For the lower FET, I'd think you would want to vary Vgs for minimal DC offset on each output terminal. How much you need will depend on the current you chose through the CCS.
The McMillan resistors are there to reduce (not eliminate) absolute offset. It will have no impact on relative offset. How much they are needed depends on how closely matched everything is, including resistors etc... Better matching yields reduced offset, both relative and absolute. I went to great pains to match everything, and it had real benefits. At turn on, without the McMillan resistors I get an absolute offset of 7 volts that quickly drops below 1 volts after a few minutes, and stabilizes less than 0.1V afte 30 minutes. With the 10k McMillan resistors the absolute offset starts at 3-4 volts, drops below 0.7 volts within a minute, and stabilizes at less tha 0.1V after 30 minutes. With 10k resistors the audible difference is pretty negligible, if you really listen carefully you can make out a slight difference. In the end I took them out and run without the, as I have my mosfets heavily heatsinked and don't have to worry about thermally overloading them during warm up.
metalman
A 10K value seems to too high to create the offset stabilization desired. If you don't mind the increased dissipation a value <100R should do better to stabilize this amp. It does look like your amp doesn't need it but did you ever try lower resistance there?
A 10K value seems to too high to create the offset stabilization desired. If you don't mind the increased dissipation a value <100R should do better to stabilize this amp. It does look like your amp doesn't need it but did you ever try lower resistance there?
Hi,
the lower the resistance the lower the sound quality. I tried 33k without problems but.....
Matching everything does only help for one value of the mains voltage. If this varies by a few volts during the day your absolute offset will vary too.
Since mains voltage seem to vary a lot where I live I tried a DC-Servo which solves the problem.
William
the lower the resistance the lower the sound quality. I tried 33k without problems but.....
Matching everything does only help for one value of the mains voltage. If this varies by a few volts during the day your absolute offset will vary too.
Since mains voltage seem to vary a lot where I live I tried a DC-Servo which solves the problem.
William
Pass DIY Addict
Joined 2000
Paid Member
The nice part about the Aleph-X amps is that you can customize their output power to suit your requirements. Have a look at my Aleph-X page (linked in my signature). It contains a protracted start-to-finish resource for building this amp.
hi, I looked at your page .. great work, congratulations. X the PSU, I already own 16 capacitors 47000uf 63v and 2.2 mh inductors on an aleph 2, if I'm not mistaken I can reuse these without having to test on Aleph X, without having to buy those 1.8 mh, right? , already in my aleph 2 I have replaced the rectification bridges with rurg3060, I think it's okay, right? one last thing; in your aleph x have you had problems with DC output on the speakers? thank you🙂
Pass DIY Addict
Joined 2000
Paid Member
Sounds like you have a great power supply already! The Rurg3060 looks good with ratings like 30A and 600v. I just used generic rectifier bridges with similar ratings and didn't worry too much about it.
The Aleph-X amps have an interesting characteristic with DC offset. Because this amp is balanced, you need to pay attention to both absolute DC offset (as measured from speaker terminal to chassis ground) and relative DC offset (measured across the speaker terminals).
My absolute DC offset has always been rock steady at zero volts. This is what is most important - it is what the speaker sees.
Relative DC offset is what you need to play with as it is affected by temperature, construction technique, choice of value for R46/R47. Everyone needed to play around with this set of variables in order to achieve lowest relative DC offset behavior in their build. Relative DC offset starts off high when the amp is first powered up, then falls as the amp warms up. A high DC offset reduces the available power output while the amp warms up. You usually adjust this so it is at zero after the amp is fully warmed up. Then, people tweak trying to get the relative offset lower when the amp is cold, or try to get it to fall to zero faster as the amp warms up.
The Aleph-X amps have an interesting characteristic with DC offset. Because this amp is balanced, you need to pay attention to both absolute DC offset (as measured from speaker terminal to chassis ground) and relative DC offset (measured across the speaker terminals).
My absolute DC offset has always been rock steady at zero volts. This is what is most important - it is what the speaker sees.
Relative DC offset is what you need to play with as it is affected by temperature, construction technique, choice of value for R46/R47. Everyone needed to play around with this set of variables in order to achieve lowest relative DC offset behavior in their build. Relative DC offset starts off high when the amp is first powered up, then falls as the amp warms up. A high DC offset reduces the available power output while the amp warms up. You usually adjust this so it is at zero after the amp is fully warmed up. Then, people tweak trying to get the relative offset lower when the amp is cold, or try to get it to fall to zero faster as the amp warms up.
Last edited:
Hi Eric,
I suppose you meant to say relative offset was rock steady? The absolutel offset is what is changing with temperature and a few other things including mains voltage.
Cheers,
William
I suppose you meant to say relative offset was rock steady? The absolutel offset is what is changing with temperature and a few other things including mains voltage.
Cheers,
William
Pass DIY Addict
Joined 2000
Paid Member
D'oh! 🙁 You are correct, William! Still use these amps several times a week - just don't think about them that much anymore...
Here is the corrected version of above:
My relative DC offset has always been rock steady at zero volts. This is what is most important - it is what the speaker sees.
Absolute DC offset is what you need to play with as it is affected by temperature, construction technique, choice of value for R46/R47. Everyone needed to play around with this set of variables in order to achieve lowest absolute DC offset behavior in their build. Absolute DC offset starts off high when the amp is first powered up, then falls as the amp warms up. A high DC offset reduces the available power output while the amp warms up. You usually adjust this so it is at zero after the amp is fully warmed up (1-2 hrs). Then, people tweak trying to get the absolute offset lower when the amp is cold, or try to get it to fall to zero faster as the amp warms up.
Here is the corrected version of above:
My relative DC offset has always been rock steady at zero volts. This is what is most important - it is what the speaker sees.
Absolute DC offset is what you need to play with as it is affected by temperature, construction technique, choice of value for R46/R47. Everyone needed to play around with this set of variables in order to achieve lowest absolute DC offset behavior in their build. Absolute DC offset starts off high when the amp is first powered up, then falls as the amp warms up. A high DC offset reduces the available power output while the amp warms up. You usually adjust this so it is at zero after the amp is fully warmed up (1-2 hrs). Then, people tweak trying to get the absolute offset lower when the amp is cold, or try to get it to fall to zero faster as the amp warms up.
on your page x Aleph X you talked about the losses of v due to loads of resistances, chap. , etc. ..so, with my power supply I have to use a 2x18v transformer like in the diagram or maybe a 2x19v or 2x20v. what can you advise me ?? thank you
Hi all
I need some advice
I have build Aleph X 60W version which I use to drive midbas (scan speak 18W8545) in active system
and have Aleph mini to drive tweeter
now I'm thinking maybe to replace Aleph mini with low power Aleph X as I have two unused PCB.
The thing is that I would like to use power supply from Aleph mini which is 2 x 15V and 150VA
by AXE 1-2 spreadsheet I would get like 10-12W on 4ohm with 18V DC and 2,5A bias.
Would that do for driving scanspeak 2904/7000
or should I forget about it and stay on Aleph Mini ?
I need some advice
I have build Aleph X 60W version which I use to drive midbas (scan speak 18W8545) in active system
and have Aleph mini to drive tweeter
now I'm thinking maybe to replace Aleph mini with low power Aleph X as I have two unused PCB.
The thing is that I would like to use power supply from Aleph mini which is 2 x 15V and 150VA
by AXE 1-2 spreadsheet I would get like 10-12W on 4ohm with 18V DC and 2,5A bias.
Would that do for driving scanspeak 2904/7000
or should I forget about it and stay on Aleph Mini ?
no way Jose .... that you need so much power
you can't go bellow 15V rails ........ and there is only way to make lower power Aleph X if you go to lower Iq
in that case you can use even IRF510 in output ....... that thing will be fast as lightning
you can't go bellow 15V rails ........ and there is only way to make lower power Aleph X if you go to lower Iq
in that case you can use even IRF510 in output ....... that thing will be fast as lightning
You lost me thare Zen Mod
I have 15V which would give me like 18V
I have to go with low Iq as my toroid is 150VA
And I have IRFP044
would that do the trick or no ?
😉
thanks
I have 15V which would give me like 18V
I have to go with low Iq as my toroid is 150VA
And I have IRFP044
would that do the trick or no ?
😉
thanks
rails are good
however , if you need small power (and I'm not believing that you need too much of it for tweeter) , better to use smaller output devices , which are happy with smaller Iq and they have smaller capacitances
IRFP044 is , so , too big part in that context
so , no need to make calculus of max power via voltage , just decide how much of Iq you need for intended power
10W or so is a killer
mentioned 2A5 Iq is way too much for that
however , if you need small power (and I'm not believing that you need too much of it for tweeter) , better to use smaller output devices , which are happy with smaller Iq and they have smaller capacitances
IRFP044 is , so , too big part in that context
so , no need to make calculus of max power via voltage , just decide how much of Iq you need for intended power
10W or so is a killer
mentioned 2A5 Iq is way too much for that
thanks ZenMode
Just one more question, although I belive you already state your preferences 😉
As I already have matched IRFP044
would there be any sound difference betwin 044 and 510 ?
As you state that 044 would be over kill at one point compered to 510 (and 510 have lower capacitance 😉)
but as I have tham, I would like to use them but that is not essential 😉
So, sound wise, would 510 sound better than 044 in this application ?
(if so, I will try to found some... 😉)
And if so, how much Iq should I place for 510 ? I don't think that thay can go like 10W each ?
Just one more question, although I belive you already state your preferences 😉
As I already have matched IRFP044
would there be any sound difference betwin 044 and 510 ?
As you state that 044 would be over kill at one point compered to 510 (and 510 have lower capacitance 😉)
but as I have tham, I would like to use them but that is not essential 😉
So, sound wise, would 510 sound better than 044 in this application ?
(if so, I will try to found some... 😉)
And if so, how much Iq should I place for 510 ? I don't think that thay can go like 10W each ?
Pass DIY Addict
Joined 2000
Paid Member
One of my Aleph-X amps runs on a 750va transformer with dual 18v secondaries. The left and right amps run a 1000va transformer that provides 19v. There is very little difference between them in terms of overall rail voltage and how this translates into max output power. Personally, I wouldn't worry about 1 volt here or there. The bigger item to watch is the temp of your output devices and the size of your sinks. If your sinks are adequate in size, a little voltage here or there and changes in bias don't make much of a difference. If your sinks are running hot (bit too small), you might want to consider CRC with 0R1 or 0R2 or so to drop little rail voltage or decrease the bias setting with the pots. All of this need to balance...
Pass DIY Addict
Joined 2000
Paid Member
I don't have any experience with the 510 devices, but I remember some discussion of IRFP244 vs IRFP044 years ago in this big thread and the consensus was that the higher gate capacitance of the 044 led to a bit of treble roll-off compared to the 244. Thus, a bit "darker" nor "not as sparkly" were typical reactions.
The current handling capability of the 044 is rather high, so they can take just about anything you can throw at them with this amp.
Thanks Zen...for you advices...I have to stop and think about everything 😉
Will get back if I decide to procide with this build 😉