Your soldering is much better than mine, that’s for sure.
I won’t get to implement the new buffering scheme until tomorrow.
I won’t get to implement the new buffering scheme until tomorrow.
Well you get some good practice when you're paid to fix electronics for.... 33 years now, and a few years of messing with car stereos as a teenager before that. The stuff at work is very challenging at times compared to my hobby stuff. It's IPC Class 3 railroad equipment that always has a thick conformal coating on the PCBs. And board damage is not an option when doing the rework. Much (but not all) of it is legacy obsolete / retired product with no replacement PCBs available.
metaphile - try HS251-ND at DigiKey
metaphile - try HS251-ND at DigiKey
. Any problems?
A little more elbow room for the big WIMAs would have been nice. You can see in my picture there how they are "out of bounds" and trying to shove some resistors around. And the lead spacing felt just a little off, meaning they didn't "fall right in", it took a little encouragement... but they're in there. I have a WIMA MKS4 1uF at C1 (just using what I had in the stash), and of course the lead spacing is different so I needed to make a couple of new holes to get those installed. The option holes would have been nice there... but just a nit-pick.
Yes, the final one I am working on but it will be close to the one on the previous page with the buffer changes included. Will also include gerbers so you can order PCBs.
@william2001 thanks for the feedback. I’ll check the footprints on the stereo board.
@william2001 thanks for the feedback. I’ll check the footprints on the stereo board.
With the parts I had in the Mouser project, everything dropped in, but @william2001 big WIMAs seem wider, and his small WIMAs have smaller lead spacing.
Humm
Would you like to know what happens when you put in 4.7 instead of 4.7K?
I was in a rush this morning and I decided to set up the new buffer arrangement…since surely it would not take more than 30 minutes.
Everything went well, except for the missing ohms.
It was still on the bench, so I realized my mistake quickly and it was only a few seconds, but yes, these JFETs are tough little buggers.
Would you like to know what happens when you put in 4.7 instead of 4.7K?
I was in a rush this morning and I decided to set up the new buffer arrangement…since surely it would not take more than 30 minutes.
Everything went well, except for the missing ohms.
It was still on the bench, so I realized my mistake quickly and it was only a few seconds, but yes, these JFETs are tough little buggers.
It probably survived if it wasn’t on for too long. With 4.7R, the source is near 0 volts and gate is near 35V. So, its Vgs is >> 0. Hard to say what would happen. It probably ran at Idss or higher for a while. If left on for a while, it will get toasted. But these are tough buggers. Put 4.7k in and see if it works.
I have some questions:
DN2335 is cascoded to let the ccs to work at higher voltages.
J113 I think is chosen because it’s low noise and works well as bottom part for ccs.
J111 as gain stage:
J113, J112, J111 all have the same transconductance; where they differ is in Vpinchoff (which results in different Idss too). So the higher the Vgsoff the better?
J175 as source follower:
Similar to the previous point for j175, j176, j177. So again the higher the Vgsoff the better?
DN2335 is cascoded to let the ccs to work at higher voltages.
J113 I think is chosen because it’s low noise and works well as bottom part for ccs.
J111 as gain stage:
J113, J112, J111 all have the same transconductance; where they differ is in Vpinchoff (which results in different Idss too). So the higher the Vgsoff the better?
J175 as source follower:
Similar to the previous point for j175, j176, j177. So again the higher the Vgsoff the better?
Yes, that's right!
Yeah, it provides a nice high impedance CCS and sounds a bit better to me. If the current were higher, we couldn't use J113 and it would need to be a single or dual DN2535 CCS.
Yes, I tried all of them and J111 has slightly more distortion, but it is H2 dominant even at high swings. The best subjective description is that it keeps its character at all volumes. It does not get congested or screechy as you turn it up.
Yeah, this one is a little hard to explain, but the subjective effect is similar to the J111. The J74 in this position gets you less distortion, but I didn't like how it sounded when you turn it up.
Thank you @ra7 , nice to see the improvements and the possibility to apply them to the pentode version as well.
Cascoded triodes (that exhibit pentode/fet/mosfet-like curves) also hybrid, can benefit from this solution as well.
Complication goes up, but this triodified-fet solution can give very good results if shunt cascoded: you can tailor the triode-like curves to your needs.
Cascoded triodes (that exhibit pentode/fet/mosfet-like curves) also hybrid, can benefit from this solution as well.
Complication goes up, but this triodified-fet solution can give very good results if shunt cascoded: you can tailor the triode-like curves to your needs.
Well, the work is never done. I put the preamp in my chain and I had some noise. Not the same kind of noise as before, but still. I was concerned something was oscillating, so I powered off.
The level of the noise increased while turning the volume know on the preamp. I switched to a 100K pot during the last stay on the bench, so I cannot know what is doing it.
However, I returned the preamp to bench and checked everything in terms of voltages and THD and everything is fine on the bench. I also tested on the bench with a cheap Fosi class D power amp, and there was no noise in that context either.
I'll update when I know more.
This project is freaking amazing, I am learning so much about so many things!
The level of the noise increased while turning the volume know on the preamp. I switched to a 100K pot during the last stay on the bench, so I cannot know what is doing it.
However, I returned the preamp to bench and checked everything in terms of voltages and THD and everything is fine on the bench. I also tested on the bench with a cheap Fosi class D power amp, and there was no noise in that context either.
I'll update when I know more.
This project is freaking amazing, I am learning so much about so many things!
A general rule of thumb for many preamp projects around here = if your source has a tube output, a 50K pot may be your best option. When your source has a solid state output, a 10K pot will work well. In other words, the Source determines what pot value you should really go with in most cases.
Noise problems can be elusive and challenging to solve. Noise from something can easily capacitively couple into low level input wiring where it happily hitches a ride into the input stage of the gain circuit. Short, high quality shielded input cabling that is far away from any potential noise source would be preferable to the opposite situation.... long, loopy, un-shielded, etc. Also, keeping the left and right channel cabling close to each other helps to reduce the total loop area and thus susceptibility to external interference pickup in the case of cross channel ground loops.
Noise problems can be elusive and challenging to solve. Noise from something can easily capacitively couple into low level input wiring where it happily hitches a ride into the input stage of the gain circuit. Short, high quality shielded input cabling that is far away from any potential noise source would be preferable to the opposite situation.... long, loopy, un-shielded, etc. Also, keeping the left and right channel cabling close to each other helps to reduce the total loop area and thus susceptibility to external interference pickup in the case of cross channel ground loops.