2.4 mm is great.
I still have a number of 300x500mm MAS boards(mascoplad) in 2.4 mm thickness.
Those who learned Gere and Timoshenko stuff at one time in the past know that rigidity multiplies by 8 if thickness is doubled.
2.4 mm is more than 4 times more rigid than 1.5 mm boards.
Almost twice more rigid than 2 mm pcb's.
I already looked at the extended leach board on the other thread closely to see how those heavy 10.000 uF Dubeliers were supported.
3 oz copper and 2.4 mm thickness at $45 each, you must be Santa, Jens.
Mind if i ask what software you use for your layouts, Jens ?
I still have a number of 300x500mm MAS boards(mascoplad) in 2.4 mm thickness.
Those who learned Gere and Timoshenko stuff at one time in the past know that rigidity multiplies by 8 if thickness is doubled.
2.4 mm is more than 4 times more rigid than 1.5 mm boards.
Almost twice more rigid than 2 mm pcb's.
I already looked at the extended leach board on the other thread closely to see how those heavy 10.000 uF Dubeliers were supported.
3 oz copper and 2.4 mm thickness at $45 each, you must be Santa, Jens.
Mind if i ask what software you use for your layouts, Jens ?
Attachments
Yea, I would go for Jacco's reccomendations as well should there be consensus by the rest as to that. I particularly like the idea of being able to use discrete power for each power stage, which in effect gives the flexibility to increase the power output of the amp without changing the compenent values of the gain stages. Am I right Jacco?
Jens, what is the power output of your extended design?
Jens, what is the power output of your extended design?
They were not my recommendations.
I just made a list of all suggestions made by people who are interested in building the Leach.
(some i am actually not in favor of
)
VL:
you are right that in my opinion it pays to split the powerlines.
For the power stage a 50 volt transformer may be enough for the super leach to deliver 250 watts.
With a 55 volts transformer about 300 watts.
Those who go for a 65 volt transformer could be talking +400 watts, if the transformer is beafy enough.
With an added small transformer, or extra secondary windings like Terry has on his toroid, all components could be kept as they are.
The only thing to do is connect the secondaries of the big and small transformer in series before rectifying.
People that like to spend cash can use a separate transformer for the high voltage, maybe even a toroid.
Even an non-regulated separate rail for the gain stages makes the amplifier more stable and the big transformer more efficient in terms of voltage rating versus power output.
Adding a voltage regulator can be a choice each makes for himself.
If you take a look at the picture i posted at the Pass monster thread it shows a pile of 24 toroid transformers, 300VA each.(post118)
I am going to use 4 Toroids per channel, 1200VA per side.
Each toroid will have its own bridge rectifier.
The powersupplies for the gain and driver stages i mentioned in my previous posting.
If the majority goes for separate rails it will be easy for me to attach the power lines.
Otherwise i will put the Dremel on the boards and make my own connections.
I am not dictating anyone to do my thing.
I just made a list of all suggestions made by people who are interested in building the Leach.
(some i am actually not in favor of
)VL:
you are right that in my opinion it pays to split the powerlines.
For the power stage a 50 volt transformer may be enough for the super leach to deliver 250 watts.
With a 55 volts transformer about 300 watts.
Those who go for a 65 volt transformer could be talking +400 watts, if the transformer is beafy enough.
With an added small transformer, or extra secondary windings like Terry has on his toroid, all components could be kept as they are.
The only thing to do is connect the secondaries of the big and small transformer in series before rectifying.
People that like to spend cash can use a separate transformer for the high voltage, maybe even a toroid.
Even an non-regulated separate rail for the gain stages makes the amplifier more stable and the big transformer more efficient in terms of voltage rating versus power output.
Adding a voltage regulator can be a choice each makes for himself.
If you take a look at the picture i posted at the Pass monster thread it shows a pile of 24 toroid transformers, 300VA each.(post118)
I am going to use 4 Toroids per channel, 1200VA per side.
Each toroid will have its own bridge rectifier.
The powersupplies for the gain and driver stages i mentioned in my previous posting.
If the majority goes for separate rails it will be easy for me to attach the power lines.
Otherwise i will put the Dremel on the boards and make my own connections.
I am not dictating anyone to do my thing.
Hi Jacco-
First, thanks for pulling the list together.
I understand that you disagree with some of the items. That's cool. I think we should be discussing the ones we do/don't like (nicely of course).
First, thanks for pulling the list together.
I understand that you disagree with some of the items. That's cool. I think we should be discussing the ones we do/don't like (nicely of course).
Wow, this is all very interesting. I'm "steppin' and fetchin'" to try and keep up with what you guys are saying.
I would love to understand how to wire my transformer and board as stated by jacco. could someone maybe draw a little diagram so I could get a better grasp on it?
I have been building a Superamp and just finished populating the boards. I made the boards using Professor Leach's design from his website. Is there a way to do the dual power hook-up with those boards? I had planned to use the TO-3's for this but I also have 25 each of the flatpacks if that would be better.
Maybe I should just finish the amp as designed by Prof. Leach and then build another one when these board become available. Don't know what I'm going to do with all of these power amps though.
Blessings, Terry
I would love to understand how to wire my transformer and board as stated by jacco. could someone maybe draw a little diagram so I could get a better grasp on it?
I have been building a Superamp and just finished populating the boards. I made the boards using Professor Leach's design from his website. Is there a way to do the dual power hook-up with those boards? I had planned to use the TO-3's for this but I also have 25 each of the flatpacks if that would be better.
Maybe I should just finish the amp as designed by Prof. Leach and then build another one when these board become available. Don't know what I'm going to do with all of these power amps though.
Blessings, Terry
Terry,
Assuming you have four separate windings, connect them all in series, hot to cold like this:
To front end PS -- 15V -- to output PS -- 65V -- ground -- 65V -- to output PS -- 15 V -- to front end ps.
You should see 160 VAC across the front end terminals. You cannot do this if your 15V windings are a single 30VCT winding.
To separate the power supplies, leave out R32 and R33. Use the pad as your connection for the front end PS. This will leave the drivers at output rail voltage.
You could also cut the trace between Q24 and Q26 collectors as well as between Q25 and Q27 collectors, leaving R32 and R33 in place.
Assuming you have four separate windings, connect them all in series, hot to cold like this:
To front end PS -- 15V -- to output PS -- 65V -- ground -- 65V -- to output PS -- 15 V -- to front end ps.
You should see 160 VAC across the front end terminals. You cannot do this if your 15V windings are a single 30VCT winding.
To separate the power supplies, leave out R32 and R33. Use the pad as your connection for the front end PS. This will leave the drivers at output rail voltage.
You could also cut the trace between Q24 and Q26 collectors as well as between Q25 and Q27 collectors, leaving R32 and R33 in place.
Thank you, Bill.
Just an example:
i am not in favor of the smallest pcb possible, it means maximum component nesting.
I like placing my components Mark Levinson style.
The component legs go through small teflon rings before soldering to the board.
The resistors keep a distance to the board, look at the way Jens soldered resistors on the extended Leach pcb.
This way the resistors have 360 degrees cooling, placing them flat on the board reflects heat back to the device.
With rings the components are dampened for vibrations, and the look is smashing.
Smaller pcb's might make it more difficult to use those rings.
Secondly, dense nesting makes cleaning boards more difficult, they gather dust more easilly. I clean amplifiers once a year.
Very dense pcb's may be at the expense of vast mass traces between signal traces. Again look at the way Jens made the pcb for the extended Leach on the other thread.
To make pcb's smaller traces may be made more narrow, i like traces wide. Professor Leach mentions a trace of 0.02" , i favor bigger. Put traces to close to eachother, they may interact.
Desiring a smaller pcb may lead to using smaller components.
This narrows the choice of the individual builder to use components of his choice.
And could lead to chosing resistors as Brian suggested.
I am 45, i can not see that well up close anymore.
I had a hard time with the 0.2" resistors Brian delivered with the GC pcb's i ordered from him. Soldering them to the boards was difficult and not fun, i like this to be a fun hobby.
Put them on a shelf,like me, and look at them, Terry.
Just an example:
i am not in favor of the smallest pcb possible, it means maximum component nesting.
I like placing my components Mark Levinson style.
The component legs go through small teflon rings before soldering to the board.
The resistors keep a distance to the board, look at the way Jens soldered resistors on the extended Leach pcb.
This way the resistors have 360 degrees cooling, placing them flat on the board reflects heat back to the device.
With rings the components are dampened for vibrations, and the look is smashing.
Smaller pcb's might make it more difficult to use those rings.
Secondly, dense nesting makes cleaning boards more difficult, they gather dust more easilly. I clean amplifiers once a year.
Very dense pcb's may be at the expense of vast mass traces between signal traces. Again look at the way Jens made the pcb for the extended Leach on the other thread.
To make pcb's smaller traces may be made more narrow, i like traces wide. Professor Leach mentions a trace of 0.02" , i favor bigger. Put traces to close to eachother, they may interact.
Desiring a smaller pcb may lead to using smaller components.
This narrows the choice of the individual builder to use components of his choice.
And could lead to chosing resistors as Brian suggested.
I am 45, i can not see that well up close anymore.
I had a hard time with the 0.2" resistors Brian delivered with the GC pcb's i ordered from him. Soldering them to the boards was difficult and not fun, i like this to be a fun hobby.
Put them on a shelf,like me, and look at them, Terry.
Terry,
your secondaries give 70 VAC in series, not 65 VAC !
After rectifying that is 7 volts DC more than designed for.
With your transformer you may need to regulate the powerline to drop the voltage to 91 VDC, the regulation comes free.
It is possible that you can run the front end of the amplifier on 98 VDC but i'd check first.
Maybe you could take a look see at Rod Elliot's for a regulator circuit.
your secondaries give 70 VAC in series, not 65 VAC !
After rectifying that is 7 volts DC more than designed for.
With your transformer you may need to regulate the powerline to drop the voltage to 91 VDC, the regulation comes free.
It is possible that you can run the front end of the amplifier on 98 VDC but i'd check first.
Maybe you could take a look see at Rod Elliot's for a regulator circuit.
This is the transformer I am hoping to use. THe add says 62+62+17+17V 1000VA .
On the side of the transformer it shows this;
If you read the add you can see the test results which show higher than 55V. I'll let those of you who know better to decide what this tranny actually puts out.
If I understand properly, 62*1.414 = 87.668VDC I know there is some loss for the rectification, but still it seems suficient to power this amp. If I add the 15V or 17V which ever it is, won't I have too much power?
62 + 17 = 79*1.14=111.706VDC
My transformer has 8 wires coming out of the secondary, 4 for the 17 + 17 and 4 for the 62 + 62.
jacco,
I wish I had read this last night when I was soldering up my boards. I was wondering to mysef if I should hold the components off of the board or not. I went ahead and set them down against the board where possible. I didn't think about heat.
Oh well, next time.
Thanks, Terry
On the side of the transformer it shows this;
An externally hosted image should be here but it was not working when we last tested it.
If you read the add you can see the test results which show higher than 55V. I'll let those of you who know better to decide what this tranny actually puts out.
If I understand properly, 62*1.414 = 87.668VDC I know there is some loss for the rectification, but still it seems suficient to power this amp. If I add the 15V or 17V which ever it is, won't I have too much power?
62 + 17 = 79*1.14=111.706VDC
My transformer has 8 wires coming out of the secondary, 4 for the 17 + 17 and 4 for the 62 + 62.
jacco,
I wish I had read this last night when I was soldering up my boards. I was wondering to mysef if I should hold the components off of the board or not. I went ahead and set them down against the board where possible. I didn't think about heat.
Oh well, next time.
Thanks, Terry
Terry,
That should work, just move the bottom 15 volt winding to the top of your picture and connect the windings in series - 0-15V to 0-55 to gnd to 0-55 to 0-15. your front end takes power from the outside windings, the output stage from the 55 volt windings.
With BJTs you really don't need to run the front end at significantly higher rail votage than the outputs like you do with MOSFETs. Use the extra voltage to allow you to regulate the front end back down to a volt or two higher than your output rails.
For a regulator you can build something like Rod Elliot's, or modify something like the one at passdiy.com in the balanced zen line stage. Just change the voltage reference zener stack to suit. For no particular reason I like the regulator used in the Pass-Thagard A75. you can change the values of the voltage divider resistors to get the output voltage you want. It's easy to build point to point on perf board if you don't want to make a pcb.
edit: I hope that your primary is really two windings or you may be in for a surprise - only half the output voltage that you expected.
That should work, just move the bottom 15 volt winding to the top of your picture and connect the windings in series - 0-15V to 0-55 to gnd to 0-55 to 0-15. your front end takes power from the outside windings, the output stage from the 55 volt windings.
With BJTs you really don't need to run the front end at significantly higher rail votage than the outputs like you do with MOSFETs. Use the extra voltage to allow you to regulate the front end back down to a volt or two higher than your output rails.
For a regulator you can build something like Rod Elliot's, or modify something like the one at passdiy.com in the balanced zen line stage. Just change the voltage reference zener stack to suit. For no particular reason I like the regulator used in the Pass-Thagard A75. you can change the values of the voltage divider resistors to get the output voltage you want. It's easy to build point to point on perf board if you don't want to make a pcb.
edit: I hope that your primary is really two windings or you may be in for a surprise - only half the output voltage that you expected.
Mains voltage is +-10% (In Denmark)
You must also take into account that the mains line can have a +10% change to its voltage…. This makes your rail voltage (62 + 17)*1.1 VAC = (79*1.14)*1.1 VDC = 111.706 VDC * 1.1 = 122VDC. That is a bit high if you ask me 😉
Oh well what do I know, you might need that kind of heat this thing will generate, I sure don’t!
\Jens
You must also take into account that the mains line can have a +10% change to its voltage…. This makes your rail voltage (62 + 17)*1.1 VAC = (79*1.14)*1.1 VDC = 111.706 VDC * 1.1 = 122VDC. That is a bit high if you ask me 😉
Oh well what do I know, you might need that kind of heat this thing will generate, I sure don’t!
\Jens
It seems there are two ways of looking at Terry's transformer. I agree with Jens that Terry probably doesn't need or want to to run the whole amp at potentially 122 volt rails.
Using the low voltage windings to allow a regulated front end, does make sense, though. Here is a sketch of the connection I tried to describe above.
With a suitable regulator Terry could end up with something like 84 volt output rails and 86 volt regulated front end rails. It's probably a good idea to mount the regulators' pass transistors on the main heat sinks - at least give them something substantial.
Using the low voltage windings to allow a regulated front end, does make sense, though. Here is a sketch of the connection I tried to describe above.
With a suitable regulator Terry could end up with something like 84 volt output rails and 86 volt regulated front end rails. It's probably a good idea to mount the regulators' pass transistors on the main heat sinks - at least give them something substantial.
Attachments
I can never decide if regulation is a good idea or not. I guess it depends on the answer to this question:
Exactly how much does the power supply get modulated when the amp is running? Does anyone have any observations?
Exactly how much does the power supply get modulated when the amp is running? Does anyone have any observations?
Modulation depends on the freq of the signal, and the number of caps in the power supply.
Regarding high voltage rails, 122V DC calls for 150V caps, very expensive and hard to find - at least where I live.
\Jens
Regarding high voltage rails, 122V DC calls for 150V caps, very expensive and hard to find - at least where I live.
\Jens
maybe Terry will build his amp so that he can easily switch from regulated to unregulated front end rails and let us know whether he finds the difference audible.
OK, OK, my head is spinning.
First of all, I just bought 4 100V 10,000uf caps for the power supply so I'm not going to be running something that takes 150V caps, at least not for this amp.
I can't seem to find any info at Rod Elliot's site about building a regulated supply so for now I think I'll stick with just using the 55V. I'm afraid I'm going to get into trouble if I start deviating too far from the proven design. And, Like Jens said, what will I do with all that heat. I just want this thing to sound great and have a little kick when I feel like it.
That running in series is a good thing to know. A lot of the transformers that the guy sells have dual secondaries like this one. Definately gives one more options.
Thanks Bob,
I think I see how to hook it up. Am I right in assuming that both sets will need their own rectifier and set of filter caps?
Thanks, Terry
First of all, I just bought 4 100V 10,000uf caps for the power supply so I'm not going to be running something that takes 150V caps, at least not for this amp.
I can't seem to find any info at Rod Elliot's site about building a regulated supply so for now I think I'll stick with just using the 55V. I'm afraid I'm going to get into trouble if I start deviating too far from the proven design. And, Like Jens said, what will I do with all that heat. I just want this thing to sound great and have a little kick when I feel like it.
That running in series is a good thing to know. A lot of the transformers that the guy sells have dual secondaries like this one. Definately gives one more options.
Thanks Bob,
I think I see how to hook it up. Am I right in assuming that both sets will need their own rectifier and set of filter caps?
Thanks, Terry
Without a current running secondary voltage increases.
That is why you need a safety margin on capacitors, i think i gave you a rundown on that some time ago.
When your amplifier is switched on there will be a bias current through the output devices.
The driver will be biased, a normal figure for front end devices is 5 to 10 mA per device totalling something like 40-50 mA.
That will make the voltage drop on the transformer.
The voltages of the transformer with different currents have to be measured, that is why the transformer data came with the sale.
If the output voltage at proper currents is 62 VAC you will need 100 volt capacitors.
The amplifier can handle a 65 Volt powersupply, you can count on it that a nominal 65 volt transformer will produce higher voltages too.
It is not a big problem if the voltage on the front end is a bit higher.
It only means that the current through the gain devices is a little higher, nothing spectacular.
It may be the difference between 5 mA or 5.5 mA.
Currents through gain devices are often determined by the optimum noise level.
A transistor will create noise, very little but with gain stages this adds up.
To keep noise level low the current going through the devices is chosen so that noise level is minimal, this can be read from the data sheet.
Currents being slightly higher because of slightly higher voltages may give a somewhat less signal to noise level.
if you use this transformer for the Leach you will have the 62vac on the output any way.
The 17 vac give you 24 vdc on top of the 62vac .
If 65 Vac is needed to deliver 300watts in 8 ohm, 49 volts, the front end needs less than16 volts, the rest is a margin on the output voltage.
Meaning, with a regulator you can drop 8 volts, from 24 to 16, to drive the amplifier and get a stable voltage.
A regulator can easilly do an 8 volt drop.
Bob's story is about right.
That is why you need a safety margin on capacitors, i think i gave you a rundown on that some time ago.
When your amplifier is switched on there will be a bias current through the output devices.
The driver will be biased, a normal figure for front end devices is 5 to 10 mA per device totalling something like 40-50 mA.
That will make the voltage drop on the transformer.
The voltages of the transformer with different currents have to be measured, that is why the transformer data came with the sale.
If the output voltage at proper currents is 62 VAC you will need 100 volt capacitors.
The amplifier can handle a 65 Volt powersupply, you can count on it that a nominal 65 volt transformer will produce higher voltages too.
It is not a big problem if the voltage on the front end is a bit higher.
It only means that the current through the gain devices is a little higher, nothing spectacular.
It may be the difference between 5 mA or 5.5 mA.
Currents through gain devices are often determined by the optimum noise level.
A transistor will create noise, very little but with gain stages this adds up.
To keep noise level low the current going through the devices is chosen so that noise level is minimal, this can be read from the data sheet.
Currents being slightly higher because of slightly higher voltages may give a somewhat less signal to noise level.
if you use this transformer for the Leach you will have the 62vac on the output any way.
The 17 vac give you 24 vdc on top of the 62vac .
If 65 Vac is needed to deliver 300watts in 8 ohm, 49 volts, the front end needs less than16 volts, the rest is a margin on the output voltage.
Meaning, with a regulator you can drop 8 volts, from 24 to 16, to drive the amplifier and get a stable voltage.
A regulator can easilly do an 8 volt drop.
Bob's story is about right.
still4given said:
Yes, you'll need two sets of rectifiers and filter caps. The good news is that the front end draws very little current, so a 1 amp bridge is more than enough. Similarly, small current means smaller filter cap requirement. For the front end, a few hundred uf before and after the regulator will be more than enough. Don't run the front end without a regulator if you boost the voltage - you don't need an extra 20V.
Jacco, "about right" implies not quite right. For my own amateur EE education, would you mind pointing out where I oversimplified or erred? Thanks. (I was thinking of my A75 front end that draws about 80 ma when I commented on heatsinking the regs. If he's got 79 VAC, unregulated voltage is around 110 and dropping that 86 volts as I suggested would mean dissipating over 2 watts in the regs - heat sinks bigger than clip ons mandatory. This front end will still draw ~40 mA a pair, so decent heat sinks on the regs are a good idea.)
- Status
- Not open for further replies.