@Nico Ras what is wrong with you? We have discussed PCB manufacturers which are good to order our PCBs for this amp from. There are not just Brian's and M0rten's approaches to create their PCBs.
And I did contribute, but you probably did not notice because you are more eager to spread your wisdom.
EDIT: "Are you just interested in the freebee, making and selling other's efforts." And what weird insinuation is that?
And I did contribute, but you probably did not notice because you are more eager to spread your wisdom.
EDIT: "Are you just interested in the freebee, making and selling other's efforts." And what weird insinuation is that?
Yeah right.
Why sneak in advertisement, it is not appropriate. I did not see anyone stop you from creating your own PCB. You are also welcome to share your skills and allow members to give some feedback on your PCB designs. You may also offer group buys, you can do absolutely anything you want except advertise, there is a Marketplace for this. I did not attack you, I did not even know of your existence until your post to me, but if you feel that I directed an attack on you I am truly very sorry.
Why sneak in advertisement, it is not appropriate. I did not see anyone stop you from creating your own PCB. You are also welcome to share your skills and allow members to give some feedback on your PCB designs. You may also offer group buys, you can do absolutely anything you want except advertise, there is a Marketplace for this. I did not attack you, I did not even know of your existence until your post to me, but if you feel that I directed an attack on you I am truly very sorry.
Last edited:
I did not advertise anything. We (soekris, lydlaug and others) were discussing PCB manufacturer options for getting PCBs for this amp (and other things) made.
Okay then accept my profuse apology and lets get on with the project. Its late we must all go to bed in my and your timer zone.
After reading the posts about 'DIY snottyness' and complete, unambigous description of mosfets, I'd like to add my 2c.
I have been selling kits for more than twenty years and I have come to this realisation.
Every person has a slight interpretation from the written word. Even simple prepositions and conjunctions can completely change the meaning, particularly where English is their second language and these small words are very different. For example, in English we describe marrying TO someone, but in Indonesian it is WITH, which of course is far more logical because it's a much younger language, not so corrupted by time.
Considering many of my customers build my kits using a second language, we should remember that I don't speak their language and in fact no other language at all, being an Australian, we live in a linguistic desert. So I try very hard to:
1. Show a graphic of polarised components, particularly hexfet, latfets and small transistors,
2. Explain gate drain source in two different sentences so there is NO possibility to make mistakes.
A few of DIYers who built the Alpha Nirvana suffered from incorrect pinouts on small transistors. Most people find it difficult to find datasheets, so you have to explain it in detail. There is a certain exasperation when things go wrong, sometimes some resentment, and people quickly lose interest in chasing down the issues. So to keep their enthusiasm, you have an obligation to really describe things clearly. Most people are well aware of their ignorance, and correspondingly they are terrified of making mistakes, knowing it might cost dollars or even threaten electrocution. You can understand that..... so identifying the facts two or three ways gives them confidence.
EVERYONE makes mistakes, even the professionals when they stuff pcb boards. I still do it occasionally even today at 73. And to consider this, realise that everyone who seems a DIY newbie usually has one or several other skills we do not know about but has been honed to perfection. It might be wrenching cars, building furniture, playing a musical instrument, cutting code, sparkling conversation - but everyone has something remarkable about them.
HD
I have been selling kits for more than twenty years and I have come to this realisation.
Every person has a slight interpretation from the written word. Even simple prepositions and conjunctions can completely change the meaning, particularly where English is their second language and these small words are very different. For example, in English we describe marrying TO someone, but in Indonesian it is WITH, which of course is far more logical because it's a much younger language, not so corrupted by time.
Considering many of my customers build my kits using a second language, we should remember that I don't speak their language and in fact no other language at all, being an Australian, we live in a linguistic desert. So I try very hard to:
1. Show a graphic of polarised components, particularly hexfet, latfets and small transistors,
2. Explain gate drain source in two different sentences so there is NO possibility to make mistakes.
A few of DIYers who built the Alpha Nirvana suffered from incorrect pinouts on small transistors. Most people find it difficult to find datasheets, so you have to explain it in detail. There is a certain exasperation when things go wrong, sometimes some resentment, and people quickly lose interest in chasing down the issues. So to keep their enthusiasm, you have an obligation to really describe things clearly. Most people are well aware of their ignorance, and correspondingly they are terrified of making mistakes, knowing it might cost dollars or even threaten electrocution. You can understand that..... so identifying the facts two or three ways gives them confidence.
EVERYONE makes mistakes, even the professionals when they stuff pcb boards. I still do it occasionally even today at 73. And to consider this, realise that everyone who seems a DIY newbie usually has one or several other skills we do not know about but has been honed to perfection. It might be wrenching cars, building furniture, playing a musical instrument, cutting code, sparkling conversation - but everyone has something remarkable about them.
HD
Last edited:
I think I close this version of the project. Last update for PCB below. Gerbers eclosed (if you don't harass me for poor routing!) Thanks to AKSA for last minute inputs! Tanks to Nico and Brien for supporting a newbie like myself. And to all else who contributed.
If anyone wants the KiCAD source (to improve said routing?), just say so!
🙂 morten
If anyone wants the KiCAD source (to improve said routing?), just say so!
🙂 morten
Just a final comment. I am not in the PCB manufacturing business, but aren't a scored board cheaper than a tab routed one. i.e. all the rounded edges? If it is single boards maybe I am wrong, but I am sure if pannelised it would be.
Up till now I have used NextPCB, and rounded corners did not increase price. Now they have begun acting strangly and requesting all sorts of additional info, so I will try one of the suggested manufacturers. If rounded corners turn out to be more expensive, I'll change to square corners.
You act like a kid not wanting to go to bed waiting for Christmas
In the datasheet for the LSJ689 I just see that there are no bands for Idss listed. Which LSK389 would you recommend byuing? I would think C could have the best overlap 🤔.
LSJ689
LSK389 bands
LSJ689
LSK389 bands
I have plenty of 2SJ109-2SK389, GR-BL and V versions. PM if interested.
R
R
I purchased 10 LJS689 and they measured in the 3.5mA to 8.5mA range. @benpe had a smaller lot that measured similarly.
Ultimately, the grade isn't that important. Nor is the current the IPS runs at. Since it doesn't matter than much, I selected 1mA simply to allow the JFETs to run cooler. Simulations also confirmed that IPS current has very little impact.
At this level, any of the Idss ranges will work. I choose "A" grade for the LSK489s to align with the measured values I had for the LSJ689's I got. However, I later found that matching between the N and P channels isn't necessary.
Bob Cordell has a paper on JFET matching that was written for his DH-220C amp. This amp is similar in design. The Ultra is sort of like a very simplified version of the DH-220C.
Complementary JFET Offset Mismatch (PDF)
Three key takeaways from the paper are:
Ultimately, the grade isn't that important. Nor is the current the IPS runs at. Since it doesn't matter than much, I selected 1mA simply to allow the JFETs to run cooler. Simulations also confirmed that IPS current has very little impact.
At this level, any of the Idss ranges will work. I choose "A" grade for the LSK489s to align with the measured values I had for the LSJ689's I got. However, I later found that matching between the N and P channels isn't necessary.
Bob Cordell has a paper on JFET matching that was written for his DH-220C amp. This amp is similar in design. The Ultra is sort of like a very simplified version of the DH-220C.
Complementary JFET Offset Mismatch (PDF)
Three key takeaways from the paper are:
- The IPS current isn't that important. He choose 2mA but admits its arbitrary.
- The N and P channels doesn't need to be matched to each other.
- Although the devices are dual matched monolithic, the matching is not perfect. In practice, sockets for the TO-71 devices help to allow the builder to flip the devices to find the direction that results in the lowest DC offset.
@m0rten , I now have done a quick check of the cascodes in LTSpice using models for 2SK146 and 2SJ74 (did not find one for 2SJ73, and this looks somewhat near).
The voltage divider when using those needed create half the rail voltage for a 50/50 voltage distribution across the cascode FETs. So you might want to check when building and maybe rather use 47k/47k instead of the earlier mentioned 47k/68k.
Looking forward to reports on your build.