Thanks both of you.
If i do some math, it look like i have to draw near 6A, if using only one psu.
My onechannel AlephJ mini draws 1,3-1,4A, so two channels will be double (2,8A)
This is meassured on the +19vdc supply rail, so the same amount will be drawn on -19vdc rail i guess.
This will be all together 5,6A; so near the limit of 6A.
Will that work?
Jesper.
Btw.: If we can discuss this in my thread it would be better. I dont want to spoil youre thread Juma.
My thread is in my signature. And i will post youre PSU schm. there Juma shortly!
If i do some math, it look like i have to draw near 6A, if using only one psu.
My onechannel AlephJ mini draws 1,3-1,4A, so two channels will be double (2,8A)
This is meassured on the +19vdc supply rail, so the same amount will be drawn on -19vdc rail i guess.
This will be all together 5,6A; so near the limit of 6A.
Will that work?
Jesper.
Btw.: If we can discuss this in my thread it would be better. I dont want to spoil youre thread Juma.
My thread is in my signature. And i will post youre PSU schm. there Juma shortly!
Google "class A amplifier". You'll find a lot of reading material on this subject at Nelson Pass' websites - DIY and FirstWatt.
As an ultimate experience I recommend to build this amp and see for yourself what's the sound and measurement difference between the class A and 40mA bias (I did it long ago).
As an ultimate experience I recommend to build this amp and see for yourself what's the sound and measurement difference between the class A and 40mA bias (I did it long ago).
Last edited:
Hello, could you please share your measurements results? I like very much the F5 simplicity. I have a F5 v3 board assembled, I need to check but I think I can adapt my board to a cubie ... but I'm very interested on a low bias .
Thanks
Sorry Bruno, but it seems that I don't have those notes with me. When I made this amp I lived in Belgrade and later that year I moved to Berlin, so I must have left them at my house in Belgrade.
Anyway, I remember that although THD was not spectacular (about 0.01) it consisted of low order harmonics only (H2 & H3), adjustable by usual tricks ("P3", small source resistor for output MOSFETs, etc...).
You'll find square wave response (90kHz at pure resistive and highly reactive load) in post #1. The amp is very fast and stable.
What I remember too is that it was very pleasant to listen with my Canton ct-1000 speakers (overall impression more on the "gentle" side) but later, on my Epos M5 (I brought them with me to Berlin because they are smaller) it sounded a bit too "gentle" so I moved on to Cubie2, 3 etc...
Have in mind that Cubie is a 10W amp so if you need more power you'll have to look somewhere else or redesign it a bit (few small interventions will turn it into F7 - all the info needed is in the F7 thread).
Anyway, I remember that although THD was not spectacular (about 0.01) it consisted of low order harmonics only (H2 & H3), adjustable by usual tricks ("P3", small source resistor for output MOSFETs, etc...).
You'll find square wave response (90kHz at pure resistive and highly reactive load) in post #1. The amp is very fast and stable.
What I remember too is that it was very pleasant to listen with my Canton ct-1000 speakers (overall impression more on the "gentle" side) but later, on my Epos M5 (I brought them with me to Berlin because they are smaller) it sounded a bit too "gentle" so I moved on to Cubie2, 3 etc...
Have in mind that Cubie is a 10W amp so if you need more power you'll have to look somewhere else or redesign it a bit (few small interventions will turn it into F7 - all the info needed is in the F7 thread).
Hello Juma. 0.01 % is a very good distortion, you know speaker drivers and room acoustics will distort the overall sound much ... much more then that. with a very good speaker driver will you maybe get 0.4-2% distortion.
I read the whole thread with great interest and I was tempted to try this little amp. My speakers have a sensitivity of 96dB, so I think there will be no problem for Cubie to drive them.
I wonder whether I could use current production lateral MOSFETs, namely the Exicon EXC10N20 and EXC10P20. I think that they come in matched pairs by the manufacturer.
I wonder whether I could use current production lateral MOSFETs, namely the Exicon EXC10N20 and EXC10P20. I think that they come in matched pairs by the manufacturer.
Giasou Kenev,
Cubie ia a nice little amp indeed, but Cubie 2 is better and it isn't much more complicated.
Also, if you look at the F7 thread you'll see that it's easy to turn a Cubie into F7.
Exicons are practically the same thing as 2sk1058/2sj162 (which are current production too and there is a groupbuy for them gong on now:
http://www.diyaudio.com/forums/group-buys/307426-sj162-sk1058-groupsbuy.html ).
You can also buy k1058/j162 on many sites (reichelt.de for example).
I don't really know if Exicons come as matched pairs from manufacturer but I doubt it...
Na'sai kala !
Cubie ia a nice little amp indeed, but Cubie 2 is better and it isn't much more complicated.
Also, if you look at the F7 thread you'll see that it's easy to turn a Cubie into F7.
Exicons are practically the same thing as 2sk1058/2sj162 (which are current production too and there is a groupbuy for them gong on now:
http://www.diyaudio.com/forums/group-buys/307426-sj162-sk1058-groupsbuy.html ).
You can also buy k1058/j162 on many sites (reichelt.de for example).
I don't really know if Exicons come as matched pairs from manufacturer but I doubt it...
Na'sai kala !
Geia sou juma,
Thank you for your immediate response! I hadn't notice the emergence of Cubie 2. Now, I will have to read the Cubie 2 thread and decide. But, since you suggest that ver. 2 is better, I have no reason not to follow your recommendations.
Regarding the lateral MOSFETs, I saw the gb thread and I might order a couple (if I'm not too late), but I quote from the last message by touchdown in that thread:
"Since the price difference is low (3.97 vs 4.19E) I vote for Exicon selected ones ...
Matched MosFets, no issues on P channel behavior, right price ... I think you can't get better Laterals in the market!
Just my opinion, though
"
So, maybe I should order from Exicon.
Thank you again,
Na' sai kala kai sy!
Vagelis
Thank you for your immediate response! I hadn't notice the emergence of Cubie 2. Now, I will have to read the Cubie 2 thread and decide. But, since you suggest that ver. 2 is better, I have no reason not to follow your recommendations.
Regarding the lateral MOSFETs, I saw the gb thread and I might order a couple (if I'm not too late), but I quote from the last message by touchdown in that thread:
"Since the price difference is low (3.97 vs 4.19E) I vote for Exicon selected ones ...
Matched MosFets, no issues on P channel behavior, right price ... I think you can't get better Laterals in the market!
Just my opinion, though
"So, maybe I should order from Exicon.
Thank you again,
Na' sai kala kai sy!
Vagelis
I checked datasheets for ECX10N20 and ECX10P20 and they show quite a difference in curves - I would not call them a matched pair (unless someone does a selection/matching job before sending them to customers).
Edit: OK, I've searched some more and found a page about their "selected" parts. it's here:
http://www.profusionplc.com/parts/ecx10n20-s
So, it's not really about matching P to N parts - it's matching N parts to N parts and P parts to P parts which makes paralleling easier, but it's not really tight - about 10% matches are offered. I can live without that because k1058/j162 come naturally in that match span. Matching P to N would be good ...
Edit: OK, I've searched some more and found a page about their "selected" parts. it's here:
http://www.profusionplc.com/parts/ecx10n20-s
So, it's not really about matching P to N parts - it's matching N parts to N parts and P parts to P parts which makes paralleling easier, but it's not really tight - about 10% matches are offered. I can live without that because k1058/j162 come naturally in that match span. Matching P to N would be good ...
Last edited:
So, it seems that your Greek is pretty good - how come this?
Well, I will better stick to the original 2SK1058/2SJ162 pair. A quick search in ebay gave plenty of results of available pairs, as, for example:
1pair Transistor HITACHI TO-3P 2SJ162/2SK1058 J162/K1058 100% Genuine and New | eBay
100% New And Genuine 2SJ162 2SK1058 Transistor K1058 J162 TO-3P | eBay
Could I trust such vendors?
I have two questions regarding those MOSFETs:
1. Is it necessary and/or possible to match P and N channel MOSFETs? And, if yes, how close this matching should be?
2. Could this 2SK1058/2SJ162 pair be used also to build an F7?
Regards,
Vagelis
I lived 4 years long in Athens kai ematha ligaki tin Elliniki glossa...
1. No need to match P to N - they are close enough by nature.
2. Yes - All modern lateral Mosfets are spin-offs of original Hitachi parts (2sk135/2SJ50) and they hardly differ one from another.
So, Exicons, Profusion or Renesas doesn'r really matter, but don't buy them on ebay, there's a 99% chance that you'll buy fake parts. Buy from big renowned sellers (Reichelt, Mouser, Digikey, Farnell and similar).
1. No need to match P to N - they are close enough by nature.
2. Yes - All modern lateral Mosfets are spin-offs of original Hitachi parts (2sk135/2SJ50) and they hardly differ one from another.
So, Exicons, Profusion or Renesas doesn'r really matter, but don't buy them on ebay, there's a 99% chance that you'll buy fake parts. Buy from big renowned sellers (Reichelt, Mouser, Digikey, Farnell and similar).
I lived 4 years long in Athens kai ematha ligaki tin Elliniki glossa...
Den emathes aplos ligo tin Elliniki glossa, tin emathes poly kala!
Thank you for your help so far. I am going to get a few pairs from the still ongoing forum group buy.
Best regards,
Vagelis
Some speakers are sensitive to humidity.. who knows what is the cause of 'gentle touch'..
