Hi Sippo,
it is important to know that there are two sorts of DC-offset in an X.
1. relative offset between + and - out
2. absolute offset between the outputs and ground.
The first is the offset seen by the speaker. Like I said this is below 20mV in both amps from startup and changes only marginal during warmup.
The second offset is not seen by the speaker, starts at 6V or so and slowly goes to zero. Having a high absolute offset reduces headroom but I wouldn´t know any other disadvantages.
I don´t use a relais, there´s no thump whatsoever an turnon (there is in my Aleph5) or turn off.
William
it is important to know that there are two sorts of DC-offset in an X.
1. relative offset between + and - out
2. absolute offset between the outputs and ground.
The first is the offset seen by the speaker. Like I said this is below 20mV in both amps from startup and changes only marginal during warmup.
The second offset is not seen by the speaker, starts at 6V or so and slowly goes to zero. Having a high absolute offset reduces headroom but I wouldn´t know any other disadvantages.
I don´t use a relais, there´s no thump whatsoever an turnon (there is in my Aleph5) or turn off.
William
more impressions
Hi,
after a lot more records and cd´s I can only repeat that it does sound very different.
Bass is very well defined and punchy, mids and highs are a lot better defined and have a somewhat cooler character. The amp is a bit less forgiving than my Aleph5 but on most records does sound a lot better (so it seems I don´t have to buy some new trannies after all)
William
Hi,
after a lot more records and cd´s I can only repeat that it does sound very different.
Bass is very well defined and punchy, mids and highs are a lot better defined and have a somewhat cooler character. The amp is a bit less forgiving than my Aleph5 but on most records does sound a lot better (so it seems I don´t have to buy some new trannies after all)
William
Hi,
X´s are still working. They even had a duration test after my wife left them on for two days
After listening to many many records and cd´s the difference between this and my aleph5 get´s bigger and bigger and I don´t want to go back anymore.
There are some problems though:
1. Mechanical hum during warmup wich is probably causes by the assymetrical loading of the transformer because of the absolute dc-offset. I will try some lower values for R46/47 like 2k7 or 2k2.
2. Mechanical hum when the neighbours use some nice electrical devices like a hairdryer. This is caused by connecting the amps directly to the mains instead of my balanced ac power transformer (wich get´s a bit too hot at 900 watts). I will order a second transformer and experiment a bit with some chokes in the primaries to get everything silent again.
I´ve attached a short excel file with the parts used in the amp and power supply. It is the original file with the values changed and some more comments. Part designators are unchanged.
William
X´s are still working. They even had a duration test after my wife left them on for two days
After listening to many many records and cd´s the difference between this and my aleph5 get´s bigger and bigger and I don´t want to go back anymore.
There are some problems though:
1. Mechanical hum during warmup wich is probably causes by the assymetrical loading of the transformer because of the absolute dc-offset. I will try some lower values for R46/47 like 2k7 or 2k2.
2. Mechanical hum when the neighbours use some nice electrical devices like a hairdryer. This is caused by connecting the amps directly to the mains instead of my balanced ac power transformer (wich get´s a bit too hot at 900 watts). I will order a second transformer and experiment a bit with some chokes in the primaries to get everything silent again.
I´ve attached a short excel file with the parts used in the amp and power supply. It is the original file with the values changed and some more comments. Part designators are unchanged.
William
Pass DIY Addict
Joined 2000
Paid Member
Lowering Bias
William,
Nice job on your amps! They look great! At some point in the future, I'll be building a set of these for my new theater and was thinking that I need to be able to deliver approx 100wpc into the speakers I'll be using. So my goal is to build something very close to what you have but with a lower bias point.
I was figuring on 22v rails and about 6A of bias per channel - using the AXE spreadsheet, this would provide about 100wpc into both 8 ohms and 4 ohms.
So, my question: To set the bias lower, it looks like I would just need to adjust V1 and V3, and maybe increase R6/R41 slightly. Is this correct?
Also, where did you find those great heatsinks?
Thanks!
Eric
William,
Nice job on your amps! They look great! At some point in the future, I'll be building a set of these for my new theater and was thinking that I need to be able to deliver approx 100wpc into the speakers I'll be using. So my goal is to build something very close to what you have but with a lower bias point.
I was figuring on 22v rails and about 6A of bias per channel - using the AXE spreadsheet, this would provide about 100wpc into both 8 ohms and 4 ohms.
So, my question: To set the bias lower, it looks like I would just need to adjust V1 and V3, and maybe increase R6/R41 slightly. Is this correct?
Also, where did you find those great heatsinks?
Thanks!
Eric
Pass DIY Addict
Joined 2000
Paid Member
Thanks for the quick reply!
There is a lot of great info in the Aleph-X wiki, but nothing about how to adjust the circuit to different voltage & bias levels. I noticed you changed many of the resistor values from Grey's original circuit, and I really have no idea how to determine what to change...
Unfortunately, I don't have any test equipment (beyond an LCR meter), so I am not sure how to adjust the ac-current gain and determine its effect on the output and bias level.
Are there any rough formulas that can be used, or is this strictly a "try it an measure" kind of thing?
For my purposes, I think I'll need about 100wpc into a 4 ohm load, possibly a little more. I'm not sure whether I should follow your schematic and just reduce the bias a little, or try to build a version that runs on 20v rails at 6 or 7 amps of bias. The PS caps I have are 25v 220,000uF and I was planning on building a choke input power supply using Hammond 156B chokes (1.5mH, 0.07 ohms, 5A).
There is a lot of great info in the Aleph-X wiki, but nothing about how to adjust the circuit to different voltage & bias levels. I noticed you changed many of the resistor values from Grey's original circuit, and I really have no idea how to determine what to change...
Unfortunately, I don't have any test equipment (beyond an LCR meter), so I am not sure how to adjust the ac-current gain and determine its effect on the output and bias level.
Are there any rough formulas that can be used, or is this strictly a "try it an measure" kind of thing?
For my purposes, I think I'll need about 100wpc into a 4 ohm load, possibly a little more. I'm not sure whether I should follow your schematic and just reduce the bias a little, or try to build a version that runs on 20v rails at 6 or 7 amps of bias. The PS caps I have are 25v 220,000uF and I was planning on building a choke input power supply using Hammond 156B chokes (1.5mH, 0.07 ohms, 5A).
Hi...
Eric I think this version have 22V and 6A
http://web.vip.hr/pcb-design.vip/alephx.html
regards
Josip
Eric I think this version have 22V and 6A
http://web.vip.hr/pcb-design.vip/alephx.html
regards
Josip
Pass DIY Addict
Joined 2000
Paid Member
HI,
I also planned for a 20V supply but somehow the losses were not as high as calculated (18V transformers, schottkys and 2,2mH Torobars)
The values should work well with a 20V supply. AC-current-gain can be measured with a multimeter and a cd player with test-disc or a funktion generator (as described many times on the forum)
William
I also planned for a 20V supply but somehow the losses were not as high as calculated (18V transformers, schottkys and 2,2mH Torobars)
The values should work well with a 20V supply. AC-current-gain can be measured with a multimeter and a cd player with test-disc or a funktion generator (as described many times on the forum)
William
Pass DIY Addict
Joined 2000
Paid Member
Thanks for the help William and Yoke!
So then, it looks like I can follow William's parts list, use 20-22v rails, adjust V1/3 for bias, and then adjust ac current gain accordingly...
This is making me feel better, thanks! Also, it looks like I might be able to source some Seifert KL271 extrusions locally...
Thanks!
So then, it looks like I can follow William's parts list, use 20-22v rails, adjust V1/3 for bias, and then adjust ac current gain accordingly...
This is making me feel better, thanks! Also, it looks like I might be able to source some Seifert KL271 extrusions locally...
Thanks!
Hi all,
after discussing the zener in the curent source that feeds the input diff pair I had a look at some data sheets and did some calculating:
a Zener 9.1V with a TK of 0,07% / K will change the current through the diff pair when heating up from 20 to 65 or 70°C by a bit over 4% (going from 9.1 to 9.4V).
Seeing the changes you can make with relativ small changes of VR2 this seems a bit too much.
I did some measurements this afternoon measuring the zener from cold (9,08V to warm9,19V) without the lid on the amp. Cooling the zener (earbud with ice spray) changed the absolute offset nicely, heating it a bit with a solder iron too.
I also tried cooling the IRF9610 with not much effect. Also cooling the diff pair wasn´t that effective.
So now I´ve ordered a pair of 1N825A (6,2V, TK 0,002% / K) and will try these this week. I´ll only have to change R17 (for 7mA through D1) and R24 (220R) and R26 (120R).
The other thing I noticed is that a 10000uF/16V cap is not enough for a dc-filter (230V / 340Watts).
I measured 0,66V AC over it wich means that the diodes will be working a bit too. Changed it to 20000uF and got 0,23V AC so that seems to be fixed too
I´ll let you know if the 1N825A changes the absolute offset behaviour.
William
after discussing the zener in the curent source that feeds the input diff pair I had a look at some data sheets and did some calculating:
a Zener 9.1V with a TK of 0,07% / K will change the current through the diff pair when heating up from 20 to 65 or 70°C by a bit over 4% (going from 9.1 to 9.4V).
Seeing the changes you can make with relativ small changes of VR2 this seems a bit too much.
I did some measurements this afternoon measuring the zener from cold (9,08V to warm9,19V) without the lid on the amp. Cooling the zener (earbud with ice spray) changed the absolute offset nicely, heating it a bit with a solder iron too.
I also tried cooling the IRF9610 with not much effect. Also cooling the diff pair wasn´t that effective.
So now I´ve ordered a pair of 1N825A (6,2V, TK 0,002% / K) and will try these this week. I´ll only have to change R17 (for 7mA through D1) and R24 (220R) and R26 (120R).
The other thing I noticed is that a 10000uF/16V cap is not enough for a dc-filter (230V / 340Watts).
I measured 0,66V AC over it wich means that the diodes will be working a bit too. Changed it to 20000uF and got 0,23V AC so that seems to be fixed too
I´ll let you know if the 1N825A changes the absolute offset behaviour.
William
Hi Hugo,
yes that´s the one. I used 10000uF per channel and never checked if this was enough
I still have some mechanical hum though I´m not shure if it´s the transformer or the torobars.
It was allright until I placed a sheet of mu-metal between transformer/chokes and caps and maybe this is vibrating....
William
yes that´s the one. I used 10000uF per channel and never checked if this was enough
I still have some mechanical hum though I´m not shure if it´s the transformer or the torobars.
It was allright until I placed a sheet of mu-metal between transformer/chokes and caps and maybe this is vibrating....
William
Hi,
just an update. After replacing the cheap 9,1V Zeners with an 1N825A absolute offset behaviour has changed although not really that much. One channel stayed almost the same, now starting at +5V and going to near zero in an hour. The other channel went from -8V to +5V also going to ca. 0V in an hour.
maybe finding a zener with the right temp. coefficient would be the better sollution but for now I will leave it as it is.
The amps have run in quite nicely now and are really sounding very nice. What´s not so nice is that the transformers are also running in and have developed a nasty buzz
You can hear it from 3meters away when everything else is quiet and it is driving me slightly mad
Until now I tried the following thing to get them quiet again:
1. put some DC filters directly in the amps (20.000uF/16V and two diodes). This reduced it a bit but not enough.
2. put a 0,22Ohm resistor in the mains (230V). A small reduction was noticable (but not enough)
3. put 3 0,22Ohm/5watt resistors in parallel between bridge and caps to reduce the current peak from a theoretical 38A to 30A. I´m not shure this worked cause the resistors started smoking after less than half a minute. I will try again with some 25W / 0,12Ohm resistors I ordered yesterday. This will also reduce the voltage at the fets from 22 to around 20V so I will have to listen if it has an influence on the sound through the speakers.
4. decoupled the transformer from the frame with rubber. No influence
Both amps are connected to a balanced ac supply (115V-0V-115V). When I connect them directly to the mains the buzz gets really loud...
I already tried putting them on different mains phases (I´m not shure this is english) but without effect.
After reading a nice article by Menno van der Veen on the Plitron website about measuring transformer noise and the causes thereof I hope I won´t have to throw away my transformers.
So what I will do next is to try the bigger resistors for a R-C-L-C supply and I will remove the mu-metall sheet (the buzz started after putting it in and although I´m almost shure the noise comes from the transformers I will just try)
Any suggestions are more than welcome!!!!
William
just an update. After replacing the cheap 9,1V Zeners with an 1N825A absolute offset behaviour has changed although not really that much. One channel stayed almost the same, now starting at +5V and going to near zero in an hour. The other channel went from -8V to +5V also going to ca. 0V in an hour.
maybe finding a zener with the right temp. coefficient would be the better sollution but for now I will leave it as it is.
The amps have run in quite nicely now and are really sounding very nice. What´s not so nice is that the transformers are also running in and have developed a nasty buzz
You can hear it from 3meters away when everything else is quiet and it is driving me slightly mad
Until now I tried the following thing to get them quiet again:
1. put some DC filters directly in the amps (20.000uF/16V and two diodes). This reduced it a bit but not enough.
2. put a 0,22Ohm resistor in the mains (230V). A small reduction was noticable (but not enough)
3. put 3 0,22Ohm/5watt resistors in parallel between bridge and caps to reduce the current peak from a theoretical 38A to 30A. I´m not shure this worked cause the resistors started smoking after less than half a minute. I will try again with some 25W / 0,12Ohm resistors I ordered yesterday. This will also reduce the voltage at the fets from 22 to around 20V so I will have to listen if it has an influence on the sound through the speakers.
4. decoupled the transformer from the frame with rubber. No influence
Both amps are connected to a balanced ac supply (115V-0V-115V). When I connect them directly to the mains the buzz gets really loud...
I already tried putting them on different mains phases (I´m not shure this is english) but without effect.
After reading a nice article by Menno van der Veen on the Plitron website about measuring transformer noise and the causes thereof I hope I won´t have to throw away my transformers.
So what I will do next is to try the bigger resistors for a R-C-L-C supply and I will remove the mu-metall sheet (the buzz started after putting it in and although I´m almost shure the noise comes from the transformers I will just try)
Any suggestions are more than welcome!!!!
William
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.