Happy Christmas, Gaborbela, Nigel and huggygood and each and every lover of JLH amp,
On my menu plate, I see many PCB designs, which I am doing my best to complete (sleepless nights), so I can enjoy holidays with relatives and enjoy my lamb and cake ....
I may probably build an updated version of basic 2003 amp with premium components, not sure yet.
On my menu plate, I see many PCB designs, which I am doing my best to complete (sleepless nights), so I can enjoy holidays with relatives and enjoy my lamb and cake ....
I may probably build an updated version of basic 2003 amp with premium components, not sure yet.
It will be lamb, roast potatoes, many vegetables. Good thing you asked as I forgot the wine, yet gave so much away. I really like white wines from the South of France. These were the work of the Austrailians in the late 1980's. It was always said the land was wrong. Not really and cold fermentation changed plenty. Good hi fi is the same, if it looks that something can work usually it will. " Terroir " is sometimes made too important. It is the husbandry. As best I know it in the folk law vines that stand high above a river that forces the roots to dig deep for minerals and water do best. If it takes 50 years that's even better. Chalk helps. This was shown to be mostly wrong as the wine being too warm was prime, obvious when thought through. The big stainless steel vats look so out of place. The wine they make can be very good. Good science always has it's place. That's all I know about wines so not an expert.
On the road now to Shaftesbury ( Very French, more so than Cornwall ). Very happy Christmas all.
YouTube
This might be usefull to some folks
Hi All,
as a JLH fan and avid experimentor I have made up a spreadsheet that does all the calculations with regards to rail voltages, power output, current different load resistances, etc, plus some on the power supply assuming the use of a capacitance multiplier.
I thought it would be useful to the folks here, but the site will not accept uplaods of .xls files.
Would anyone be intrested in it, If so any ideas on how to upload it?
If I can get it uploaded, you are welcome to use it with no warrantry and at your own risk.
Cheers
Mike
Hi All,
as a JLH fan and avid experimentor I have made up a spreadsheet that does all the calculations with regards to rail voltages, power output, current different load resistances, etc, plus some on the power supply assuming the use of a capacitance multiplier.
I thought it would be useful to the folks here, but the site will not accept uplaods of .xls files.
Would anyone be intrested in it, If so any ideas on how to upload it?
If I can get it uploaded, you are welcome to use it with no warrantry and at your own risk.
Cheers
Mike
I think you'll find a lot of data on suitable power supplies, power output and voltages prepared by JLH himself and published in Wireless World, back in the day. The material was all archived some years ago on Geoff Moss' TCAA Site, along with updates and several later versions. The PDFs are now upgraded and hosted on Rod Elliott's ESP site, take a look at the output rating tables there: http://sound-au.com/tcaas/jlh1969.pdf
Fred,
can you protect the sheets with a "Blank" password. That way any Member who wants to unprotect don't need a password to open up the formulae.
But unprotect the data input cells before you protect. This allows the Members to use the spreadsheet without risk of corrupting all your formulae and links.
If you want you can format all the data input cells to a different colour so that we can see where we are required to make a choice.
can you protect the sheets with a "Blank" password. That way any Member who wants to unprotect don't need a password to open up the formulae.
But unprotect the data input cells before you protect. This allows the Members to use the spreadsheet without risk of corrupting all your formulae and links.
If you want you can format all the data input cells to a different colour so that we can see where we are required to make a choice.
thank's !!!
good idea !
just to share my christmas with you, i just got a couple tight air ATM1S / carySLP98 in repair. I am super happy, I love this set :RÉ
Hi Andrew,
I hadn't realised that I had protected the sheets and cells
I wrote the original in Office 2010 and then exported it as an Excel '97 spreadsheet as I though this would be the format that most people would be able to read. If there is another format that would be better, please let me know - So I think I have now done what you suggested.
My thoughts when I did the spreadsheet were as follows:
Take the first sheet which is '15 Ohms' and is based on Geoff Moss's articles, but is equally valid for the original JLH - also please note that it is all based on approximations to give a feel for what is going on and is for a single channel. Your mileage may vary, batteries are not included, conditions may apply
At the top of the page the 'Number of output transistors' is the output configuration, which is 2 - a single pair (Tr1 and Tr2) and the 'Load' is the intended speaker impedance - 15 Ohms
In the main table, working left to right:
The 'Rail Voltage' is the +/- supplies that you intend to use, so let's take '18 volts' - that is +/- 18 volts, but could also be used for a single supply of 36 volts.
Vce (sat) is the saturation voltage of the output transistors, which I nominally put at 1 Volt - although this is more likely to be about 0.6 to 0.7v.
From this we can get the theoretical maximum voltage swing, which is the rail to rail voltage minus 2 of the Vce(sat) - in this case about 34 volts.
Then we can get the 'RMS value' of this output swing. In this case 12 volts.
The load resistance is stated as 15 ohms from above.
Delivering 12 volts RMS into 15 ohms gives us 9.6 watts and needs 0.8 Amps RMS. so we need to set Iq to about this.
With no signal and the amp just sitting there idling, the transistors are dissipating power at the max value, which is the voltage accross them multipled by the current going through them. So we have the full rails multiplied by the standing current, which in this case is 36 volts multipled by 0.8 amps, which is 28.9 watts. Assuming that the midpoint of the output pair is at 0 Volts for this example, each transistor is dissipating this power divided by the number of transistors(2) which is 14.4 watts. This is usful for latter sheets, where we may want greater power output with more output pairs and also for specifying heatsinks.
On this point I have used the quiescent set up with a current limted regulated power supply to measure/calibrate heatsinks by putting a known power through a pair of output transistors and measuring the rise in temperature about ambient. This is a useful excercise in itself.
The idea of doing this for a plethora of rail voltages is so that I could see what the rails and Iq would be for various output powers - a bit ar*e about face I know, but we are usually constrained by transformer voltages etc.
The example of the '+/- 18' volt rails hold true for the +36 volt single rail and matches up with the figures in JLH's original paper - so it seems to be about right
The second table below is a simple reference of useful RMS voltages into the specified load impedance.
Geoff Moss in his exemplary update article did the calculation a bit more direct with a specific formula, which I used at the bottom of the page as a double check which matched up with the main table.
The individual sheets are then presets of various numbers of output transistors (2, 4, 8 etc.) into the various load impedances. For example, Geoff's design for driving the ESL uses 4 ouput transistors (2 pairs) with 0.1 Ohm current sharing resistors.
If you want to really spread the power dissipation for a high-power version, you could try using 8 output transistors and a 6 Ohm load which I am playing with on the bench.
The Power Supply table of the right hand side was (again) just something useful to calculate the various voltage drops and power dissipation etc. when using a capacitor multiplier as shown in Geoff's article. - note again, this is for a single channel.
Hope this explains it enough and folks find it useful.
Best regards for the Christmas season to all.
Mike
I hadn't realised that I had protected the sheets and cells
I wrote the original in Office 2010 and then exported it as an Excel '97 spreadsheet as I though this would be the format that most people would be able to read. If there is another format that would be better, please let me know - So I think I have now done what you suggested.
My thoughts when I did the spreadsheet were as follows:
Take the first sheet which is '15 Ohms' and is based on Geoff Moss's articles, but is equally valid for the original JLH - also please note that it is all based on approximations to give a feel for what is going on and is for a single channel. Your mileage may vary, batteries are not included, conditions may apply
At the top of the page the 'Number of output transistors' is the output configuration, which is 2 - a single pair (Tr1 and Tr2) and the 'Load' is the intended speaker impedance - 15 Ohms
In the main table, working left to right:
The 'Rail Voltage' is the +/- supplies that you intend to use, so let's take '18 volts' - that is +/- 18 volts, but could also be used for a single supply of 36 volts.
Vce (sat) is the saturation voltage of the output transistors, which I nominally put at 1 Volt - although this is more likely to be about 0.6 to 0.7v.
From this we can get the theoretical maximum voltage swing, which is the rail to rail voltage minus 2 of the Vce(sat) - in this case about 34 volts.
Then we can get the 'RMS value' of this output swing. In this case 12 volts.
The load resistance is stated as 15 ohms from above.
Delivering 12 volts RMS into 15 ohms gives us 9.6 watts and needs 0.8 Amps RMS. so we need to set Iq to about this.
With no signal and the amp just sitting there idling, the transistors are dissipating power at the max value, which is the voltage accross them multipled by the current going through them. So we have the full rails multiplied by the standing current, which in this case is 36 volts multipled by 0.8 amps, which is 28.9 watts. Assuming that the midpoint of the output pair is at 0 Volts for this example, each transistor is dissipating this power divided by the number of transistors(2) which is 14.4 watts. This is usful for latter sheets, where we may want greater power output with more output pairs and also for specifying heatsinks.
On this point I have used the quiescent set up with a current limted regulated power supply to measure/calibrate heatsinks by putting a known power through a pair of output transistors and measuring the rise in temperature about ambient. This is a useful excercise in itself.
The idea of doing this for a plethora of rail voltages is so that I could see what the rails and Iq would be for various output powers - a bit ar*e about face I know, but we are usually constrained by transformer voltages etc.
The example of the '+/- 18' volt rails hold true for the +36 volt single rail and matches up with the figures in JLH's original paper - so it seems to be about right
The second table below is a simple reference of useful RMS voltages into the specified load impedance.
Geoff Moss in his exemplary update article did the calculation a bit more direct with a specific formula, which I used at the bottom of the page as a double check which matched up with the main table.
The individual sheets are then presets of various numbers of output transistors (2, 4, 8 etc.) into the various load impedances. For example, Geoff's design for driving the ESL uses 4 ouput transistors (2 pairs) with 0.1 Ohm current sharing resistors.
If you want to really spread the power dissipation for a high-power version, you could try using 8 output transistors and a 6 Ohm load which I am playing with on the bench.
The Power Supply table of the right hand side was (again) just something useful to calculate the various voltage drops and power dissipation etc. when using a capacitor multiplier as shown in Geoff's article. - note again, this is for a single channel.
Hope this explains it enough and folks find it useful.
Best regards for the Christmas season to all.
Mike
You did not protect any of the cells.
That's why I suggested you should protect but without a password.
It helps new or unfamiliar users from accidentally corrupting your work.
Those that are familiar will know how to "tools - protect document - unprotect sheet" to get access to the formulae when they need to. And will know to hold a spare protected copy in a separate folder, just in case.
That's why I suggested you should protect but without a password.
It helps new or unfamiliar users from accidentally corrupting your work.
Those that are familiar will know how to "tools - protect document - unprotect sheet" to get access to the formulae when they need to. And will know to hold a spare protected copy in a separate folder, just in case.
Just about to cook the lamb. Gascogne wines are excellent and not expensive here, good idea. I gave away all my best wine, it's Chardonnay Pays D'Oc today ( I think Oc means yes from very old times, people who say yes ). I was given some Riesling as I protest it can be very good. This one only 10.5% so might be a bit sweet. Luxembourg ones can be wonderful. The Pays D'Oc might not be a Chablis, at 13% it will be a bit mighty ( not in oak I think ). We Brits make supurb Champagne clones. The climate is if anything better than the Champagne as global warming might be real and might be helping, we get a good rain and day length balance. We get slightly less frost in coastal chalk areas, same chalk as Champagne which continues through southern England. Grapes the Germans make table wine with do better here, Muller Thurgau for example . Far more complexity. Far too expensive alas. If you visit Madrid see the botanic gardens to see their vines of the world collection. In fact see Madrid in May, great city and not expensive.
5 Best English Sparkling Wines | ShortList
If my laziness doesn't overtake me I might be able to log many variations of the JLH. My first priority is to make the 27V version. My hi fi is packed away for house moving. I might bring it out of the garage if results look OK.
BTW. I will cook Yorkshire pudding with the lamb, Usually it's with beef. It's very much a pancake mix put into the fat at high temperature. I suspect back in time it was a common recipe of all Europe. If you never tried it have a go. It is usualy for it to go dark brown at the edge and less cooked in the middle. Keep the oven door shut for at least 20 minutes. Like most things best when just cooked.
5 Best English Sparkling Wines | ShortList
If my laziness doesn't overtake me I might be able to log many variations of the JLH. My first priority is to make the 27V version. My hi fi is packed away for house moving. I might bring it out of the garage if results look OK.
BTW. I will cook Yorkshire pudding with the lamb, Usually it's with beef. It's very much a pancake mix put into the fat at high temperature. I suspect back in time it was a common recipe of all Europe. If you never tried it have a go. It is usualy for it to go dark brown at the edge and less cooked in the middle. Keep the oven door shut for at least 20 minutes. Like most things best when just cooked.
Last edited:
hi nigel
there is very good wine in the "country of OC" but you have to know the producers because there is also a lot of no good.
when champagne france, we do not drink any more, we prefer a cremant of Loire or Montlouis and white wine hillside of the layon.
I'm going to taste an English sparkling wine, I did not know.
we did an Orloff roast with a lussac st emilion castle white beard, a real treat!
I would like to taste an English pudding one day, I have often heard about it but never eaten (yet we have English friends but do not know how to cook)
Happy New Year and we talk about it after the holidays
ps: I bought a strange JLH with mosfet on aliexpress, just to see what they could do with this schema
there is very good wine in the "country of OC" but you have to know the producers because there is also a lot of no good.
when champagne france, we do not drink any more, we prefer a cremant of Loire or Montlouis and white wine hillside of the layon.
I'm going to taste an English sparkling wine, I did not know.
we did an Orloff roast with a lussac st emilion castle white beard, a real treat!
I would like to taste an English pudding one day, I have often heard about it but never eaten (yet we have English friends but do not know how to cook)
Happy New Year and we talk about it after the holidays
ps: I bought a strange JLH with mosfet on aliexpress, just to see what they could do with this schema
These are real values tried today ( they look about right ). Worked first time! The parts are what I had easilly to hand. 100K fixed resistor for now, already reasonably near to the centre voltage. BC327/337 -40 seem OK. 120R/620R is what I had. Now to try to find some test gear that is working.The PSU is 30V 3A Voltcraft lab PSU for now. 27V 1.2A seems a nice idea. Even though the heat sink is generous it gets warm. My bootstrap cap is only 25V rating. I think it will do for now. The build is untidy, again OK for now. I am rather impresssed already. The only time I found it reacting was testing BC327 base, that's to be exspected.
