Hi
I have been following this thread for a while and awaited the results. I,m disappointed!
56Vrail to give 156w (50Vpk) is good for 2 pair of outputs but from 5 pair - seems to lose too much voltage somewhere.
BUT what happened to the PSU? 7v lost when loaded at 156w and 10v lost when loaded 275w and all from 800VA.
Something seriously amiss here? This transformer should be able to support at least 400w and maybe 500w at a push.
Can you give more detail of the PSU and wiring losses, caps etc.
I have been following this thread for a while and awaited the results. I,m disappointed!
56Vrail to give 156w (50Vpk) is good for 2 pair of outputs but from 5 pair - seems to lose too much voltage somewhere.
BUT what happened to the PSU? 7v lost when loaded at 156w and 10v lost when loaded 275w and all from 800VA.
Something seriously amiss here? This transformer should be able to support at least 400w and maybe 500w at a push.
Can you give more detail of the PSU and wiring losses, caps etc.
If you followed the thread on the Super Leach you may have read my suggestion for splitting powerlines for front end gain stages and the output stage.
This is the reason why.
Separate powerlines, use higher voltage for the front end, and the amplifier will deliver much more out of a 45 volt transformer.
45 volts rectified is 63 vdc, 50 volts out is a 13 volt drop.
156 and 275 are quite good numbers i think.
This is the reason why.
Separate powerlines, use higher voltage for the front end, and the amplifier will deliver much more out of a 45 volt transformer.
45 volts rectified is 63 vdc, 50 volts out is a 13 volt drop.
156 and 275 are quite good numbers i think.
It's easy to split the supply to the input section, there is a resistor in series with that voltage 🙂
Rails loaded by 8 ohm are 56V @ 156W Output = 50V peak
Rails loaded by 4 ohm are 53V @ 275W Output = 47V peak
These numbers are based on RMS output power.
\Jens
Rails loaded by 8 ohm are 56V @ 156W Output = 50V peak
Rails loaded by 4 ohm are 53V @ 275W Output = 47V peak
These numbers are based on RMS output power.
\Jens
R1, R23, right ?
I did an amplifier from a magazine article once.
On a 30 VAC/300VA toroid and an added 9 VAC transformer for the front end it did 100 watt rms in 8, 175 in 4 Ohms.
(28.3 V and 26.5 V)
Mosfet amplifier, go figure.
I did an amplifier from a magazine article once.
On a 30 VAC/300VA toroid and an added 9 VAC transformer for the front end it did 100 watt rms in 8, 175 in 4 Ohms.
(28.3 V and 26.5 V)
Mosfet amplifier, go figure.
jacco vermeulen said:
How much higher front end voltage is recommended? and how much is the highest voltage we can go within the SOA for power transistors?
I believe that for the Super Duper only something like 6 resistor values were altered, the rest were add-ons.
Everything above Q20 are extra's, the part recommendation list for Q1 till Q20 is the same.
I havent compared them exactly, so correct me if i am wrong.
Jens posted earlier that he did a Spice-sim. for 75 volts.
With the 10 devices Jens uses raising voltage isnt a problem, just needs a bigger heatsink.
At 350 volts theoretically you could build a 1500 watt amplifier with them.( just the thought)
I wasnt suggesting using a higher voltage on the output stage, merely pointing to the effect a single powersupply has.
Under load the voltage drops, so voltage drops on the front stage too.
With lower voltage on the front end it is not able to give the same voltage swing on the output stage and the amplifier reaches 1% distortion on a lower output level.
With a separate powersupply on the front end that can maintain a constant voltage of 63 volts i would expect the output stage to deliver a higher rms power level.
btw:
bigger toroids have a lower regulation factor, so voltage will drop less under load.
But any toroid can deliver twice the nominal output in peaks, an 800 VA model delivers 1600 if needed.
Everything above Q20 are extra's, the part recommendation list for Q1 till Q20 is the same.
I havent compared them exactly, so correct me if i am wrong.
Jens posted earlier that he did a Spice-sim. for 75 volts.
With the 10 devices Jens uses raising voltage isnt a problem, just needs a bigger heatsink.
At 350 volts theoretically you could build a 1500 watt amplifier with them.( just the thought)
I wasnt suggesting using a higher voltage on the output stage, merely pointing to the effect a single powersupply has.
Under load the voltage drops, so voltage drops on the front stage too.
With lower voltage on the front end it is not able to give the same voltage swing on the output stage and the amplifier reaches 1% distortion on a lower output level.
With a separate powersupply on the front end that can maintain a constant voltage of 63 volts i would expect the output stage to deliver a higher rms power level.
btw:
bigger toroids have a lower regulation factor, so voltage will drop less under load.
But any toroid can deliver twice the nominal output in peaks, an 800 VA model delivers 1600 if needed.
Power supply update
Hello,
I just reconfigured my test supply slightly.
Rail voltage at no load = +-93V
Rail voltage @ 8ohm = 76V, Vpeak = 70V => 315W into 8 ohms
Rail voltage @ 4ohm = 66V, Vpeak = 60V => 452W into 4 ohms
This was done with the same voltage on the input section as on the output section.
I have a vario transformer in front of the transformer, this makes the mains change more than it would be in a normal situation and therefor the supply has higher regulation than normal.
\Jens
Hello,
I just reconfigured my test supply slightly.
Rail voltage at no load = +-93V
Rail voltage @ 8ohm = 76V, Vpeak = 70V => 315W into 8 ohms
Rail voltage @ 4ohm = 66V, Vpeak = 60V => 452W into 4 ohms
This was done with the same voltage on the input section as on the output section.
I have a vario transformer in front of the transformer, this makes the mains change more than it would be in a normal situation and therefor the supply has higher regulation than normal.
\Jens
Re: Power supply update
Wow, Jens,
You ran the rails up to 93VDC and no ill effects? Do you think it could live at that voltage? Aren't there some resistors who's value is determined by the projected voltage?
Sounds like you have one tough design working there.
Blessings, Terry
JensRasmussen said:
Wow, Jens,
You ran the rails up to 93VDC and no ill effects? Do you think it could live at that voltage? Aren't there some resistors who's value is determined by the projected voltage?
Sounds like you have one tough design working there.
Blessings, Terry
NO! do not run the amp at more than +-80V
for a longer period of time. I's better to get a transformer that has better regulation and use it at a lower voltage.
Yes you need to recalculate a resistorvalue by means of the formula give on Mr. Leach's hompage.
Again, do not run the amp at more than 80V!!!!
\Jens
for a longer period of time. I's better to get a transformer that has better regulation and use it at a lower voltage.Yes you need to recalculate a resistorvalue by means of the formula give on Mr. Leach's hompage.
Again, do not run the amp at more than 80V!!!!
\Jens
Re: Power supply update
This was with the 2 times 10.000 uF on the board, Jens ?
Are you going to do some more dummy load testing?
I am curious what peak output in 4 Ohm is with 4 * 10.000.
JensRasmussen said:
This was with the 2 times 10.000 uF on the board, Jens ?
Are you going to do some more dummy load testing?
I am curious what peak output in 4 Ohm is with 4 * 10.000.
Re: Re: Power supply update
Hello,
Yes this was done with the 2 x 10.000µF on board.
The peak output voltage will be about 6V below the railvoltage. the RMS power output will depend on the regulation of the transformer used.
I will try to make more measurements tonight.
\Jens
jacco vermeulen said:
Hello,
Yes this was done with the 2 x 10.000µF on board.
The peak output voltage will be about 6V below the railvoltage. the RMS power output will depend on the regulation of the transformer used.
I will try to make more measurements tonight.
\Jens
Jens, apart from the measurements have you , by any chance, managed to arrange for a group buy of parts? or at least perhaps the semiconductors which sometimes are more difficult to trace?
Do you have any recommendations for heatsinks as to a supplier specific product?
Do you have any recommendations for heatsinks as to a supplier specific product?
Hello,
I just send the parts list to my local parts pusher here in Denmark.
They will look into the whole thing tomorrow.
I will post the result here.
If anybody is interested I can post the list, or I can order extra components.
I have not decided on heat sinks yet, but they will need to be in the region of 0.3 – 0.4 k/W
On my prototype amp I used a profile called S245
\Jens
I just send the parts list to my local parts pusher here in Denmark.
They will look into the whole thing tomorrow.
I will post the result here.
If anybody is interested I can post the list, or I can order extra components.
I have not decided on heat sinks yet, but they will need to be in the region of 0.3 – 0.4 k/W
On my prototype amp I used a profile called S245
\Jens
My two cents:
0.40 C/W with the MJL's Jens uses is good for real low impedance loads. That value is good for driving 3 Ohm hard on 45VAC mains.
I just bought 50 Toshiba devices, the ones they dont make anymore.
These have higher C/W than the MJL's so i will have to go for the 0.30 heatsink.
0.40 C/W with the MJL's Jens uses is good for real low impedance loads. That value is good for driving 3 Ohm hard on 45VAC mains.
I just bought 50 Toshiba devices, the ones they dont make anymore.
These have higher C/W than the MJL's so i will have to go for the 0.30 heatsink.
Yikes, NO !
These are the mommy's and daddy's of the 2SA1943/2SC5200.
In TO247/ TOP3L clothing, so they'll fit Jens's extended boards.
These are the mommy's and daddy's of the 2SA1943/2SC5200.
In TO247/ TOP3L clothing, so they'll fit Jens's extended boards.
Wait, you mean these are 5200/1943s? They still make those. I am confused once again. Never mix babysitting and electronics.
Hi Jens
Any chance you can check the numbers without the variac in front of the PSU?
The outputs @ 4R & 8R from the loaded Vrail values quoted are pretty good.
It appears to be the PSU that is strangling the output.
Any chance you can check the numbers without the variac in front of the PSU?
The outputs @ 4R & 8R from the loaded Vrail values quoted are pretty good.
It appears to be the PSU that is strangling the output.
Bill, the 1943/5200 devices are the replacements for transistors as the 2SA1301, 2SA1302, 2SC3280, 2SC3281 and some other numbers.
The latter are the mommy and daddy's.
The 1943/5200 are the preferred now because Toshiba stopped production of the other ones years ago, and the world is flooding with Fakes.
The Chinese Shing Jing Co. makes a couple of those under the original number, dont know if they are considered counterfits.
I bought 2 tubes of the original guys.
Those devices have been used in so many top notch amplifiers in the past that i'll probably need a week to recollect all of them.
One that springs to my mind is the Luxman M07 from the 90s.
Around 125 watts in 8 Ohm , 6 of the Toshiba's per channel.
At that time a msrp of around $15.000.- (msrp in Europe, not US)
How old are your chickens?
The latter are the mommy and daddy's.
The 1943/5200 are the preferred now because Toshiba stopped production of the other ones years ago, and the world is flooding with Fakes.
The Chinese Shing Jing Co. makes a couple of those under the original number, dont know if they are considered counterfits.
I bought 2 tubes of the original guys.
Those devices have been used in so many top notch amplifiers in the past that i'll probably need a week to recollect all of them.
One that springs to my mind is the Luxman M07 from the 90s.
Around 125 watts in 8 Ohm , 6 of the Toshiba's per channel.
At that time a msrp of around $15.000.- (msrp in Europe, not US)
How old are your chickens?
- Status
- Not open for further replies.