I am going to make an enclosure for my tubelab SE. I would like to know what tweaks do you think are worth, or at least won't hurt, as some of them have impact on the space I need.
1) bypassing the rectifier diodes of the B- with caps. What would be a good value ?
2) bypassing the choke with two 0.22 uF caps
3) Replace the electrolitic caps of the B- with film caps.
4) What value for the choke ? 5H or 10H ?
I am using a Solen cap as reservoir.
5) I was thinking of putting a 100 uF film cap as filter cap, and adding two 47uF attached to the B+ of the OPT.
6) should I bypass the big caps with smaller ones ? 1/100 of the value ?
I think I will make a temporary chassis before drilling aluminum
Thanks,
Davide
1) bypassing the rectifier diodes of the B- with caps. What would be a good value ?
2) bypassing the choke with two 0.22 uF caps
3) Replace the electrolitic caps of the B- with film caps.
4) What value for the choke ? 5H or 10H ?
I am using a Solen cap as reservoir.
5) I was thinking of putting a 100 uF film cap as filter cap, and adding two 47uF attached to the B+ of the OPT.
6) should I bypass the big caps with smaller ones ? 1/100 of the value ?
I think I will make a temporary chassis before drilling aluminum
Thanks,
Davide
My two cents:
1) The top diode package is a half-wave Schottky, so there would not be much improvement. The bottom two are more run-of-the-mill, so it would help. However the whole supply is regulated anyway, so the only thing it would do is reduce the amount of switching noise that might be coupling into other circuits. I have replaced those two discrete diodes for the negative side with a pair of Schottky's as well. They don't run as hot when in 300B configuration because of the lower drop.
2) If I am not mistaken, this is more important on choke input filters. Since this is CLC, I am not sure what the benefit might be. If your choke buzzes, it might help with that.
3) No idea.
4) I went with 10H. The more henries the better regulation will be.
5) Probably won't hurt, but the amount of current on the B+ rails is small. I assume you are trying to bypass coupling between the channels. Have you measured any?
6) I have done this on the cathode bypass caps of the input/driver stage. i didn't notice any difference, by my ears are not that sensitive.
1) The top diode package is a half-wave Schottky, so there would not be much improvement. The bottom two are more run-of-the-mill, so it would help. However the whole supply is regulated anyway, so the only thing it would do is reduce the amount of switching noise that might be coupling into other circuits. I have replaced those two discrete diodes for the negative side with a pair of Schottky's as well. They don't run as hot when in 300B configuration because of the lower drop.
2) If I am not mistaken, this is more important on choke input filters. Since this is CLC, I am not sure what the benefit might be. If your choke buzzes, it might help with that.
3) No idea.
4) I went with 10H. The more henries the better regulation will be.
5) Probably won't hurt, but the amount of current on the B+ rails is small. I assume you are trying to bypass coupling between the channels. Have you measured any?
6) I have done this on the cathode bypass caps of the input/driver stage. i didn't notice any difference, by my ears are not that sensitive.
My SE has a 10H choke @ ~80 DCR.
I have considered bypass caps but using a low ESR motor run fufills most of that purpose. One of the things I am still interested experimenting with is:
1) What is the correct size reservoir cap. I have 80uF, mt PSUDII sim says it may too large for quick recovery.
2) bias current for dirvers. I am presently at 12mA
When I finish my soundcard voltmeter I would like to look at cross talk, Distortion, and noise from other supplies. In these cases I want to measure improvement with data.
I am curious on LED's on the driver as well as choke loaded.
I have considered bypass caps but using a low ESR motor run fufills most of that purpose. One of the things I am still interested experimenting with is:
1) What is the correct size reservoir cap. I have 80uF, mt PSUDII sim says it may too large for quick recovery.
2) bias current for dirvers. I am presently at 12mA
When I finish my soundcard voltmeter I would like to look at cross talk, Distortion, and noise from other supplies. In these cases I want to measure improvement with data.
I am curious on LED's on the driver as well as choke loaded.
Recovery from what, exactly? A class A SE amp current draw is pretty much constant unless the amp goes into clipping. I've got a 220uF in the C5 position. No motor run cap...still on the breadboard.
I haven't done any measurements of this amp yet as it sits.
Fair Question.
Not many of us enjoy ultra high efficiency speakers. Mine are only 91dB+/-. Clipping occurs more often than I would care to omit. A simple hard hit drum for example.
There are several schools of thought on this so the path that I am pursuing "today" is to have the power supply able to recover from these clipping events quickly. In other words when the having it optimally dampened. Larger capacitance will result in over dampened, and to small of capacitance will result in ringing.
I don't know the right answer hence it is a "tweek".
Not many of us enjoy ultra high efficiency speakers. Mine are only 91dB+/-. Clipping occurs more often than I would care to omit. A simple hard hit drum for example.
There are several schools of thought on this so the path that I am pursuing "today" is to have the power supply able to recover from these clipping events quickly. In other words when the having it optimally dampened. Larger capacitance will result in over dampened, and to small of capacitance will result in ringing.
I don't know the right answer hence it is a "tweek".
Actually I was thinking of bypassing the HV diodes with caps, but the idea of replacing the two LV diodes is not bad. I was really wondering what was the point in not having four identical diodes.
The 80 uF for the first cap of the PSU is a bit too much. As I like to try various rectifiers I prefer to put a 22 uF Solen, as the limitations on capacitance can be a constrain. I did not really noticed a significant decrease in ripple if I put bigger ones.
Now I have only 10 uF, as I have a very high voltage on AC (770 V) but this is only for testing.
BR,
Davide
The 80 uF for the first cap of the PSU is a bit too much. As I like to try various rectifiers I prefer to put a 22 uF Solen, as the limitations on capacitance can be a constrain. I did not really noticed a significant decrease in ripple if I put bigger ones.
Now I have only 10 uF, as I have a very high voltage on AC (770 V) but this is only for testing.
BR,
Davide
I need to proof my comments a little better.
You are right 80uF for the first cap is too big for most rectifiers. I meant that I have 80uF in C5 or the second cap, after the choke. I have 47uF stock electrolytic for C4.
I should have mentioned rolling the coupling cap is also a interesting tweek.
You are right 80uF for the first cap is too big for most rectifiers. I meant that I have 80uF in C5 or the second cap, after the choke. I have 47uF stock electrolytic for C4.
I should have mentioned rolling the coupling cap is also a interesting tweek.
I am looking in all these details, as I am spending a lot on the transformers. Here in Japan they are good, but really expensive. And as I need universal primary, it will add 130 $ more. As I have 4 ohm speakers, the only models available are very expensive :-(
I will use Tango for all transformers and chokes.
Anyway, I think I will assemble the amp as follow, based also on many components that are in my drawer. Please tell me if there is something that will harm the sound instead of improving it, or if I could make better use of the components:
C4 22uF Solen polypropilene
C5 100uf Unicon Electrolitic bypassed with 1 poly
C6 47 uF Standard electrolitic bypassed with 0.47 poly
C7 100uf Unicon Electrolitic bypassed with 1 poly
Choke 10H 126ohm
Additionally I will put two 47 uF Solen caps on each B+ of the OPT. These will form a RC filter with the tuning resistors (R18-R29).
D4 and D5 replaced with schotky diodes (less heat and noise)
0.039 1KV cap on the AC input (for RF filtering)
All diodes (HT and LT) bypassed with 0.1 polypropilene
I was thinking of bypassing also C8 and C10, but if I apply the 1/100 rule the bypass cap is still too big. I could just put a 10uF. Putting diodes become a nightmare to tune it as these 5842 do not seem too consistent, although I had mine matched.
If I get it right before the transformers arrive I will try to replace the CCS with a cascode CCS.
I should get a B+ around 380 V with 70mA for each 300B.
Look forward to hearing your opinion.
Best Regards,
Davide
I will use Tango for all transformers and chokes.
Anyway, I think I will assemble the amp as follow, based also on many components that are in my drawer. Please tell me if there is something that will harm the sound instead of improving it, or if I could make better use of the components:
C4 22uF Solen polypropilene
C5 100uf Unicon Electrolitic bypassed with 1 poly
C6 47 uF Standard electrolitic bypassed with 0.47 poly
C7 100uf Unicon Electrolitic bypassed with 1 poly
Choke 10H 126ohm
Additionally I will put two 47 uF Solen caps on each B+ of the OPT. These will form a RC filter with the tuning resistors (R18-R29).
D4 and D5 replaced with schotky diodes (less heat and noise)
0.039 1KV cap on the AC input (for RF filtering)
All diodes (HT and LT) bypassed with 0.1 polypropilene
I was thinking of bypassing also C8 and C10, but if I apply the 1/100 rule the bypass cap is still too big. I could just put a 10uF. Putting diodes become a nightmare to tune it as these 5842 do not seem too consistent, although I had mine matched.
If I get it right before the transformers arrive I will try to replace the CCS with a cascode CCS.
I should get a B+ around 380 V with 70mA for each 300B.
Look forward to hearing your opinion.
Best Regards,
Davide
C4: with the lower ESR of the film cap, it will make life for the rectifier a little harder. You are using a smaller value which will help, but if you end up needing to go with 47uF, I personally wouldn't do it with a film cap. Just my two cents.
Other than that, it all sounds good to me. As to bypassing C8 and C10, any film cap will help here. On paper, these Panasonic electrolytic caps look pretty good and the 1/100 rule seems overkill. 1/1000 seems more reasonable...even 1/10000.
Other than that, it all sounds good to me. As to bypassing C8 and C10, any film cap will help here. On paper, these Panasonic electrolytic caps look pretty good and the 1/100 rule seems overkill. 1/1000 seems more reasonable...even 1/10000.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.