Shaun, The wiki that you have made reference to is for future interest only. We are in the process of EVALUATING the 16 output device Super Leach.
The current group buy -- which is now full -- was for the 10 transistor Low TIM Leach amplifier.
When the prototype beta tests are done on the Super Leach me or someone else will start a group buy for it.
Tad
Bob, I just looked at the Super Leach wiki. There is tremendous interest there. Over 200 boards --- at 20.00 plus each-- ouch. Plus all of the 200 volt rated parts. What a monster.
Tad
The current group buy -- which is now full -- was for the 10 transistor Low TIM Leach amplifier.
When the prototype beta tests are done on the Super Leach me or someone else will start a group buy for it.
Tad
Bob, I just looked at the Super Leach wiki. There is tremendous interest there. Over 200 boards --- at 20.00 plus each-- ouch. Plus all of the 200 volt rated parts. What a monster.
Tad
AndrewT,
What is considered a proper method of testing zeners for matching? I have not seen this in any of the material. Tad
What is considered a proper method of testing zeners for matching? I have not seen this in any of the material. Tad
series connect the Zener (DUT) with a 1k0 resistor.tryonziess said:
Apply a DC voltage to the pair and measure the current passing the 1k0 resistor. Adjust the DC supply voltage till the current is approx 10% of max Zener rating. eg 400mW 20V Zener, Imax=20mA, Itest=2mA.
Measure the Zener volts drop.
Now swap Zeners to find matching Zeners.
For the Leach you simply want pairs that add up to ~40V and the pairs do not need to be matched, it's the totals that must be matched. eg. 19.5+20.6=19.7+20.4=40.1V on the cascode and CCS. I'm sure Leach mentions this in his papers.
samoloko said:
Your call Sam. Prof Leach suggests around 7,000 uf for a stereo amp, IIRC. I've got a 4 channel Leach amp with 39,000 uf total that sounds fine. Some say at least 10,000 uf on each ral per 100W. For class A operation 10,000 uf per amp of output is reasonable.
The only catch is when you get much over 50,000 uf you need to look at an inrush limiting. That can be as simple as a CL30 in series with the transformer primary.
http://www.diyaudio.com/forums/showthread.php?postid=1640179#post1640179
"The rest" refers to C22-26, 28-32. Just a little extra decoupling that wasn't in the original Leach design. Your caps are fine, as would be no caps.
I used Panasonic FC for some, Nichicon low ESR for other electrolytics. Epcos film caps do nicely for bypass duty. You could go crazy with Wima FKP in addition to the main bypass, but I doubt you'll hear a difference.
"The rest" refers to C22-26, 28-32. Just a little extra decoupling that wasn't in the original Leach design. Your caps are fine, as would be no caps.
I used Panasonic FC for some, Nichicon low ESR for other electrolytics. Epcos film caps do nicely for bypass duty. You could go crazy with Wima FKP in addition to the main bypass, but I doubt you'll hear a difference.
Thanks, Bob, for your input.
I have some new/old stock Panasonic HF & HFS 330uf/25v left over from an old CD player "POOGE" 10-15 years ago- Do you think these would be large enough to substitute for the 470uf/16v at C7 & C12, or does one truly need 470uf (and the parts list specifiesC7/12 as "Not-Polarised"; is this correct) ?
Also, anyone have advice on which parts' quality are most crucial to best sound in this amp? So far I've identified C6/7, R11, R6/18, R9/14, R10/15, R12 and R13, as being directly in the signal path...
Thanks,
-chas
I have some new/old stock Panasonic HF & HFS 330uf/25v left over from an old CD player "POOGE" 10-15 years ago- Do you think these would be large enough to substitute for the 470uf/16v at C7 & C12, or does one truly need 470uf (and the parts list specifiesC7/12 as "Not-Polarised"; is this correct) ?
Also, anyone have advice on which parts' quality are most crucial to best sound in this amp? So far I've identified C6/7, R11, R6/18, R9/14, R10/15, R12 and R13, as being directly in the signal path...
Thanks,
-chas
Also consider the feedback path resistors. Rs 7, 8, 12, 13, 16, 19.
If you are going to include R9/10 and 14/15 in the signal path, why not R5 and R20? Why not The emitter resistors in the VAS? (R24, 25, 33,35) Why not the differential's load resistors R2 an 21? Just to get carried away, the signal touches everything but the protection circuit resistors.
I am not an advocate of exotic parts, but I've heard that feedback path resistors make the most difference. Use high quality parts if you wish, but Professor Leach said 5% resistors were good enough. I used 1% Panasonic's and it sounds great.
C7 and 12 must be non polarized - using a polarized cap would have a larger detrimental effect than low grade resistors. Leach used 220 uf for C12 and didn't use C7. Larger values mean lower rolloff at the bottom end. Nichicon Muse caps are 1 mm bigger than the space allotted, but you could probably squeeze them in.
If you are going to include R9/10 and 14/15 in the signal path, why not R5 and R20? Why not The emitter resistors in the VAS? (R24, 25, 33,35) Why not the differential's load resistors R2 an 21? Just to get carried away, the signal touches everything but the protection circuit resistors.
I am not an advocate of exotic parts, but I've heard that feedback path resistors make the most difference. Use high quality parts if you wish, but Professor Leach said 5% resistors were good enough. I used 1% Panasonic's and it sounds great.
C7 and 12 must be non polarized - using a polarized cap would have a larger detrimental effect than low grade resistors. Leach used 220 uf for C12 and didn't use C7. Larger values mean lower rolloff at the bottom end. Nichicon Muse caps are 1 mm bigger than the space allotted, but you could probably squeeze them in.
Good news.
It looks like our boards have shipped a few weeks early. I just received an email from Advanced Circuits stating that the order had shipped. Of course we can not count this, for sure, until they hit my doorstep. Just maybe.
On another note there are still individuals who have not confirmed there order totalling 24 pcb's. Please expedite so I can contact those individuals on the waiting list. And thanks to all of you who have already paid. My bill just arrived yesterday and I should be able to pay.
Tad
It looks like our boards have shipped a few weeks early. I just received an email from Advanced Circuits stating that the order had shipped. Of course we can not count this, for sure, until they hit my doorstep. Just maybe.
On another note there are still individuals who have not confirmed there order totalling 24 pcb's. Please expedite so I can contact those individuals on the waiting list. And thanks to all of you who have already paid. My bill just arrived yesterday and I should be able to pay.
Tad
Good News......
Tad,
Would you please resend invoice. I've had problems with email and have not been able to find it.
many thanks,
Ray (aka: brutepuppy)
Tad,
Would you please resend invoice. I've had problems with email and have not been able to find it.
many thanks,
Ray (aka: brutepuppy)
C12 non-polar?
Hey, Bob
Just to clarify... while I understand C7 needs to be non-polar as it's directly in the signals path, is C12(which is in shunt between R16 and GND) also part of the high-pass filter? 😕
I'm thinking of using a Blackgate "N" series at C7 and a "Muse" at C12, both with good film bypasses (not to mention PRP 1% MF resistors in all critical locations)... 😎
Hey, Bob
Just to clarify... while I understand C7 needs to be non-polar as it's directly in the signals path, is C12(which is in shunt between R16 and GND) also part of the high-pass filter? 😕
I'm thinking of using a Blackgate "N" series at C7 and a "Muse" at C12, both with good film bypasses (not to mention PRP 1% MF resistors in all critical locations)... 😎
Brutepuppy,
I have sent numerous emails to you at all of the addresses you have provided. I will not send anymore. Tad
I have sent numerous emails to you at all of the addresses you have provided. I will not send anymore. Tad
CLM -
C12 is in the feedback path - a most critical spot for good quality components according to some. It and R16 form a high pass filter that blocks DC amplification. The larger you make C12, the lower the cutoff frequency of the amp. Professor Leach gives a pretty good explanation (easy for non EEs to understand) of the various parts of the circuit. Suggested reading.
If you are sure that your source will NEVER send a DC signal (e.g. output caps on pre), you can jumper it and C7 out. If you do jumper it, be aware that any DC input will be amplified by a factor of 22. (1 VDC at the input = 22 VDC out)
Again, C7//C6 is an addition Jens made to the circuit. hundreds if not thousands of Leach amps work fine without a blocking cap on the input. Using C12 brings the DC gain to unity. (1 VDC at the input = 1 VDC out) If you are of the school that says the best cap is no cap, jumper it out.
Jens' philosophy is that decent quality parts are all it takes. He allowed space for resistors with double the power rating that Professor Leach specified. That should minimize the thermal effects (if any). Notice that the space allotted for caps is tight for most exotics and that you cannot use Caddock emitter resistors in the output without some serious lead bending.
Unless your speakers are top notch I doubt you'll notice any difference in resistors or caps as long as you use appropriate values. Even then, could you tell the difference blind?
Don't over think this stuff. I'm not looking to get into an argument about audibility of caps and resistors. Just pointing out is easy to drive yourself crazy with what is the best part for this position decisions. Pick something and enjoy the project. Built correctly the Leach amp will sound great whatever brand of components you use.
For everyone who agrees with your choice there will be at least two people who will tell you that you should have use brand ZZZ since your choice sounds terrible. A lot like arguing religion.
Could it be improved with different components? Possibly. Then again some people would prefer that the differential was loaded by a current mirror rather than resistors and fed with constant current source/sink.
Build a few in different configurations and decide what if any sounds best and put that version in a good case. That's the joy of this hobby. I've got a few variations in the works that I think will have more impact on the sound than exotic parts - mosfet input and output, lateral mosfet output, jfet input, all FET... Something tells me I'll keep them all. 😉
C12 is in the feedback path - a most critical spot for good quality components according to some. It and R16 form a high pass filter that blocks DC amplification. The larger you make C12, the lower the cutoff frequency of the amp. Professor Leach gives a pretty good explanation (easy for non EEs to understand) of the various parts of the circuit. Suggested reading.
If you are sure that your source will NEVER send a DC signal (e.g. output caps on pre), you can jumper it and C7 out. If you do jumper it, be aware that any DC input will be amplified by a factor of 22. (1 VDC at the input = 22 VDC out)
Again, C7//C6 is an addition Jens made to the circuit. hundreds if not thousands of Leach amps work fine without a blocking cap on the input. Using C12 brings the DC gain to unity. (1 VDC at the input = 1 VDC out) If you are of the school that says the best cap is no cap, jumper it out.
Jens' philosophy is that decent quality parts are all it takes. He allowed space for resistors with double the power rating that Professor Leach specified. That should minimize the thermal effects (if any). Notice that the space allotted for caps is tight for most exotics and that you cannot use Caddock emitter resistors in the output without some serious lead bending.
Unless your speakers are top notch I doubt you'll notice any difference in resistors or caps as long as you use appropriate values. Even then, could you tell the difference blind?
Don't over think this stuff. I'm not looking to get into an argument about audibility of caps and resistors. Just pointing out is easy to drive yourself crazy with what is the best part for this position decisions. Pick something and enjoy the project. Built correctly the Leach amp will sound great whatever brand of components you use.
For everyone who agrees with your choice there will be at least two people who will tell you that you should have use brand ZZZ since your choice sounds terrible. A lot like arguing religion.
Could it be improved with different components? Possibly. Then again some people would prefer that the differential was loaded by a current mirror rather than resistors and fed with constant current source/sink.
Build a few in different configurations and decide what if any sounds best and put that version in a good case. That's the joy of this hobby. I've got a few variations in the works that I think will have more impact on the sound than exotic parts - mosfet input and output, lateral mosfet output, jfet input, all FET... Something tells me I'll keep them all. 😉
You're the man!
Thanks again, Bob, for the "sound" advice(pun intended)! I'll re-read Leach's papers again(it's been a while).
Sometimes that 2nd cup of coffee with dinner stirs up the audiophool in me,and I start ruminating about this or that "super" part (what if...)!
BTW I've been having a hard time finding a big snap-in cap for C21/27*; I'd like to avoid the various fakes on ebay and noticed Digikey has some Panasonic 4700uf/100v jobs- these large enough(the parts list says 10,000uf-!)?
Regards to all,
-chas
*I have some 4700uf/80v, but my rails are +/-75v(too close for comfort).
Thanks again, Bob, for the "sound" advice(pun intended)! I'll re-read Leach's papers again(it's been a while).
Sometimes that 2nd cup of coffee with dinner stirs up the audiophool in me,and I start ruminating about this or that "super" part (what if...)!
BTW I've been having a hard time finding a big snap-in cap for C21/27*; I'd like to avoid the various fakes on ebay and noticed Digikey has some Panasonic 4700uf/100v jobs- these large enough(the parts list says 10,000uf-!)?
Regards to all,
-chas
*I have some 4700uf/80v, but my rails are +/-75v(too close for comfort).
Re: You're the man!
big caps here and in the front end will blow the PCB fuses at start up.clm811 said:
Check out www.apexjr.com for caps. He has some 68,000 uf 100V computer grade caps at $25 that would make a great main bank - At that rate you could easily use smaller caps on the board.
Steve is a great guy and has all sorts of cool surplus stuff. He doesn't have the time to put everything he has on the website. Call and see if he has some appropriate caps or heat sinks. I bought a couple dozen 15,000 uf/100V snap ins a while back.
For lower voltage applications he has some 15,000 uf/63V snap ins for $2.99. Yes, I'd use them with 40VAC transformers ~63V rails.
Oops, I'd better order my caps before everyone else steals them. 😀
EDIT: I use slow blow fuses to avoid surge issues.
Steve is a great guy and has all sorts of cool surplus stuff. He doesn't have the time to put everything he has on the website. Call and see if he has some appropriate caps or heat sinks. I bought a couple dozen 15,000 uf/100V snap ins a while back.
For lower voltage applications he has some 15,000 uf/63V snap ins for $2.99. Yes, I'd use them with 40VAC transformers ~63V rails.
Oops, I'd better order my caps before everyone else steals them. 😀
EDIT: I use slow blow fuses to avoid surge issues.
Bob, I got a great deal on 37000 uf caps from JEA capacitors on ebay. I have 4 per rail. Works great. Tad
Tad,
Thank you for your concern. Would someone here send me the address address so I can send in my G__D____ forty bucks for the boards. I really appreciate your help, Tad. It's good that you let everyone else here know what a nice guy you are in print. Instead of trying to help to solve a problem, you are one.
regards
Thank you for your concern. Would someone here send me the address address so I can send in my G__D____ forty bucks for the boards. I really appreciate your help, Tad. It's good that you let everyone else here know what a nice guy you are in print. Instead of trying to help to solve a problem, you are one.
regards