So I fixed a few issues with the board.
I wound myself a spring probe, started testing
I was running 40mA through the ccs with a 20mA load. ( which is about the plan for where I will run it)
Started using the probe. Then TICK and gone, my probe slipped. the smoke was released.
I did get a quick look at the output, but unable to capture. The same level of noise was there, but somewhat less 50 or 100Hz based. A bit more random.
I will see if I can fix it. I only have a bunch of irf840 and some 2sk369, having said that I did order more parts yesterday so I will have complete sets.
Does the IDSS of the jfet have an effect on stability/performance. The one I put in was 9mA for an lsk170, the 2sk369s I have are9 to 15mA.
Regards
Anthony
I wound myself a spring probe, started testing
I was running 40mA through the ccs with a 20mA load. ( which is about the plan for where I will run it)
Started using the probe. Then TICK and gone, my probe slipped. the smoke was released.
I did get a quick look at the output, but unable to capture. The same level of noise was there, but somewhat less 50 or 100Hz based. A bit more random.
I will see if I can fix it. I only have a bunch of irf840 and some 2sk369, having said that I did order more parts yesterday so I will have complete sets.
Does the IDSS of the jfet have an effect on stability/performance. The one I put in was 9mA for an lsk170, the 2sk369s I have are9 to 15mA.
Regards
Anthony
So I got my hands on some other shunt regs that were at a friends place and put them on the scope.
After all the guests from our children’s birthday party left I did some testing. Now I only had one beer, but that was hours ago.
I am now feeling that I may be chasing my tail a little here. I was just going to measure the output of each unit, but I also wanted to see what the difference was between the 2 probes. The sprung one and the longer ground probe.
I first hooked up the standard probe, and the usual noise on the trace. I then hooked up the spring probe and the first trace cleaned up quit allot and reduced a little in amplitude. But the big thing I noticed was both traces were not identical nor were they mirrored. They were both bobbling about at a similar frequency, but both appeared to be independent of each other. Seems to me that this instability that I have been chasing down. Could be a figment of my oscilloscopes imagination.
I have arranged for another crt scope and different probes to test this theory.
After all the guests from our children’s birthday party left I did some testing. Now I only had one beer, but that was hours ago.
I am now feeling that I may be chasing my tail a little here. I was just going to measure the output of each unit, but I also wanted to see what the difference was between the 2 probes. The sprung one and the longer ground probe.
I first hooked up the standard probe, and the usual noise on the trace. I then hooked up the spring probe and the first trace cleaned up quit allot and reduced a little in amplitude. But the big thing I noticed was both traces were not identical nor were they mirrored. They were both bobbling about at a similar frequency, but both appeared to be independent of each other. Seems to me that this instability that I have been chasing down. Could be a figment of my oscilloscopes imagination.
I have arranged for another crt scope and different probes to test this theory.
Oscillation in regs like that is a sinewave resembling waveform, usually few hundreds of mV at few MHz. For LF nuisances like ripple residual or periodic thermal drift even a DMM in AC voltage measurement mode can help compare. But there are also external things that can plague a measurement loop. Common mode noise is easily invasive interference when a high impedance is probing a low impedance. Some SMPS bricks especially when hooked up in vicinity can introduce a lot. It shows up as RFI noise.
Ok so I had a mate rock up with a crt scope.
What we did and what we found.
The flouro lights are injecting noise into the scopes. Not injecting noise into the phono preamp and then into to trace. Verified by using 2 scopes and switching lights in and out. Then turning preamp off.
My cheap scope is just generating its own noise. Not picking anything near as much up on the crt scope, but getting 6 to 15mV on the digital scope. I will go through it’s instructions and see if it has any noise canceling options.
The digital scope is injecting noise into the crt scope. When you turn the dso off, the noise on the crt goes away.
In a way I am relieved that there does not appear to be a noise issue, but concerned that my test equipment is ******* into the wind.
I still have a low frequency stability issue with my phono preamp, but it will be hard for me to trace and prove or disprove.
I am thinking of going with the sshv2 now anyway so it could be a magic bullet
I am also going to fiddle with my riaa high pass filter. At 14hz for 3db. Might go to 20 or 25
What we did and what we found.
The flouro lights are injecting noise into the scopes. Not injecting noise into the phono preamp and then into to trace. Verified by using 2 scopes and switching lights in and out. Then turning preamp off.
My cheap scope is just generating its own noise. Not picking anything near as much up on the crt scope, but getting 6 to 15mV on the digital scope. I will go through it’s instructions and see if it has any noise canceling options.
The digital scope is injecting noise into the crt scope. When you turn the dso off, the noise on the crt goes away.
In a way I am relieved that there does not appear to be a noise issue, but concerned that my test equipment is ******* into the wind.
I still have a low frequency stability issue with my phono preamp, but it will be hard for me to trace and prove or disprove.
I am thinking of going with the sshv2 now anyway so it could be a magic bullet
I am also going to fiddle with my riaa high pass filter. At 14hz for 3db. Might go to 20 or 25
Maybe it has a digital filter menu to set a drastic low-pass.
If you put the probe's cable near to the screen of some digital scopes the noise tends to go up. Especially if the wire is folded or looped. Try that. So to know what to avoid if that's also an issue. Run the scope very slow set to DC coupling measuring your phono output and watch the line if moving up and down and how much. That should be revealing of VLF instability. Then you can experiment with its RIAA filter and internal couplings or PSU possible issues by watching something actually moving to improve.
If you put the probe's cable near to the screen of some digital scopes the noise tends to go up. Especially if the wire is folded or looped. Try that. So to know what to avoid if that's also an issue. Run the scope very slow set to DC coupling measuring your phono output and watch the line if moving up and down and how much. That should be revealing of VLF instability. Then you can experiment with its RIAA filter and internal couplings or PSU possible issues by watching something actually moving to improve.
So I have been looking into my couplingof my pre in anoth thread looking for advice, and I keep getting steered back to the power supply.
I was asked these questions over in the tube and valve section about coupling capacitor selection.
#8
DF96 is offline DF96 England
diyAudio Member
Join Date: May 2007
I am not familiar with the Salas type regulators, but a quick look at the circuit makes me wonder whether they are primarily intended to reduce noise and ripple rather than lower frequency/DC variations. Voltage stability seems to rely on small Early effect in the CCS BJTs. Has anyone done any tests on DC regulation or output impedance?
Have you monitored the supply rail to the cascode stage? If that is jumping up and down by more than a few mV then you will have problems.
I will say that if I can eliminate the artifact from the trace. There will still be some VLF movement.
I do have multiple preamps with multiple ccs/shunts, most have 220uF/10H/220uF as input filters. Could these be causing the issue.
I have also ordered more parts and waiting on them, but I was just trying to get my head around other mechanisms of my pre.
Is the SSHV2 just a better choice?
I was asked these questions over in the tube and valve section about coupling capacitor selection.
#8
DF96 is offline DF96 England
diyAudio Member
Join Date: May 2007
I am not familiar with the Salas type regulators, but a quick look at the circuit makes me wonder whether they are primarily intended to reduce noise and ripple rather than lower frequency/DC variations. Voltage stability seems to rely on small Early effect in the CCS BJTs. Has anyone done any tests on DC regulation or output impedance?
Have you monitored the supply rail to the cascode stage? If that is jumping up and down by more than a few mV then you will have problems.
I will say that if I can eliminate the artifact from the trace. There will still be some VLF movement.
I do have multiple preamps with multiple ccs/shunts, most have 220uF/10H/220uF as input filters. Could these be causing the issue.
I have also ordered more parts and waiting on them, but I was just trying to get my head around other mechanisms of my pre.
Is the SSHV2 just a better choice?
Experiment comes first. See if the rail on the phono moves, and if it does how it moves up and down, by running the scope very slow in time. Is it synchronized to the phono's breathing? Then you know its related to the PSU.
A. Try one PSU for all stages per channel. Some circuits don't couple well through many PSU.
B. Try adequate voltage 2W Zener instead of your 120K R9 ref resistor and open your R6 trimmer to the max. Position the Zener with its anode to ground. See if something changes before you throw away your K&Ks. It may end up with more hiss but see if the breathing stops.
That version I don't know if it does something different thermally due to some different active parts and passive values but its based on SSHV1 and its surely many times used on K&K's customers projects. It still has a Norton Vref doing I/V to a JFET's current. Its not directly based on a BJT's CCS parameter. Maybe the reference on BJT Early effect is based on the very early sketches on posts #2,3.
I have added the original SSHV1 on a high gain Canary USA tube phono back then and it did not breath.
A. Try one PSU for all stages per channel. Some circuits don't couple well through many PSU.
B. Try adequate voltage 2W Zener instead of your 120K R9 ref resistor and open your R6 trimmer to the max. Position the Zener with its anode to ground. See if something changes before you throw away your K&Ks. It may end up with more hiss but see if the breathing stops.
That version I don't know if it does something different thermally due to some different active parts and passive values but its based on SSHV1 and its surely many times used on K&K's customers projects. It still has a Norton Vref doing I/V to a JFET's current. Its not directly based on a BJT's CCS parameter. Maybe the reference on BJT Early effect is based on the very early sketches on posts #2,3.
I have added the original SSHV1 on a high gain Canary USA tube phono back then and it did not breath.
Attachments
A. still does it with 1 per channel, it is right on the limits to do so.
And the undulating/pulsing happens at the same time on the mains. We have the 2nd largest goldmine for ore processing in the world on our grid. The mine has a “heart beat” around every second or 2.
I will continue experimenting.
And the undulating/pulsing happens at the same time on the mains. We have the 2nd largest goldmine for ore processing in the world on our grid. The mine has a “heart beat” around every second or 2.
I will continue experimenting.
Hi Nick
I'm building sshv1 to try the high voltage version
I'm using it for a preamplifier which have a design of clclc. And very high rdc. From what I read it's a good idea to remove the chokes? Rdc first choke is 1k25 and 2nd is 2k5
The starting voltage is around 280v and drop till 250v at the tube.
I'm building sshv1 to try the high voltage version
I'm using it for a preamplifier which have a design of clclc. And very high rdc. From what I read it's a good idea to remove the chokes? Rdc first choke is 1k25 and 2nd is 2k5
The starting voltage is around 280v and drop till 250v at the tube.
Last edited:
We don't care about high RDC here because this is a CCS input reg which is by definition a very high impedance barrier on its own. So since you got them leave them in to eat the EMI nasty stuff. That's for rectification filtering chokes. If they are located after the regulator at the audio circuit's rail, skip them.
Hello all,
Made a BJT version SSHV for my Dynaco PAS3 phono + pre with great result, in comparison to conventional CRCRC PSU. But could not proceed to the Jfet version due non availability of 2sk170 locally. Can I use J113 Jfet in place of 2sk170? what component values need to be adjusted, if any.
Presently planning SSHV2 with J113 (if possible) for 20mA Phono load at 210V B+. 40mA CCS load. SSHV input voltage = 250V.
Regards,
Bhaskar
Made a BJT version SSHV for my Dynaco PAS3 phono + pre with great result, in comparison to conventional CRCRC PSU. But could not proceed to the Jfet version due non availability of 2sk170 locally. Can I use J113 Jfet in place of 2sk170? what component values need to be adjusted, if any.
Presently planning SSHV2 with J113 (if possible) for 20mA Phono load at 210V B+. 40mA CCS load. SSHV input voltage = 250V.
Regards,
Bhaskar