pcb width: 90mm (3540 mils)
pcb current length: 175mm (6890mils) PS section will be added to the end so its overall length will be somewhat longer. you will though be able to separate the PS section if desired; (I myself will be doing dual mono and maybe keep them together if the case can package the final length).
pcb current length: 175mm (6890mils) PS section will be added to the end so its overall length will be somewhat longer. you will though be able to separate the PS section if desired; (I myself will be doing dual mono and maybe keep them together if the case can package the final length).
Hi folks
I can match FETs for this project. I should be able to do them right after the new year. I am retiring at that point and can dedicate time to the job. According to Koja that should dovetail with the arrival of the boards. I have what I need to set up for matching large quantities. A ZIF socket, meters, regulated supplies etc. I can do this for cost + shipping. The cost is a little hard to nail down due to not knowing the yield and the number of sets, but my guess would be somewhere in the 20$ area for a stereo set of 12 FETs + shipping. If you are interested post here please and we'll see how much interest there is.
I can match FETs for this project. I should be able to do them right after the new year. I am retiring at that point and can dedicate time to the job. According to Koja that should dovetail with the arrival of the boards. I have what I need to set up for matching large quantities. A ZIF socket, meters, regulated supplies etc. I can do this for cost + shipping. The cost is a little hard to nail down due to not knowing the yield and the number of sets, but my guess would be somewhere in the 20$ area for a stereo set of 12 FETs + shipping. If you are interested post here please and we'll see how much interest there is.
Halo,
It has been a nice preamp but quite some years ago. I do not see a reason not to go to later J-fet preamps that should be better sonically. Actually as I have built them all in the past I preferred Aleph P 1.0 better with the IRFP240 devices. I also preferred it better (actually both) powered from batteries. And for the last I do not see any reason for using those Wima caps in signal path. The signal path caps make so much diffrance that using the Wimas is just giving up a whole lot out of what can be achived. There are so many other much better choices
Cheers
It has been a nice preamp but quite some years ago. I do not see a reason not to go to later J-fet preamps that should be better sonically. Actually as I have built them all in the past I preferred Aleph P 1.0 better with the IRFP240 devices. I also preferred it better (actually both) powered from batteries. And for the last I do not see any reason for using those Wima caps in signal path. The signal path caps make so much diffrance that using the Wimas is just giving up a whole lot out of what can be achived. There are so many other much better choices
Cheers
Brian: I am also interested in matched fets for 4 channels please; (I may also just ship you the fets I have since otherwise I will not need them).
-----------
Zygi: I remember your comments about AlephP way back, (I think I got the service manual for v.1.0 from you
). I remember you saying P was a great improvement over L for you, and then 1.7 not that much over 1.0. I think Papa bumped up the gain with v.1.7 only because he had F4 coming up at that time and wanted folks to have more choices. AlephP is still unique for unb-to-bal and back conversion and also for its easy gain adjustment (single point pot with no significant bearing on Dc offset). i am also trying to save those SK/SJ fets (for example I have the pcb for Juma's LSK-pre but I am still building the Aleph preamps).
-----------
Zygi: I remember your comments about AlephP way back, (I think I got the service manual for v.1.0 from you
). I remember you saying P was a great improvement over L for you, and then 1.7 not that much over 1.0. I think Papa bumped up the gain with v.1.7 only because he had F4 coming up at that time and wanted folks to have more choices. AlephP is still unique for unb-to-bal and back conversion and also for its easy gain adjustment (single point pot with no significant bearing on Dc offset). i am also trying to save those SK/SJ fets (for example I have the pcb for Juma's LSK-pre but I am still building the Aleph preamps).btw interesting that Zygibajt made a comment about improved SQ with a quiet PS (battery). if anyone has an idea for an improved regulator at 60V we could give it a thought I guess. an input from a guru like Salas might be called for? (otherwise, I started this weekend adding the original PS section to the pcb).
fyi I asked Salas if his highly acclaimed LVregulator could be taken as high as 60V and this is what he answered:
"The bib (sslv1.1) can scale up to 60 Vo spec if Q102 (202, 302) will be exchanged for a resistor. VDS MAX of that JFET is the limiting factor in the CCS voltage rating. Rsub=(Vin-5V)/3mA. A 24K 1W resistor will be alright for 70V rectified and 60V regulated output. Q104 (204, 304) should be 65V BJTs in this case. Like BC546-556."
so there you have it: if you want to do better than the "stock" regulator, you can always break those off and do sslv instead. (I actually have those pcbs from TeaBag's gb and have been saving them for something special; but already have xformers for 80Vrectified so would first have to throw some voltage away I guess).
"The bib (sslv1.1) can scale up to 60 Vo spec if Q102 (202, 302) will be exchanged for a resistor. VDS MAX of that JFET is the limiting factor in the CCS voltage rating. Rsub=(Vin-5V)/3mA. A 24K 1W resistor will be alright for 70V rectified and 60V regulated output. Q104 (204, 304) should be 65V BJTs in this case. Like BC546-556."
so there you have it: if you want to do better than the "stock" regulator, you can always break those off and do sslv instead. (I actually have those pcbs from TeaBag's gb and have been saving them for something special
; but already have xformers for 80Vrectified so would first have to throw some voltage away I guess).
Last edited:
I have been talking to the more knowledgeable about the PS: apparently papa's simple reg could possibly be improved by using the less noisy "buried" zeners (10 in 6.3V value instead of the plain vanila 9 in 9.1V value), but the larger package and pcb real estate are not something i am looking for). with charging caps to 80V rectified and then taking voltage way down at 64V (before loosing another 4V at the regulator fet), it makes you wonder how much ripple is even picked up by the following circuit (so even that CRC Papa put in there may be doing very little).
I am making two AlephPs and I have the Salas LV reg pcb; so with his help I may be able to use it in one of the APs, just for the heck of it to see if any improvement over Papa's simple approach is subjectively appreciable (I doubt with my speakers since I am not so much into the exotics, but then DIY is also about sleeping well because it measures better, right?).
-------------
mrIT: I got your line: with the rectifier on the same board I will carve out the PwrGnd from SigGnd and force the current down from cap to cap as it progressively gets cleaner and then connect it to SignalGND at one ref point where the pre load is attached. they are schematicaly all marked as analog ground to avoid error messages in the pcb software. satisfied?
I am making two AlephPs and I have the Salas LV reg pcb; so with his help I may be able to use it in one of the APs, just for the heck of it to see if any improvement over Papa's simple approach is subjectively appreciable (I doubt with my speakers since I am not so much into the exotics, but then DIY is also about sleeping well because it measures better, right?
).-------------
mrIT: I got your line: with the rectifier on the same board I will carve out the PwrGnd from SigGnd and force the current down from cap to cap as it progressively gets cleaner and then connect it to SignalGND at one ref point where the pre load is attached. they are schematicaly all marked as analog ground to avoid error messages in the pcb software. satisfied?
Last edited:
the plan is to finish the pcb layout by the end of month, publish it here, wait a week and then place the order.
so please check back for updates. (especially (!!) if you still think you may want to cancel your interest, for whatever reason).
I just took a preliminary count here and it looks like we may be good for about 50 or so pieces of the board; (reminder: 1 pcb = 1 PS (60Vreg) + 1 channel).
I will know the final price only once I have uploaded the files on the supplier's website, but my expectation for the cost to you is about US$10/per piece plus the cost of my mail to you.
current count:
Koja (4 pieces)
Bob Ellis (??)
yoke (6)
zashoomin (4)
gwrskien (6) or 8?
Mr.dB (2)
Ginum (2)
dudaindc (2)
bk856er (2) ?
nopants (2)
Macfixer (2)
Bfpca (2)
skunark (4)
tome (2)
DaveM (2)
Myint67 (4)
jdkJake (4)
so please check back for updates. (especially (!!) if you still think you may want to cancel your interest, for whatever reason).
I just took a preliminary count here and it looks like we may be good for about 50 or so pieces of the board; (reminder: 1 pcb = 1 PS (60Vreg) + 1 channel).
I will know the final price only once I have uploaded the files on the supplier's website, but my expectation for the cost to you is about US$10/per piece plus the cost of my mail to you.
current count:
Koja (4 pieces)
Bob Ellis (??)
yoke (6)
zashoomin (4)
gwrskien (6) or 8?
Mr.dB (2)
Ginum (2)
dudaindc (2)
bk856er (2) ?
nopants (2)
Macfixer (2)
Bfpca (2)
skunark (4)
tome (2)
DaveM (2)
Myint67 (4)
jdkJake (4)
My best guess would be 20$ max. for a set of 12 fets. Shipping extra of course. A lot depends on how many I order and the yield. That will depend on how close we want them matched. The more mosfets I test the closer the match will be. I will start getting my test setup put together soon. I will have a temp. controlled heatsink to mount the mosfets. I don't think these will run that hot with even a small heatsink but I will test at whatever temp. the group decides on. I'd suggest somewhere in the 30 to 40C range.
I don't see a reason to match the mosfets channel to channel. Loose matching of each quad of current sources/sinks makes some sense. (Does a bit of DC offset matter in a capacitor coupled circuit? Note that in the X-amps matching each quadrant is nice to have, but not necessary) Primarily the gain pair should be matched as tightly as possible.
Reducing the number of devices to be matched together will greatly increase the yield. Ordering in full tubes does as well, since the devices in each tube are likely to be from the same wafer. I easily got four matched sextets and several triples out of two tubes of 25 IRFP240s. A $35 tube of 50 IRF610s should yield enough matched sets for this and many more projects if you want to match them yourself.
Reducing the number of devices to be matched together will greatly increase the yield. Ordering in full tubes does as well, since the devices in each tube are likely to be from the same wafer. I easily got four matched sextets and several triples out of two tubes of 25 IRFP240s. A $35 tube of 50 IRF610s should yield enough matched sets for this and many more projects if you want to match them yourself.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.