:
Is it posible to have a look here post 61? http://www.diyaudio.com/forums/equi...g-e-mu-0404-usb-measurements-arta-rmaa-7.html
I want your opinion.
Thanks.
Is it posible to have a look here post 61? http://www.diyaudio.com/forums/equi...g-e-mu-0404-usb-measurements-arta-rmaa-7.html
I want your opinion.
Last edited:
Hi Guys,
I finally scrapped my Aleph X so now I have the heatsinks to finish my Slewmaster A/B setup. I mounted all 4 OPS boards. I have an identical transformer to the one in the picture, I just have to remember which amp I used it in. They are 55-0-55vac 400VA. They won't be "Party duty" but should be fine for comparison purposes.
I have a question about the Symasym IPS you see in the picture. I noticed when running it that Q7 and Q8 are running pretty warm. when I checked them I realized that I had lerf the BC546B in those positions. This is running at +-73V after the cap multiplier so they are probably at the limit. I should replace them but wonder if it would be better to use KSC1845 or 2SN5551? Any suggestions? Also what current should the TO-92 devices be run?
Thanks, Terry
I finally scrapped my Aleph X so now I have the heatsinks to finish my Slewmaster A/B setup. I mounted all 4 OPS boards. I have an identical transformer to the one in the picture, I just have to remember which amp I used it in. They are 55-0-55vac 400VA. They won't be "Party duty" but should be fine for comparison purposes.
I have a question about the Symasym IPS you see in the picture. I noticed when running it that Q7 and Q8 are running pretty warm. when I checked them I realized that I had lerf the BC546B in those positions. This is running at +-73V after the cap multiplier so they are probably at the limit. I should replace them but wonder if it would be better to use KSC1845 or 2SN5551? Any suggestions? Also what current should the TO-92 devices be run?
Thanks, Terry
Attachments
Last edited:
Your Q7 is running 2X what the CCS on the wolverine or symasui is (7+ma).
You are also over Vceo using 546b with 73V. Lucky no magic smoke !!
Use 1k for R5/6 , 1845's for Q7/8 and replace R11 with a 500R trimmer.
with R11=400R+ your total LTP current will be 3.5-4ma (instead of 7).
Using the ksa1845 easily allows for 75+ V rails (120V devices).
You could use the 2N5551's as well (just a different pinout /higher vceo).
Your question of the best currents for the to-92's .....
2ma gives the best compromise of performance/derating , I run all my
LTP's between 1.5-2ma for each device (3-4ma total CCS current).
Most of the 100's of OEM BJT input stages I've seen also run
1-2ma per device.
OS
Q8 shouldn't be warm at all as it should have very little voltage across it, but Q7 basically 'sees' a full rail and needs to be rated appropriately, you are exceeding Q7's voltage rating. You could likely just replace Q7 with a 2N5551 (mind the pin-out - they are different) and see how it goes. The tail current source is only flowing ~3mA, not the 7mA+ as OS suggested so I wouldn't be too quick to make the other changes. You could make R12 more like 47kΩ though with the higher rails.
Looks like if you wanted to go full-on party mode you are set to bridge or bi-amp
.
In addition to my prior comment, the CCS (Q7) is dissipating ~220mW and can be expected to rise ~45C above ambient which for open air could mean a device operating at 70C which would be uncomfortable to touch. A 1W rated TO-92L device would run a little cooler but a higher gain TO-126 device would be much better suited. I try to keep my TO-92 devices to 150mW or less.
Hi Guys,
I changes out R11 to 270R. That dropped the temp some. Still several of the TO-92 transistors get pretty warm to the touch including Q8. It sims at 3.2mA but maybe I have something wrong. Looks like THD goes way up with the R11 change too. I was surprised at how much bias adjustment was required after changing R11.
I am attaching the spice model I am using so you can check my figures.
Blessings, Terry
I changes out R11 to 270R. That dropped the temp some. Still several of the TO-92 transistors get pretty warm to the touch including Q8. It sims at 3.2mA but maybe I have something wrong. Looks like THD goes way up with the R11 change too. I was surprised at how much bias adjustment was required after changing R11.
I am attaching the spice model I am using so you can check my figures.
Blessings, Terry
Symasym is not the "symasui".
Like trying to get the Toyota to run like the Ferrari.
Mike B. made the sym to run on normal rails(50V). I make my IPS's to be
"scalable". This was my design goal.
35-85V is for real ... all my IPS's perform the same across that whole
range.
This is why I am "cascode" happy with regulated low voltage input
stages.
I am quite happy you even got an "alien" IPS to work with the "slew",
this shows how good it is.
When you get your CFA-Xh boards , you will see what I mean ... all
will run super cool , you could even omit the heatsink and be fine.
PS - you are really learning fast !!
OS
Like trying to get the Toyota to run like the Ferrari.
Mike B. made the sym to run on normal rails(50V). I make my IPS's to be
"scalable". This was my design goal.
35-85V is for real ... all my IPS's perform the same across that whole
range.
This is why I am "cascode" happy with regulated low voltage input
stages.
I am quite happy you even got an "alien" IPS to work with the "slew",
this shows how good it is.
When you get your CFA-Xh boards , you will see what I mean ... all
will run super cool , you could even omit the heatsink and be fine.
PS - you are really learning fast !!
OS
Hi OS,
I'm not trying to make the SymaSym IPS to out do your IPS. It is just something to play with so I can learn. It is amazing how much can be learned from building spice models and getting them to run. You guys have been so much help. It is also interesting to see how much difference there is in THD measurements between the different designs. Can't wait to get started on some of your designs.
Blessings, Terry
I'm not trying to make the SymaSym IPS to out do your IPS. It is just something to play with so I can learn. It is amazing how much can be learned from building spice models and getting them to run. You guys have been so much help. It is also interesting to see how much difference there is in THD measurements between the different designs. Can't wait to get started on some of your designs.
Blessings, Terry
Member
Joined 2009
Paid Member
Hi BV,
Cool! I'll give that a try when I'm at my computer. I see you removed part of the circuit. Please explain the reasons so I can learn something. Thanks for fixing the text file. I just copied and pasted the models from spice and then attempted to mimic the format of the other models in the file. Guess I split it up wrong. I will try to look at the changes you made so I will understand how to do it in the future. I was able to add your Cordell files to the 'standard bjt' file in Ltspice. I was kind of tickled about that though I couldn't' find where to add the diode files.
I may try pulling the parts you removed in your schematic from my boards and see how things sound.
Blessings, Terry
Member
Joined 2009
Paid Member
Hi Terry,
But!!! It is only simulation, so be carefull, startat first with protective resistors (33-100R/10-15W) instead of +-Ub fuses and without output load. It can save several transistors.
2 Thimios
Your measurments are very contaminated with mains. Try at first card loopback and than connect ground of this loopback cable to amplifier ground (e.g. to ground output terminal, try with amplifier switched on, off, but with connected mains cable). If will appear this mains contamination, then here are serious grounding issues.
I do not like load at output of VAS. It only unnecesary worsens basic VAS linearity and decrease loop gain , so I omit two resistors and capacitive loading. Instead of this i tried to compensate in local feedback loop
Whole mistake was excessive * in front of .model statement line
In component folder of LTspice is LTspice diode file (D standart), edit this file and add diode models here.
But!!! It is only simulation, so be carefull, startat first with protective resistors (33-100R/10-15W) instead of +-Ub fuses and without output load. It can save several transistors.
2 Thimios
Your measurments are very contaminated with mains. Try at first card loopback and than connect ground of this loopback cable to amplifier ground (e.g. to ground output terminal, try with amplifier switched on, off, but with connected mains cable). If will appear this mains contamination, then here are serious grounding issues.
Last edited: