Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › A DIY electrostatic amp for intermediate DIYers?
New Posts  All Forums:Forum Nav:

A DIY electrostatic amp for intermediate DIYers? - Page 9

post #121 of 593
2/5/09 at 8:39pm
Hey, Pabbi1, stop giving away the secrets.
Reply
post #122 of 593
2/5/09 at 9:51pm
Stealing your thunder - just wondering when someone was gonna cue you with 'Where is my psu'? Always saving the best for last. Of course, anyone would have figured it out, but, since I use mine daily / nightly, just thought you deserved a shameless plug - the psu is ROCK solid.
Reply
post #123 of 593
2/6/09 at 1:40am
Pabbi1, your wish is my command.

Here's a PS that should work. Maybe needs a bit of refinement. Those are 400V complementary mosfets. They will need heatsinking depending on O/P voltage and transformer voltage. It will be difficult to make this variable for wide range of voltages without very large heatsinks. But it should be easy to have several versions where some of the resistor values change depending on O/P voltages and a particular HV secondary is chosen.

Reply
post #124 of 593
2/6/09 at 1:47am
Aww, Pabbi1 - I thought you were gonna clue me about this {whine whine}

How do you keep up with all this! More reading for me in the AM!!
Reply
post #125 of 593
2/6/09 at 2:14am
Quote:
Originally Posted by wiatrob View Post
Aww, Pabbi1 - I thought you were gonna clue me about this {whine whine}

How do you keep up with all this! More reading for me in the AM!!
Worry not - this isn't what I am keeping you posted about...

Runeight, your work never ceases to amaze... to think that the base of a design has come within a week is, well, pretty astonishing.
Reply
post #126 of 593
2/6/09 at 3:06am
So... a PSU like that... would it scale well to supply say... the serious EL34 circuits you had on the previous pages? With higher voltage silicon I suppose.
Reply
post #127 of 593
2/6/09 at 10:24am
Quote:
Originally Posted by runeight View Post
Here's a PS that should
Alex, what's the output impedance?
Reply
post #128 of 593
2/6/09 at 11:49am
Quote:
Originally Posted by TimmyMac View Post
So... a PSU like that... would it scale well to supply say... the serious EL34 circuits you had on the previous pages? With higher voltage silicon I suppose.
In principle it scales, in practice scaling is hard. This for several reasons.

In order for the cap mulitplier and regulator to work there needs to be some voltage drop in the filter/regulator section. Let's arbitrarily say 50V. Thus, if we are regulating to 350V we need to have 400V at the first filter caps. This means, of course, that the first caps are 450V caps and this is cutting it close. But, I have done this before with no ill effects over time.

But, any higher output from the reg means higher input at the first cap. 500V at the reg means 550V at the first filter cap and this means, unless you have big 630V caps around, that we have to put caps in series with voltage equalizing resistors. Furthermore, the rest of the electrolytics will also have to be two in series. It's been done. Lots of times. But it's many more components and complexity.

The higher voltage fets are not a problem, but The higher voltages require much more careful wiring and spacing.

And, lastly, the energy storage difference between 350V and 500V is not trivial. Dump one of these caps into a piece of wire and it will vaporize. So, the sucker is dangerous. It's dangerous at 350V, but much more dangerous at 500V.

Having said all that, it can be done.
Reply
post #129 of 593
2/6/09 at 11:54am
Quote:
Originally Posted by luvdunhill View Post
Alex, what's the output impedance?
You mean the Zo of the PS?
Reply
post #130 of 593
2/6/09 at 5:31pm
It looks like the Blue Hawaii PSU might be cheaper in parts and since we already have the layout it would be very simple to have some boards made once the errors have been fixed. There is a cheaper alternative for the 2SA1968's so that's not an issue.

We should perhaps also have small +/-15v supply on the board as well so the same PSU can be used to drive pretty much any one of the ESP amps, KGSS, BH, Egmont, SRX etc.
Reply
post #131 of 593
2/6/09 at 6:39pm
Quote:
Originally Posted by spritzer View Post
It looks like the Blue Hawaii PSU might be cheaper in parts and since we already have the layout it would be very simple to have some boards made once the errors have been fixed. There is a cheaper alternative for the 2SA1968's so that's not an issue.

We should perhaps also have small +/-15v supply on the board as well so the same PSU can be used to drive pretty much any one of the ESP amps, KGSS, BH, Egmont, SRX etc.
Including the iron to run it? The Bijou psu uses a $50 Hammond (270DAX, from Angela), where I thought the BH needed a pricier ($100+) toroid - and, the Bijou power boards have been in production for almost a year, already in stock at GlassJar. The BH psu advantage is the lack of a tube.

But, your point is well taken that there are psu alternatives, pretty well known, and with proven track records.
Reply
post #132 of 593
2/6/09 at 6:59pm
Quote:
Originally Posted by holland View Post
Like nikongod, I think this can be done with some routing (and silkscreen work?) and jumpers/wires in place.

SNIP

I'll definitely be in, but it'll be a few months for me to save up enough cash to build the project and buy the ESP-950. Tough economic times.
There is something undeniably appealing to a single board with multiple configurations (flexibility? economy?).

I am looking forward to next weekend when I will listen to a bunch of 'statics and decide whether to succumb to the temptation!
Reply
post #133 of 593
2/6/09 at 7:47pm
It's just a matter of trade offs, as it always is.

The BH supply is not regulated. It uses a CCS to provide current for a zener string. The zener string sets the gate voltage for the mosfet which sets the source voltage. If the mosfet is loaded lightly the voltage will stay pretty constant. And in an electrostatic amp the dynamic current draw appears to be in uA ranges. So this supply is perfectly alright.

If you want to use this supply for more general purposes the BH supply will not provide as good regulation as one with feedback through an error amplifier (as in the one I drew).

Once again it comes down to what the need is. If we are certain that the PS is only for electrostatic amps with low dynamic current requirements or if we really don't care about super regulation then the BH supply (or variant thereof) is perfectly alright and a known quantity.

To make the BH a bit simpler I would make it a true complementary supply (easy to do) so that we can use a transformer with a CT secondary if someone wants to do that. OTOH, KG points out that NFets are cheaper then PFets.

luvdunhill, if you mean the Zo of the PS that I drew, for typical values it should be around 25mOhms at 1KHz.
Reply
post #134 of 593
2/6/09 at 8:12pm
It's always a trade off... especially given the price we have to work with. I for one would love to try your PSU on a few amps I have kicking around but they aren't limited by price in any way. This brings up the issue of having more then one design to choose from with clear ways to upgrade as not everybody wants just a cheap amp to get them going. I'm sure that people would love to be able to build the SP ES amps at a fraction of the price...

Quote:
Originally Posted by pabbi1 View Post
Including the iron to run it? The Bijou psu uses a $50 Hammond (270DAX, from Angela), where I thought the BH needed a pricier ($100+) toroid - and, the Bijou power boards have been in production for almost a year, already in stock at GlassJar. The BH psu advantage is the lack of a tube.

But, your point is well taken that there are psu alternatives, pretty well known, and with proven track records.
Well we (aka non 117v users) can't use the 270 Hammond series so that's either a 370 or something else. I'd rather try some of the R-cores from China which are cheap as dirt. The Bijou PSU is also not +/- so it won't work on a stat amp. Since it uses a tube rectifier using full wave there is very little difference to just using a non-CT transformers with two extra diodes.
Reply
post #135 of 593
2/6/09 at 8:51pm
Quote:
Originally Posted by pabbi1 View Post
Including the iron to run it? The Bijou psu uses a $50 Hammond (270DAX, from Angela), where I thought the BH needed a pricier ($100+) toroid - and, the Bijou power boards have been in production for almost a year, already in stock at GlassJar. The BH psu advantage is the lack of a tube.
The BH transformer is only more expensive because it requires the 36v winding which makes it a custom job. There are a couple companies that make suitable transformers (without the 36v winding) in the $50 ballpark. A bulk order of BH transformers would probably bring the price into the $65 range.

Anyways, on the thread of modifying/adapting the bijouish PSU to this project:
Try adding a CRC filter at the input of the PS. It will solve several problems: the R can be adjusted cheaply to dump off a few volts without stressing the regulators. The HF spikes of a big cap input supply can be reduced. Both are good things IMO.

Maybe Im prejudiced (aren't we all towards our favorite designs?), but im not a fan of making the mosfets do *all* the work of cleaning the HV lines all by themselves. Even a simple & cheap CRC network makes their life so much easier.
Reply
New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › A DIY electrostatic amp for intermediate DIYers?