20mA surplus is good to feed own reg needs and keep output impedance low enough. More produces heat fast on the sink due to there is high voltage involved, and in your case when approaching 100mA+ at some point soon you are going to meet the Supertex depletion MOSFET's current generation limit. It is called IDSS and is not the same in each production sample.
Hi,
I am sure questions like this get covered a lot, so please direct me to the post if you can.
I have my power supply feeding 135V my SSHVII board which is just sitting on my table, which then feeds my #26 preamp in a seperate chassis.
I went through the power on steps - read about 25mA at the test points.
Connected to my preamp, still reading steady current but no HT. It was about 3V.
There is some kind of interference because when I place a meter probe on the test points the LED cuts out, this also happens when I touch the heat sinks - even though these are definitely isolated from the board. Is this a grounding issue? Or maybe a blown device?
Does this sound common? I am checking over everything again and again because 9/10 it's wiring issue with me, but I thought I would ask too.
Thanks
Charlie
I am sure questions like this get covered a lot, so please direct me to the post if you can.
I have my power supply feeding 135V my SSHVII board which is just sitting on my table, which then feeds my #26 preamp in a seperate chassis.
I went through the power on steps - read about 25mA at the test points.
Connected to my preamp, still reading steady current but no HT. It was about 3V.
There is some kind of interference because when I place a meter probe on the test points the LED cuts out, this also happens when I touch the heat sinks - even though these are definitely isolated from the board. Is this a grounding issue? Or maybe a blown device?
Does this sound common? I am checking over everything again and again because 9/10 it's wiring issue with me, but I thought I would ask too.
Thanks
Charlie
Did you check with dummy load first, was it working OK for setting the voltage also? Are the 4 wires to load routed as per instruction sheet?
One thing to know is to have set sufficient ccs current. Try 35mA for two #26 channels because the reg itself will need +20mA on top of those. Then it is going to need 10V minimum input to output difference, 20V is recommended. Is that preamp set for only 105V B+?
One thing to know is to have set sufficient ccs current. Try 35mA for two #26 channels because the reg itself will need +20mA on top of those. Then it is going to need 10V minimum input to output difference, 20V is recommended. Is that preamp set for only 105V B+?
I followed the instructions for the wiring, and it was working OK with the dummy load.
Consumption for per channel is 6mA plus 20mA across the reg as you said. I do think now that I may be too conservative with my voltage margains - I am aiming for about 105V for the #26, but maybe I need to up the input HT...
It seemed curious that by placing the probe on the board it gave different readings and made the LED shine. Is the LED meant to shine constantly?
Cheers
Charlie
Consumption for per channel is 6mA plus 20mA across the reg as you said. I do think now that I may be too conservative with my voltage margains - I am aiming for about 105V for the #26, but maybe I need to up the input HT...
It seemed curious that by placing the probe on the board it gave different readings and made the LED shine. Is the LED meant to shine constantly?
Cheers
Charlie
Last edited:
So 12mA assuming 2 channels since you mentioned 1 SSHV2, +20mA allowance, then set for 35mA. The Led is in series with the voltage reference. Should always shine when there is current there. Something's wrong outside the reg since it worked on dummy. Verify that its still good by using the dummy once again and set higher ccs than before. Maybe an oscillation flashes some energy when you touch probes. Still it should be shining anyway when all is well.
2 things may trouble you on first approach. 1. Too much current went to the pre for your CCS setting, not enough left to power the reg's innards. Can it be it draws more than 6mA each valve as usual for #26 preamps due to low B+ bias point, and 25mA was truly marginal, leaving the reg breathless? 2. The 4 wire system maybe is not routed correctly to the preamp and the reg's sense reference circuitry sees no power from the main force lines not meeting them properly to the load's side. Verify that +sense & +force go together to B+, same for the other two to ground. At least the 0sense and 0force wires should meet at a symmetric as possible grounding node between the preamp's channels.
2 things may trouble you on first approach. 1. Too much current went to the pre for your CCS setting, not enough left to power the reg's innards. Can it be it draws more than 6mA each valve as usual for #26 preamps due to low B+ bias point, and 25mA was truly marginal, leaving the reg breathless? 2. The 4 wire system maybe is not routed correctly to the preamp and the reg's sense reference circuitry sees no power from the main force lines not meeting them properly to the load's side. Verify that +sense & +force go together to B+, same for the other two to ground. At least the 0sense and 0force wires should meet at a symmetric as possible grounding node between the preamp's channels.
If the shunt FET is designed to survive the worst case operation. Survive not operate perfectly during the fault condition, but return to perfect working after the fault condition is removed.
This Shunt FET worst case survival is Tj < Tjmax when the output is open circuit, i.e. the fuse is blown or the output has simply been disconnected.
If your Shunt meets this requirement then you don't need any dummy load.
The Shunt will pull down the output voltage to the set point and hold it there while the Shunt Tj rises far above normal. The output voltage will drift as this happens. But you only need a few seconds to confirm that all the lights are on and that the set point voltage is about right.
I consider survival in the two worst case operation to be an absolute requirement for any regulated PSU that I build.
Shorted output when mains voltage is maximum
and
Open circuit output. Mains voltage does not matter.
This Shunt FET worst case survival is Tj < Tjmax when the output is open circuit, i.e. the fuse is blown or the output has simply been disconnected.
If your Shunt meets this requirement then you don't need any dummy load.
The Shunt will pull down the output voltage to the set point and hold it there while the Shunt Tj rises far above normal. The output voltage will drift as this happens. But you only need a few seconds to confirm that all the lights are on and that the set point voltage is about right.
I consider survival in the two worst case operation to be an absolute requirement for any regulated PSU that I build.
Shorted output when mains voltage is maximum
and
Open circuit output. Mains voltage does not matter.
Those are things that shunts in general may survive if the sinks are big and there will be no arcs, but I don't recommend to nobody non trained electrician to try test such conditions in SSHV regs, i.e. doing shorting tests when a reg is handling high voltage or to leave it open circuit when his sink is not calculated for max. Builder please be conservative and follow the pre testing procedure on helper dummy for the power and heat to be divided between reg and load. Be super careful, high voltage can kill. When wiring and probing think your moves twice. If it will ever accidentally short or stay long without a load and will get too hot it may survive depending on the severity and byproducts of a fault, your settings and sinks, just retest it by the original procedure if in doubt. Always fuse the primary of the power transformer to its mains draw mA spec.
Went back to the dummy load and it was working well again. So I rebuilt the raw supply to give it some more breathing space and reconnected. Everything is working fine now so I was just not allowing for enough voltage for both reg and circuit. Actually, everything is working really well....
Thank you so much for designing this, it has really helped me understand these complex regulators and am looking forward to using them more.
Cheers
Charlie
Thank you so much for designing this, it has really helped me understand these complex regulators and am looking forward to using them more.
Cheers
Charlie
Did you set 35mA also?
This sounds like you are suggesting that the HV version of the Salas Shunt does not suit survivability in the event of the output becoming open circuit.
That type of error whether by accident (fuse blowing, or wire breaking loose) or human error in leaving the output open when powering up, could be quite common.
I don't agree that a failure of the Shunt regulator due to this worst case operation is a risk that many builders will want to take.
I can see that the shorted output when full mains voltage is applied is a much less likely worst case operation and many builders may be prepared to accept this risk, of shunt failure, in this situation.