Hi Cheff
Build the amp and you hear what I mean.
I'm not bashing your amp and always looking for some improvement because building something is a part from the hobby.
Rob
It's the firs time I did a review about sound and maybe the last time because its very difficult to describe what you hear.CheffDeGaar said:
Build the amp and you hear what I mean.
I'm not bashing your amp and always looking for some improvement because building something is a part from the hobby.
Rob
Very interesting listening tests
and good work from you who have designed and built the UGS module! Hats off.
Now,
to get the most out of the UGS module I think that some issues have not been dealt with.
I am thinking about the PS and the passive component quality.
IMHO - the UGS module is worth the best passive components one can afford and the best internal wiring and connectors and so on.
Dynamcs can be related to the capacitors used in the PS. Resolution has not been mentioned but resistors show different degrees of resolution.
Just to use standard metal film resistors and a simple PS is not going to give the best performance from the UGS module.
Dual mono setup is of course mandatory as are good transformers and excellent rectifying diodes. and active ground.
A regulated PS like this
or like this
is what I would use per channel.
And a setup like this for the transformers and AC mains filter
or like this
using this rectifying scheme (big caps not shown):
Just my 2 cents.
Sigurd
and good work from you who have designed and built the UGS module! Hats off.
Now,
to get the most out of the UGS module I think that some issues have not been dealt with.
I am thinking about the PS and the passive component quality.
IMHO - the UGS module is worth the best passive components one can afford and the best internal wiring and connectors and so on.
Dynamcs can be related to the capacitors used in the PS. Resolution has not been mentioned but resistors show different degrees of resolution.
Just to use standard metal film resistors and a simple PS is not going to give the best performance from the UGS module.
Dual mono setup is of course mandatory as are good transformers and excellent rectifying diodes. and active ground.
A regulated PS like this
An externally hosted image should be here but it was not working when we last tested it.
or like this
An externally hosted image should be here but it was not working when we last tested it.
is what I would use per channel.
And a setup like this for the transformers and AC mains filter
or like this
An externally hosted image should be here but it was not working when we last tested it.
using this rectifying scheme (big caps not shown):
Just my 2 cents.
Sigurd
Rob Dingen said:
I have impression that nobody criticize you,especially not your fast work;
and especially not me-just because you use tube preamp VS UGS one;
besides-I don't agree that tubes and 3 legged fuses are apples and oranges;
good gadget is good,doesn't matter with what is constructed;
what I say,and most of others replies on your test is that you didn't probably used best setup for UGS....you can freely disregard comparison with tube preamp......
develop tube preamp further
do the same with UGS
I'm sure that you have more than one room-to fill with music producing devices
and ,please,keep telling us your impressions
Hi
Today I changed the feedback resistors from 56k to 150k so that I have a gain of 10x.
Also the 1k resistor to ground is now 10k.
The problems I have now is that it’s not so stable anymore and offset is wandering around.
Also with the higher gain I get overshoot and I compensate it with a 10pf cap over the 150k feedback resistor and limited the bandwidth to 100Khz.
I think it sound better now with more dynamics and it’s more musical.
Maybe Cheff has some more spare boards so that he can switch between the 2 and tell his listening impressions.
I don’t now what causes the difference in the sound but I think the feedback resistors.
When I have time I change the 10k to ground to 4k7 or 2k2.
And maybe later change the Jfets to types with lower input capacitane.
Rob
Today I changed the feedback resistors from 56k to 150k so that I have a gain of 10x.
Also the 1k resistor to ground is now 10k.
The problems I have now is that it’s not so stable anymore and offset is wandering around.
Also with the higher gain I get overshoot and I compensate it with a 10pf cap over the 150k feedback resistor and limited the bandwidth to 100Khz.
I think it sound better now with more dynamics and it’s more musical.
Maybe Cheff has some more spare boards so that he can switch between the 2 and tell his listening impressions.
I don’t now what causes the difference in the sound but I think the feedback resistors.
When I have time I change the 10k to ground to 4k7 or 2k2.
And maybe later change the Jfets to types with lower input capacitane.
Rob
Hi Rob
No problemo. I wasn't thinking your were bashing the pre, but rather
feeling that you were honestly reporting what you felt.
And you happen to prefer your tube pre, that's life
Anyone should thank you for a) building it so quickly, b) performing the tests,
and c) reporting. So many threads suddenly stop when it comes to
reporting listening results...
And I perfectly understand that it could be difficult to describe what
you're hearing. And the same goes for me, except it was just the
opposite global sensations I think nobody's wrong or right here,
it's just a matter of taste... Can we explain why some people only
swear by tubes, or class D and so on ? Surely no
I just hope it won't stop anybody to try this pre, experiment and report,
would it be good or bad...That's what this forum is for, and where
the fun is.
As far as I know, there's been only 3 ugs built using this schematic
and listened to. Too few to get a global picture, so we're waiting for
you, gentlemen. And the first ones I built were not top notch (single
supply, unshielded everywhere, etc...) but I already felt something
unusual
So I'll stick to ugs... until I build your pre
Cheers
Rob Dingen said:
No problemo. I wasn't thinking your were bashing the pre, but rather
feeling that you were honestly reporting what you felt.
And you happen to prefer your tube pre, that's life
Anyone should thank you for a) building it so quickly, b) performing the tests,
and c) reporting
. So many threads suddenly stop when it comes toreporting listening results...
And I perfectly understand that it could be difficult to describe what
you're hearing. And the same goes for me, except it was just the
opposite global sensations
I think nobody's wrong or right here, it's just a matter of taste... Can we explain why some people only
swear by tubes, or class D and so on ? Surely no
I just hope it won't stop anybody to try this pre, experiment and report,
would it be good or bad...That's what this forum is for, and where
the fun is.
As far as I know, there's been only 3 ugs built using this schematic
and listened to. Too few to get a global picture, so we're waiting for
you, gentlemen. And the first ones I built were not top notch (single
supply, unshielded everywhere, etc...) but I already felt something
unusual
So I'll stick to ugs... until I build your pre
Cheers
Rob,
Sorry, I was writing while you posted
Thanks for experimenting, and for this report.
I don't think you need to have a 10K Rout. Just leave it 1K, and keep
the 150K FB resistor. These 10K do not help stabilizing the offset (x10)
and the CL gain won't be much affected if you go back to 1K.
I've no spare boards with current mirrors, but I'll have soon, so I'll compare.
Just tell us how it goes, but thanks again for keeping it up.
I guess it's still less musical than tubes, but how does it improved ?
Regards
Sorry, I was writing while you posted
Thanks for experimenting, and for this report.
I don't think you need to have a 10K Rout. Just leave it 1K, and keep
the 150K FB resistor. These 10K do not help stabilizing the offset (x10)
and the CL gain won't be much affected if you go back to 1K.
I've no spare boards with current mirrors, but I'll have soon, so I'll compare.
Just tell us how it goes, but thanks again for keeping it up.
I guess it's still less musical than tubes, but how does it improved ?
Regards
I have followed this thread with great interest - and desire to build a UGS - But I have a problem I think I share with a lot of people around the world - and that is the availability of the double Jfets 109/389 - I have found the 389 in UK @ 5 GBP a piece !! - thats something like 10 USD.....and then they are not in stock...
I think that if anyone who live in a part of the world where these part are easily available would like to start up a "helpline" supplying these small devices, many more UGS would be build.
So - if anyone knows any good, reliable places to buy them - please do tell me
Cheers !
Buhl
I think that if anyone who live in a part of the world where these part are easily available would like to start up a "helpline" supplying these small devices, many more UGS would be build.
So - if anyone knows any good, reliable places to buy them - please do tell me
Cheers !
Buhl
Buhl -
I ordered some SK389V and SJ109V from Borbely Audio last week, and he still have some - in stock.
Price is good and service excellent
http://www.borbelyaudio.com/pics/Compprice.pdf
Sigurd
I ordered some SK389V and SJ109V from Borbely Audio last week, and he still have some - in stock.
Price is good and service excellent
http://www.borbelyaudio.com/pics/Compprice.pdf
Sigurd
Hi spencer
You guys are amazing !
You didn't specially make the enclosure to host the ugs, did you ?
Great work ! Is the supply the one I described, or another one ?
It would be better I guess to increase RFB to increase the gain (try 68K
to 100K). That's strange, I never felt the gain was low in my system...
At my normal listening levels at home, the ouput is around 30dB from
full scale, with the 15K/56K gain resistors, but I use it fully balanced...
(Edit : If I remember, you use 47k as RFB ?)
If I may, I have some questions :
What is the gain of the Borbely preamp your'e using ?
Do you use the ugs unbal to bal or fully unbalanced ?
And what is the impedance of the stepped attenuator you're using ?
Did you put it at output, or input ?
Anyway, thanks for teststing, and keep us to date
Cheers,
You guys are amazing !
You didn't specially make the enclosure to host the ugs, did you ?
Great work ! Is the supply the one I described, or another one ?
It would be better I guess to increase RFB to increase the gain (try 68K
to 100K). That's strange, I never felt the gain was low in my system...
At my normal listening levels at home, the ouput is around 30dB from
full scale, with the 15K/56K gain resistors, but I use it fully balanced...
(Edit : If I remember, you use 47k as RFB ?)
If I may, I have some questions :
What is the gain of the Borbely preamp your'e using ?
Do you use the ugs unbal to bal or fully unbalanced ?
And what is the impedance of the stepped attenuator you're using ?
Did you put it at output, or input ?
Anyway, thanks for teststing, and keep us to date
Cheers,
Hi
I made another change in terms of stability.
Now I have 150k feedback resistors and 10k from output to ground.
Also like in the AX amps I put 2 10k resistors from the +/- outputs to the RS1 and RS2 connection and the amp is very stable in terms of offset.
I think it sounds a lot better like this but I don’t now if you put the Volume pot at output you get problems with clipping.
The voltage swing could be higher because it now has 5k at the output in stat of 1k.
Rob
I made another change in terms of stability.
Now I have 150k feedback resistors and 10k from output to ground.
Also like in the AX amps I put 2 10k resistors from the +/- outputs to the RS1 and RS2 connection and the amp is very stable in terms of offset.
I think it sounds a lot better like this but I don’t now if you put the Volume pot at output you get problems with clipping.
The voltage swing could be higher because it now has 5k at the output in stat of 1k.
Rob
Hi Rob,
As I stated in the first post of this thread, one of my concerns was to
get rid of these common mode feedback resistors, that impede on the
unbal to bal function, but they sure help a lot to stabilize the absolute offset.
As you own a double trace scope (alas I don't...), could you test
the unbalance between the two outputs when feeding the UGS
unbalanced ? The lower the comon mode fb resistors, the worse
the results. If you can plot V(out+) + V(out-), you should normally
have zero when the unbal to bal conversion is perfect. And in sims,
it was around few tens of mV when the resistors where present.
And that's what I measured with a PC soundcard scope...
You could also test higher resistors values (100K, 200K). Their effect
is still there on offset stability (slightly worse, but workable), and the
unbal to bal conversion is better...
And as I said previoulsy, perhaps you could test back with a 1K Rout,
keeping the "high" feedback resistors. The effect on the closed
loop gain will not be high, and it also might help you stabilizing the
offset.
I will try 100K or 150K resistors as RFB this WE (keeping Rout as 1K),
and report back my impressions.
Thanks for all your tests, and for keeping it up.
Cheers
As I stated in the first post of this thread, one of my concerns was to
get rid of these common mode feedback resistors, that impede on the
unbal to bal function, but they sure help a lot to stabilize the absolute offset.
As you own a double trace scope (alas I don't...), could you test
the unbalance between the two outputs when feeding the UGS
unbalanced ? The lower the comon mode fb resistors, the worse
the results. If you can plot V(out+) + V(out-), you should normally
have zero when the unbal to bal conversion is perfect. And in sims,
it was around few tens of mV when the resistors where present.
And that's what I measured with a PC soundcard scope...
You could also test higher resistors values (100K, 200K). Their effect
is still there on offset stability (slightly worse, but workable), and the
unbal to bal conversion is better...
And as I said previoulsy, perhaps you could test back with a 1K Rout,
keeping the "high" feedback resistors. The effect on the closed
loop gain will not be high, and it also might help you stabilizing the
offset.
I will try 100K or 150K resistors as RFB this WE (keeping Rout as 1K),
and report back my impressions.
Thanks for all your tests, and for keeping it up
.Cheers
Hi Cheff,
I give up my Aleph 1.7 casing and put in the UGS. I only hv two casings with XLR and RCA sockets.
The supply is just a zener and mosfet series regulator, follow NP design. Voltage at about 22.3V.
I use 3.3k Rin now and Rfb 47k, the gain is higher a bit but still not up to my expection. I need to turn to 12:00 for loud level.
The Borbely SE Jfet preamp is about 21 dB gain, I turn to 9:00 for same volume level. Gain is a bit high to me. I believe 10:30 for my loud level is best position to me.
I use 10k ohm volume at input for both amp.
My DAC is unbal output, UGS output is bal to my Aleph 60 diy.
I believe your power amp has higher gain than mine.
Spencer
I give up my Aleph 1.7 casing and put in the UGS. I only hv two casings with XLR and RCA sockets.
The supply is just a zener and mosfet series regulator, follow NP design. Voltage at about 22.3V.
I use 3.3k Rin now and Rfb 47k, the gain is higher a bit but still not up to my expection. I need to turn to 12:00 for loud level.
The Borbely SE Jfet preamp is about 21 dB gain, I turn to 9:00 for same volume level. Gain is a bit high to me. I believe 10:30 for my loud level is best position to me.
I use 10k ohm volume at input for both amp.
My DAC is unbal output, UGS output is bal to my Aleph 60 diy.
I believe your power amp has higher gain than mine.
Spencer
