AEM6000 Based 50W Amp

And here's where the TO-3 version is at. I think that's probably a goer. I'll check over it in the morning and then post up Gerbers tomorrow. There's something particularly weird about mixing TO3 with SMD.
 

Attachments

  • Screen Shot 2021-02-16 at 11.42.32 pm.png
    Screen Shot 2021-02-16 at 11.42.32 pm.png
    231.2 KB · Views: 477
  • Screen Shot 2021-02-16 at 11.44.15 pm.png
    Screen Shot 2021-02-16 at 11.44.15 pm.png
    246.3 KB · Views: 471
  • Screen Shot 2021-02-16 at 11.46.19 pm.png
    Screen Shot 2021-02-16 at 11.46.19 pm.png
    153 KB · Views: 465
  • Screen Shot 2021-02-16 at 11.46.38 pm.png
    Screen Shot 2021-02-16 at 11.46.38 pm.png
    201 KB · Views: 469
Very nice. Again, don't bother if it's too weird or doesn't make sense to do this one! 🙂

Couple of questions:
1) Are the gate resistors back beside the bias pot? Is that close enough to the gate pins? (I've often read for laterals to have them as close as possible, but I'm not an expert. And, it probably is pretty close as this is such a small pcb.
2) I believe the pins on the TO-3 are gate and drain, not gate and source, or am I misinterpreting the diagrams.

Thank you for what you've done so far, and don't feel bad if you don't want to continue on this one.

Steve.
 
Hey Steve, very nice catch - thatnks for picking that up. I'd assumed the pinout of the TO3 was the same as the TO3P (1 gate, 2 drain, 3 source), but it makes sense that the body is source as it's a lateral.

I've also moved the gate resistors, plus added pours for the +40V and -40V terminals on board top, just to ensure no other pours get in between the pins.

This new version of KiCad is _so fast_ to design in, as is no doubt evident by how quickly I've been producing boards in the last few weeks. It's like I have an idea, I draw the idea. The tools help rather than hinder. I'm probably sounding like a broken record here, but I've used many different PCB CAD systems over my career, and this is by far the best one I've ever used. And it's FREE!!!

Also I found a cool resistor that I can use for sources. It's allegedly 7W in a 7.1 x 4.6mm package, and really low inductance to boot. See https://au.mouser.com/datasheet/2/427/wshm2818-1762134.pdf

And here's the Gerbers, as promised.

AEM6000 based 50W Amp TO-3 Gerbers.zip - Google Drive

Again, please bear in mind that it's completely untested, as with the TO-264 version.

Also as with the TO-264 version, it may do slightly more than 50W.
 

Attachments

  • Screen Shot 2021-02-17 at 7.03.53 am.png
    Screen Shot 2021-02-17 at 7.03.53 am.png
    228.5 KB · Views: 366
  • Screen Shot 2021-02-17 at 7.04.06 am.png
    Screen Shot 2021-02-17 at 7.04.06 am.png
    215.2 KB · Views: 172
  • Screen Shot 2021-02-17 at 7.04.23 am.png
    Screen Shot 2021-02-17 at 7.04.23 am.png
    144.7 KB · Views: 166
  • Screen Shot 2021-02-17 at 7.04.36 am.png
    Screen Shot 2021-02-17 at 7.04.36 am.png
    172.1 KB · Views: 156
Last edited:
I must give KiCAD another go.. but the past few times ive tried it, i've found it to be clunky garbage and gone straight back to EAGLE (pre-Autodesk)

It's a pity 2SK2145 has one side of the JFETs commoned... it looks like a good little part otherwise
 
Hi spind,

yours is a timely post as I have a few 2SK135/2SJ50.

As you seem to be knowledgeable, would I have to change anything as far as the output transistor circuitry (C9, C11, R44) is concerned? I am asking because Borbely Servo 100 using these transistors has 330 pf in position of C9 and nothing as far as the other FET is concerned.

Kindest regards,

M

M
 
I’m doing a board run in the next week or so. I shall include some of the double-die TO-264 and TO-3 ones.

However I don’t have any TO-3 transistors to prototype with. I’ll buy some Exicon ones, but would like to check performance with some older Hitachi parts. Anyone willing to lend me some for development?