Hi Susan,
I've followed your thread on a daily basis and find these recent distortion figures to be most impressive. I also note Mike's very favourable independent comments following his audition.
Do you have any plans for developing higher output, say 20 to 25Vrms across 8 ohms, and have you yet measured the output impedance and angle at say 10kHz ?
Cheers ........... Graham.
I've followed your thread on a daily basis and find these recent distortion figures to be most impressive. I also note Mike's very favourable independent comments following his audition.
Do you have any plans for developing higher output, say 20 to 25Vrms across 8 ohms, and have you yet measured the output impedance and angle at say 10kHz ?
Cheers ........... Graham.
Hi Graham,
Nice to hear from you
I appreciate you spending the time, thank you.
I am pleased that he liked the system
Yes...
For any given bias setting there is a trade off between transformer mode (2:1 or 4:1 step down) output power, impedance and distortion.
I can get higher powers by increasing the PSU voltage. But this increases the quiescent dissipation unless I decrease the bias (but this increases the distortion levels) or requires bigger heatsinks.
I know you know this one
High powers in the hundreds of watts are possible, as has been demonstrated, but this is not my personal interest (100 watts peak into 4 ohms for a low bass driver is probably as much as I would ever need).
Test setup summary:
Maximum Output (PSU approx 34 Vac).
4:1 Tx = 10 Vrms = 12.5 watts
2:1 Tx = 20 Vrms = 50 watts into 8 ohms.
THD+N, 1 kHz at 2 watts into 8 ohms:
4:1 Tx = 0.0058%
2:1 Tx = 0.027%
Output impedance, 1 kHz:
4:1 = 0.25 ohms
2:1 = 0.99 ohms
N.B. I measured the TPA6120 line driver THD+N = 0.005% so I am at the limit of my current setup (and the 4:1 levels may be even lower).
Residual noise is down near the minimum levels capable of being measured by a 24 bit converter i.e. -146.25 dB (N.B. most of the graphs have test signal at -6 dB). (I need a higher resolution converter.)
Still sorting out some of the methodologies for the testing.
I am building an amp as per my 3D sketch (I have the boxes and a heatsink). I will then repeat the current tests and do what I can of the others.
I will then build a pair of "definitive" amps to this design...
... then some serious listening tests.
Something to keep me busy over the holidays.
Best wishes,
Susan.
Nice to hear from you
Graham Maynard said:
I appreciate you spending the time, thank you.
I am pleased that he liked the system
Yes...
For any given bias setting there is a trade off between transformer mode (2:1 or 4:1 step down) output power, impedance and distortion.
I can get higher powers by increasing the PSU voltage. But this increases the quiescent dissipation unless I decrease the bias (but this increases the distortion levels) or requires bigger heatsinks.
I know you know this one
High powers in the hundreds of watts are possible, as has been demonstrated, but this is not my personal interest (100 watts peak into 4 ohms for a low bass driver is probably as much as I would ever need).
Test setup summary:
Maximum Output (PSU approx 34 Vac).
4:1 Tx = 10 Vrms = 12.5 watts
2:1 Tx = 20 Vrms = 50 watts into 8 ohms.
THD+N, 1 kHz at 2 watts into 8 ohms:
4:1 Tx = 0.0058%
2:1 Tx = 0.027%
Output impedance, 1 kHz:
4:1 = 0.25 ohms
2:1 = 0.99 ohms
N.B. I measured the TPA6120 line driver THD+N = 0.005% so I am at the limit of my current setup (and the 4:1 levels may be even lower).
Residual noise is down near the minimum levels capable of being measured by a 24 bit converter i.e. -146.25 dB (N.B. most of the graphs have test signal at -6 dB). (I need a higher resolution converter.)
Still sorting out some of the methodologies for the testing.
I am building an amp as per my 3D sketch (I have the boxes and a heatsink). I will then repeat the current tests and do what I can of the others.
I will then build a pair of "definitive" amps to this design...
... then some serious listening tests.
Something to keep me busy over the holidays.
Best wishes,
Susan.
Choosing the preamp output impedence?
Hi all,
I know we've discussed this before, but it has all gotten lost in the 40+ pages of posts.
I decided on building Nelson Pass's balancedAleph 1.7 preamp which uses two 2k input pots to set the gain and two output pots to set the final output(see link).
Everyone who has built the preamp uses either 5k or 10k, but someone had mentioned that a 1M linear carbon pot law-faked(resistor 10-20% of pot value) sounds much better.
Would that setup give "Zeus" any trouble?
Would 5k or 10k be better?
Glennb gave this explanation for law-faked output resistence:
So, using a parallel resistance calculator - max volume resistence of the 1M pot will be 100k resistor/1M pot/<1-2k source = 1-2k ohms
Or, a 10k pot would equal 475-625Ohms
Seem right?
Hi all,
I know we've discussed this before, but it has all gotten lost in the 40+ pages of posts.
I decided on building Nelson Pass's balancedAleph 1.7 preamp which uses two 2k input pots to set the gain and two output pots to set the final output(see link).
Everyone who has built the preamp uses either 5k or 10k, but someone had mentioned that a 1M linear carbon pot law-faked(resistor 10-20% of pot value) sounds much better.
Would that setup give "Zeus" any trouble?
Would 5k or 10k be better?
Glennb gave this explanation for law-faked output resistence:
So, using a parallel resistance calculator - max volume resistence of the 1M pot will be 100k resistor/1M pot/<1-2k source = 1-2k ohms
Or, a 10k pot would equal 475-625Ohms
Seem right?
Zeus Balanced Line Driver
Hi Darkmoebius and All,
This is what I have been working up:
http://www.susan-parker.co.uk/zeus-lateral+pre-amp-testbench-0_022-thd-full-1-600.jpg
And it's FFT:
http://www.susan-parker.co.uk/z-lat...m-2watt-ip-6db-0_021-thd-45hz-96khz-1-600.gif
THD for the line driver is about 0.01% with IRFP140 mosfets and a 15VA toroid.
Will be working up a "proper" version over the holidays with matched mosfets.
Line Driver = IRFP140 bias at 250 mA total.
Input = Gardners 1:4 transformer
Output line drive to power amp input transformer across mosfets/transformer windings.
Power amp = Lateral FETS
2:1 Transformer:
0.022% at 2 watts
4:1 transformer:
0.0146% at 1 watt
0.0172% at 2 watts
0.0216% at 3 watts
0.0251% at 4.5 watts
0.0302% at 6.5 watts
I will also make a full set of tests etc.
Best wishes,
Susan.
P.S. I would also point out the 50 Hz spike and it's harmonics marching right across the graph. Which is why I have been looking at better PSU designs and also the use of steel enclosures (not just aluminium).
Hi Darkmoebius and All,
This is what I have been working up:
An externally hosted image should be here but it was not working when we last tested it.
http://www.susan-parker.co.uk/zeus-lateral+pre-amp-testbench-0_022-thd-full-1-600.jpg
And it's FFT:
An externally hosted image should be here but it was not working when we last tested it.
http://www.susan-parker.co.uk/z-lat...m-2watt-ip-6db-0_021-thd-45hz-96khz-1-600.gif
THD for the line driver is about 0.01% with IRFP140 mosfets and a 15VA toroid.
Will be working up a "proper" version over the holidays with matched mosfets.
Line Driver = IRFP140 bias at 250 mA total.
Input = Gardners 1:4 transformer
Output line drive to power amp input transformer across mosfets/transformer windings.
Power amp = Lateral FETS
2:1 Transformer:
0.022% at 2 watts
4:1 transformer:
0.0146% at 1 watt
0.0172% at 2 watts
0.0216% at 3 watts
0.0251% at 4.5 watts
0.0302% at 6.5 watts
I will also make a full set of tests etc.
Best wishes,
Susan.
P.S. I would also point out the 50 Hz spike and it's harmonics marching right across the graph. Which is why I have been looking at better PSU designs and also the use of steel enclosures (not just aluminium).
i agree....
i have been reading this thread for awhile. A lot of good information. I do have a question for those out there:
I am looking to build an OPT.... here are my required specs.
1) 64:8 Z ratio
2) high primary inductance to match to BUZ900
3) handles at least 2A of unbalanced DC on primary
good for about 30-40W output.
any takers.. or point me in the right direction?
also looking to build a passive PS for 40vdc and 2A-3A.
Thinking 4700-10,000uf then 2mH then 4700-10,000uf?
Anyway I have a design up my sleeve and I will build it.
finally, what is the BEST way to protect MOSFETS from ESD and over dissapation??
THANKS!!!
i have been reading this thread for awhile. A lot of good information. I do have a question for those out there:
I am looking to build an OPT.... here are my required specs.
1) 64:8 Z ratio
2) high primary inductance to match to BUZ900
3) handles at least 2A of unbalanced DC on primary
good for about 30-40W output.
any takers.. or point me in the right direction?
also looking to build a passive PS for 40vdc and 2A-3A.
Thinking 4700-10,000uf then 2mH then 4700-10,000uf?
Anyway I have a design up my sleeve and I will build it.
finally, what is the BEST way to protect MOSFETS from ESD and over dissapation??
THANKS!!!
Thanks and Seasons Greetings
Hi Tony
Thank you for your kind words and interest.
=================
May I take this opportunity to thank everyone both viewers and posters for all your interest and contributions.
You have rekindled my enthusiasm for audio and given my own past efforts renewed value.
So I would like to extend to everyone seasons greetings and best wishes for the New Year (whenever that may be for you).
Susan.
Hi Tony
Thank you for your kind words and interest.
TonyF said:
=================
May I take this opportunity to thank everyone both viewers and posters for all your interest and contributions.
You have rekindled my enthusiasm for audio and given my own past efforts renewed value.
So I would like to extend to everyone seasons greetings and best wishes for the New Year (whenever that may be for you).
Susan.
Hi
I´ve just found a source for cheap 500VA toroids with 2x110V primarys and 2x38V secondarys.
Would such transformers be of any use in this kind of circuits?
I recently became bass guitar player in a heavy metal band, so cheap watts is of great interest
BTW I don´t need anything below, say, 35Hz. I guess that´s a good thing when it comes to transformer size versus output power?
Merry Christmas!
I´ve just found a source for cheap 500VA toroids with 2x110V primarys and 2x38V secondarys.
Would such transformers be of any use in this kind of circuits?
I recently became bass guitar player in a heavy metal band, so cheap watts is of great interest
BTW I don´t need anything below, say, 35Hz. I guess that´s a good thing when it comes to transformer size versus output power?
Merry Christmas!
Hi Fuling,
Yes, you should be able to use these.
Try using the primarys as per the normal amp and wire the secondaries in series (for 8 ohm speakers) or parallel (for 4 ohm speakers).
This should give you better distortion figures although at lower output powers. With about a 50 volt supply you should get about 100 watts.
If you need more power for 4 ohms then secondaries in series.
Some experimentation may be required.
Then you can use the secondaries only, but the distortion figures get higher.
I hope you have success in this
Yes.
Thank you. I hope yours has been too
Best wishes,
Susan.
Fuling said:
Yes, you should be able to use these.
Try using the primarys as per the normal amp and wire the secondaries in series (for 8 ohm speakers) or parallel (for 4 ohm speakers).
This should give you better distortion figures although at lower output powers. With about a 50 volt supply you should get about 100 watts.
If you need more power for 4 ohms then secondaries in series.
Some experimentation may be required.
Then you can use the secondaries only, but the distortion figures get higher.
I hope you have success in this
Yes.
Thank you. I hope yours has been too
Best wishes,
Susan.
Hi Fuling,
500 VA toroid output:
For your application you will probably need a heatsink a little larger than this.
Main heat depends on the bias setting and the overall voltage. If you can abide the slight noise fan cooling does dramatically reduce the heatsink size.
Large fan(s) running slowly to reduce tip speed is best.
Thermal tracking bias is recomended for higher powers.
Best wishes,
Susan.
Fuling said:Thanks for the reply!
Seems like I better catch one or two of those transformers before they go out of stock then!
I already have one huge 750VA 2x24V toroid at home that might be useful in the PSU, some 65-70VDC should give about 200W or so I guess.
Smells like fan cooling...
Cheers!
500 VA toroid output:
For your application you will probably need a heatsink a little larger than this.
Main heat depends on the bias setting and the overall voltage. If you can abide the slight noise fan cooling does dramatically reduce the heatsink size.
Large fan(s) running slowly to reduce tip speed is best.
Thermal tracking bias is recomended for higher powers.
Best wishes,
Susan.
Hi Fuling,
One pair of matched IRFP260N should be fine for these power levels.
Bias probably about 250 mA per mosfet to start with as you have a higher supply voltage.
Use a third mosfet for the tracking bias.
What sensitivity, impedance speaker drive units will you be using?
Best wishes,
Susan.
Fuling said:Looks good!
For once, fan noise is not an issue!
No living creature would be able to hear a 120mm fan through the sounds we create when we rehe****
Would four pieces of IRFP260N be sufficient here, or do I need more?
One pair of matched IRFP260N should be fine for these power levels.
Bias probably about 250 mA per mosfet to start with as you have a higher supply voltage.
Use a third mosfet for the tracking bias.
What sensitivity, impedance speaker drive units will you be using?
Best wishes,
Susan.
Hi Susan
One pair seems a little puny for this job, the average power level is quite high when playing bass guitar.
Better safe than sorry I guess.
Haven´t decided on the speaker yet, but I´m leaning towards a 200 litre bass reflex box with two 96dB 8 ohms 15" drivers.
Should be able to deliver quite impressive SPL´s
One pair seems a little puny for this job, the average power level is quite high when playing bass guitar.
Better safe than sorry I guess.
Haven´t decided on the speaker yet, but I´m leaning towards a 200 litre bass reflex box with two 96dB 8 ohms 15" drivers.
Should be able to deliver quite impressive SPL´s
Hi Fuling
Should be VERY impressive.
Each mosfet is rated at 300 watts, 35 amps at 100 degrees centigrade (chip temp) continious, 200 amps pulsed.
Using a pair you are dividing the load across two mosfets. Then with a 2:1 (ish) step down you are trading current for voltage.
Main limitation is the bias level (which will also depend on the wire diameter of the 120 volt windings).
Hope this helps.
Best wishes,
Susan.
Fuling said:Hi Susan
One pair seems a little puny for this job, the average power level is quite high when playing bass guitar.
Better safe than sorry I guess.
Haven´t decided on the speaker yet, but I´m leaning towards a 200 litre bass reflex box with two 96dB 8 ohms 15" drivers.
Should be able to deliver quite impressive SPL´s
Should be VERY impressive.
Each mosfet is rated at 300 watts, 35 amps at 100 degrees centigrade (chip temp) continious, 200 amps pulsed.
Using a pair you are dividing the load across two mosfets. Then with a 2:1 (ish) step down you are trading current for voltage.
Main limitation is the bias level (which will also depend on the wire diameter of the 120 volt windings).
Hope this helps.
Best wishes,
Susan.
Tube variant
Hello Suzan, all,
Interesting design.
I'm curious about the comparison made with the tubed WE25.
Somewhere in this thread it was mentioned that the design of Suzan can also be made with tubes. I guess the tube must be in a kind of follower configuration.
Alas I'm unable to design this myself.
Any interest of others to give it a try?
Best regards, Arjen.
Hello Suzan, all,
Interesting design.
I'm curious about the comparison made with the tubed WE25.
Somewhere in this thread it was mentioned that the design of Suzan can also be made with tubes. I guess the tube must be in a kind of follower configuration.
Alas I'm unable to design this myself.
Any interest of others to give it a try?
Best regards, Arjen.
Apologies for that last comment being so out of place.
I have been having terrible probs with my e-mail / Internet connection, with outages of digital signal continuity upsetting my computer.
I was answering Eva's post about playing loud piano music, which somehow appeared as if it was the last entry in this column, and thus I thought that it had been repeated.
Now that I am back on normal line again I see different and normally progressive postings.
Cheers ............ Graham.
I have been having terrible probs with my e-mail / Internet connection, with outages of digital signal continuity upsetting my computer.
I was answering Eva's post about playing loud piano music, which somehow appeared as if it was the last entry in this column, and thus I thought that it had been repeated.
Now that I am back on normal line again I see different and normally progressive postings.
Cheers ............ Graham.