I ran it at full speed for about 4-5 hours buy now, but SQ is not the same clarity, airiness and details level as I have on my DCG3 via headphones. Probably it need more burn-in for Ele caps to be in spec and I'm considering C6/7 & C13/14 upgrades. Did not touch T3/5 for H2/H3 advised setup and I'm running with 5.4V plate voltage. Will try to adjust that right after all caps will be upgraded. Considering well regarded MIFLEX KPCU-01 0.1uF 600V and relatively affordable ClarityCap Capacitor CSA 10uF 250V. Other brands 10uF nice Caps are way-way too pricy.
Those polyesters in signal's way should contribute to a more dull sound, less details, less air, etc so in order to be more fair try changing them to at least their Wima PP/ FF equivalent though it would be hard to fit them in. Wima's MKP10 /FKP are very big for radials. One step up are the Clarity CMRs.
I'd still expect for KB1 to sound fuller than DCG3, my DCB1 certainly does but what other qualities you gonna loose it's something to hear by yourself.
I'd still expect for KB1 to sound fuller than DCG3, my DCB1 certainly does but what other qualities you gonna loose it's something to hear by yourself.
Hi Savvas,
Thanks for tips. I presented my subjective and strictly personal experience based on original BOM in my “wired” two-floors setup. No any upgrades are made yet and plate voltage is also as manual directed equal to +5.4VDC tight for all relevant TPs.
Now, I already ordered Clarity Caps and Miflex caps. So, let’s see how it will sound with these caps installed. During waiting time, I must do something with my 60Hz toroids flux radiation and that junk pickup. That noise is below -95dB, so I can’t hear it even on my very high SPL headphones. But any way, make sense to try to reduce it with all its harmonics.
Later on I’ll also try what Nash did for plate voltages and H2/H3 changes. Ha, I have big plans….
Looks messy with wiring, but functional.
Now, it is day and night difference.
Details are came back and it is very juicy. I can say it is more “meat” in sound now.
I just started to run it 4 hours ago and I listened first time with new Caps now.
Let’s see what would be in a day or two of "burn in" with these Caps.
One question is remained and it is about better SQ substitute for almost obsolete OPA551PA.
Thank you a lot for your help.
Now, it is day and night difference.
Details are came back and it is very juicy. I can say it is more “meat” in sound now.
I just started to run it 4 hours ago and I listened first time with new Caps now.
Let’s see what would be in a day or two of "burn in" with these Caps.
One question is remained and it is about better SQ substitute for almost obsolete OPA551PA.
Thank you a lot for your help.
Attachments
Glad to hear that the big caps made a good difference. Do you have space to lift up the PCB enough so as to tuck them under?
Cant say what's better than the OPA551. I thought you were planning on using it primarily as a line stage wwith the Jfets you had on previously?
nash
No, I do not have any space to place these Caps beneath my amp PCB. Amp PCB is in my second floor when first floor is copied by PSU. Just to remind you: I use 4 UltBIBs PCBs there, means two pairs with +/- outputs and two 25VA toroids.
About your second question: OPA551 choice is related to usage of Low Z headphones. I have 200 Ohm headphones and probably can use classic OPAMPs with way less than 200mA current. Will test several include single discrete units like Sparkos Lab SS3601.
Also, I was naïve and tried to use it for both purposes: as headphone amp and line preamp. I am afraid it is not possible with current configuration.
Finally, I got several "audio" opamps" to test and I ended up with SparkoS Labs SS3601. It is totally enough juice to run my 80 and 200 Ohm headphones. The sound changed from to be blunt (with BOM OPA551PA) to detailed, cristal-clear and very open.
Very happy with the that change.
Again, I still run my amp based on designer advised BIAS equal to 5.4V and did not play with H2/3/4 yet and etc. That might be my next step. Lets see...
Very happy with the that change.
Again, I still run my amp based on designer advised BIAS equal to 5.4V and did not play with H2/3/4 yet and etc. That might be my next step. Lets see...
Attachments
alexkosha, I'm in the process of sourcing parts for the Millett Machine and I want to thank you for sharing your build experience. I'm a premium parts addict and always appreciate when someone puts something special through its paces.
Also learning a lot about the packaging approach from what you're doing and I think I'll go for 2 separate chassis, one for the pre boards, and one for the Salas UltraBibb (which will have 2 transformers).
Good luck, and keep it comin'.
Also learning a lot about the packaging approach from what you're doing and I think I'll go for 2 separate chassis, one for the pre boards, and one for the Salas UltraBibb (which will have 2 transformers).
Good luck, and keep it comin'.
alexkosha, I'm in the process of sourcing parts for the Millett Machine and I want to thank you for sharing your build experience. I'm a premium parts addict and always appreciate when someone puts something special through its paces.
Also learning a lot about the packaging approach from what you're doing and I think I'll go for 2 separate chassis, one for the pre boards, and one for the Salas UltraBibb (which will have 2 transformers).
Good luck, and keep it comin'.
My pleasure. About your idea of two chases, be careful since UltBIB has no Kelvin since and long output wires might be a issue. I would consult with Salas about umbilical cord for remote UltBIB...
Alex
Hi Nash, Thank you. I have 10R soldered to PCB for headphones and 100R is soldered directly on XLR output for line out. No, I did not try it yet as line out. Need to finish my Nelson's F5T Balance to SE input setup first.
Alex
Now, I have "$1M" question. What is more important for SQ:
a. to stay at recommended BIAS (which is my case equal to 5.4V tight).
b. better match for advised in manual H2/3/4 by deviating from BIAS and paying by some degradation for CMRR
Pete and guys, please let me know your opinions.
I measured initial FR and H1/2/3s via 100mV (did not go for 1V yet, but assume that situation there will be similar) input and see quite big difference in behave for L vs. R channels for harmonics. No difference for FR.
So, please advise.
P.S. Disregard 60Hz spikes and all their harmonics. I believe it is some artifact of my Focusrite setup. I can't hear any humm, buzz or hiss via my high SPL headphones. It is totally quite with no playing music and in MAX volume.
a. to stay at recommended BIAS (which is my case equal to 5.4V tight).
b. better match for advised in manual H2/3/4 by deviating from BIAS and paying by some degradation for CMRR
Pete and guys, please let me know your opinions.
I measured initial FR and H1/2/3s via 100mV (did not go for 1V yet, but assume that situation there will be similar) input and see quite big difference in behave for L vs. R channels for harmonics. No difference for FR.
So, please advise.
P.S. Disregard 60Hz spikes and all their harmonics. I believe it is some artifact of my Focusrite setup. I can't hear any humm, buzz or hiss via my high SPL headphones. It is totally quite with no playing music and in MAX volume.
Attachments
Last edited:
Hi Pete,
Thanks for you feedback. I’ll try to play with both channels tomorrow for min similar Harmonics on both. Lets listen how that sounds first.
I believe it is time consuming process since I need to go through several input voltages. Will report soon.
I compared my FR to what I found in your manual and seems it is very close behave, if not the same as you posted there. So, that part is not bothering me.
Thank you.
Alex
Thanks for you feedback. I’ll try to play with both channels tomorrow for min similar Harmonics on both. Lets listen how that sounds first.
I believe it is time consuming process since I need to go through several input voltages. Will report soon.
I compared my FR to what I found in your manual and seems it is very close behave, if not the same as you posted there. So, that part is not bothering me.
Thank you.
Alex
I set T3/T5 on both channels to be very close to each other. You can see how Hs looks like in attached image. Through 0.25V and 0.5V inputs, outputs Hs signals seems very close to each other too. For 1V in, it is hard to judge harmonics due to too many of them on screen, but again, it's looks close on both channels..
I like how it is sounding now, but I have some question. It seems it is quite shy on highs.
Is any way to add some high frequency by adjusting harmonics?
I like how it is sounding now, but I have some question. It seems it is quite shy on highs.
Is any way to add some high frequency by adjusting harmonics?
Attachments
Hi Alex,
Reduce the THD shown at 0.4v out from 0.19 to 0.05% and let me know. As PM has said H2 is a matter of taste. See my previous posts with THD as a possible guide. Roughly 0.05%THD at 1v in and 0.4vout sounds best to me in line use with my Sony Vfet power amps. This using the Focusrite gen3 setup I described earlier. Absolute numbers will vary on setup used. Lovely highs no shyness.
nash
Reduce the THD shown at 0.4v out from 0.19 to 0.05% and let me know. As PM has said H2 is a matter of taste. See my previous posts with THD as a possible guide. Roughly 0.05%THD at 1v in and 0.4vout sounds best to me in line use with my Sony Vfet power amps. This using the Focusrite gen3 setup I described earlier. Absolute numbers will vary on setup used. Lovely highs no shyness.
nash