Edit: My thread title was supposed to be a question not a statement, but I can't edit it accordingly.
Hi All,
Firstly, let me apologise if what I am about to say is completely incorrect. I am very new to this and know nothing
Always keen to be set straight.
I have been informed that modern IRF9610s are not as linear as the older versions due to differing manufacturing techniques. Whether this is significant or not, I have also been thinking that perhaps the 9610 is over rated as an input fet for the Aleph amplifiers, and that a smaller signal fet may be more linear in this application.
It seems to me that using any fet that is significantly overrated for a task means also dealing with more gate capacitance which is a bad thing. NP pushes the running of fets at high currents ideology, and it seems that the 9610s are also not being pushed in this application. That said, I know NP has his reasons for everything, and I am sure there is a good reason why he chose the 9610 that I am not aware of.
Anyway, I have been looking for a direct drop in improvement that will not require any resistor/bias changes. This rules out any JFETs.
I have stumbled across the ZVP3310, which is cheap, readily available and has less than half the input capacitance of the 9610 (50 pf vs 170pf). It can easily be dropped into existing boards by simply twisting the pins around and should require no resistor changes. NP also used it in ZV4, so I guess it can't be too bad. Edit 2: I also see in the ZV4 article that NP actually tried several fets and found the 3310 to be superior to the 9610 in that application, which is promising
Any thoughts on the suitability of the ZVP3310 to this task, and how it would perform both objectively and subjectively?
Your advice is appreciated!
Greg.
Hi All,
Firstly, let me apologise if what I am about to say is completely incorrect. I am very new to this and know nothing
Always keen to be set straight.I have been informed that modern IRF9610s are not as linear as the older versions due to differing manufacturing techniques. Whether this is significant or not, I have also been thinking that perhaps the 9610 is over rated as an input fet for the Aleph amplifiers, and that a smaller signal fet may be more linear in this application.
It seems to me that using any fet that is significantly overrated for a task means also dealing with more gate capacitance which is a bad thing. NP pushes the running of fets at high currents ideology, and it seems that the 9610s are also not being pushed in this application. That said, I know NP has his reasons for everything, and I am sure there is a good reason why he chose the 9610 that I am not aware of.
Anyway, I have been looking for a direct drop in improvement that will not require any resistor/bias changes. This rules out any JFETs.
I have stumbled across the ZVP3310, which is cheap, readily available and has less than half the input capacitance of the 9610 (50 pf vs 170pf). It can easily be dropped into existing boards by simply twisting the pins around and should require no resistor changes. NP also used it in ZV4, so I guess it can't be too bad. Edit 2: I also see in the ZV4 article that NP actually tried several fets and found the 3310 to be superior to the 9610 in that application, which is promising
Any thoughts on the suitability of the ZVP3310 to this task, and how it would perform both objectively and subjectively?
Your advice is appreciated!
Greg.
Last edited:
I vaguely remember an Aleph shootout posted a couple years back where an amp built with ZVP3310s was preferred over the irf9610 version. So, yes, it is a suitable alternative in this application.
here's the thread http://www.diyaudio.com/forums/pass-labs/75685-pass-shootout-salt-lake-city-part-2-a.html
here's the thread http://www.diyaudio.com/forums/pass-labs/75685-pass-shootout-salt-lake-city-part-2-a.html
Last edited:
For an Aleph input pair you don't need the TO-220 dissipation, as the dissipation is only ~10mA x Rail voltage. You should keep an eye on the CCS dissipation, though. ~20 mA x (rail - 9)
Mouser carries them, too if you want to avoid the minimum order charge.
Use insulation stripped from hookup wire to prevent accidents with lead routing. Sort of forces you to keep them long - and allows you to glue the pair together to help with thermal drift.
Mouser carries them, too if you want to avoid the minimum order charge.
Use insulation stripped from hookup wire to prevent accidents with lead routing. Sort of forces you to keep them long - and allows you to glue the pair together to help with thermal drift.
Oh dear, I didn't notice the digikey minimum order was for a bulk quantity, only that the number was one!
So I assume that with my aleph 30 and +/-25v rails it will be fine for the ccs?
Great idea with the insulation.
Optimum operating point?
Hi All,
Is anyone able to offer some advice regarding the optimum operating point for this device in an Aleph input?
Bill F. and I have been discussing the possibility that it might be better to drop the current slightly to bring the operating point more in line with the Zen V4?
Any thoughts? I still have trouble with transistor operating curves.
Cheers,
Greg.
Hi All,
Is anyone able to offer some advice regarding the optimum operating point for this device in an Aleph input?
Bill F. and I have been discussing the possibility that it might be better to drop the current slightly to bring the operating point more in line with the Zen V4?
Any thoughts? I still have trouble with transistor operating curves.
Cheers,
Greg.
Last edited:
As I am sure Bill mentioned, besides hitting the best transfer function for the ZVPs, the tradeoff is being conservative in the ZVP power dissipation vs. having enough current to drive the output devices' gate capacitance. Since you're not driving huge arrays of IRFP240s, you should be able to drop the current without ill effects. The AJ runs a little under 5 mA on the inputs with no apparent ill effects.
I'd go with Bill's advice.
I'd go with Bill's advice.
Hello all! concerning the A 30, I built a diy clone a while back and have been seeking to get the best sound out of it. This is with the P. Daniel boards. What I have found is that the cap at "C1-close to input side of pcb is a place that different caps can be tried to get the sound you most like. I use 50v 100-220 mfd. I don't know but I prefer the higher voltage rated caps. Also the input wire, I use some anaconda HF wire, it is as thick as human hair, probably the same diameter as wire used in a phono cartridge.
I ask this question because search has not found: what is the ideal dc offset on the A30? In my amp one channel is 0 mv-not steady, it hunts up and down a little and the other channel -37 mv, what is the effect of this negative dc offset and how do you fix it?
I there is an article about this, I would like to read it. I really would like to tune this amp up.
I there is an article about this, I would like to read it. I really would like to tune this amp up.
Ideal offset is zero, but anything under 50-100mV is good (depending on who you listen to). Slight tweaking of the input CCS is how you adjust this (assuming that you have a pot there) I'd leave well enough alone.
The offset is temperature sensitive - is the channel that drifts closed up? Someone walking past 5' away from an open amp has caused all sorts of havoc for me. Ditto for the heat kicking on. Adjust, put the top on and let it settle for 30 minutes or so before attempting to adjust out half of the offset at a time.
There used to be a wiki page for this, but I can't find it. It may have been lost in the move to new software.
The offset is temperature sensitive - is the channel that drifts closed up? Someone walking past 5' away from an open amp has caused all sorts of havoc for me. Ditto for the heat kicking on. Adjust, put the top on and let it settle for 30 minutes or so before attempting to adjust out half of the offset at a time.
There used to be a wiki page for this, but I can't find it. It may have been lost in the move to new software.
Last edited:
ZVP3306A would be OK for you as the LTP devices never see more than 2 or 3 volts in operation. Farnell and CPC have them in E-Line (basically TO-92) package.
Either ZVP3306 or ZVP3310 are adequate as far as working voltage. Look at the transfer curve though. The ZVP3310 curve is more linear at lower currents/ Vgs - down where we are going to run it. If you use the ZVP3306 you might want to run it at higher bias, and change the load resistor to get the ~4V needed by the output devices.
http://www.diodes.com/datasheets/ZVP3310A.pdf
http://www.diodes.com/datasheets/ZVP3306A.pdf
http://www.diodes.com/datasheets/ZVP3310A.pdf
http://www.diodes.com/datasheets/ZVP3306A.pdf
Hifinerd - did you ever get around actually using the 3310 as input fet in your Aleph30??
Did you use it for the CCS also, it did you stick with the 9610 here? (higher current draw and more heat in this fet?)
I just ordered a couple of matched 3310s so will be trying them out in my 30 very soon....
Best regards
Hans
Did you use it for the CCS also, it did you stick with the 9610 here? (higher current draw and more heat in this fet?)
I just ordered a couple of matched 3310s so will be trying them out in my 30 very soon....
Best regards
Hans
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.