I'm using the reg for this lineamp schematic (Rp 15K, Rk100R, +Vs 180 & without 80K optional) but I have a doubt: the +Vs 180 is before Rp as specified in the schematic or is after Rp? because if it's after Rp the anode have only 40V, it's right?
I have these questions because the reg without Rp can reach 180V directly to the anode.
An externally hosted image should be here but it was not working when we last tested it.
I have these questions because the reg without Rp can reach 180V directly to the anode.
Last edited:
Bypass c2 with 0.022-0.047u Teflon if you got any proper in VDC around. Can you hear more detail in music?
Problem remains
Hi,
I have successfully implemented your cascoded version of the regulator. Everything was working fine for few hours, but when I hit the power off switch the leds and resistor around leds blew again. It is weird, because before this problem the regulator was working fine for two months. Clamping zeners were installed.
Any clue what should I be looking for?
Thank you.
Hi,
I have successfully implemented your cascoded version of the regulator. Everything was working fine for few hours, but when I hit the power off switch the leds and resistor around leds blew again. It is weird, because before this problem the regulator was working fine for two months. Clamping zeners were installed.
Any clue what should I be looking for?
Thank you.
When operated Vin is 350V, without load Vin is 370V,so no overvoltage. No spark on mains switch, although I do not have suppression capacitor installed - it could be that. The load is aikido amplifier. I forgot to mention, that when I switch the power off I switch 2k 5W bleeder resistor in to drain filter caps before regulator and this bleeder is connected from last C of CRC before regulator (PSU is CRC-regulator) - this could be the problem? But it worked fine before.
If it is the overvoltage spike created by the switch that is killing the reg then the spike must be very big in terms of voltage, right?
I am trying to guess, because especially when going off, things are supposed to be milder than when going on. The switching to bleeder is something not commonly practised. Lets solder the bleeders permanently first. No need to bleed more than 1-2mA so they can be there all the time as in most. If it will be blowing the leds still, lets take away the RC parallel to them. Did you notice any subjective difference with the cascoded by the way?
Well I have quite big capacitance in preamp PSU: 470uF - 270R - 1000uF and it would take very long time do discharge bleeding just 1-2mA and in that time the valves are already cold and they would have still high voltage on them, so I discharge quickly with 2k bleeder. But with bleeder resistor permanently in psu, the sounds is noticeably worse!.
Whats the purpose of RC parallel to leds? To shunt noise? Why would be the problem with RC?
I have the same regulator installed in aikido mic preamp and in another aikido preamp besides the one which is blowing now and the others are working properly (the old not cascoded version).
As for the new cascoded version I cant compare much, because in cascoded version I had 10mA less left for shunt then before and increased gate stoppers from 220R to 330R. But I would guess the difference in sound was very little if any.
Must say I have your low voltage (version 1) of regulator in my headphone amplifier running for almost a half year on daily basis without any fault and sound superb!
If anything I can do to increase the safety of high voltage regulator even when sacrificing little bit of performance I would do it.
This is the second weird problem. First one was the exactly same but during the powering on. The leds and RC resistor burned.
Whats the purpose of RC parallel to leds? To shunt noise? Why would be the problem with RC?
I have the same regulator installed in aikido mic preamp and in another aikido preamp besides the one which is blowing now and the others are working properly (the old not cascoded version).
As for the new cascoded version I cant compare much, because in cascoded version I had 10mA less left for shunt then before and increased gate stoppers from 220R to 330R. But I would guess the difference in sound was very little if any.
Must say I have your low voltage (version 1) of regulator in my headphone amplifier running for almost a half year on daily basis without any fault and sound superb!
If anything I can do to increase the safety of high voltage regulator even when sacrificing little bit of performance I would do it.
This is the second weird problem. First one was the exactly same but during the powering on. The leds and RC resistor burned.
1000uF! That can give quite some Joules. Go low there, you don't need that with the reg. Very possible it does stuff when switching things on it. For cascode it does not change much in this circuit with those BJTs, so your impression ties. Still good for more Vomax range.
Hi,
I am looking at building an over 300V version the reg, with up to 370V input, which should be able to provide output voltage up to 340-350V. It will power an Aikido line stage on octal tubes. The expected power consumption is 15-25 mA per channel. Eventually it is going to be a separate regulator for each channel, although initially I may try one for both channels.
So I am looking for higher voltage alternatives for MJE350 in a schematic from post #967, or perhaps a cascoded version from post #1750.
What are the benefits/drawbacks of the cascoded vs non-cascoded schematic (apart from the voltage limitations)?
What parameters to look for when choosing a higher voltage substitute for MJE350? Vceo is obvious, but what else? A search for PNP transistors with Vceo at least 375V at Farnell shows a few possible choices available "locally":
MJE5731AG (http://www.onsemi.com/pub/Collateral/MJE5730-D.PDF)
MJE5852G (http://www.onsemi.com/pub/Collateral/MJE5850-D.PDF)
STX93003 (http://www.st.com/stonline/products/literature/ds/9077/stx93003.pdf)
ZTX558 (http://www.zetex.com/3.0/pdf/ZTX558.pdf)
ZTX758 (http://www.farnell.com/datasheets/73105.pdf)
The last three are small TO-92 or similar packages, so could there be issues with heat dissipation?
There are also a few SMD devices available, some of which seem to be analogous to the above STX and ZTX devices, and some are only available as SMD, such as PBHV9540Z (http://www.nxp.com/documents/data_sheet/PBHV9540Z.pdf) and PBHV9050Z (http://www.nxp.com/documents/data_sheet/PBHV9050Z.pdf)
How do I pick the right one?
I am looking at building an over 300V version the reg, with up to 370V input, which should be able to provide output voltage up to 340-350V. It will power an Aikido line stage on octal tubes. The expected power consumption is 15-25 mA per channel. Eventually it is going to be a separate regulator for each channel, although initially I may try one for both channels.
So I am looking for higher voltage alternatives for MJE350 in a schematic from post #967, or perhaps a cascoded version from post #1750.
What are the benefits/drawbacks of the cascoded vs non-cascoded schematic (apart from the voltage limitations)?
What parameters to look for when choosing a higher voltage substitute for MJE350? Vceo is obvious, but what else? A search for PNP transistors with Vceo at least 375V at Farnell shows a few possible choices available "locally":
MJE5731AG (http://www.onsemi.com/pub/Collateral/MJE5730-D.PDF)
MJE5852G (http://www.onsemi.com/pub/Collateral/MJE5850-D.PDF)
STX93003 (http://www.st.com/stonline/products/literature/ds/9077/stx93003.pdf)
ZTX558 (http://www.zetex.com/3.0/pdf/ZTX558.pdf)
ZTX758 (http://www.farnell.com/datasheets/73105.pdf)
The last three are small TO-92 or similar packages, so could there be issues with heat dissipation?
There are also a few SMD devices available, some of which seem to be analogous to the above STX and ZTX devices, and some are only available as SMD, such as PBHV9540Z (http://www.nxp.com/documents/data_sheet/PBHV9540Z.pdf) and PBHV9050Z (http://www.nxp.com/documents/data_sheet/PBHV9050Z.pdf)
How do I pick the right one?
From the above list the first one has been applied successfully by a member already (MJE5731AG). That, and the cascoded MJE350. No significant differences in OLG. See to scale the Vref resitor(s) dissipation properly. The CCS as a whole can be substituted by a DN2540 TO-220 Supertex depletion also. Don't use heavy CRC/CLC pre-filtering at such voltages and see that your bridge diodes have the proper PIV.
Thanks for the suggestions, Salas. I was also thinking of using MJE5731AG, although I will probably also order a couple of STX93003 just to try them out.
I want to try this first with a CRC filter using a couple of 40uF MKP1848 film caps. We'll see if that is enough. 150uF electrolytes are standing by
I have put together the CRC filter and CCS on IRF9620 to test the actual supply voltages under load. The Vin is 362V at 40mA using a 6.9k resistor load for the CCS.
I used some low current LEDs with nominal If=2mA, so R3 is 2x 68k resistors in series, which makes the LED current around 2.6mA. That takes care of excessive power dissipation nicely. I've got a couple of DN2540's lying around, will probably try them out too once I get the whole thing working.
I want to try this first with a CRC filter using a couple of 40uF MKP1848 film caps. We'll see if that is enough. 150uF electrolytes are standing by
I have put together the CRC filter and CCS on IRF9620 to test the actual supply voltages under load. The Vin is 362V at 40mA using a 6.9k resistor load for the CCS.
I used some low current LEDs with nominal If=2mA, so R3 is 2x 68k resistors in series, which makes the LED current around 2.6mA. That takes care of excessive power dissipation nicely. I've got a couple of DN2540's lying around, will probably try them out too once I get the whole thing working.