For anyone considering using non-soft recovery diodes or bridges, here is a paper that explains RC damping, or "snubbing", and how to calculate the RC values.
It could also show what is needed for a board design that incorporates a snubber.
http://www.hagtech.com/pdf/snubber.pdf
It could also show what is needed for a board design that incorporates a snubber.
http://www.hagtech.com/pdf/snubber.pdf
I haven't studied the ground routes on the new boards, because I didn't get in on the buy. But, in principle, keeping them separate as far as possible would help. The +- wires of each cap bank could be twisted together right up to the amp board.
There is no reason that I can think of to have the star connection at the transformer or cap bank. In fact, there is good reason not to.
Imagine if you had four PS boards, two for each channel, i.e., front end and output. It gets a lot more complex about where to star.
Pooge,
I was going over those hexfred diodes again and I do not need 16 just 4. Duh. It is a bridge device. Having the 68 amp capabiliy is well worth 23.00 per item. Install it and forget it. With a few thermistors a soft start relay and a very large capacitor bank would be easy to light up.
Now we need a nice frontend setup. Maybe 50va rails plus 10 volts. What would be a good bridge for this area. Does IXYS have some smaller items? I was reading here somewhere that on lower current circuits the benefit of a toroid are not as critical. There is more choice in iron in the EI devices.
I have some Aleph P 1.7 shunt power supplies that net 84 volts that should work here.
I had a chat with Jens today. He has found a job. That is good. Maybe in the fall he will have time for DIY audio and we can get the Super Leach up and running.
Tad
I was going over those hexfred diodes again and I do not need 16 just 4. Duh. It is a bridge device. Having the 68 amp capabiliy is well worth 23.00 per item. Install it and forget it. With a few thermistors a soft start relay and a very large capacitor bank would be easy to light up.
Now we need a nice frontend setup. Maybe 50va rails plus 10 volts. What would be a good bridge for this area. Does IXYS have some smaller items? I was reading here somewhere that on lower current circuits the benefit of a toroid are not as critical. There is more choice in iron in the EI devices.
I have some Aleph P 1.7 shunt power supplies that net 84 volts that should work here.
I had a chat with Jens today. He has found a job. That is good. Maybe in the fall he will have time for DIY audio and we can get the Super Leach up and running.
Tad
Tad, I'm confused. Are you talking about the Vc and Vs rails?
I realized that Jen didn't separate Vc and Vcc on the 10 output board. I wonder what is the purpose on doing that. I'm really lost on that matter. If this was already discussed here, would someone please link me any paper or thread that I can read?
Thank you!
Reg front end
Miguel,
In response to request, Jens gave the boards the option to separate the front end power supply from the output. This allows the use of regulated power on the front end.
Miguel,
In response to request, Jens gave the boards the option to separate the front end power supply from the output. This allows the use of regulated power on the front end.
Last edited:
Thank you for your response Ed 🙂
Well, If we decide to use a regulated PSU on the frontend, what would be the voltage? I assume that by using a reg psu there, we need to remove the 100 ohms resistors that link the output devices' voltage and the frontend voltage, don't we? I'm talking about R1 and R76.
Thank you again for your kindness, Ed.
Well, If we decide to use a regulated PSU on the frontend, what would be the voltage? I assume that by using a reg psu there, we need to remove the 100 ohms resistors that link the output devices' voltage and the frontend voltage, don't we? I'm talking about R1 and R76.
Thank you again for your kindness, Ed.
Miguel,
I'd say you are correct. With the deletion of the resistors, those locations would become the insertion points for regulated power.
I'm just now beginning a serious review of:
http://users.ece.gatech.edu/~mleach/lowtim/
deja vu, I built version 4.4 many years ago.
Last edited:
To all,
I have acquired most all of my parts for the build. I have the BCxx for the differentials, some Sanyo and Toshiba parts for the VAS and predriver.
Can someone explain to me what type of environment the protection transistors will be in. Will the standard 80 volt BCxx devices suffice here at 74 volt rails. They are rated 80 volt devices. I just do not know what type of conditions must be met here. I hope they never get used.
I was also wondering if a .33 emitter might be better served here. The .47 is going to steal some power. Of course stability is a concern with 12 output devices and higher rail voltages than the designer planned.
Luke,
You had mentioned the posibility of Class A on your Leach with the large sinks and all. How far are you planning to bias the rails. In excess of 2 amps?
I hope to be smelling solder fumes this weekend. 🙂🙂
Like I need another amp!!!!
Tad
I have acquired most all of my parts for the build. I have the BCxx for the differentials, some Sanyo and Toshiba parts for the VAS and predriver.
Can someone explain to me what type of environment the protection transistors will be in. Will the standard 80 volt BCxx devices suffice here at 74 volt rails. They are rated 80 volt devices. I just do not know what type of conditions must be met here. I hope they never get used.
I was also wondering if a .33 emitter might be better served here. The .47 is going to steal some power. Of course stability is a concern with 12 output devices and higher rail voltages than the designer planned.
Luke,
You had mentioned the posibility of Class A on your Leach with the large sinks and all. How far are you planning to bias the rails. In excess of 2 amps?
I hope to be smelling solder fumes this weekend. 🙂🙂
Like I need another amp!!!!
Tad
Hi Tad,
your bcXX parts are fine. Heres a diagram with the dc values, its just a couple of volts.
Either 0.33 or .47 will do fine for emmitter resistors. The difference is 0.14 Ohm. With 6 amps output (300Watts into 8 ohm) you will lose 6 amps/6 transistors, or one amp through each. the difference in voltage drop across each resistor is V=IR or 1*0.14 or 0.14V or 140mV. This is neglidgeable spelt incorrectly🙂
I would pick a value your more likely to use more often if your buying more than you need. Stability is probably better with 0.47, but thats what Ive read I cant back it with an explanation.
Im not sure about class A now. If I have 80V rails, 1.875 amps bias thats over 300 watts dissipation and about 20 watts RMS into 8 ohms.
I could drop the rails to 50V and have a higher class A output, but then it wouldnt be a grunty amp. This is a good oportunity to build a high power amp, so Im leaning to higher rails.
Got plenty of time to decide, still have a while before the ugs is done.
Show us pics Tad, your the first for this build I think.
Luke, Thanks for sharing.
Linear Technology Design Simulation & Device Models
Linear Technology - Design Simulation and Device Models
There is a link provided above in post #527. At the top of the page linked, Dr. Leach shares a spice model that has been provided by one of the amp builders, Valerio Maglietta.
I'm totally a newb at spice 🙂smash: pop a cherry
). Is this an effort I should explore to increase my confidence with 75V rails? Can someone familiar with spice characterize the learning curve?
Linear Technology Design Simulation & Device Models
Linear Technology - Design Simulation and Device Models
There is a link provided above in post #527. At the top of the page linked, Dr. Leach shares a spice model that has been provided by one of the amp builders, Valerio Maglietta.
I'm totally a newb at spice 🙂smash: pop a cherry
). Is this an effort I should explore to increase my confidence with 75V rails? Can someone familiar with spice characterize the learning curve?
Last edited:
I'm totally a newb at spice 🙂smash: pop a cherry
That would be something like this:
Difficult to get started
Difficult to use external models (for semis)
Easy to run when models work
Easy to adabt to new circuits
The really hard part (at least for me) is the external models - someone might know more.
All the work is worth it!
\\\Jens
Hello Jens! Hope all is well with you and family.
Well, I've downloaded the SW, & the referenced file opens up in it...looks like a Leach amp circuit...I don't know when I'll get into it.
Well, I've downloaded the SW, & the referenced file opens up in it...looks like a Leach amp circuit...I don't know when I'll get into it.
Last edited:
Ed,
You are tredding in some deep water. I hope you have luck with the Spice simulations. I looked at some of this about 2 years ago and all I can say is I looked. I am too old to try this. I just have gotten use to using a computer.
If you come up with some useful projections please share here with us. Some of the members here have had some very good work done with spice. It saves a lot in actual parts and time.
Try and sim a Leach with cascode parallel jfets in the differentials and VAS. That would be something to build.
The 75 volt rail build is not too difficult. Just check everything 3 times and then once more. It is one sweet little -- big -- amp. I liked mine enough to sell off all of my Mcintosh items. I did not part with the MR78 tuner easily.
Tad
You are tredding in some deep water. I hope you have luck with the Spice simulations. I looked at some of this about 2 years ago and all I can say is I looked. I am too old to try this. I just have gotten use to using a computer.
If you come up with some useful projections please share here with us. Some of the members here have had some very good work done with spice. It saves a lot in actual parts and time.
Try and sim a Leach with cascode parallel jfets in the differentials and VAS. That would be something to build.
The 75 volt rail build is not too difficult. Just check everything 3 times and then once more. It is one sweet little -- big -- amp. I liked mine enough to sell off all of my Mcintosh items. I did not part with the MR78 tuner easily.
Tad
Hi Jens, nice to see you again and good work getting a new job🙂
Ed, your welcome. Good luck with the spice, I must admit its something Ive wanted to do for a while. Also thank for the link, I might take a look this beautiful morning. Weve had a week of rain, but its a nice sunny Autumn morning here today.
Tad, you have quite a fascination with Leach amps🙂 But I agree jfet front end would be nice to try.
Was your Mcintosh amp valve? I have never heard a mcintosh, they dont seem to get imported here, but they sure look proper. Are you still planning to use a black glass front panel for you leach? That sounds quite nice, but I dont know how you would attach glass to metal. I think it could work if you use lots of araldite and glue an L bracket to the glass and them screw the bracket into the heatsinks.
I have considered doing this as well, but I dont know how well it would work.
Ed, your welcome. Good luck with the spice, I must admit its something Ive wanted to do for a while. Also thank for the link, I might take a look this beautiful morning. Weve had a week of rain, but its a nice sunny Autumn morning here today.
Tad, you have quite a fascination with Leach amps🙂 But I agree jfet front end would be nice to try.
Was your Mcintosh amp valve? I have never heard a mcintosh, they dont seem to get imported here, but they sure look proper. Are you still planning to use a black glass front panel for you leach? That sounds quite nice, but I dont know how you would attach glass to metal. I think it could work if you use lots of araldite and glue an L bracket to the glass and them screw the bracket into the heatsinks.
I have considered doing this as well, but I dont know how well it would work.
Luke,
Never got around to owning any tube Mac equipment all solid state. At the prices they are bringing now I could have turned a good profit if they had been tube. Seems that the members here are not too keen on Mcintosh amps. Something to do with the prodigious feedback or whatever. I had owned a nice modest system in the early 70's which was liberated from me by some thieving SOB. I had forgotten about it for about 30 years until one day on Ebay. I bought a new setup. When it needed repairing I ended up here. That is history. Now I roll my own.
I plan on either a nice fat glass front or maybe a piece of black marble. If I make a slot the thickness of the glass on each side plate of the amp I can just slide the glass into it. That way there is no need for a mechanical joint.
I came upon this Leach thing while reading John Curl's thread on the blowtorch preamp. One of the modestly famous amp designers name Charles ?? steered me toward a Leach build. He gave it very high marks and he was one who should know -- I guess, It sounds as good or better than anything else I have and much cheaper, plus everything inside is new and matched. I have a relay attenuator and FM tuner kit from Dantamax which I have to assemble and add to this project. Plus the F5, Aleph JX and so on. I hope I never run out of future projects. That way there is always something to do.
Plus, we have to keep trying to get the Super Leach amp thing together. I still have hope for that project. When Jens gets settled in to his new job and access to all that computer software he will come onboard. Jens is quite talented in this area.
I also have a huge woofer to build using a ribbon speaker design and some very large neodymium magnets. That way there is no coil to mess with. A 100 percent resistive bass speaker with no back EMF.
Tad
Never got around to owning any tube Mac equipment all solid state. At the prices they are bringing now I could have turned a good profit if they had been tube. Seems that the members here are not too keen on Mcintosh amps. Something to do with the prodigious feedback or whatever. I had owned a nice modest system in the early 70's which was liberated from me by some thieving SOB. I had forgotten about it for about 30 years until one day on Ebay. I bought a new setup. When it needed repairing I ended up here. That is history. Now I roll my own.
I plan on either a nice fat glass front or maybe a piece of black marble. If I make a slot the thickness of the glass on each side plate of the amp I can just slide the glass into it. That way there is no need for a mechanical joint.
I came upon this Leach thing while reading John Curl's thread on the blowtorch preamp. One of the modestly famous amp designers name Charles ?? steered me toward a Leach build. He gave it very high marks and he was one who should know -- I guess, It sounds as good or better than anything else I have and much cheaper, plus everything inside is new and matched. I have a relay attenuator and FM tuner kit from Dantamax which I have to assemble and add to this project. Plus the F5, Aleph JX and so on. I hope I never run out of future projects. That way there is always something to do.
Plus, we have to keep trying to get the Super Leach amp thing together. I still have hope for that project. When Jens gets settled in to his new job and access to all that computer software he will come onboard. Jens is quite talented in this area.
I also have a huge woofer to build using a ribbon speaker design and some very large neodymium magnets. That way there is no coil to mess with. A 100 percent resistive bass speaker with no back EMF.
Tad
I dont want to plan that far ahead, cos I want to build so many things, but the super Leach sounds good to me. Go Jens.
Are the Aleph-JX and F5 future projects? I think the Aleph-X is quite an undertaking. Seems many pcbs are made but few amps come to fruition. But it is very doable if you plan right, match properly and folllow the right advice.
Never heard of ribbon woofers you got pics?
If the numbers are to be believed
I was having a lot of concern for how to measure the emitter resistors for matching. It dawned on me that my WT3 has the capability for resistance measuring. It derives the Re for each reading from impedance. I started it up to see what it could do. The user guide says it does fine (1%) down to 1 ohm.
I wanted this to be a casual exercise. All the same I didn't want it to be a waste of time. Each resistor was placed in the test leads and measured a minimum of 3 times. This isn't too difficult. It requires a couple of mouse clicks for each reading. When results from one resistor showed a relatively large variation more readings were taken. Care was taken to seat the resistors consistently in the test clips
I have 2 batches of resistors I plan to use for the emitters. They include 56 pieces of 5W Yageo wire wound of 0.51 ohms and 50 pieces of 2W Yageo metal oxide film of 2.0 ohms.
Once the readings were collected I returned to the parallel resistor calculator linked in post 348.
By inverting one set of resistors to the other (one set is sorted by value ascending, the other set is sorted by value descending) I find I can make 2 sets of 24 resistors with a nominal value of 0.416 ohm. The tolerance within the sets is ~1/2%
I'll use ~0.42 ohm for further consideration of amp components.
I was having a lot of concern for how to measure the emitter resistors for matching. It dawned on me that my WT3 has the capability for resistance measuring. It derives the Re for each reading from impedance. I started it up to see what it could do. The user guide says it does fine (1%) down to 1 ohm.
I wanted this to be a casual exercise. All the same I didn't want it to be a waste of time. Each resistor was placed in the test leads and measured a minimum of 3 times. This isn't too difficult. It requires a couple of mouse clicks for each reading. When results from one resistor showed a relatively large variation more readings were taken. Care was taken to seat the resistors consistently in the test clips
I have 2 batches of resistors I plan to use for the emitters. They include 56 pieces of 5W Yageo wire wound of 0.51 ohms and 50 pieces of 2W Yageo metal oxide film of 2.0 ohms.
Once the readings were collected I returned to the parallel resistor calculator linked in post 348.
By inverting one set of resistors to the other (one set is sorted by value ascending, the other set is sorted by value descending) I find I can make 2 sets of 24 resistors with a nominal value of 0.416 ohm. The tolerance within the sets is ~1/2%
I'll use ~0.42 ohm for further consideration of amp components.
hi,
to measure sub-ohm resistors you can use the 4 wire method, you need a constant current source of say 1 ampere and a millivolt-meter....
a 0.33 ohm resistor then will have 330millivolts dropped across....
to measure sub-ohm resistors you can use the 4 wire method, you need a constant current source of say 1 ampere and a millivolt-meter....
a 0.33 ohm resistor then will have 330millivolts dropped across....
Hi Tony,
do you think 1% emmitter resistors warrants this level of matching? I've never done this in the past and wonder if I should consider it.
- Status
- Not open for further replies.