--> Amp Changes 2 - running
Hi Susan,
thanks a lot for your replies !
I bought only 50 IRF 530N over Ebay for about 12 EURO.
And i match any parts, so i have many IRF 530N from 4,19V - 4,22V.
I think qty of 25 my bee enough.
The TR1 with 2x 12V 30VA is only a test Transformator, i read the tread about the "bifilar windings", but i have no dealer today.
I don't know if the TR1 has "bifilar 12 volt windings", i thing about to use the "E. J. Sowter #8160".
Do you know how mutch the #8160 cost ?
Fantastic, no i understand more and more how to drive the Zeus Amp -
i am excited !
Where can I buy the OP275 or the TPA6120 in Germany or Europe ?
a nice day - to day - uwe
Hi Susan,
thanks a lot for your replies !
Susan-Parker said:
I bought only 50 IRF 530N over Ebay for about 12 EURO.
And i match any parts, so i have many IRF 530N from 4,19V - 4,22V.
I think qty of 25 my bee enough.
The TR1 with 2x 12V 30VA is only a test Transformator, i read the tread about the "bifilar windings", but i have no dealer today.
I don't know if the TR1 has "bifilar 12 volt windings", i thing about to use the "E. J. Sowter #8160".
Do you know how mutch the #8160 cost ?
Fantastic, no i understand more and more how to drive the Zeus Amp -
i am excited !
Where can I buy the OP275 or the TPA6120 in Germany or Europe ?
a nice day - to day - uwe
Re: --> Amp Changes 2 - running
Hi Uwe,
And thank you for your interest and enthusiasm.
50 IRF530N for 12 Euro sounds a good buy to me - opening another browser window quickly and seeing if there are any more going.
It's difficult to know definitely unless the transformer manufacturer actually specifies it.
The 8160 "interstage" transformer, Standard model 8160 with M6 core will be approximately UKP 70 each, plus VAT as appropriate, and shipping.
Discounts:
6 off = 10%
13 off = 15%
25 off = 20%
A M6/Mumetal 50/50% core version will be about 25% more expensive.
Note however that Mumetal saturates at a lower level than M6 (40%?) so a 100% Mumetal core is not advised.
(I have some Mumetal laminations to try out but still need to set up to be able to test these - unfortunately life has kept me busy.)
N.B. I have added this pricing info to the input transformer page.
Excellent
Digikey - http://de.digikey.com
OP275 are Euro 1.64 each.
TPA6120 are Euro 3.45 each.
If you don't feel up to doing a special on PTH micro board PCBs for the TPA6120 are available from
http://home.swipnet.se/~w-50719/hifi/qrv07/index.html
Note to All: Personally it is relatively easy for me to make up the TPA6120 driver the hard way on the micro board, but I would NOT recommend this to anyone who isn't considerably experienced in surface mount technology and has a really good temperature controlled soldering iron with very fine tips.
Hope this helps.
Best wishes,
Susan.
Hi Uwe,
dx.master said:
And thank you for your interest and enthusiasm.
50 IRF530N for 12 Euro sounds a good buy to me - opening another browser window quickly and seeing if there are any more going.
It's difficult to know definitely unless the transformer manufacturer actually specifies it.
The 8160 "interstage" transformer, Standard model 8160 with M6 core will be approximately UKP 70 each, plus VAT as appropriate, and shipping.
Discounts:
6 off = 10%
13 off = 15%
25 off = 20%
A M6/Mumetal 50/50% core version will be about 25% more expensive.
Note however that Mumetal saturates at a lower level than M6 (40%?) so a 100% Mumetal core is not advised.
(I have some Mumetal laminations to try out but still need to set up to be able to test these - unfortunately life has kept me busy.)
N.B. I have added this pricing info to the input transformer page.
Excellent
Digikey - http://de.digikey.com
OP275 are Euro 1.64 each.
TPA6120 are Euro 3.45 each.
If you don't feel up to doing a special on PTH micro board PCBs for the TPA6120 are available from
http://home.swipnet.se/~w-50719/hifi/qrv07/index.html
Note to All: Personally it is relatively easy for me to make up the TPA6120 driver the hard way on the micro board, but I would NOT recommend this to anyone who isn't considerably experienced in surface mount technology and has a really good temperature controlled soldering iron with very fine tips.
Hope this helps.
Best wishes,
Susan.
Re: Balanced Line Drivers
Aaaah, glad to see we have finally breeched the oh-so-important topic of how to get a balanced/differential signal to feed the Zeus amp. There a few other threads around here on the subject .
Peter Daniel started a similar thread on this subject in which Carlosfm suggested the Analog Devices AD815 as great sounding basis for differential or single-ended to differential preamps. Although, it looks like that thread is fading to an end, maybe we can move the subject forward.
The datasheet has three interesting designs for SE to Diff line drivers.
But, Wensan posted today that:
So, that brings us to the final design:
Thoughts anyone?
I'm ordering up some samples(including AD826) and hopefully the great evaluation board they have for the AD815.
Susan-Parker said:
Aaaah, glad to see we have finally breeched the oh-so-important topic of how to get a balanced/differential signal to feed the Zeus amp. There a few other threads around here on the subject .
Peter Daniel started a similar thread on this subject in which Carlosfm suggested the Analog Devices AD815 as great sounding basis for differential or single-ended to differential preamps. Although, it looks like that thread is fading to an end, maybe we can move the subject forward.
The datasheet has three interesting designs for SE to Diff line drivers.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
But, Wensan posted today that:
So, that brings us to the final design:
An externally hosted image should be here but it was not working when we last tested it.
Thoughts anyone?
I'm ordering up some samples(including AD826) and hopefully the great evaluation board they have for the AD815.
You can create an extremely accurate and low distortion diff output from an SE input using a split load emitter follower with a floating power supply.
A conventional emitter follower is set up using an NPN BJT with a supply to the collector, emitter resistor which is split into two equal, series parts, and the negative supply to the bottom of these two resistors. The center of the two series emitter resistors is grounded. The base is biased from a divider string between the positive and negative supply, but the supply is fully floating, which means the winding/frame capacitance should be low.
Output signal is taken capacitively from the emitter and from the negative supply rail. These are exactly and precisely balanced, and both of very low impedance. The circuit, pioneered by a Frenchman whose name escapes me, is eminently suited to tubes as well. The one difficulty is the fully floating supply, but with SS this is not terribly difficult.
Don't ask me how it sounds. I haven't a clue but the idea is extremely elegant and simple.
Cheers,
Hugh
A conventional emitter follower is set up using an NPN BJT with a supply to the collector, emitter resistor which is split into two equal, series parts, and the negative supply to the bottom of these two resistors. The center of the two series emitter resistors is grounded. The base is biased from a divider string between the positive and negative supply, but the supply is fully floating, which means the winding/frame capacitance should be low.
Output signal is taken capacitively from the emitter and from the negative supply rail. These are exactly and precisely balanced, and both of very low impedance. The circuit, pioneered by a Frenchman whose name escapes me, is eminently suited to tubes as well. The one difficulty is the fully floating supply, but with SS this is not terribly difficult.
Don't ask me how it sounds. I haven't a clue but the idea is extremely elegant and simple.
Cheers,
Hugh
Hi bobo1on1,
Hum, not sure why this would happen.
How are you driving the inputs (your schematic shows one input as IN-GND although I assume this is not actually the case)?
What is the lowest frequency you are driving at full power?
What is the real power supply voltage?
Looking at the irf540 data sheet you should be good for more than 30 amps per device but perhaps this is being exceeded?
0.35 ohms and 30 volts is 86 amps at DC.
I would have thought the fuse after the storage capacitors would have shriveled long before this.
Um.
More info....
Best wishes,
Susan.
bobo1on1 said:
Hum, not sure why this would happen.
How are you driving the inputs (your schematic shows one input as IN-GND although I assume this is not actually the case)?
What is the lowest frequency you are driving at full power?
What is the real power supply voltage?
Looking at the irf540 data sheet you should be good for more than 30 amps per device but perhaps this is being exceeded?
0.35 ohms and 30 volts is 86 amps at DC.
I would have thought the fuse after the storage capacitors would have shriveled long before this.
Um.
More info....
Best wishes,
Susan.
Susan-Parker said:
Hi Susan and bobo1on1,
Ur Mosfets are damaging because their gate to Source voltage is
exceeds the VGS limit.
Ur implementation of zeners is totally wrong. instead u must use 11V back to back Zeners and their is no need of 80V zeners. Simply lower the rail voltage or use IRF640N 200V mosfet .
The Transformer is creating a sort of TRICKLE FLYBACK although Body diodes are present but they also avalanche due to this effect and the mosfets are damaged.
Hope this Helps
Schematic Update
Hi,
Update schematic showing zener protection as per IR recommendations.
I have mentioned this in previous posts but I should have updated this schematic. The previous version was as drawn in 1994.
I apologize for this as I have come to realize that once the posts move on to the next page it become quickly quite difficult to remember where things are.
I spent a while yesterday trying to find the post with the Sowter input transformer price info as I couldn't find the email - then realized that I had had this from my phone conversation - oops!
The image (once one has refreshed the browser cache) should now appear correctly in the
http://www.susan-parker.co.uk/zeus.htm
and other pages as well.
However I have run configurations like the previous schematic with the zener across the mosfet directly and whilst I have managed to get the mosfet to take off there are only little squibs of RF on part of the cycle and the mosfets themselves haven't minded.
Of course with the output transformer capable of radiating up into the many megahertz region this is not a good idea from the EMC/RFI point of view.
Because the mosfet is a follower it has a voltage gain just under unity - so parasitic oscillation doesn't amplify unlike mosfets used as voltage gain.
However I am sure that there are still lots of possibilities to let the magic smoke out. Partly why I have been big on cheepie mosfets like the IRFP140s which whilst not quite throw away items like fuses don't cause massive financial grief unlike nice lateral or RF mosfets
I do applaud bobo1on1's efforts and hope we can get this sorted out before discouragement sets in.
I am meant to be working at the moment, hence lack of replies on other less pressing posts.
Best wishes,
Susan.
Hi,
Update schematic showing zener protection as per IR recommendations.
I have mentioned this in previous posts but I should have updated this schematic. The previous version was as drawn in 1994.
I apologize for this as I have come to realize that once the posts move on to the next page it become quickly quite difficult to remember where things are.
I spent a while yesterday trying to find the post with the Sowter input transformer price info as I couldn't find the email - then realized that I had had this from my phone conversation - oops!
The image (once one has refreshed the browser cache) should now appear correctly in the
http://www.susan-parker.co.uk/zeus.htm
and other pages as well.
However I have run configurations like the previous schematic with the zener across the mosfet directly and whilst I have managed to get the mosfet to take off there are only little squibs of RF on part of the cycle and the mosfets themselves haven't minded.
Of course with the output transformer capable of radiating up into the many megahertz region this is not a good idea from the EMC/RFI point of view.
Because the mosfet is a follower it has a voltage gain just under unity - so parasitic oscillation doesn't amplify unlike mosfets used as voltage gain.
However I am sure that there are still lots of possibilities to let the magic smoke out. Partly why I have been big on cheepie mosfets like the IRFP140s which whilst not quite throw away items like fuses don't cause massive financial grief unlike nice lateral or RF mosfets
I do applaud bobo1on1's efforts and hope we can get this sorted out before discouragement sets in.
I am meant to be working at the moment, hence lack of replies on other less pressing posts.
Best wishes,
Susan.
Well they seem to blow up when I disconnect the speaker while it's playing at some serious power, I know I shouldn't do this, but it can happen sometime.
I think it creates a big spike on the transformer just high enough to blow up the mosfet, I am using some 800 volt mosfets right now and there is no problem at all.
So I will put in the zeners, maybe it won't do any good, but the cost is next to nothing.
Oh and I forgot to put in the fuse, again!
Right now I am using a 2x 6 v 120 VA toroid with a 16v 55000 µF cap.
The in_gnd is not the same as the psu gnd, it is the gnd coming from the signal source (my soundcard).
I can get 48 75345P power mosfets for 4 euro, I might use those, although the gate capacitance is a bit high, 4 nF.
Ehm what are back to back zeners? I did put in a 12 volt zener between the gate and the source, that should make sure the Vgs does not exceed 12 volts, or am I wrong?
I think it creates a big spike on the transformer just high enough to blow up the mosfet, I am using some 800 volt mosfets right now and there is no problem at all.
So I will put in the zeners, maybe it won't do any good, but the cost is next to nothing.
Oh and I forgot to put in the fuse, again!
Right now I am using a 2x 6 v 120 VA toroid with a 16v 55000 µF cap.
The in_gnd is not the same as the psu gnd, it is the gnd coming from the signal source (my soundcard).
I can get 48 75345P power mosfets for 4 euro, I might use those, although the gate capacitance is a bit high, 4 nF.
amp_man_1 said:
Ehm what are back to back zeners? I did put in a 12 volt zener between the gate and the source, that should make sure the Vgs does not exceed 12 volts, or am I wrong?
bobo1on1 said:
Yes it does. One of the perils of development.
BTW what is your definition of serious power?
More than 35 watts I assume
Hum, you should be driving a diferential signal into the mosfets - see darkmoebius's very helpful post:
http://www.diyaudio.com/forums/showthread.php?postid=508517#post508517
However an input transformer drive will allow you to drive the gates over the nominal supply rail which could give you a few more watts.
You need to look at the capacitance from the gate to the drain, not the source.
http://www.fairchildsemi.com/ds/HU/HUFA75345S3S.pdf
Figure 12 - Crss.
However if you are runing off automotive supply i.e. 14 volts then you are still up in the 1000 pF region.
Best wishes,
Susan.
A big inductanse will keep on raising the voltage until the energy from the collapsing magnetic field is wasted somehow. Usually you have a spark through the isolation of the windings. I got a jolt from a big transformer that was several thousands of volts once....
I got rid of the problem with fets blowing up when i connected 18 volts 1,3 watts zeners back to back between gate and source..
My amp is very happily putting out 115 watts of clean dynamic power into 4-8 ohms, without a trace of fatigue or distorsion.
-AND THE SOUND IS WONDERFUL!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
For all of you that have read Allen Wright`s Tube preamp......: "DDR" is outstanding!!!!!
Spead and dynamics is WONDERFULL!!!
I love this amp!!!!
Thank you Susan!!!!!!!!!!!
Hi Circlomanen ,
Have you any information about the Tr1 an Tr2 ?
And any pictures ?
Which Mosfet do you use and what about the Biascontrol ?
How mutch current go through the Mosfets ?
I use now 0,65A through every Mosfet, that sounds good.
Look at the Schematic:
Changes: Rx = 7,5Ohm.
- uwe
Circlomanen said:
Have you any information about the Tr1 an Tr2 ?
And any pictures ?
Which Mosfet do you use and what about the Biascontrol ?
How mutch current go through the Mosfets ?
I use now 0,65A through every Mosfet, that sounds good.
Look at the Schematic:
An externally hosted image should be here but it was not working when we last tested it.
. 
Changes: Rx = 7,5Ohm.
- uwe
Tr 1 is a standard 50 VA toroid vith a singel 24 volt secondary and a split 115+115 volt primary coupled backwards.
Tr2 is a centertapped Choke winded on a E150 laminationstack with normal plastic isolated 1mm2 coupling wire, 71 turns trifilar at the moment,i will rewind it as soon as i get any time to do it.
The mosfets are IRFP250/IR with 300 ohm gatestoppers.
Bias is 355 mA / Fet.
Bias-controll is a 7,5 volt Zener as a voltage-referece, a 100 K pot and a P-chanel IRF9620 as a buffer/thermal feedback- to give the transformer midpoint a low impedans and a stabel V-ref. Cold it runns 320 mA of current/Fet, after 2 hours and a temp-rise of 15 degrees celcius it runns 355 mA, and seems as it has stabilised.
No Im sorry, It looks like a birdnest in the midle of a tornado, its just a prototyp to check out the sound and how it works.
Susan-Parker said:
Yes it does. One of the perils of development.
BTW what is your definition of serious power?
More than 35 watts I assume
Hum, you should be driving a diferential signal into the mosfets - see darkmoebius's very helpful post:
http://www.diyaudio.com/forums/showthread.php?postid=508517#post508517
However an input transformer drive will allow you to drive the gates over the nominal supply rail which could give you a few more watts.
Best wishes,
Susan.
It was running at about 40 watts into a 4 ohm speaker, which is alot for an old philips speaker from 1974.
The output from my soundcard is a differential signal, it just depends on how you look at it, remember the psu from the amplifier is completely seperated from the computer's psu, the only electrical connection between them is the sound connection.
So in that way the soundcard puts a voltage over the 1k resistors, which has the same effect as using a transformer, the only difference beeing the galvanic separation.
Oh and if I am going to use those cheap fets with high capacitance I will probably drive them with an opamp like transistor circuit which can deliver a reasonable amount of current to drive the gates, maybe something in class A with 500 mA idle current or something.
