Goldmund Wiki and build 2017

Jump to Latest
#223
Last edited:
#224
Attacked for easy view.
 

Attachments

  • 20180116_233151.jpg
    20180116_233151.jpg
    1,014.5 KB · Views: 1,105
  • 20180116_233206.jpg
    20180116_233206.jpg
    991.9 KB · Views: 1,035
  • 20180116_232146.jpg
    20180116_232146.jpg
    995.5 KB · Views: 963
  • 20180116_232823.jpg
    20180116_232823.jpg
    1,001.8 KB · Views: 978
#228
carlmart said:
Availability and price are an issue with the LSK389.

On my Goldmund LTSpice sim. THD is improved with the 2SK170, compared to the 2N5565.

But of course, to get that result the 170s have to be matched.
Click to expand...
2SK170 are in separate houses, so even when matching them, you do not get the performance of the 2N556x as they share the same chip.
It would be nice to have the transductance of the 2SK170 inside a 2N556x though.
The input capacitance is the double in the 2SK170. You could try to simulate with two parallelled 2N556X and I guess you will get the same reduction in THD...!
 
#230
schillg11 said:
Of course perfect matched 2SK170 will improve the THD, because you used 2 identical JFETs for your simulation. In real life they are alltimes different. Maybe you' ll find 2 very similar ones, when you have tested 100 pcs. The main problem here is to find 100 genuine parts. Most of the offered one's on the internet are fakes, For a vry good matched pair (1%) , from $30 upwards you can count.

BR
Günni

PS: at Trensetter.com you can order in different case SOIC8 or TO-71/6
Click to expand...
As stated in the above response I think the reduction in THD comes from the fact that 2SK170 has higher transduction and therefore higher open loop gain. You can get almost the same transduction from parallelling two 2n556x and then you also get the same input capacitance. So it is not the device itself per. se. that makes the difference.
And then we are back to the point of the two 2SK170 not tracking each other as well as the two half parts in the 2N556X. IMHO it is one of the reasons the Goldmund sounds so good. The topology has been used, more or less exactly the same, in many other amplifiers, but with separate input fets.
 
#235
I puchase a lot of these 2sk43, they are mached + -0.1mA.

The 2N556x is a two die put together in same case, which combines heat more closely. I will try to tie 2sk43. hope to achieve good.
The Sony 2SK97 (dual) very match to 2N556x, but they are hard to find now.
 

Attachments

  • FB_IMG_1503633335613.jpg
    FB_IMG_1503633335613.jpg
    92.1 KB · Views: 976
  • i-img1200x991-1521859159whsr9g8970.jpg
    i-img1200x991-1521859159whsr9g8970.jpg
    150.5 KB · Views: 971
  • i-img1200x915-1521859159gsr83o8970.jpg
    i-img1200x915-1521859159gsr83o8970.jpg
    106.2 KB · Views: 924
  • laserdiode56-img1200x804-1521366288gwcjbs28221.jpg
    laserdiode56-img1200x804-1521366288gwcjbs28221.jpg
    71.8 KB · Views: 932
Last edited:
#237
Junm said:
Hi Erikovsky, do you have PDF of the TO-3 PCB you have posted. I have a good stock of K176 and J56 to use with and happened to find 2 pcs 2n5566 for these.

I also have plenty of dual jfet of K194 and K58 and some K389 and uph68h as alternatives. also good quantity of k43. k44 etc.
Click to expand...

I will start a Group-buy recently, the PCB will be gold-plated.

Regards,

Eric
 
You must log in or register to reply here.
Top Bottom