Using the lowest common denominator should work.
I don't know where to look for now.
I think 1mm would be pushing it. But if what we have is somewhat already compliant, then all the better.
Besides, this is not aimed at being sold commercially, so we can allow ourselves some slack!
I use eagle, then export data with eagleup and use sketchup to get a 3D assembly.
Then I have access at work to solidworks, so I rebuild the same thing in it and render with it, so every detail can be checked and rechecked, to make sure what's designed can be made exactly right.
That's why I mentioned that I'm working right now with a metal work shop in china with whom I cooperate for work, and they're very well equipped to make whatever we want. They can laser cut that case to my specs and do whatever bending, folding, welding to get it done as the drawing says.
Being able to virtually design everything before making it is a nice engineering feature that can not only save time, but it can save money and efforts.
I use the image output generated by sketchup's eagleup plugin to make it even better in solidworks, so the pcbs look as close as possible to the real thing, to the right dimensions, including all the components, so that can be used to check for sizes, clearances etc...
Then I have access at work to solidworks, so I rebuild the same thing in it and render with it, so every detail can be checked and rechecked, to make sure what's designed can be made exactly right.
That's why I mentioned that I'm working right now with a metal work shop in china with whom I cooperate for work, and they're very well equipped to make whatever we want. They can laser cut that case to my specs and do whatever bending, folding, welding to get it done as the drawing says.
Being able to virtually design everything before making it is a nice engineering feature that can not only save time, but it can save money and efforts.
I use the image output generated by sketchup's eagleup plugin to make it even better in solidworks, so the pcbs look as close as possible to the real thing, to the right dimensions, including all the components, so that can be used to check for sizes, clearances etc...
This for sure is the future of Engineering/Manufacturing.
I wish the cost of Solidworks was more in the reach of non-commercial users.
Thanks for sharing your process work flow.
Me too, it's outrageously priced, but do like I do, find some others who have it and get access when they're not using it and it's sitting idle.
@spookydd:
Very impressive! I just finished two Leach v4.5 monoblocks in Hafler DH200 chassis but sure are not works of art.
I have many Krell branded TO-3 transistors (MJ3281/1302 type) but ended up using 4 pairs each channel NJL3281 / 1302 ThermalTrak (sp?) flatpaks. I really like the easy thermal bias diode hookups and stable biasing. That might be a good reason to think about those.
Very impressive! I just finished two Leach v4.5 monoblocks in Hafler DH200 chassis but sure are not works of art.
I have many Krell branded TO-3 transistors (MJ3281/1302 type) but ended up using 4 pairs each channel NJL3281 / 1302 ThermalTrak (sp?) flatpaks. I really like the easy thermal bias diode hookups and stable biasing. That might be a good reason to think about those.
I did think about those thermalTrak ones, and I was seriously considering them, but Onsemi is retiring them and they don't even sell them directly any more. You can only find left over stock from few places and that's a pitty.
I went on designing this leach with TO3s because I have plenty in stock, but my plans include more amps and I will switch to the flat packs later.
For now, I want to get in a group buy for Jen's amp, he's the greatest!
Our pal is Jens, so that should be Jens'
OOPS!!!! Mea culpa! I know better. Typo.
I did not know OnSemi was retiring these devices...
That would be a pity, since these devices have high hfe and good matching out of the box between NPN-PNP.
They already have retired the NJL21193/94s,I don't see why. That's stupid. Those are popular devices.
They're old, "Newer is better" and it costs a lot of money to manufacture similar parts with relatively minor specification differences. Stinks when we find a part we like and are comfortable with and it gets phased out, but we aren't even a drop in the bucket part of their business.
I understand that it was a rhetorical "I don't know why", spooky, but I felt like venting having dealt with a lot of other bureaucratic stupidity today.
I understand that it was a rhetorical "I don't know why", spooky, but I felt like venting having dealt with a lot of other bureaucratic stupidity today.
Last edited:
AMEN!
I understand that it was a rhetorical "I don't know why", spooky, but I felt like venting having dealt with a lot of other bureaucratic stupidity today.
"I feel your pain!" <<<with slick willy's voice>>>
I was really getting interested in those parts, but now I don't even dare touching them. And those suckers won't even have anything to replace them with.
you may want to look here...http://www.diyaudio.com/forums/soli...stors-todays-market-any-experience-these.html
I was quite surprised when Elektor was claiming 200W with single TO-3P device and then I looked further and seen that they were offering a 400W device in a TO-264 package. I commented that I thought Elektor should have used 2-pairs,
I do wonder about the ability to get the heat off the die to keep reliability up.
Need to do some temp calc's too.
Semelab's do not publish reliability data. Are they hiding something or just unprofessional? I even asked about Spice models, I was wondering if the sales contacts even knew what I was talking about, There is no way to to get any tech support from what I can tell. Makes you wonder about this organization workings and that they have more names than types of ducks.
I do wonder about the ability to get the heat off the die to keep reliability up.
Need to do some temp calc's too.
Semelab's do not publish reliability data. Are they hiding something or just unprofessional? I even asked about Spice models, I was wondering if the sales contacts even knew what I was talking about, There is no way to to get any tech support from what I can tell. Makes you wonder about this organization workings and that they have more names than types of ducks.
Last edited:
I am wary of the flat packs when it comes to power dissipation. You need a larger number of them compared to the metal cases.
Eventually I will make a version of my design with flat packs, as they are more convenient and easier to mount on heatsinks. Plus we can design for a forced cooled tunnel heatsink and that is one of my plans for the bigger amps for my low frequency speakers.
Here is an updated schematic with some of the protection resistors calculated.
I've been busy on other things, so this one advances a little slower, but I just ran most of the calculations on those protections and I wouldn't mind getting some feedback about this.
The resistors R41-43, 60-62, and R28/30/26/34 are the ones in question. And I still don't know how to proceed about the R74/75 ones.
Any hints welcome.
I've been busy on other things, so this one advances a little slower, but I just ran most of the calculations on those protections and I wouldn't mind getting some feedback about this.
The resistors R41-43, 60-62, and R28/30/26/34 are the ones in question. And I still don't know how to proceed about the R74/75 ones.
Any hints welcome.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.