Dr. Leach in his original article claims 5 percent resistors will work just fine in his amp. Of course at that time DIY audio or even the web was not even a spark in anyones mind.
Since you usually only match NPN to NPN and PNP to PNP a match of 6 should work here. Finding a larger group within 1 percent could mean quite a few units. All of the Mills resistors in my last Leach checked to within .5 ohm.
You can always buy precision metal films and parallel like Nelson and Upopa do on there amps. You could get to within half of a percent. It is also much cheaper than power output devices and they are non-inductive.
Has anyone noticed that there is a 10 ohm output resistor on the ground and speaker output. Are both to be used? This is the first time I have encountered the resistor on the ground side. I guess it is for balanced output.
Tad
Since you usually only match NPN to NPN and PNP to PNP a match of 6 should work here. Finding a larger group within 1 percent could mean quite a few units. All of the Mills resistors in my last Leach checked to within .5 ohm.
You can always buy precision metal films and parallel like Nelson and Upopa do on there amps. You could get to within half of a percent. It is also much cheaper than power output devices and they are non-inductive.
Has anyone noticed that there is a 10 ohm output resistor on the ground and speaker output. Are both to be used? This is the first time I have encountered the resistor on the ground side. I guess it is for balanced output.
Tad
that is to avoid ground loops, frankly i never used them, i just used a central ground and my amps are very very quiet....
I started with cheap resistors. I spent less than $10 for all of them...with shipping.
Both batches of resistors are 5% rated. The readings I gathered say each batch falls into a 2% range. I can discard the odd resistor that is out-of-range.
That increases confidence.
Combining them as described above will narrow the range they fall into.
My use of metal oxide film as trimmers will decrease the effective inductance of the combination.
A link to the WT3:
http://www.parts-express.com/pe/sho... woofer tester&CFID=30426388&CFTOKEN=60956987
Both batches of resistors are 5% rated. The readings I gathered say each batch falls into a 2% range. I can discard the odd resistor that is out-of-range.
That increases confidence.
Combining them as described above will narrow the range they fall into.
My use of metal oxide film as trimmers will decrease the effective inductance of the combination.
A link to the WT3:
http://www.parts-express.com/pe/sho... woofer tester&CFID=30426388&CFTOKEN=60956987
power supply that does not sag
Incorporating a regulated frontend for the Leach 12 clone is one of my objectives for this build. However, would it be possible to build a main supply for the output devices that will not sag more than 5 volts. Overall output is projected at 300 watts so a stout 1.5kw transformer per channel should more than supply the rails with stable amperage. With a projected capacitance of 120,000uf per channel will this be enough.
I fail to see why the rails sag so much if the power supply is designed in advance to handle the output. This constant shifting of the voltage up and down can not be good for the sonic quality.
Tad
Incorporating a regulated frontend for the Leach 12 clone is one of my objectives for this build. However, would it be possible to build a main supply for the output devices that will not sag more than 5 volts. Overall output is projected at 300 watts so a stout 1.5kw transformer per channel should more than supply the rails with stable amperage. With a projected capacitance of 120,000uf per channel will this be enough.
I fail to see why the rails sag so much if the power supply is designed in advance to handle the output. This constant shifting of the voltage up and down can not be good for the sonic quality.
Tad
Hi Tad,
voltage drop can be quite high. I did some modeling using duncans PSU designer and an Xcel spreadsheet. I suggest you try doing this, its quite an interesting excercise.
Download
All capacitance does is reduce the ripple, it wont have any effect on rail sag. Rail sag, to the best of my knowledge is caused by the losses in the copper wire. As the current increases, more voltage is lost, its basic ohms law stuff.
Ripple is difference and can be managed by capacitance, having said that, the ripple effect is more profound at higher currents.
With my trasformer fully tapped out and delivering approx 500W/ch Peak, my rails sagged from 91V down to 76. I used a 65 + 65 1KVA Tx.
I guess the way to deal with this is to have a front end reg that can deal with 20V overhead.
Take a look how levinson have done, it's on page two. they have allowed over 30 volts, I assume, so at full output the frontend will still be regulating. Fancy breadboarding this, its a little on the hard side I admit.
Feel free to email me if you need help with this.
BTW nice amount of capacitance, looks like a quality build.
Last edited:
yes, the transformer can be simplified to it's Thevenin equivalent circuit and the secondary resistance (referred to primary) can be measured to determine the regulation. that is why bigger traffo with higher current density will have lower equivalent resistance and thus have better regulation.
Luke, I assume Levinson is using the elevated voltage for the frontend from an auxiliary tap on the main tranny. If I use a totally independent power source for the frontend -- tranny and all just a 5 volt increase will always be above the rails. With the voltages we are using having over 100 volts on these little frontend devices is quite stressing. Maybe a little heatsink flag on the TO-92 devices.
I will continue using my two 1kw trannies with 50-0-50 and approximately 74 volt rails and place more emphasis on the frontend power supply. A 50 to 100 VA tranny will be more than enough for the frontend.
I did not get any soldering done. I was running conduit and wire to the junction box in the shop all weekend to power up my TIG welder. 6 gauge copper wire is expensive. Bought the thing a year ago and never plugged it in. Figured on doing some aluminum TIG welding on this chassis and milling out some nice cases. Comes in handy when you break off a tap.
Tad
I will continue using my two 1kw trannies with 50-0-50 and approximately 74 volt rails and place more emphasis on the frontend power supply. A 50 to 100 VA tranny will be more than enough for the frontend.
I did not get any soldering done. I was running conduit and wire to the junction box in the shop all weekend to power up my TIG welder. 6 gauge copper wire is expensive. Bought the thing a year ago and never plugged it in. Figured on doing some aluminum TIG welding on this chassis and milling out some nice cases. Comes in handy when you break off a tap.
Tad
Rail tracking.
(Technics Class A+, SE-A1 and Glen Kleinschmidt' 1024 RMS Class A amp)
Jens & Bob Ellis collaborated on a regulated bi-polar supply. It was fit for +/-65V and enough current for this front end.
Ed,
I think with some adjustments it will go to about 110 volts. Bob included a nice large heatsink in the kits, I have 4 of them, which should help dissipate some of the high voltage heat.
Yet another project to complete to finish this up.
Ed know where you can buy some extra hours in a day without going to the chain gang. Never enough time.
Jacco, I can not get the Glen K. website up. Is it still working? Would like to see this amp layout.
Tad
I think with some adjustments it will go to about 110 volts. Bob included a nice large heatsink in the kits, I have 4 of them, which should help dissipate some of the high voltage heat.
Yet another project to complete to finish this up.
Ed know where you can buy some extra hours in a day without going to the chain gang. Never enough time.
Jacco, I can not get the Glen K. website up. Is it still working? Would like to see this amp layout.
Tad
...skipping down the path of diminishing returns...
I have decided to get more of the 2 ohm, 2W metal oxide film resistors. This will be with the intent to arrive at an effective Re of ~.33 ohm, and reduce the effective inductance by almost half.
As a result, I am able to offer "close tolerance" sets of emitter resisters to others. Let me know by pm if you would be interested in this...I will work for food...I'll also barter for it...😀
I have decided to get more of the 2 ohm, 2W metal oxide film resistors. This will be with the intent to arrive at an effective Re of ~.33 ohm, and reduce the effective inductance by almost half.
As a result, I am able to offer "close tolerance" sets of emitter resisters to others. Let me know by pm if you would be interested in this...I will work for food...I'll also barter for it...😀
Last edited:
...from the back row...
OK, class is in session...
Where should the V for the front end be set? 😕 Or is current the thing to be concerned with?
OK, class is in session...
Where should the V for the front end be set? 😕 Or is current the thing to be concerned with?
Hi Tad,
a separate transformer would do it. Where can you get +/-60V tx in a 50 to 100VA?
Why do you needa TIG Welder to make an amplifier? I would love more tools, but I havent welded since school and Im sure the littel I learnt is lost. Night classes are no longer free here, bummer.
Ed, by my calcs to get 0.33 Re you would need a 0.17 ohm resistor to go in parallel with your 2 ohm one. Is that what your doing?
Hi Ed, 5 - 10 V above your outputs.
How are you doing this, with a dedicated transformer or are you taking a tap of your outputs?
Dont forget to calculate R4 = R7 = (Vrail - 40)/8.2, this the resistance in Kohms you want.
I made some conservative estimates based on the min hfe of MJL2193 (25) that the front end would need to deliver about 300mA when the amp is maxed out.
I would allow for double that to be sure to be sure.
Luke,
I get my aluminum in various lots and shapes and sizes thus I must weld a few things together to get what I am planning to build. I have the heatsinks already layed out. I can weld a flange here and there for mounting the top and bottom plates.
The main reason I got this new TIG welder was to build a frame for my motorcycle project. It also comes in handy when I weld carbide pieces on various cutters I make here for more projects.
I would like to replicate a chassis like the Mcintosh units. They are kind of an open layout with a military style to them. All of the individual components are housed and compartmentalized with the main board access from the bottom. I can fab up some aluminum boxes for each of the toroids.
I am getting ahead of myself. I must first get the boards soldered up and operational plus the frontend boards and on and on. Never enough time.
What is the weather like down in New Zealand? I always figured the lifestyle there to be relaxed and efficient without too many neighbors kind of like the states before most became snobs.
Tad
I get my aluminum in various lots and shapes and sizes thus I must weld a few things together to get what I am planning to build. I have the heatsinks already layed out. I can weld a flange here and there for mounting the top and bottom plates.
The main reason I got this new TIG welder was to build a frame for my motorcycle project. It also comes in handy when I weld carbide pieces on various cutters I make here for more projects.
I would like to replicate a chassis like the Mcintosh units. They are kind of an open layout with a military style to them. All of the individual components are housed and compartmentalized with the main board access from the bottom. I can fab up some aluminum boxes for each of the toroids.
I am getting ahead of myself. I must first get the boards soldered up and operational plus the frontend boards and on and on. Never enough time.
What is the weather like down in New Zealand? I always figured the lifestyle there to be relaxed and efficient without too many neighbors kind of like the states before most became snobs.
Tad
I almost forgot. I will probably have to use an EI frame tranny to achieve the voltage I need for the frontend. Most toroids in this voltage range are huge for a frontend. With all of the solid state components in the regulator any gain from a toroid would probably be miniscule. I think the Parasound 400 amp has a separate set of trannies for its frontend. It is a BJT amp and got very good reviews. I need this amp to be the crown jewel because my amp building has probably come full circle. On to some exotic ribbon speakers. More money!!!!
Of course Talema has a few good trannies but they are imported and expensive.
Tad
Of course Talema has a few good trannies but they are imported and expensive.
Tad
I started with Yageo 0.51 ohm, 5W ww resistors.
A 2 ohm, 2W in parallel with that gets me to 0.41 ohm.
I have ordered more of the 2 ohm, in order to obtain better matches.
I'll have enough to use 2 in parallel with the 0.51 ohm to get to ~0.33 ohm.
Some of what I do the first time I will never do again. 😱
Sounds like an ambitious build. Compartments for the TX at least sound like a very good idea.
A bike project sounds like a good idea, what are you restoring? I was going to restore an old Vespa once, but I know my limits. Most mechanical repairs Ive done did not go smoothly and frustration creeps in of you dont know what your doing.
Weather here sucks at the moment, we've had a week of rain and another is on the horizon. Had a nice Sunday bar that its been poor.
NZ is 1600 Km long, weather goes from freezing down south to almost tropical up north. Wellington is notorius for its wind. I once woke up to the sound of my neighbors roof getting blown away sheet by sheet. The next morning he was nailing it back on🙂
NZ is a nice place, and having travelled alot, I can't really imagine living any where else.
- Status
- Not open for further replies.