I'm well aware of the WJ variant - it improves the ripple rejection from -150dB to -160dB at low frequencies, however both figures are very close to measurable limits and certainly not audible in this use case. The modified variant only has 3 extra components but one of these is another fairly large capacitor - I may go down this route later but my primary area of research currently is using an alternative for M1 as it isn't dissipating nearly so much energy and there are better option.
The real story here isn't about the CCS anyway - that's only being used to set the operating point of the zeners - it's all about regulator noise.
The real story here isn't about the CCS anyway - that's only being used to set the operating point of the zeners - it's all about regulator noise.
Hi Nick... Can you possibily sell me one completed and tested regulator?
I've screw-up all ten pcs. of my DN2540 and am wondering if they're actually fake or imitation stuff!? Why is it so fragile? Beleives there must be a rule of handling. How do you test it out before connecting to a high voltage supply?
Only success I've was using a single DN2540 to drive a pair of OD3.
What I really need is a regulated and stable 300vdc @ 50mA.
Thanks!
Zekk,
just saw your post now. I fried some of mine DN2540 because the soldering iron's tip wasn't properly grounded, and the ESD killed the chips. Now, every time I deal with SS stuff, I do check if the tip is grounded
Best, Erik
regarding the WJ variant, with the additional resistors + cap.
Will it also do higher currents in comparison to the standard cascode? I do think so, but want to confirm.
With the standard cascode, when setting higher currents, the VGS of the upper element (actually of both elements) will become smaller and smaller. As VGS of upper element = VDS of lower element, the lower element does lose all of its VDS. With the WJ trick, even with a closer to saturation upper element (lower VGS), one can still give the lower element some more VDS to work properly? ?
Many thanks,
Erik
Will it also do higher currents in comparison to the standard cascode? I do think so, but want to confirm.
With the standard cascode, when setting higher currents, the VGS of the upper element (actually of both elements) will become smaller and smaller. As VGS of upper element = VDS of lower element, the lower element does lose all of its VDS. With the WJ trick, even with a closer to saturation upper element (lower VGS), one can still give the lower element some more VDS to work properly? ?
Many thanks,
Erik
Yes - the reference voltage there is set by the CCS controlling the current through R4.
Note that the MJSR voltage is only adjustable in deltas of the chosen zener voltage, i.e. a typical granularity of 5.6V (*), but there are ways round that for such a low-current voltage reference - the CCS in the 01a G2 is only set to about 500uA/0.5mA.
(*) with 21 x 5V6 zeners under typical operating conditions you should get about 120.5V of regulated output.
FWIW, initial feedback from knowledgeable beta testers I've been using on another forum (mentioned above) has been extremely positive...
Note that the MJSR voltage is only adjustable in deltas of the chosen zener voltage, i.e. a typical granularity of 5.6V (*), but there are ways round that for such a low-current voltage reference - the CCS in the 01a G2 is only set to about 500uA/0.5mA.
(*) with 21 x 5V6 zeners under typical operating conditions you should get about 120.5V of regulated output.
FWIW, initial feedback from knowledgeable beta testers I've been using on another forum (mentioned above) has been extremely positive...
Last edited:
Ah - you have your eye on the 01A Gen 2 as well!
That's something I'm building, having heard Ale's and been mighty impressed.
That's something I'm building, having heard Ale's and been mighty impressed.
Starting with 6v6 and will progress to 301a, 10y, 801a, and maybe 45. Gonna use Ale's hybrid mu load on all in an attempt at a fair comparison. This LED referemce should make sure that noise is kept at bay, at least in the load components.
I think that will be highly dependent on the circuit it's attached to. The main raison d'etre is low noise, and that it indeed does achieve. The use of a Kelvin connection cap is icing on the cake.
Sy, How would this differ from a bypass cap across the diodes?
Sorry for the bother, but is it a foolish question to ask if LED's would outperform the zeners in terms of noise? I apologize as I do not have the article and am unaware if he mentions this.
Sy, How would this differ from a bypass cap across the diodes?
The Kelvin connection cap essentially eliminates the effect of lead inductances.
Re: noise, Morgan compared the noise of the regulator with LEDs and Zeners and demonstrated that the former were quieter.
I will have to look up an example of the Kelvin connection. It is new to me.
I had laid out the board for LED's when I realized it was a mistake. I just assumed that an LED would be quieter than a zener based on previously shared information here on the forum.
I had laid out the board for LED's when I realized it was a mistake. I just assumed that an LED would be quieter than a zener based on previously shared information here on the forum.
the noise is a little scary
Hello,
Has anyone built one of these?
The composite Zener chain is like one big long voltage reference?
I like the potential PSRR, the noise is a little scary.
DT
Hello,
Has anyone built one of these?
The composite Zener chain is like one big long voltage reference?
I like the potential PSRR, the noise is a little scary.
DT
Member
Joined 2009
Paid Member
- Home
- Amplifiers
- Tubes / Valves
- Statistical regulator from VA 4rd edition