Small R is usually found in series with that small C across rectification diodes. R series probably offers the same amount of rectification's EMI damping in your schematic but better use Duncan's PSUD2 to analyze transient smoothness quality between the two methods and to evaluate max current through R.
I had an SSHV2 regulator go thump in the night, and i am totally at a loss as to why. I found Q6 (2SK117GR) and Q5 (KSA1381) were both dead from the + side of the voltage regulator. Replaced both parts and everything is working again. My question is what conditions could cause a failure of those two components?
In troubleshooting always look at the last thing changed right?
I changed the main power transformer earlier that day. The regulator has run for the last few months with a 200v B+, current set to 25ma, and an output of 150v into a load of 3-4ma.
With the new setup i have 230v B+, current set to 30ma, and an output of 160v into a load of 3-4ma. Plan to eventually dial it up to 175-185v.
A small change, and all well within the limits of the components and heat dissipation. Q3 is bolted to the heat sink of a 2U Modushop enclosure, Q1 has a taller 50mm heat sink, and both Q4 & Q5 have small heat sinks attached to them. All overkill for the application.
I just need to figure out why it failed, as it ran for 4-5 hours before it failed, and only one of the two went. Any ideas?
In troubleshooting always look at the last thing changed right?
I changed the main power transformer earlier that day. The regulator has run for the last few months with a 200v B+, current set to 25ma, and an output of 150v into a load of 3-4ma.
With the new setup i have 230v B+, current set to 30ma, and an output of 160v into a load of 3-4ma. Plan to eventually dial it up to 175-185v.
A small change, and all well within the limits of the components and heat dissipation. Q3 is bolted to the heat sink of a 2U Modushop enclosure, Q1 has a taller 50mm heat sink, and both Q4 & Q5 have small heat sinks attached to them. All overkill for the application.
I just need to figure out why it failed, as it ran for 4-5 hours before it failed, and only one of the two went. Any ideas?
Solid state parts are sensitive to HV transients. The DN2540s for instance are notoriously sensitive to stray inductance that can exacerbate transients. Not tolerating fast spikes above their max VDS spec. It can happen to a degree for the rest of parts too.
The new transformer maybe has less output impedance than the one's everything was running smoothly with for months of operation. So the turn on transient maybe got stronger. Semis can be shot gradually from transient strain also, not always instantly blown.
In any case you can replace Z3 with a Schottky diode for faster tighter threshold reaction and also put a series RC like 0.1uF 400V film 10R 2W across the raw DC input connector.
The new transformer maybe has less output impedance than the one's everything was running smoothly with for months of operation. So the turn on transient maybe got stronger. Semis can be shot gradually from transient strain also, not always instantly blown.
In any case you can replace Z3 with a Schottky diode for faster tighter threshold reaction and also put a series RC like 0.1uF 400V film 10R 2W across the raw DC input connector.
When you talk about transients, do you mean the sudden application of B+ power, or the application of a load?
New transformer does have lower impedance as its larger, something like 60ohms vs 82ohms for the old. Both are have 170v secondarys, filter section is a basic 270uf-10H(250r)-270uf.
I also use a delay start for the HV supply to give the filaments a head start, so the load should be there as soon as power is applied to the regulators, therefore the B+ isn't peaking at a higher voltage waiting for the tubes to warm up. The SSHV2 board is supplying one of Ales Gyrators boards for a 01a tube.
I should have the parts for the input RC so will try that. I don't follow what you mean by Z3?
Thanks for the help!
New transformer does have lower impedance as its larger, something like 60ohms vs 82ohms for the old. Both are have 170v secondarys, filter section is a basic 270uf-10H(250r)-270uf.
I also use a delay start for the HV supply to give the filaments a head start, so the load should be there as soon as power is applied to the regulators, therefore the B+ isn't peaking at a higher voltage waiting for the tubes to warm up. The SSHV2 board is supplying one of Ales Gyrators boards for a 01a tube.
I should have the parts for the input RC so will try that. I don't follow what you mean by Z3?
Thanks for the help!
Last edited:
Hi
Better avoid oil including caps near heat sources like this reg's sink. They tend to dry faster with use than when mounted in cooler areas of the enclosure. Wima MKP 10 400V will do very well in this application because pulse rated. DC-Link capacitor types are very good here too. C1 can range from 0.27uF to 0.47uF if you can't find 0.33uF exactly.
Better avoid oil including caps near heat sources like this reg's sink. They tend to dry faster with use than when mounted in cooler areas of the enclosure. Wima MKP 10 400V will do very well in this application because pulse rated. DC-Link capacitor types are very good here too. C1 can range from 0.27uF to 0.47uF if you can't find 0.33uF exactly.
Thanks for the quick response!
I'm going to try it out in the near future, and I'm going to use it at the preamplifier. The amplifier has a 75 db gain. I have three questions for you:
1. Does the SSHV2 shunt power supply have an advantage over the conventional series power supply in terms of noise suppression?
2， The amplifier has three stages: + Power High voltage-ECC82 buffer (47uf electrolytic capacitor)-1.5k-ecc83(22uf electrolytic capacitor)-4.7k-ecc83(22uf electrolytic capacitor) . The original power source is 5Y3-4uf-20H-220uf-5,1K-RCA OD3X2(series 300V) , now cancel the series of OD3 plus SSHV2, will enlarge the line of 47uf, 22uf, 22uf need to change it?
3, two amplifier units (left and right channels) , one channel 7.5 ma, can you share one SSHV2? Or are the two sshv2s independently powered? Is there a significant difference in the final sound quality between a SSHV2 shared by the left and right channels and a single SSHV2 reflected in a single channel?
Look, I asked a few questions, it's a chatterbox, sorry to bother you,
thanks again!
I'm going to try it out in the near future, and I'm going to use it at the preamplifier. The amplifier has a 75 db gain. I have three questions for you:
1. Does the SSHV2 shunt power supply have an advantage over the conventional series power supply in terms of noise suppression?
2， The amplifier has three stages: + Power High voltage-ECC82 buffer (47uf electrolytic capacitor)-1.5k-ecc83(22uf electrolytic capacitor)-4.7k-ecc83(22uf electrolytic capacitor) . The original power source is 5Y3-4uf-20H-220uf-5,1K-RCA OD3X2(series 300V) , now cancel the series of OD3 plus SSHV2, will enlarge the line of 47uf, 22uf, 22uf need to change it?
3, two amplifier units (left and right channels) , one channel 7.5 ma, can you share one SSHV2? Or are the two sshv2s independently powered? Is there a significant difference in the final sound quality between a SSHV2 shared by the left and right channels and a single SSHV2 reflected in a single channel?
Look, I asked a few questions, it's a chatterbox, sorry to bother you,
thanks again!
1. It has unity gain voltage reference based on passive I to V conversion of a low noise JFET CCS so its not producing elevated noise of its own. Speaking relatively to the high voltage level its working at.
Series or shunt its the voltage reference self noise and its amplification rate inside the circuit that counts. Broadband noise that is.
For suppressing rectification ripple type of noise its adequate enough but not like comparing to low voltage low noise modern SMD chip regulators PSRR spec. Good raw DC pre-filtering helps.
2. No need to change the existing filter capacitors values in the rail's RC filter cells.
3. Can be shared. Dual mono can enhance stereo imaging to a degree. But try shared SSHV2 first. To know if it works without some compatibility problems with your particular high gain application and to estimate the subjective tonal synergy. Before fully adopting the solution.
Series or shunt its the voltage reference self noise and its amplification rate inside the circuit that counts. Broadband noise that is.
For suppressing rectification ripple type of noise its adequate enough but not like comparing to low voltage low noise modern SMD chip regulators PSRR spec. Good raw DC pre-filtering helps.
2. No need to change the existing filter capacitors values in the rail's RC filter cells.
3. Can be shared. Dual mono can enhance stereo imaging to a degree. But try shared SSHV2 first. To know if it works without some compatibility problems with your particular high gain application and to estimate the subjective tonal synergy. Before fully adopting the solution.
Hi Mr. Salas,
I'm in the design phase of a new 6SN7 line preamplifier, and I came across your shunt regulator.
The preamplifier wiIl be based on a MU-Followe design, each channel drawing 4ma for a total of 8mA. The power supply transformer has 0-270V @30mA output, requested B+ is 310V. With a solid state rectifier I manged to get ~370V raw B, which I trimmed down to 330-340V. I have couple of questions:
1. I have a generic red led, do I need to look for some specific type?
2. Is the raw B too high? The voltage changes quite a lot (~10V) over the day. I'm planning to use the chassis side panel as a heatsink for Q3 and going to install on Q1 the one recommended in your guide.
3. Given that the preamplifier current consumption is 8mA, what is the current drawn from the SSHV2 by itself ? Will the transformer be ok ?
Thank you very much.
I'm in the design phase of a new 6SN7 line preamplifier, and I came across your shunt regulator.
The preamplifier wiIl be based on a MU-Followe design, each channel drawing 4ma for a total of 8mA. The power supply transformer has 0-270V @30mA output, requested B+ is 310V. With a solid state rectifier I manged to get ~370V raw B, which I trimmed down to 330-340V. I have couple of questions:
1. I have a generic red led, do I need to look for some specific type?
2. Is the raw B too high? The voltage changes quite a lot (~10V) over the day. I'm planning to use the chassis side panel as a heatsink for Q3 and going to install on Q1 the one recommended in your guide.
3. Given that the preamplifier current consumption is 8mA, what is the current drawn from the SSHV2 by itself ? Will the transformer be ok ?
Thank you very much.
SSHV1 had two opposite going Zeners for gate protection not a Zener and a diode. One Zener is generally sufficient.