Terry - nicely done. Glad it all went well for you. I would also be interested to hear your impressions of regulated versus not.
This is your first Leach I presume - how would you rate it against other amps you have built? I really like mine and can't wait to hear what the updated components and layout do for it.
What regulation did you go with in the end?
This is your first Leach I presume - how would you rate it against other amps you have built? I really like mine and can't wait to hear what the updated components and layout do for it.
What regulation did you go with in the end?
Hi Terry,
could you make a commercial propositon out of making cases for us?
That's the nicest looking Leach I have seen.
Glad to see you went for 4*27mF smoothing caps. Are they four separate rails or two // rails?
Waiting to hear comparisons. I'm guessing Leach on bass, Krell on top/mid and P101 in the garage.
could you make a commercial propositon out of making cases for us?
That's the nicest looking Leach I have seen.
Glad to see you went for 4*27mF smoothing caps. Are they four separate rails or two // rails?
Waiting to hear comparisons. I'm guessing Leach on bass, Krell on top/mid and P101 in the garage.
Terry your case looks big and cool enough for five channels! I can see putting two boards per side and one across the rear and add some heatsinking for the rear across the top, uprating the transformer to a 1000 to 1500VA and putting the caps above the transformer. Would make a top notch HT amp!
Or did you plan on using this case for the Super Leach?
Or did you plan on using this case for the Super Leach?
Hi guys,
Thanks for the kind words. I won't have time to do it today but I will try it with and without regulation to see if it is detectable. I am interested as well. I think I can do it with a two pole switch so it would be quick. I'll get the stuff together so I can do it sometime this week.
I used the regulation circuit for the A75 that Bill had suggested. I just built it on a perf board. Another first for me. It was easier than I thought it would be. Seems to work fine.
Hi Andrew,
Yes I went with three bridges. One for the regulation and two for the back end, so there are four separated rails.
The Krell is smoother on top but much less powerful. After all, the Krell is 39V rails and the Leach is 64V. My P101 is 64V rails as well with 700vA. It falls somewhere in between the two. It may have a slightly smoother top end than the Leach. All three are a pleasure to listen to. It would be interesting to bi-amp the Leach and the Krell and see how it sounds.
I'm not at the point where I would feel comfortable making cases for others just yet. This case was made from some scrap aluminum I had been saving. I would not go with such thick plate for the rear panel again. It was a real pain mounting things in it. None of the jacks I had would work with out relieving some of the metal. It's just what I had on hand. The case itself weighs 14 lbs. I prefer a shorter case but I had these heatsinks set up for the Superamp I had built that didn't work. I think now that I have built this amp I may understand the circuit enough to find the problem with the superamp. I have a 55-0-55 1000vA transformer for that one. Of course, now I I need some more heat sinks.
Oh, before I forget, what size fuse should I use at the front?
Thanks again, Terry
Thanks for the kind words. I won't have time to do it today but I will try it with and without regulation to see if it is detectable. I am interested as well. I think I can do it with a two pole switch so it would be quick. I'll get the stuff together so I can do it sometime this week.
I used the regulation circuit for the A75 that Bill had suggested. I just built it on a perf board. Another first for me. It was easier than I thought it would be. Seems to work fine.
Hi Andrew,
Yes I went with three bridges. One for the regulation and two for the back end, so there are four separated rails.
The Krell is smoother on top but much less powerful. After all, the Krell is 39V rails and the Leach is 64V. My P101 is 64V rails as well with 700vA. It falls somewhere in between the two. It may have a slightly smoother top end than the Leach. All three are a pleasure to listen to. It would be interesting to bi-amp the Leach and the Krell and see how it sounds.
I'm not at the point where I would feel comfortable making cases for others just yet. This case was made from some scrap aluminum I had been saving. I would not go with such thick plate for the rear panel again. It was a real pain mounting things in it. None of the jacks I had would work with out relieving some of the metal. It's just what I had on hand. The case itself weighs 14 lbs. I prefer a shorter case but I had these heatsinks set up for the Superamp I had built that didn't work. I think now that I have built this amp I may understand the circuit enough to find the problem with the superamp. I have a 55-0-55 1000vA transformer for that one. Of course, now I I need some more heat sinks.
Oh, before I forget, what size fuse should I use at the front?
Thanks again, Terry
Hi Terry,
400VA divided by 110V = 3.6A at max draw. Transformers need a fuse rating about 3 times max draw to allow for starting current giving you a 12A fuse. Choose the next highest available.
Do you see the problem?
The amp and anything near it will be on fire before that fuse blows. A fuse blows quickly only when large currents pass through it (10 to 100 times rating). At a current of about 2 times rating the fuse may pass for 1 second to 30 seconds and at 125% of rating could pass current for hours, cooking everything within reach.
A soft start will allow a fuse rating about the same as max draw. In this case choose the next size down.
When you do this a fault that starts to overdraw current will blow the fuse before the amp has too much time to seriously overheat.
A good general rule for fusing is to put in the smallest fuse that allows the item to operate normally without the fuse blowing every few weeks.
400VA divided by 110V = 3.6A at max draw. Transformers need a fuse rating about 3 times max draw to allow for starting current giving you a 12A fuse. Choose the next highest available.
Do you see the problem?
The amp and anything near it will be on fire before that fuse blows. A fuse blows quickly only when large currents pass through it (10 to 100 times rating). At a current of about 2 times rating the fuse may pass for 1 second to 30 seconds and at 125% of rating could pass current for hours, cooking everything within reach.
A soft start will allow a fuse rating about the same as max draw. In this case choose the next size down.
When you do this a fault that starts to overdraw current will blow the fuse before the amp has too much time to seriously overheat.
A good general rule for fusing is to put in the smallest fuse that allows the item to operate normally without the fuse blowing every few weeks.
Hi Andrew,
Thanks for the explanation. Would maybe a lower value slow blow fuse work better then? Maybe slow enough to get past start-up but low enough to blow if there is a problem? I did use a thermistor however, when starting it at first when using the light bulb in the mains line the bulb burned pretty brightly for the first few seconds. I'll try smaller fuses and see what I can get away with. One of the reasons I used the kind of fuse holder that I did was to make fuse changes easier.
Hi lgreen,
Thanks for the compliment.
I did look at the Aleph X a while ago but I think it was still in morph stage at that time. I suppose folks have built successfully by now. Maybe I will take a look at them.
I would like to maybe build a DIY preamp first. I have some pretty good pres but would still like to build one.
Also, after doing a lot of listening back and forth between the three amps I've built, the Krell was the clear loser to my ears. It is now residing in the garage with a PAS 3 pre and a tuner. I gave it some good speakers and it sounds really good out there. I still can't decide between the P101 and the Leach. They are so close it is very tough to decide.
Blessings, Terry
Thanks for the explanation. Would maybe a lower value slow blow fuse work better then? Maybe slow enough to get past start-up but low enough to blow if there is a problem? I did use a thermistor however, when starting it at first when using the light bulb in the mains line the bulb burned pretty brightly for the first few seconds. I'll try smaller fuses and see what I can get away with. One of the reasons I used the kind of fuse holder that I did was to make fuse changes easier.
Hi lgreen,
Thanks for the compliment.
I did look at the Aleph X a while ago but I think it was still in morph stage at that time. I suppose folks have built successfully by now. Maybe I will take a look at them.
I would like to maybe build a DIY preamp first. I have some pretty good pres but would still like to build one.
Also, after doing a lot of listening back and forth between the three amps I've built, the Krell was the clear loser to my ears. It is now residing in the garage with a PAS 3 pre and a tuner. I gave it some good speakers and it sounds really good out there. I still can't decide between the P101 and the Leach. They are so close it is very tough to decide.
Blessings, Terry
Hi Terry
Practically a lower value slow blow fuse should work. IE 3.5A. This is in line with Leach's original recommendation and I haven't heard of too many of his student's houses burning down. You could even go lower, and when encountering a regular problem - go for a slightly higher value
I run a large toroid with no inrush limiting and slow blow fuses rated for transformer VA and have had no problems for the past 2 years.
Was the Krell the loser simply for lower power - or in terms of tonal quality?
Practically a lower value slow blow fuse should work. IE 3.5A. This is in line with Leach's original recommendation and I haven't heard of too many of his student's houses burning down.
You could even go lower, and when encountering a regular problem - go for a slightly higher valueI run a large toroid with no inrush limiting and slow blow fuses rated for transformer VA and have had no problems for the past 2 years.
Was the Krell the loser simply for lower power - or in terms of tonal quality?
Hi Ross,
Thanks for the conformation. I will try some small value slow blow fuses and see how it does. I've been told that E-core transformers aren't as hard on the fuses.
As far as the Krell goes. I set up the amps with seperate preamps so I could adjust the volumes to the same audible loudness. The Krell doesn't sound bad but it just sounded congested compared to the others. The sound stage was just much larger with the others.
Now having said that, it could very easily be the way I built the Krell and the types of parts I used. I didn't do any of the protection circuit nor the componants past the emitter resistors. I could probably try those as well as the by-pass caps that someone suggested.
Hi pooge,
I set the bias at 100mA. I got that value off of The Leach website. If I should make it higher or lower I would like to hear about that.
Thanks, Terry
Thanks for the conformation. I will try some small value slow blow fuses and see how it does. I've been told that E-core transformers aren't as hard on the fuses.
As far as the Krell goes. I set up the amps with seperate preamps so I could adjust the volumes to the same audible loudness. The Krell doesn't sound bad but it just sounded congested compared to the others. The sound stage was just much larger with the others.
Now having said that, it could very easily be the way I built the Krell and the types of parts I used. I didn't do any of the protection circuit nor the componants past the emitter resistors. I could probably try those as well as the by-pass caps that someone suggested.
Hi pooge,
I set the bias at 100mA. I got that value off of The Leach website. If I should make it higher or lower I would like to hear about that.
Thanks, Terry
Leach Bias
Since the Leach amp has an emitter-follower output stage, D.Self and R.Slone are adamant that optimal bias is a very narrow range.
It is ideally 47-54mV across both Re of ONE output pair. In this case, with Re = 0.47R, Vq = 54mV. So Iq = 59mA. Since this version of the Leach has three output pairs instead of the original two, I'd say bias current should be set to 59X3 = 177mA.
Give that a try and see if you like it (unless you have gear to actually measure optimum bias). Have fun.
Since the Leach amp has an emitter-follower output stage, D.Self and R.Slone are adamant that optimal bias is a very narrow range.
It is ideally 47-54mV across both Re of ONE output pair. In this case, with Re = 0.47R, Vq = 54mV. So Iq = 59mA. Since this version of the Leach has three output pairs instead of the original two, I'd say bias current should be set to 59X3 = 177mA.
Give that a try and see if you like it (unless you have gear to actually measure optimum bias). Have fun.
Transformer Soft-Start
Psst, this is off-topic but since there was some talk about suitable input fusing for the power transformer(s), I just want to let you know I am working on a project to address that.
It has the following features:
1) Compact layout
2) Soft-start (can support > 1kVA loads)
3) 12V DC trigger support
4) DC power button (no dangerous AC going to power switch)
5) Configurable power button (can support either SPST momentary or toggle action switches)
6) Support for LED illuminated power buttons
7) Configurable for 110/220AC operation
8) DC trap (to prevent transformer from humming)
9) Filtered AC with Class X capacitors
Disclaimer: This is not an advertisement coz I won't be selling anything for profit. It's for my personal use and I'll gladly share it with the community.
Back to the Leach....
Psst, this is off-topic but since there was some talk about suitable input fusing for the power transformer(s), I just want to let you know I am working on a project to address that.
It has the following features:
1) Compact layout
2) Soft-start (can support > 1kVA loads)
3) 12V DC trigger support
4) DC power button (no dangerous AC going to power switch)
5) Configurable power button (can support either SPST momentary or toggle action switches)
6) Support for LED illuminated power buttons
7) Configurable for 110/220AC operation
8) DC trap (to prevent transformer from humming)
9) Filtered AC with Class X capacitors
Disclaimer: This is not an advertisement coz I won't be selling anything for profit. It's for my personal use and I'll gladly share it with the community.
Back to the Leach....
still4given said:
The value of bias that Leach specified is for two sets of output transistors, so the value you used is definitely low for three sets. I have found that the Leach sounds smoother with a higher bias than he specified. Should be no trouble with your heat sinks, although it might take quite some time to adjust with their size. I would definitely try to adjust by measuring the voltage across a series pair of output emitter resistors, and adjusting for 54.8 mv across the pair, as Arius stated, if you used .47ohm resistors. This value would be 47.6mv for .33ohm resistors.
With .33 ohm resistors, the bias current would be approximately 74mA per transistor. With .47ohms, it would be 59mA per. With my current version using .33ohm resistors and 74mA per, there is no problem. Leach himself indicated you could try up to 150ma or more, so don't worry.
So with .47ohm resistors, your total bias current should be in the 3 x 59mA neighborhood, or about 180mA. I found that adjusting by measuring voltage was a LOT easier to do than with current. The voltage seems to wander a lot less, making it an easier target. It wandered even less when I added heat sinks to the VAS and the predriver. The overall sound got smoother and more "effortless". As you don't presently have the bias current as high in each transistor as Leach recommends, I think you will hear an improvement.
jacco vermeulen said:
Terry,
Since you have the heat sink to play with, it would be interesting to see whether you think Pass style heavily biased Class AB (As much as it can handle while maintaining reasonable junction temperatures) or Self/Slone's "optimum" bias AB sounds better, or makes a noticeable change.
When Jacco gets his class A Leach up and running, we'll have two opinions.
FYI, I set the total bias on my standard Leach amps at 150 mA. It sounds better than 100 mA, but not quite as smooth as my A75 at 650 mA total, the sound of which did not change noticeably when reduced from 1.3 A.
I'll run the junction temp calcs for you when I get a chance.
Hey guys, thanks for all of the advice.
I will try to raise it today and see how it sounds.
I'm not sure what you mean by across a series pair. are you talking about clipping the leads to the outside legs of a pair of opposing emitter resistors and measuroing the drop across both of them? On the Krell I measured the drop across only one at a time. It may be easier to measure the current since the emmiter resistors are at the bottom of the case, but if it is nore accurate to measure the voltage drop I will try that.
Blessings, Terry
I will try to raise it today and see how it sounds.
I'm not sure what you mean by across a series pair. are you talking about clipping the leads to the outside legs of a pair of opposing emitter resistors and measuroing the drop across both of them? On the Krell I measured the drop across only one at a time. It may be easier to measure the current since the emmiter resistors are at the bottom of the case, but if it is nore accurate to measure the voltage drop I will try that.
Blessings, Terry
Maximum current based on 50C heat sink temperature calculations:
50C junction temp rise above heat sink. (100C junction temperature limit for reliability, 50C below max rating)
Isolator thermal resistance, roughly 1.4 K/W (silpad)
Device Juntion to case thermal resistance MJL 3281 - .625, MJL4281 - .54 (I cannot remember which you used)
Junction to heat sink thermal resistance -
3281 - 2.025 (.625 + 1.4)
4281 - 1.94 (.54 + 1.4)
Junction rise above ambient = Thermal resistance * power
rearanging, to solve for power
Power per device = 50C rise/ Thermal resistance
3281 = 24.6W
4281 = 25.7W
Solve for current - I=P/V (V=64 for Terry)
3281 - .380A = 24.6W/64V
4281 - .401A = 25.7W/64V
Solve for voltage across emitter resistors V=IR
0R33
3281 - .127V
4281 - .132V
0R47
3281 - .179V
4281 - .188V
Class A power at this bias setting = 1/2 (2*I)^2/R (push pull output stage can provide twice bias current)
3281 - 20W
4281 - 23.4 W
Slowly increase the bias and allow the heat sink temperature to stabilize for an hour or so between adjustments (top on the case) with no input signal. If the heat sink temperature reaches 50C before you get to the bias current target, stop as you are heat sink limited. (I suspect you will be, as big as they are, those sinks actually look small for the ~150W called for by the target current)
If you reach the target bias before you reach 50C heat sinks, you can squeeze a little more out of it. If you want to try, we can calculate your new target current.
The reason for no input signal is that in class A operation, Idle is the worst case - power delivered to the speaker reduces the heat disspated on the sinks.
50C junction temp rise above heat sink. (100C junction temperature limit for reliability, 50C below max rating)
Isolator thermal resistance, roughly 1.4 K/W (silpad)
Device Juntion to case thermal resistance MJL 3281 - .625, MJL4281 - .54 (I cannot remember which you used)
Junction to heat sink thermal resistance -
3281 - 2.025 (.625 + 1.4)
4281 - 1.94 (.54 + 1.4)
Junction rise above ambient = Thermal resistance * power
rearanging, to solve for power
Power per device = 50C rise/ Thermal resistance
3281 = 24.6W
4281 = 25.7W
Solve for current - I=P/V (V=64 for Terry)
3281 - .380A = 24.6W/64V
4281 - .401A = 25.7W/64V
Solve for voltage across emitter resistors V=IR
0R33
3281 - .127V
4281 - .132V
0R47
3281 - .179V
4281 - .188V
Class A power at this bias setting = 1/2 (2*I)^2/R (push pull output stage can provide twice bias current)
3281 - 20W
4281 - 23.4 W
Slowly increase the bias and allow the heat sink temperature to stabilize for an hour or so between adjustments (top on the case) with no input signal. If the heat sink temperature reaches 50C before you get to the bias current target, stop as you are heat sink limited. (I suspect you will be, as big as they are, those sinks actually look small for the ~150W called for by the target current)
If you reach the target bias before you reach 50C heat sinks, you can squeeze a little more out of it. If you want to try, we can calculate your new target current.
The reason for no input signal is that in class A operation, Idle is the worst case - power delivered to the speaker reduces the heat disspated on the sinks.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.