They are incomparable, output impedance and performance are surely very very very poor compared to the SS version, but still yet it works as I want. And I'm happy with it.
Last edited:
That's good. There is no substitute for happiness.
Now what would a solid state 220 V DC 20 mA regulator bring you besides smaller size, lower cost, less power consumption, less heat, better specs and a device it feeds that performs better ? Not much I think ...
Now what would a solid state 220 V DC 20 mA regulator bring you besides smaller size, lower cost, less power consumption, less heat, better specs and a device it feeds that performs better ? Not much I think ...
Last edited:
Because I had designed it to give regulated power supply to a 12AX7 and 12AU7´s DC coupled preamplifier, and it will be all-tube from the rectifier itself. No sand at all. =D
If "no sand at all" is the leading criterium maybe learning to play an instrument is a better option 😀 AFAIK there is sand in the glass of the tubes.
OK, I'm interested in your design. How can I obtain it? Do you have boards or what?
I was not advertising. We did a LM723 design for our ES9023 DAC. It is in the Group Buy section. No idea if boards still can be ordered. I don't arrange the Group Buy. It is a 5 V low current power supply. Nothing special, just a reuse of old technique for use with a relatively new DAC chip.
Certainly.
In a non far future, I'll try to make a small SMPS using tubes, and thinking about a 6BN6 as voltage comparator. I have about a dozen of them.
A 6S4 as power tube, a 6X4 as a catch diode, a 12AU7 as the frequency multivibrator (cathode coupled), and the DC output from this buck tube SMPS feeding my preamp. This is only a pre-idea. No problem with cathode isolated from heaters, trafos are done by my own as I want. Frequency, about 20KHz in first attempt.
Totally off topic but why would one make a tubed SMSP on 20 kHz so in the audible range ? SMPS are often made because of efficiency etc. Using tubes with heaters for again a mediocre performing device has not much to do with efficiency. I could be wrong but I think a marriage of a good solid state regulator with a tube preamp is nothing to be ashamed of.
http://www.tubecad.com/2006/11/blog0087.htm
http://www.next-tube.com/articles/hvs/hvrEn.pdf
One of these per channel:
http://www.wescomponents.com/datasheets/LR8.pdf
http://www.tubecad.com/2006/11/blog0087.htm
http://www.next-tube.com/articles/hvs/hvrEn.pdf
One of these per channel:
http://www.wescomponents.com/datasheets/LR8.pdf
Last edited:
Again you are true, but I ma moved by the experience of doing something rare, new, and imagined from the start. I can or can't work, but still I want to experiment it.
Learn to fly.
I hope your tubed SMPs will perform as you like it. Please show a picture when it's ready.
I hope your tubed SMPs will perform as you like it. Please show a picture when it's ready.
Last edited:
Not really: the reference is a zener, which unbypassed is rather noisy.
The spec says 86µVrms for a 100Hz to 10KHz bandwidth.
The LM317 is just a bit younger, but uses a bandgap reference. In normalized equivalent conditions except bandwidth, the noise (also typical) is 15µV, but the bandwidth is one decade wider (10Hz to 10KHz).
http://www.ti.com/lit/ds/symlink/lm723.pdf
http://www.ee.buffalo.edu/courses/elab/LM117.pdf
That 723's onboard reference - it is really rather good. It's a proper buried zener, not a bandgap - and it has terminals that allow you to hang a decent reference decoupling cap off it too. Which also means you can pour a bit of current through it ... on occasion I've used them instead of LM329s for a 6.9v reference and have no reason to change. Can be very, very quiet...
By modern standards the onboard opamp is a bit duff, and you'll want to use an external pass transistor anyway. But as a cheap a way to experiment with the super-reg approach its educational.
By modern standards the onboard opamp is a bit duff, and you'll want to use an external pass transistor anyway. But as a cheap a way to experiment with the super-reg approach its educational.
AFAIK both buried zeners and band gaps are used by different manufacturers of the LM723 so trying out different makes can be beneficial. Anyhow, the measured results in the earlier posted table speak for themselves.
LM723 is less noisy than LM317 whatever one wants to debate about it or whatever ones religious beliefs are. BTW only the DIP version has the Zener leads connected to the outside.
Now where is my S2 "double crown" stamp ?
LM723 is less noisy than LM317 whatever one wants to debate about it or whatever ones religious beliefs are. BTW only the DIP version has the Zener leads connected to the outside.
Now where is my S2 "double crown" stamp ?
Last edited:
Ah, I didn't know that. I only have a tube of old 14-pin DIL 723s I picked up somewhere... for pennies
'S2 Double Crown stamp' - tee hee 😀
'S2 Double Crown stamp' - tee hee 😀
What really surprised me is that LM340 also is less noisy (in a decade wider bandwidth too) compared to LM317. Few people use LM340 but it seems to be a good reg. I have a lot of them but never used them as I thought/assumed (dangerous!) that it would not be OK for audio. Few devices use this reg so .. Assumption is the mother of f*ck ups I see.
Martin, even new LM723 (still in current production, who uses them in 2012 ?) cost nil to nothing so price/quality ratio is unbeatable. There is quite good info that can be found on the interweb on the innards of LM723 (almost all from the seventies) but it takes some time to find it.
Just finished stamping a full rail of Motorola MC723C 1985 S2 double crowns....Any takers ?
Martin, even new LM723 (still in current production, who uses them in 2012 ?) cost nil to nothing so price/quality ratio is unbeatable. There is quite good info that can be found on the interweb on the innards of LM723 (almost all from the seventies) but it takes some time to find it.
Just finished stamping a full rail of Motorola MC723C 1985 S2 double crowns....Any takers ?
Last edited:
BUt most fixed voltage regs can be quieter than teh '317, not a secret. The thing about teh 17 is that its universal, cheap, and its modest-to-really quite good performance 9which is dirt cheap) can be improved slightly to suit teh circumstance - cheaply.
Beyond that - I think noise specs, in and of themselves, are only a very small part of the whole, which possibly get over-emphasised because it is a headline spec / easy to measure. The noisiest damn thing in the circuit is always the load and its dependencies, not the regulator. It is the big picture that matters, always.
PS Jean-Paul - could you make me a DoubleCrown bumper sticker please? It will make my car faster and, well, everything better.
Beyond that - I think noise specs, in and of themselves, are only a very small part of the whole, which possibly get over-emphasised because it is a headline spec / easy to measure. The noisiest damn thing in the circuit is always the load and its dependencies, not the regulator. It is the big picture that matters, always.
PS Jean-Paul - could you make me a DoubleCrown bumper sticker please? It will make my car faster and, well, everything better.
- Status
- Not open for further replies.