Hi Joey,
two single pole switches would do for the hi current rectifiers. Four double pole is a waste.
There are four options for amp switching that occur to me.
1. all off
2. front end on, Hi current off
3. front end on, Hi current end switched to low bias.
4. all on.
Option 3 can be used with Jan's boards and with a little piggybacking can be made to work on All's as well.
two single pole switches would do for the hi current rectifiers. Four double pole is a waste.
There are four options for amp switching that occur to me.
1. all off
2. front end on, Hi current off
3. front end on, Hi current end switched to low bias.
4. all on.
Option 3 can be used with Jan's boards and with a little piggybacking can be made to work on All's as well.
Bratislav said:
A train is as fast as the thing that is pulling it. To make the train run smoothly or go up the hill without loosing speed, the locomotive needs to be able to go much faster and accelerate much faster.
That has no correllation to steady state conditions.
As long as there is a steady state condition, i see a class A output stage as a kind of regulator, setting the voltage constant for the front end.
Like a big mountain stream supplying water to a lake, the lake is very large, through an opening there is a constant flow of water out of the lake .
The mountain is the transformer, the lake the capacitor, the opening is the output device.
All very well if nothing changes, untill autumn hits.
A class A amplifier may have lesser need for regulators, nonetheless a regulated voltage supplied front end of a class AB amplifier functions more accurate.
I am sure others like Per Anders and my countryman Janneman Didden can vouch how a regulator that is not up to the task can take the easyness and life out.
Like in the 80s, i fumbled around with regulator IC's, to notice that it isn't always an improvement. Just sticking a lifted 78/79**
regulator on a rail was no garanty.
The KSA50 is from the early 80s, a time when it was common not to use regulation. In the US the Levinson ML2s were the team to beat in those days, a total regulated design.
If you are talking dollars and cents, my money is on the KSA50/KSA100.
I remember the headache i got viewing the pricetag of ML2s seeing them at an audio show stand.
For a diy person the added investment is often more time than cash. Building something like ML2s would not be that much more expensive, but a load of work.
How an updated DIY ML2 would do against a modernised DIY KSA50, that would be interesting.
I just tried the KSA as a headphone amplifier, just one channel and not the output stage.
Who would opt for regulated rails for such use ?
awpagan said:
Here's a useful tool for modeling power supply performance:
http://www.duncanamps.com/psud2/index.html
shocking to see the size of current spikes through the bridge, shows how careful you have to be to keep them away from the amplifier.
C- R - C - R - C looks promising, with R around 0.1 ohm and C between 10mF and 60mF.
As for regulating the front end, how much current does it need? Maybe a diode, small series R and a lot of C per rail could be used to isolate it from the main rails?
tried to do an ascii schematic here but couldn't get it to format correctly in preview, sorry
jacco vermeulen said:
Can you make it a bit simpler please, my head hurts
😀
Don't be afraid to use engineering terms, you are among friends here. I may be HiFi rookie but have 20+ years experience as electronics engineer.
Why do you need speed in a regulator when current draw from the amp is as constant as it gets (after the initial surge) ?
Bratislav
PS if Jan Didden or Per Anders or whoever wants to chime in, please do. Just don't use colorful language - I'm a stupid engineer 🙁
PPS I've seen LM317 in some rather demanding appliactions (as a regulator of a Peltier cooler for an astronomical CCD camera). It didn't do too bad - augmented with a T03 it gave ruler flat regulation at up to 10A. Not too shabby in my books.
jacco
I just tried the KSA as a headphone amplifier, just one channel and not the output stage.
Who would opt for regulated rails for such use ?
mono, dual mono?
allan
I just tried the KSA as a headphone amplifier, just one channel and not the output stage.
Who would opt for regulated rails for such use ?
mono, dual mono?
allan
Bratislav
Why do you need speed in a regulator when current draw from the amp is as constant as it gets (after the initial surge)
?
If current is set, what is the gain of massive caps?
just to filter ripple or provide extra current when needed?
bremen nacht
seen duncanamps very usefull
CRCRC or passive or reg'd
everyone has there personal preference.
If i reg'd front end i would stll need big caps on output.
defeats the idea of not having to get the large caps?
anyone tried the different types of psu on the same ksa50?
allan
Why do you need speed in a regulator when current draw from the amp is as constant as it gets (after the initial surge)
?
If current is set, what is the gain of massive caps?
just to filter ripple or provide extra current when needed?
bremen nacht
seen duncanamps very usefull
CRCRC or passive or reg'd
everyone has there personal preference.
If i reg'd front end i would stll need big caps on output.
defeats the idea of not having to get the large caps?
anyone tried the different types of psu on the same ksa50?
allan
Bratislav said:
I had the impression you are a lot more than that.
An EE engineer i am not, and others may find something worth reading without fancy electronic talk.
I consider myself a hick who finger read through 445 pages of Dhe Self-talk and Dhe others, who has been slapped with a textbook by NP himself long time ago for overcomplicating crimes.
Like Bob i've done the A75, done the Thagard A100, and noticed the difference regulation makes.
Currently i am finishing the modified A40, adding regulation and 50VA toroids for the front end of each channel.
Regulators need speed, Walt Jung, Jan didden and Per Anders explain that far better than i can. Why bother reading a quote if a thread search can hand over the original words.
The simplicity of the kind of amplifiers NP designs make it a lot harder to implement voltage regulators, probably much harder than the investment is worth.
Therefore the arguement that it is of no use because Nelson Pass does not have it is not valid.
Again just my thoughts, i am not standing on a box in the park.
Not sure if Per Anders has it mentioned on his regulator pages, the error circuit in LM***s is no better than that of the 741.
Just a standard LM3** regulator circuit may not be an improvement, no matter the capacitor size that is in front of it.
jacco
The simplicity of the kind of amplifiers NP designs make it a lot harder to implement voltage regulators, probably much harder than the investment is worth.
farnel aus$ prices CAPACITOR, 68000UF 63V; $176.01 each need 4 for dual mono
regulated expensive?
allan
The simplicity of the kind of amplifiers NP designs make it a lot harder to implement voltage regulators, probably much harder than the investment is worth.
farnel aus$ prices CAPACITOR, 68000UF 63V; $176.01 each need 4 for dual mono
regulated expensive?
allan
A commercial designer needs to get his parts straight from the distributor for ease of production. Both waiting for a cheap deal or carrying the risk and costs of having a full storage is bad business. JIT, Demming and Duran stuff.
For commercial reasons he has to make a compromise between added benefits and added costs.
A diy person can buy cheap through saving parts, as it is a OOAK project. With lower parts costs and without production cost multiplication factors the choice for going bigger is easier.
Single ended class A amplifiers are as inefficient as can be, the bias is much higher than needed for pushpull operation for the same output. The output stage needs those big cans, needs CRC or CLC too or the chassis would need to be big enough to house capacitors the size of submarine battery cells.
Those capacitor cans are needed, with or without regulation.
Unless you go fully regulated, but for a commercial enterprise the added costs of parts and labour would be so much higher than just stuffing a chassis with big caps and coils.
NP might very well put himself out of business going up that alley, and a businessman he is. There are no fully regulated amplifiers on the market, they would costs too much.
e.g. , the regulation of the ML2 was equivalent to a capacitor size of 1.000.000 uF for each channel. It did 25 watts, the pricetag had 4 zeros in the mid 80s.
Aluminium is cheap down under, electrolytics are not.
68.000/63V do AUS60 in the shops overhere. Wim picked up his KSA boards some time ago, and made me happy by selling me a tray of his 63V Rifa's. For like AUS20 in an equivalent 68.000uF size. A Scandinavian i am not but the Rifas are as good as you can get.
With low prices for heatsinks in Australia constructing fully regulated might be cheaper. But the work, parts and boards involved is the same as building 2 amplifiers, better make that 3.
To top it off, you risk waking up with Mr Greg Ball attacking you with an axe, Nightmare on Ball street.
For commercial reasons he has to make a compromise between added benefits and added costs.
A diy person can buy cheap through saving parts, as it is a OOAK project. With lower parts costs and without production cost multiplication factors the choice for going bigger is easier.
Single ended class A amplifiers are as inefficient as can be, the bias is much higher than needed for pushpull operation for the same output. The output stage needs those big cans, needs CRC or CLC too or the chassis would need to be big enough to house capacitors the size of submarine battery cells.
Those capacitor cans are needed, with or without regulation.
Unless you go fully regulated, but for a commercial enterprise the added costs of parts and labour would be so much higher than just stuffing a chassis with big caps and coils.
NP might very well put himself out of business going up that alley, and a businessman he is. There are no fully regulated amplifiers on the market, they would costs too much.
e.g. , the regulation of the ML2 was equivalent to a capacitor size of 1.000.000 uF for each channel. It did 25 watts, the pricetag had 4 zeros in the mid 80s.
Aluminium is cheap down under, electrolytics are not.
68.000/63V do AUS60 in the shops overhere. Wim picked up his KSA boards some time ago, and made me happy by selling me a tray of his 63V Rifa's. For like AUS20 in an equivalent 68.000uF size. A Scandinavian i am not but the Rifas are as good as you can get.
With low prices for heatsinks in Australia constructing fully regulated might be cheaper. But the work, parts and boards involved is the same as building 2 amplifiers, better make that 3.
To top it off, you risk waking up with Mr Greg Ball attacking you with an axe, Nightmare on Ball street.
awpagan said:
I get my 50V 68,000uF caps from Steve at APEX.JR for $6 (six bucks!) each...
I buy 10 at a time...
ripple...
Here's my 3c worth...
I constructed a CLC filter section for an aleph drawing 2.5 amps, each C was 68k, the L was 2.7mH. The ripple on the back end was simulated to be <5mv, and measured to be 5-10mv. The current consumption of the Krell is largely constant, unless you push it into class B, when it does exactly what you would expect...
You really don't need great PSRR when the ripple is already in the single digit millivolts, but it definitely can't hurt, and if you decide to regulate it's a great place to start from...
Stuart
Here's my 3c worth...
I constructed a CLC filter section for an aleph drawing 2.5 amps, each C was 68k, the L was 2.7mH. The ripple on the back end was simulated to be <5mv, and measured to be 5-10mv. The current consumption of the Krell is largely constant, unless you push it into class B, when it does exactly what you would expect...
You really don't need great PSRR when the ripple is already in the single digit millivolts, but it definitely can't hurt, and if you decide to regulate it's a great place to start from...
Stuart
rabstg said:
I have wondered the logic of sets of ten filter caps for audio. I have yet to build an amp that needed 10 filter caps. For most of my amps I used either 6 or 8 caps. What kind of PSU setup uses 10?
By the way, twice I have bought 10 caps in a set from ebay. Now I have two pairs of caps that are different values.
Blessings, Terry
still4given said:
I BUY 10 at a time, I don't use ten in a single project...
I just figure I may as well have spares around for other projects..
Thinking about heatsinks,maybee it´s been discussed before.If I found heatsinks with c/w 0,25 will I manage whitout a fan if i squeze 45-50w out of the amp?And if I have a heatsink with c/w 0.5 and a 120mm fan at lowest rpm so i wont hear it per chanel,will this work?
Ryssen
Du har säkert snubblat över annonsen innan men jag har 2st av dessa leksakerna över i Göteborg.
Hör av dig om du är intresserad. (finns med samma nick på hififorum med)
Sorry for the swedishcook thing guys 😀
To keep it short:
I just told him to write a line or to if he needed those in the link.
Du har säkert snubblat över annonsen innan men jag har 2st av dessa leksakerna över i Göteborg.
Hör av dig om du är intresserad. (finns med samma nick på hififorum med)
Sorry for the swedishcook thing guys 😀
To keep it short:
I just told him to write a line or to if he needed those in the link.
Jacco
No, they aren't. 🙂
I got ahold of them some months ago but I havn't got any use for them myself now when I found some other sinks so...
No, they aren't. 🙂
I got ahold of them some months ago but I havn't got any use for them myself now when I found some other sinks so...
Im thinking CLCLC
C=1000uF
L=1mH
http://www.jaycar.co.nz/productView...&pageNumber=&priceMin=&priceMax=&SUBCATID=770
these things are great in cd player supplies. And Ive had 2 experiences with regulated supplies where passive was better, both in JLH and gainclone.
seeya ab
C=1000uF
L=1mH
http://www.jaycar.co.nz/productView...&pageNumber=&priceMin=&priceMax=&SUBCATID=770
these things are great in cd player supplies. And Ive had 2 experiences with regulated supplies where passive was better, both in JLH and gainclone.
seeya ab