematic for this board yet. I understand it's a Schoeps variant and probably does a good job with the 2SK107 in place.

This is what my my board looks like. It's taken from a thread on V57m mods: (https://gearspace.com/board/geekzone/104906-mod-mxl-v57m.html)
View attachment 105600
There's also a bias board (photo of it in that thread), but I'm not sure how to change the bias.

Don't confuse the biasing of the JFET, with the DC bias voltage applied to the capsule. That second board deals strictly with the latter.

 

[Екатерина]

I have used a lot of 2SK170, which is really great. It often brings more low-frequency harmonics when used for simple replacement, but the high-frequency will be a little dim

 

[doorunrun]

Don't confuse the biasing of the JFET, with the DC bias voltage applied to the capsule. That second board deals strictly with the latter.

Thanks for making that distinction! I haven't come across a method, but I'm beginning to suspect it requires more test equipment than I have. I'll look more in that direction. Saw a YT video that looked like the process; I can go back and take another look.
Thanks again!!!

 

[doorunrun]

While I have the 2SK170 removed from the circuit board I tested it with my offshore component tester and was struck by its testing as an NPN transistor but with a very high hFE and low forward voltage (Vf. right?).


Here's how it was oriented in the tester's socket:


I tested a couple of the 2N3819s and they show up a "real" JFETs on that tester.
Anyone familiar with this type of result? Maybe I zapped the 2SK170 in the process of removing it?
I haven't it back in the circuit to try it out yet.

Thanks!

 

[Drosselmeier]

Maybe it needs to be connected in a different way

 

[doorunrun]

Maybe it needs to be connected in a different way

The tester I'm using is set up, going left to right: 1-2-3-1-1-1. Also, the 2N3819 was inserted in the same slots and I got the expected JFET display result.
Although, I do have another one of these testers, an older model, probably just older firmware, I can try testing with it and see. Thanks!

 

[kingkorg]

While I have the 2SK170 removed from the circuit board I tested it with my offshore component tester and was struck by its testing as an NPN transistor but with a very high hFE and low forward voltage (Vf. right?).
View attachment 106195

Here's how it was oriented in the tester's socket:
View attachment 106196

I tested a couple of the 2N3819s and they show up a "real" JFETs on that tester.
Anyone familiar with this type of result? Maybe I zapped the 2SK170 in the process of removing it?
I haven't it back in the circuit to try it out yet.

Thanks!

I have actually seen fake fets that were bipolars

 

[charlesaudio]

I have the same issue testing fets with one of this same tester. They test like bjt's or low value resistors. They test ok on another tester of same model.

 

[doorunrun]

@charlesaudio Thanks for that information!