Luke,
I have a pair of 50-0-50 1kw toroids from my A-75 that never got finished. I smoked the board and cannot find any to replace it. However, I really would rather this Leach be built with something approaching 85-90 volt rails. I would need some 65 volt toroids for that. The other thing is the transistors on the Bob Ellis/Jens frontend regulator are getting pretty hot going into 100 volts. Maybe just a dedicated frontend transformer and no reg. Something to mull over. I am thinking about a pile of 8200uf 100 volt caps from Apex Jr. for this build. I cannot find any cheap screw top devices. Maybe 24 arranged 2c-R-2c-R-2c or there about. The caps are only 3.00 each. The resistors will be more than the caps.
Does a bjt amp benefit as much from a higher voltage frontend like a mosfet amp does?
I want a really nice chassis and case for this amp. I have a big pile of aluminum and the mill has not been used in quite sometime. Maybe a nice thick black glass front like my old Mcintosh amps. They always looked respectable and fit in to the house decor better. Though the seven foot speakers kind of stand out.
I have most of my shopping done. I have Sanyo, Toshiba and BC546/547 transitors to choose from. The output resistors should be here today or Monday. I have to set aside some money for the MJL4281/4302 On Semi's they are pricey.
Thanks again for the civil response to my filter question. Tad
I have a pair of 50-0-50 1kw toroids from my A-75 that never got finished. I smoked the board and cannot find any to replace it. However, I really would rather this Leach be built with something approaching 85-90 volt rails. I would need some 65 volt toroids for that. The other thing is the transistors on the Bob Ellis/Jens frontend regulator are getting pretty hot going into 100 volts. Maybe just a dedicated frontend transformer and no reg. Something to mull over. I am thinking about a pile of 8200uf 100 volt caps from Apex Jr. for this build. I cannot find any cheap screw top devices. Maybe 24 arranged 2c-R-2c-R-2c or there about. The caps are only 3.00 each. The resistors will be more than the caps.
Does a bjt amp benefit as much from a higher voltage frontend like a mosfet amp does?
I want a really nice chassis and case for this amp. I have a big pile of aluminum and the mill has not been used in quite sometime. Maybe a nice thick black glass front like my old Mcintosh amps. They always looked respectable and fit in to the house decor better. Though the seven foot speakers kind of stand out.
I have most of my shopping done. I have Sanyo, Toshiba and BC546/547 transitors to choose from. The output resistors should be here today or Monday. I have to set aside some money for the MJL4281/4302 On Semi's they are pricey.
Thanks again for the civil response to my filter question. Tad
Leach's version of the protective output limiter is a two slope current limiter.
It is not a VI limiter.
All of Jens' Leach clone PCBs have the extra resistor added to create a VI limiter from the standard Leach limiter.
I found it very difficult and not at all intuitive to calculate the resistor values in the VI limiter circuit. I ended up modeling the limiter in a spreadsheet and even then it took a while to learn how to manipulate values to get the IV curves that protected against DC faults (i.e. dc currents & volts inside the DC SOAR) and could be extended to allow high current transients to pass provided they were of short duration and still inside the temperature de-rated SOAR curves.
It is not a VI limiter.
All of Jens' Leach clone PCBs have the extra resistor added to create a VI limiter from the standard Leach limiter.
I found it very difficult and not at all intuitive to calculate the resistor values in the VI limiter circuit. I ended up modeling the limiter in a spreadsheet and even then it took a while to learn how to manipulate values to get the IV curves that protected against DC faults (i.e. dc currents & volts inside the DC SOAR) and could be extended to allow high current transients to pass provided they were of short duration and still inside the temperature de-rated SOAR curves.
the Maths is so simple WE can all do it in our heads.
First.
RC time constant is a way of specifying a single pole filter.
The filter value can be high to just allow the Audio signal through or it can be lower to ensure all the Audio signal gets through with minimal phase distortion or it can be very low to allow everything (well almost) through except the DC.
Typical RC values for these three examples would be
Audio only RC ~ 8milliseconds (ms)
minimal phase distortion RC ~ 90ms
let everything through RC ~ 500ms
Take Rin value in kohms. For the Leach 22k
Take selected RC value.
Let's go with 8ms
Now the maths (arithmetic) part:
Blocking capacitor is C = RC / Rin = 8ms / 22k = 0.36uF
That's it. Just one division operation.
The skill is in selecting the data values. eg. Rin=22k and RC=8ms
These are needed for whichever system you use for calculating the Blocking Capacitor value.
The formula RC / Rin also works for the NFB blocking cap. The resistor (RL) in the lower leg of the NFB is used.
RL=1k, RC = 50ms
C = 50ms / 1k = 50uF
I prefer than the NFB cap be increased by a factor of SQRT(2). So that 50uF becomes >70.7uF I would use 100uF in this case.
Two jobs done.
It really is that simple.!!!!
Last edited:
Hi Tad,
you don't need to thank me, if it were not for you I wouldn't be planning this project now🙂
I know what you mean, I have an old torroidal 55 0 55 that was given to me, its about 500VA by the looks and at a pinch it would do the trick, but I have these huge heat sinks and I want to use them. I also want to regulate the output. I will look around for a 2 X 65V Tx.
What Im going to do for the front end is join the secondaries from a 2 X 12v 20VA Tx in series with the 65V and just build a simple series pass reg like Nelson does. I expect this is well worth doing.
Fortunately I have access to a local scrap yard that dismantles old Mainframes. The power supplies in these have all sorts of goodies. I get 4 and 8 mm sheets of aluminium from him, caps, switches and power filters, power resistors etc. Some of the boards I have have high voltage caps and power resistors, so I keep them and allways take a look before I order anything.
Im going pay him a visit some time soon to see if he has any 100V electrolytics. Most his stuff is lower voltage, ie 25, 35, 50 V.
My ugs is using 4 x 100,000 35v Caps i got from him for 5 dollars a piece. If you need any of these, let me know. I imagine the freight would be more than the caps but.
I'm going to do some research on the limiting soon so I'll share if I find anything useful.
you don't need to thank me, if it were not for you I wouldn't be planning this project now🙂
I know what you mean, I have an old torroidal 55 0 55 that was given to me, its about 500VA by the looks and at a pinch it would do the trick, but I have these huge heat sinks and I want to use them. I also want to regulate the output. I will look around for a 2 X 65V Tx.
What Im going to do for the front end is join the secondaries from a 2 X 12v 20VA Tx in series with the 65V and just build a simple series pass reg like Nelson does. I expect this is well worth doing.
Fortunately I have access to a local scrap yard that dismantles old Mainframes. The power supplies in these have all sorts of goodies. I get 4 and 8 mm sheets of aluminium from him, caps, switches and power filters, power resistors etc. Some of the boards I have have high voltage caps and power resistors, so I keep them and allways take a look before I order anything.
Im going pay him a visit some time soon to see if he has any 100V electrolytics. Most his stuff is lower voltage, ie 25, 35, 50 V.
My ugs is using 4 x 100,000 35v Caps i got from him for 5 dollars a piece. If you need any of these, let me know. I imagine the freight would be more than the caps but.
I'm going to do some research on the limiting soon so I'll share if I find anything useful.
Last edited:
Hi Luke,
Just ran a few numbers through Bensen's spreadsheet,
using 50+50Vac, 3.5% regulation transformer,
with +-75mF/channel,
Re=0r33,
Vre=24mV,
6pair MJL4281
Tc<=50degC.
The model predicts 286W into 8ohm @ 60degree phase angle and just on the DC SOAR. i.e. easy load and reliable.
540W into 4ohm @ 60degree phase angle and just over the 100ms SOAR. i.e. difficult load and right on most builders' reliability limit.
57.8Apk into 1r0 just on the DC SOAR. it can drive a test load as long as Tc<=50degC.
I would suggest that the 50V+50Vac 1kVA transformer is perfect for a 4ohm capable amplifier.
If you use a much higher voltage then it will not properly drive 4ohm loading.
If your worst case loading is 6ohm and/or 8ohm then the transformer voltage can be higher.
Quiescent conditions.
Re=0r33
Vre=24mV
Ib=73mA
Total Ib=436mA
ClassA current=870mApk
ClassA power 3W into 8r0.
Max power 286W into 8r0.
ClassA power ~-20dBW ref max.
Pq=66.6W plus the driver and front end dissipation.
How big will the heatsink need to be to dissipate 66W and keep Tc<=50degC.
What is your Ta?
Just ran a few numbers through Bensen's spreadsheet,
using 50+50Vac, 3.5% regulation transformer,
with +-75mF/channel,
Re=0r33,
Vre=24mV,
6pair MJL4281
Tc<=50degC.
The model predicts 286W into 8ohm @ 60degree phase angle and just on the DC SOAR. i.e. easy load and reliable.
540W into 4ohm @ 60degree phase angle and just over the 100ms SOAR. i.e. difficult load and right on most builders' reliability limit.
57.8Apk into 1r0 just on the DC SOAR. it can drive a test load as long as Tc<=50degC.
I would suggest that the 50V+50Vac 1kVA transformer is perfect for a 4ohm capable amplifier.
If you use a much higher voltage then it will not properly drive 4ohm loading.
If your worst case loading is 6ohm and/or 8ohm then the transformer voltage can be higher.
Quiescent conditions.
Re=0r33
Vre=24mV
Ib=73mA
Total Ib=436mA
ClassA current=870mApk
ClassA power 3W into 8r0.
Max power 286W into 8r0.
ClassA power ~-20dBW ref max.
Pq=66.6W plus the driver and front end dissipation.
How big will the heatsink need to be to dissipate 66W and keep Tc<=50degC.
What is your Ta?
Last edited:
Hi Ed,
ApexJr.com has 100V caps, but with 75V rails you could get away with 80V caps. It seesm to get expensive when you go to 100V. I would also check out this forum. Do a search in the swapmeet using advanced search for capacitors. You may find something.
A 50Vac 4% regulation on +6% mains voltage can use 80V capacitors.
However, a 230:50+50Vac run on 240Vac+6% is just over 80V if the amps disconnect (blown fuses) from the PSU.
AndrewT,
Thanks for that little bit of information on the transformer selection. I may now use the two trannies I have. The net is close to 74 volts per rail with the available voltage at my residence. Anything in excess of 250 watts will just be nice but not really necessary. My main speaker load is around 5.5 to 6 ohms.
Sounds like the current limiter could be quite a challenge. Hmmm. Any guess as to what the default limit is when built to the BOM?
Tad
Thanks for that little bit of information on the transformer selection. I may now use the two trannies I have. The net is close to 74 volts per rail with the available voltage at my residence. Anything in excess of 250 watts will just be nice but not really necessary. My main speaker load is around 5.5 to 6 ohms.
Sounds like the current limiter could be quite a challenge. Hmmm. Any guess as to what the default limit is when built to the BOM?
Tad
Too much juice?
Luke, I appreciate the thought.
Andrew, I appreciate the caution.
My line voltage is ~122 VAC.
I need the larger (100V) caps.
My transformers are (rated) 55 V rails.
If any of you helpful folk want to incorporate wood into the project, let me know by pm.
Luke, I appreciate the thought.
Andrew, I appreciate the caution.
My line voltage is ~122 VAC.
I need the larger (100V) caps.
My transformers are (rated) 55 V rails.
If any of you helpful folk want to incorporate wood into the project, let me know by pm.
Andrew,
Just to clarify: the transformer size calculations made regards to one channel only?
That is: to properly use 4 ohm loads and a 50-0-50VAC, we need 1KVA per channel?
Thanks!
Just to clarify: the transformer size calculations made regards to one channel only?
That is: to properly use 4 ohm loads and a 50-0-50VAC, we need 1KVA per channel?
Thanks!
Hi,
The predictions are based on 3.5% regulation, about what I would expect a 1kVA transformer to be.
You can see that 2channels driving 8ohms+8ohms has an output of 286+286W.
I see no problem getting close to those powers with both channels driven simultaneously.
2channels driving 40hms+4ohms has an output of 540+540W.
I seriously doubt that driving both channels together will give this predicted power. But will that matter? 540W from one channel and 270W from the other or 500 from both seem good results to me for a 50Vac transformer.
The part that did apply per channel is the smoothing capacitance. I ran the model with +-75mF, that's 300000uF for a 2channel amplifier. I hope I got the numbers of zeros right. I hate using 000000s of uF when 300mF tells everyone, who is interested, what they need to know.
The predictions are based on 3.5% regulation, about what I would expect a 1kVA transformer to be.
You can see that 2channels driving 8ohms+8ohms has an output of 286+286W.
I see no problem getting close to those powers with both channels driven simultaneously.
2channels driving 40hms+4ohms has an output of 540+540W.
I seriously doubt that driving both channels together will give this predicted power. But will that matter? 540W from one channel and 270W from the other or 500 from both seem good results to me for a 50Vac transformer.
The part that did apply per channel is the smoothing capacitance. I ran the model with +-75mF, that's 300000uF for a 2channel amplifier. I hope I got the numbers of zeros right. I hate using 000000s of uF when 300mF tells everyone, who is interested, what they need to know.
the other fuse holder
441-EPD2005 Eagle Plastic Devices Fuseholders, Clips, & Hardware
from Mouser, an Eagle product
441-EPD2005 Eagle Plastic Devices Fuseholders, Clips, & Hardware
from Mouser, an Eagle product
Hi Tad,
I have looked and could not find anything on the forum for setting the VI settings for Jens's Leach, but ther is constant reference toa spread sheet which I could not find.
I did however find it on my drive🙂
All you have to do is endter the yellow fields to get results. I have updated it with the resistor numbers on Jens's latest diagram and BOM.
Dont forget the last yellow field requires input, put the closest value resistor from the calculated one. If you dont your results will be wrong. You will need to know your max rail voltage, and device details from the Datasheet. I dont knwo how accurate this is, but with out access to any thing else,I think Im going with this.
If you need any help, just tell me your rails and Device. From memory your looking at 70V rails and Mjl1302?
Good luck.
I have looked and could not find anything on the forum for setting the VI settings for Jens's Leach, but ther is constant reference toa spread sheet which I could not find.
I did however find it on my drive🙂
All you have to do is endter the yellow fields to get results. I have updated it with the resistor numbers on Jens's latest diagram and BOM.
Dont forget the last yellow field requires input, put the closest value resistor from the calculated one. If you dont your results will be wrong. You will need to know your max rail voltage, and device details from the Datasheet. I dont knwo how accurate this is, but with out access to any thing else,I think Im going with this.
If you need any help, just tell me your rails and Device. From memory your looking at 70V rails and Mjl1302?
Good luck.
Back on post #354 I promised a general purpose power supply PCB that could be used with this project. I now have the prototype PCBs and am ready to show them. The board is setup to accept both 35 and 40mm snap-in capacitors. The prototype is fitted with 8 10,000uF @ 100v Panasonic capacitors (ie: 40,000uF per rail). The PCBs are thick, double sided 2oz copper and silk screened. They measure 3.25 x 8.5 inches.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
A Pic of the above described PCB ...
An externally hosted image should be here but it was not working when we last tested it.
And a shot without the caps. Anyone interested in organizing a group buy? They would be $8 to $10 each dependent upon the number we buy.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Another intersting point is 65V 1kva Tx had rails of 90V at no load, at max 16Amp load the rails had sagged to 63V. Kinda makes me wonder if its worth regulating the output stage, and if so at what voltage, maybe 60 V?
Hi Andrew, so you quote peak or avg power?
A 1KVA Transformer will deliver 1000 Watts Peak between two channels thats 500 Watts a side. That is about 350 Watts Avg. With higher rails you can get maximum power output into both 8 and 4 ohms.
This has been quite an interesting excercise, I wonder if those Mark levinson amps that claim to deliver 1200 Watts into 2 ohms can, and if so what sort of power supplies and transistors do they have?
- Status
- Not open for further replies.