K-amps said:
We are all at a loss, but I would not expect the 240/120V conversion to be significant.
The impedance reflected from primary to secondary should remain about the same for the two different voltage set-ups.
I guess the reflected impedance may be slightly lower for the 240Vac wiring but this will depend on the house and distribution board wiring.
Now, bias setting, that could be VERY significant. This has been reported by many others with different ClassA and ClassAB amplifiers.
LineSource said:
Moi.
a: i've already got more than a few
b: i've still got enough beautifull Avid-Thermalloys, for which i still owe Mark a big favor.
c: i hate to cut those up again for 1/2-duty on Jappon devices.
I've got 4 spare KSA50 output boards as my fellow Dutch KSA50 GB member on second thought did not require the second batch Pinky boards.
If there's a way of getting another 8 KSA50 output boards or the KSA50 Gerbers i'm a happy camper.
Mr Watts Sir,
Given that Mark's reaction to his K-100 was written in very BOLD lettering, i suggest to set yours to comply with his first to see if there's a difference.
The 32/33s are far superior to 30/31s, 110/115 to 220/230 should be irrelevant.
i'd definitely rip the BG out.(no criticism of BGs intended)
Hi Linesource,
Leach went for three series diodes and changed to four to suit physical constraints.
I have found that four (1n4004 or 1n4148) diodes overcompensates (reduces bias as sink heats up) but it is a nice, easy & free, protection mechanism.
On-chip diodes will be more effective (both in speed and deltaT) and so I guess three will be sufficient.
Which three (from 8 or 12) should be chosen?
how many would you (or anyone else) use in the Leach style Vbe multiplier?
Leach went for three series diodes and changed to four to suit physical constraints.
I have found that four (1n4004 or 1n4148) diodes overcompensates (reduces bias as sink heats up) but it is a nice, easy & free, protection mechanism.
On-chip diodes will be more effective (both in speed and deltaT) and so I guess three will be sufficient.
Which three (from 8 or 12) should be chosen?
I run large cinema sub woofer amps on 240 instead of 120 and it's always with a definate improvement in cleanliness on the extreme lows.
I think BG's are almost a joke.
I'd run the bias back up where it belongs and also let the amp break in for several days without shutting it down. Also let it play music even when no one is around.... I've found bias level to have the largest overall efect on how an amp actually sounds.
I used the 21193/94 combo in both the 50's I've built and in the 100. Keep in mind that when I get back home that I am going to revamp the 100 power supply and install that 1.5 kva toroid I have and have another listening session. In theory it should improve alot over how it sounds now... which is alot better than my 50 does.
Mark
I think BG's are almost a joke.
I'd run the bias back up where it belongs and also let the amp break in for several days without shutting it down. Also let it play music even when no one is around.... I've found bias level to have the largest overall efect on how an amp actually sounds.
I used the 21193/94 combo in both the 50's I've built and in the 100. Keep in mind that when I get back home that I am going to revamp the 100 power supply and install that 1.5 kva toroid I have and have another listening session. In theory it should improve alot over how it sounds now... which is alot better than my 50 does.
Mark
As it is the 100 stabilizes in about 20 minutes. If one uses a smaller "Krell" size heatsink for the drivers it would stabilize even faster. Mine is quite a bit heftier than theirs was. With out the VBE multiplier on the actual O.P. device heatsink it's very stable over the long haul. I did not note any more than a few MV drift after a half hour warm up once things are set up... thats pretty good if you ask me....
Mark
Hi Jacco,
Is that a NO then?
Mark,
there's those Mega (MV) again.
I would not call that stable, dangerous more like!
I will come down on the side that the bigger sink will stabise more quickly (to within an acceptable tolerance).
Is that a NO then?
Mark,
there's those Mega (MV) again.
I would not call that stable, dangerous more like!
I'm not sure this follows. The lower case temperature might imply getting there sooner and less change when it does get there. The smaller sink has to let the case temp rise higher and this may take longer and will involve a bigger change in parameters while it is heating up.
I will come down on the side that the bigger sink will stabise more quickly (to within an acceptable tolerance).
Well Mark, there's only one way to satisfy our (let's keep it personal since no else joined in) curiosity.
Measure it with your bigger sink,
then measure it with a smaller sink and post the two sets of results.
Starting bias and output offset,
bias and offset at 5minute intervals,
final bias and offset.
Try to arrange for ambient temperature and supply voltage to be similar for both tests.
Then we can try to identify how long it took each version to reach, within some selected tolerance (not too wide), the finish line.
Measure it with your bigger sink,
then measure it with a smaller sink and post the two sets of results.
Starting bias and output offset,
bias and offset at 5minute intervals,
final bias and offset.
Try to arrange for ambient temperature and supply voltage to be similar for both tests.
Then we can try to identify how long it took each version to reach, within some selected tolerance (not too wide), the finish line.
Hi
Some quick feedback re the sound quality following the work done by Pierre.
I have removed the 2uF cap from the mains voltage at the transformer primaries at the wire harness. That made a 20-30% improvement in restoring the bass tightness the Krell. However I still find it lacking in realism, overall transparency and extreme dynamics. The attached picture refers to the cap change on the PSU.
I am hoping that the above sound attributes could be due to the fact that some burn-in is still required. For the meantime the amp stays as is. If there is no improvement my gut feeling is that if one might need to revert back to the original PCB drivers?
At the moment I am playing the amp without a cover and despite the lower class A rating the output transistor heat sinks still do get pretty hot..
I keep the group updated.
Jozua
Some quick feedback re the sound quality following the work done by Pierre.
I have removed the 2uF cap from the mains voltage at the transformer primaries at the wire harness. That made a 20-30% improvement in restoring the bass tightness the Krell. However I still find it lacking in realism, overall transparency and extreme dynamics. The attached picture refers to the cap change on the PSU.
I am hoping that the above sound attributes could be due to the fact that some burn-in is still required. For the meantime the amp stays as is. If there is no improvement my gut feeling is that if one might need to revert back to the original PCB drivers?
At the moment I am playing the amp without a cover and despite the lower class A rating the output transistor heat sinks still do get pretty hot..
I keep the group updated.
Jozua
Attachments
Me recollects Mark posted he's a hands-On type, and prefers to leave the calculating to nerds and terds.
I remember doing civil engineering classes called System Dynamics, some 15+ years ago.
Part of the course was Dynamo training, the oldest simulation software.
As a civil engineer you'd have to be familiar with Dynamo, given that Great Britain was the first old world country to introduce it in the Euro zone.
Although Dynamo already was quite outdated when i got to know it, it didn't/doesn't do graphic displays, it did help me to understand the dynamic thermal behaviour of heatsinks better.
www.vensim.com/sdmail/sdsoft.html
Surprise:
an ideal heatsink (one that halves thermal resistance with 2 times the size and mass) takes exactly the same time to reach 1/2 the temperature difference with regard to ambient than a heatsink that has a thermal resistance which is twice as high.
Which isn't that difficult to understand, as in the end state both heatsinks have the same amount of stored heat with regard to ambient temperature and both have the same heat flow rate to ambient.
(of course this is not the case if the mass/heat capacity of the heatsink is low compared to the mass of the heat dissipating device)
Stuffing a thermal differential equation in a simple software package is easy.
In real mode, a heatsink that has 1/2 the thermal resistance has more than twice the mass.
This results in the bigger heatsink having a larger amount of stored energy when it reaches the steady state end temperature than the smaller heatsink.
So the bigger heatsink takes longer to reach the end state temperature.
There is a second way if you're that curious: you could do the measuring yourself and/or stuff a dozen command lines in a simulation software package.
Let's not interrupt your tête à tête, i'm just a terd passing by.
I've got the quote back and the DFM is fine but I still have to call Advanced today for the exact pricing on 200 boards, they only gave me vague pricing on boards at the 50 board quantity and 4 week delivery.
Since there are several of us that appreciate the look of seeing all of the gold plating what I am going to do is order 50 boards that have no solder mask at all but have gold plated traces and white silkscreen and the rest with gold traces and full soldermask... for those that still can't solder. These will be scheduled for 4 week delivery or what ever the cheapest lead time is when I call. I am pretty sure they can just pull 50 boards and skip the soldermask step and go directly to silkscreen. I will post the prices and info later today or first thing tommrrow.
Once this is posted I will give payment instructions on the WIKI and on this thread so you can make payments, it will be the same as the KSA-50 G.B.
I am only going to accept PAYPAL on this G.B. so if you are not set up with PAYPAL you may want to do so or ask a friend. If you are in the U.S. I will accept all other forms of payment as well however my Credit Union does not accept out of country checks of any sort. Since my Credit Union has been soo nice to me over the years I have no intention of opening a bank account just to satisfy a few out of country bank checks. If there is a volunteer out there that would be willing to launder the funds from those out of country individuals with real bank accounts and then forword that payment to me via PayPal would also be ok. Perhaps I can get LuckyLyndy to do this……?
Mark
Since there are several of us that appreciate the look of seeing all of the gold plating what I am going to do is order 50 boards that have no solder mask at all but have gold plated traces and white silkscreen and the rest with gold traces and full soldermask... for those that still can't solder. These will be scheduled for 4 week delivery or what ever the cheapest lead time is when I call. I am pretty sure they can just pull 50 boards and skip the soldermask step and go directly to silkscreen. I will post the prices and info later today or first thing tommrrow.
Once this is posted I will give payment instructions on the WIKI and on this thread so you can make payments, it will be the same as the KSA-50 G.B.
I am only going to accept PAYPAL on this G.B. so if you are not set up with PAYPAL you may want to do so or ask a friend. If you are in the U.S. I will accept all other forms of payment as well however my Credit Union does not accept out of country checks of any sort. Since my Credit Union has been soo nice to me over the years I have no intention of opening a bank account just to satisfy a few out of country bank checks. If there is a volunteer out there that would be willing to launder the funds from those out of country individuals with real bank accounts and then forword that payment to me via PayPal would also be ok. Perhaps I can get LuckyLyndy to do this……?
Mark
AndrewT
I did the initial listening test with a transistor pre-amp that was fired up cold with the amp. The pre-amp has now been on one and half days and listening to the amp this afternoon I could detect a further improvement in speed and low end extention. I now suspect that the cold-start of the pre-amp could have had a impact on the overall sound.
It is summer in Cape Town and I was very surprised how hot the large pcb driver heatsinks became. My gut feeling is that if the case is not a constraint nothing can be lost by using larger heatsinks.
For the moment I intend to leave the PSU cap out.
Jozua
I did the initial listening test with a transistor pre-amp that was fired up cold with the amp. The pre-amp has now been on one and half days and listening to the amp this afternoon I could detect a further improvement in speed and low end extention. I now suspect that the cold-start of the pre-amp could have had a impact on the overall sound.
It is summer in Cape Town and I was very surprised how hot the large pcb driver heatsinks became. My gut feeling is that if the case is not a constraint nothing can be lost by using larger heatsinks.
For the moment I intend to leave the PSU cap out.
Jozua
