Agreed
Ya, it's pretty amazing the accuracy too. I was looking at getting maybe an older HP 3585 or something like that, but unless I'm reading the specs wrong what we have already has way lower noise floor ability.
I would still like to get one for sanity check and because I think they are cool and look nice in the lab. lol I haven't quite got enough to look like a starship, but getting there. 😛
Ya, it's pretty amazing the accuracy too. I was looking at getting maybe an older HP 3585 or something like that, but unless I'm reading the specs wrong what we have already has way lower noise floor ability.
I would still like to get one for sanity check and because I think they are cool and look nice in the lab. lol I haven't quite got enough to look like a starship, but getting there. 😛
R124,126
R124-R126 -I tried to locate the post XRK mentioned these, I seem to recall he said these were to be carbon film and not carbon comp. I am still try to sort out my congested highs on the bench before casing things up.
Bill
R124-R126 -I tried to locate the post XRK mentioned these, I seem to recall he said these were to be carbon film and not carbon comp. I am still try to sort out my congested highs on the bench before casing things up.
Bill
R124, R126, R133, R135
R124 is not as critical and leave as metal thin film is ok
Do not use metal oxide thick film under any circumstance for R126
R124 is not as critical and leave as metal thin film is ok
Do not use metal oxide thick film under any circumstance for R126
Why don't you add a Lu1014 to the nirvana making it a LuAN?
Maybe it could be a winning combo!
Maybe it could be a winning combo!
The LU device is no longer produced, although you can buy them from Ebay sellers but I'm guessing not for long. There is no p channel as far as I know, and the n type has a max dissipation of 57W. In the AN the mosfets are run at 45 watts, and this is a bridge too far for the LU!
HD
HD
The LuFo makes the same power as the AN at half the rail voltage. However, the AN probably has better manners at max power.
my question was just academic, I knew the answer
great place to try LU1014D is F1J, F2, Aleph J and others with just n-channel
but new stuff should be fun too
do not forget, you can parallel the matched jfets
great place to try LU1014D is F1J, F2, Aleph J and others with just n-channel
but new stuff should be fun too
do not forget, you can parallel the matched jfets
Of course I meant cascoded LU for the top part of the output stage of AN and keeping the mu follower at the bottom part.
Just to avoid the LuFo MOT that has too high DCr here in 230v countries...
No problem if efficiency of AN is lower...
Just to avoid the LuFo MOT that has too high DCr here in 230v countries...
No problem if efficiency of AN is lower...
Try rewinding the MOT with 14ga magnet wire. The core is what is important and I think they are the same for 120vac or 240vac. Removing the old wire is easy. Hammer and chisel. If not using the secondaries, plenty of space to work with.
LuFo Amp - 39w SE Class A from 28v Rail
Cheap 'N Easy ZVS Build 2/5: Rewinding a microwave oven transformer for an induction heater - YouTube
LuFo Amp - 39w SE Class A from 28v Rail
Cheap 'N Easy ZVS Build 2/5: Rewinding a microwave oven transformer for an induction heater - YouTube
Last edited:
how many turns does it need?
If the same as the 120V primary it is nearly impossible to wind it passing the magnet wire in the holes
you need to open the core to insert the coil....
If the same as the 120V primary it is nearly impossible to wind it passing the magnet wire in the holes
you need to open the core to insert the coil....
There are formulae for inductance, dimensions and turns. But pivotal is the magnetic permeability, and you really don't know unless you know the metal of the laminations. And when you set the gap - typically 1-2 thou - the inductance takes a huge drop. You need that gap so the inductor cannot saturate at 3.2A, but it makes the winding LARGE and that drives up the DCR. We need around 0.5R, no more..... so THICK wire is mandated.
The best option is to build one, then measure. You need to figure out the gap too, and it should be set so that the flux is one half of the max value of the metals in the laminations, so that used in this amplifier it can give you the full negative swing at the specified current.
Inductors are essentially simple, but the actual engineering to build one to spec is fiddly. We have found that MOTs set for 110-120Vac mains give around 68mH at 3.2A. Building a custom inductor with very high quality materials can easily cost more than $US200, hence the use of a MOT which is $30 and far the best option for trimming costs. As it happens, audio is not high frequency, and MOTs work very, very well. If the 68mH specified is not high enough, the bass will suffer, nothing more........
HD
The best option is to build one, then measure. You need to figure out the gap too, and it should be set so that the flux is one half of the max value of the metals in the laminations, so that used in this amplifier it can give you the full negative swing at the specified current.
Inductors are essentially simple, but the actual engineering to build one to spec is fiddly. We have found that MOTs set for 110-120Vac mains give around 68mH at 3.2A. Building a custom inductor with very high quality materials can easily cost more than $US200, hence the use of a MOT which is $30 and far the best option for trimming costs. As it happens, audio is not high frequency, and MOTs work very, very well. If the 68mH specified is not high enough, the bass will suffer, nothing more........
HD
Last edited:
at the end of the day it will be easier to insert the "LU cascoded triode cell" on the top part of the AN amp rather than building the choke for the LUFO
would it make sense? and give a nice H2 H3 relation?
would it make sense? and give a nice H2 H3 relation?
Well, they are very different amps. One is a minimalist choke loaded SE Class A with a SIT device and AC coupled output. The other is a DC coupled Aleph based topology with built in front end.
Both have dominant second harmonic profile with descending higher orders. The LuFo is zero global feedback and 0dB gain.
Both have dominant second harmonic profile with descending higher orders. The LuFo is zero global feedback and 0dB gain.
Hello. I was very interested in the AN amplifier. I'm going to build it for myself. Before that, I built a lot of class AB amplifiers. Pure Class A never did. To begin with, there were two questions. My speaker system is not the best Dali Spector 2, with a resistance of 6 ohms, not very high sensitivity. Which option will suit me better 4 ohms or 8 ohms? And the second question. I plan to use AN with a pre-amplifier, also class A with an integrator to maintain zero at the output. In this case, can I exclude the input separation capacitor (10mkf) from the AN circuit? Thanks. Sincerely, Crazoff.
Last edited:
I would use the 8R schematic, and set a slightly higher current on the quiescent, say 15%.
If you have sufficient heatsink area you will have to dissipate about 110W per channel. This points you at a 0.2C/watt heatsink, sometimes people like to use four large heatsinks for a stereo AN.
It's a very good amplifier with very good sound quality. You will be very pleased but you should have at least 88+ db/watt speakers.
Are you in Moscow? You should talk to Valery, he is very au fait with the AN (and many others too!).
Hugh
If you have sufficient heatsink area you will have to dissipate about 110W per channel. This points you at a 0.2C/watt heatsink, sometimes people like to use four large heatsinks for a stereo AN.
It's a very good amplifier with very good sound quality. You will be very pleased but you should have at least 88+ db/watt speakers.
Are you in Moscow? You should talk to Valery, he is very au fait with the AN (and many others too!).
Hugh
Thank you for the quick answer, although the second question remains. Could you point me to Valery's post so I can contact him? At the moment I am on vacation and it is not very convenient to scroll through such a large volume on a smartphone. Thanks
Hi Hugh
I hope you are warm in this mid winter weather!
I also have some speakers which are “in between“, ~ 6 ohms
I understand the heatsinking would need to increase. Would that be pro rata, ie also by 15%? If not, in the same “indicative” metric, by what likely percentage?
I think(?) a higher quiescent is the same as biasing “a little more into class A”
Is that correct?
What effect would that have on the output?
Much appreciated