Question about Regulated HT for Tube circuit

  Could someone school my Newb butt why most regulated Tube heater circuits , loaded with tubes or not hooked up to any load stay at the correct voltage - 6.3  or 12.6    but as with the Gyraf  regulated HT circuit for the G9 and one other I found in the forum , the voltage rises well over the desired voltage for a few seconds while the heaters heat up then comes down and settles at the desired voltage. 

  I mean I know that the heaters need to heat up to load the HT - but why does the heater circuit stay set at the voltage wanted with or without a load  and the HT / B+  doesn't  ???

 

audiophreak said:

  Could someone school my Newb butt why most regulated Tube heater circuits , loaded with tubes or not hooked up to any load stay at the correct voltage - 6.3  or 12.6    but as with the Gyraf  regulated HT circuit for the G9 and one other I found in the forum , the voltage rises well over the desired voltage for a few seconds while the heaters heat up then comes down and settles at the desired voltage. 

Click to expand...

Well if a so called regulated HT output rises well over the desired voltage then settles down as the tubes warm up as you suggest, then it ain't a regulated HT supply.

However, it is not uncommon for a regulated supply to be further decoupled before feeding sensitive low level circuits, and thus you would expect the HT voltage at these circuits to behave as you describe since they are downstream of the regulator.


Cheers

Ian

Thanks Ian 

  ..but I'm still confused ,  I've built a couple of G9s and they all react in this manner , the raw HT is about 312vdc , and with my dmm at a test point after the " Regulated HT circuit " the HT goes up to about 306vdc from when I turn the unit on, then drops to about 246vdc after 8-15 seconds after the heaters in the tubes heat up. 
    I built a quad of NYDave One Bottles , perf boarded the G9 HT circuit and ran in to the same situation however the PT I am using for that build the raw HT is about 350vdc - so that first few seconds after switching the unit on, the HT was going up to 318vdc or so before the heaters heated up and it settles around 240vdc - in order to get around that I inserted a time delay circuit to delay the HT 30 seconds so when the HT hits the heaters are warmed up and HT instantly loads to the 240vdc - if there is a slight spike its imperceptible with my dmm attached.

    I am building a few other Tube projects and would like to eliminate that time delay circuit and have a solid steady regulated HT , would you know of a true Regulated HT circuit that stays at desired voltage from switch on,  or point me in a direction ?

  Thanks,
            Chip 

You did not provide a link to the specific plan you use.

However for a _microphone_, it would be good practice to have an R-C filter between the regulator and the mike. The best regulators make some hiss, R-C filtering cleans this up. There is also some voltage loss from no-load (cold tube) and normal load (hot tube).

I don't know why you care what the B+ voltage is while the tube is warming up (as long as not so high it will endanger parts). You usually power-up tube gear several to many minutes before it will be used.

Thank you PRR,
        I perf boarded the Gyraf G9 HT circuit , unfortunately the PT I am using has a 250vac sec , so  rectified 350vdc , so when the heaters are heating up the B+ goes up to almost 320vdc for a few seconds and in the case of the NYDave One Bottles , 300vdc is the max plate voltage on the tube spec sheet - just not wanting to damage the tubes - the projects I'm working on are tube mic pre's - a single tube - as with the G-Pultec,  the voltage swing isnt that large with just the one tube, but when multiple tubes are added the voltage swing from first start up to 10-15 seconds later gets a lot larger ,( I'm guessing from the additional load ) - just looking for a solution - I guess I could just try and find a PT better suited  ;D

... as well as still not understanding why a heater circuit with a LM317 stays at the set voltage loaded or un-loaded , but cant seam to find the same results for an B+ circuit

I must ask the larger question:  what does it matter?  Is it sending your caps over voltage?  That's the only reason I can think of to care, if we're talking about largely self-regulating class A small signal tube circuits that draw constant current and establish a fixed load.  I understand it's not behaving as expected, but there's no downside that I see.    The tube spec sheet concerns max voltage for a duration, under operating current.  You are over voltage because the tubes aren't drawing current yet. 

I had a glance at the G9, and it looks regulated, and says so.  It also mentions using 250V caps for a 245V supply(!).  That seems like a fool's dream for certain.  Standard rule of thumb there is 350V caps minimum, and 24/7 duty broadcast built with 450V caps for sub-300V B+.  Starting from scratch, I'd be inclined to go the other direction, and build a simple correct unregulated supply that delivers the correct voltage with expected load. 

regulation difference between fil and B+: probably a question of current swing potential.  I bet the fil circuit can handle a wider range of swing, and the B+ can't, so it shows up as voltage swing in the B+.  It's not capable of maintaining spec with no load, it expects to be loaded a minimum amount.  You might slug it's output with a power bleeder resistance for minimum load. 

Thanks emrr ,
            "  The tube spec sheet concerns max voltage for a duration, under operating current.  You are over voltage because the tubes aren't drawing current yet.  "  ... I hadn't quite looked at it from that perspective  ;D
  I have used 450v caps in both G9s and in my NYDave One Bottles.

Thank you all for taking the time and schooling me.

audiophreak said:

Thanks Ian 

  ..but I'm still confused ,  I've built a couple of G9s and they all react in this manner , the raw HT is about 312vdc , and with my dmm at a test point after the " Regulated HT circuit " the HT goes up to about 306vdc from when I turn the unit on, then drops to about 246vdc after 8-15 seconds after the heaters in the tubes heat up. 

Click to expand...

I have now found an online copy of the G9 HT supply, and as I suspected it has an RC smoothing network directly after the regulator so the behaviour you see is precisely what should happen.

I do not know what the designer was thinking but a regulator gets rid of a lot of ripple in one stage  and stabilised the HT against mains voltage variations, but as PRR says they can be noisy. So you follow it with an RC filter and you have a nice economical HT supply.

Cheers

Ian

630V film caps are cheap, small and plentiful these days.

Well if a so called regulated HT output rises well over the desired voltage then settles down as the tubes warm up as you suggest, then it ain't a regulated HT supply.

Click to expand...

Ian, could you clarify here.  I have always thought of regulated supplies (for tube) as supplies that maintain a range of voltage under normal operating conditions, meaning after warm up when the circuit is amplifying, etc.  I've never thought of them in the capacity of keeping turn on peak rise suppressed.  For older style regulators using several tubes it doesn't seem like that would be possible.

lassoharp said:

630V film caps are cheap, small and plentiful these days.

Well if a so called regulated HT output rises well over the desired voltage then settles down as the tubes warm up as you suggest, then it ain't a regulated HT supply.

Click to expand...

Ian, could you clarify here.  I have always thought of regulated supplies (for tube) as supplies that maintain a range of voltage under normal operating conditions, meaning after warm up when the circuit is amplifying, etc.  I've never thought of them in the capacity of keeping turn on peak rise suppressed.  For older style regulators using several tubes it doesn't seem like that would be possible.

Click to expand...

Yes, semiconductor regulators will fire up much more quickly than a tube based regulator so in that sense they are certainly better placed to avoid turn on peaks. Having said that, in a tube regulator the basic voltage reference, be it a tube or zener, will be up and stable well before the regulating tubes are heated up.  I would also expect the larger series pass tube to heat up more slowly than any smaller tubes involved in regulation so all the output voltage determining circuits should be fully operational before the series pass tube starts to pass current. Thus I see no reason in principle for a tube regulator to have a turn on peak.

On a separate but related topic, when considering class A only stages such as in mic pres, it is worth remembering that the average power drawn by a class A circuit is constant. It does vary instantaneously with the signal voltage but, and this is a very important but, if the stage is RC decoupled then that instantaneous power is supplied by the decoupling capacitor and the power supply sees an essentially constant load. So in  class A circuits, one of the primary objectives of a regulator - to keep the voltage constant in the face of a varying load - is not needed. In my view its main purpose it to keep the HT supply constant in the face of inout voltage variations i.e.mains.

Cheers

Ian

TL783C data says that it needs at least 10mA load or so to start regulating, so it might be normal that at startup, when cold tubes don't draw any current yet, B+ rises (unregulated) . Emrr suggestion to add a power bleeder resistance for minimum load should solve this.
regards,
Laurent.

Yes, semiconductor regulators will fire up much more quickly than a tube based regulator so in that sense they are certainly better placed to avoid turn on peaks. Having said that, in a tube regulator the basic voltage reference, be it a tube or zener, will be up and stable well before the regulating tubes are heated up.  I would also expect the larger series pass tube to heat up more slowly than any smaller tubes involved in regulation so all the output voltage determining circuits should be fully operational before the series pass tube starts to pass current. Thus I see no reason in principle for a tube regulator to have a turn on peak.

Click to expand...

Thanks for explaining that.

emrr said:

You might slug it's output with a power bleeder resistance for minimum load.

Click to expand...

pyjaman said:

TL783C data says that it needs at least 10mA load or so to start regulating, so it might be normal that at startup, when cold tubes don't draw any current yet, B+ rises (unregulated) . Emrr suggestion to add a power bleeder resistance for minimum load should solve this.
regards,
Laurent.

Click to expand...

lassoharp said:

Thanks for explaining that.

Click to expand...

      I will try that tomorrow !!  THANKS  !!   

pyjaman said:

TL783C data says that it needs at least 10mA load or so to start regulating, so it might be normal that at startup, when cold tubes don't draw any current yet, B+ rises (unregulated) . Emrr suggestion to add a power bleeder resistance for minimum load should solve this.
regards,
Laurent.

Click to expand...

No need for a power bleed resistor. I have already done some experiments by modifying my TL783 based phantom power PCB to become a regulated HT board. If you use a 150R from the output to the ref pin and a 33K 5W resistor from the ref pin to ground then you get the required minimum current and a 275V output.

Cheers

Ian

Thank you Ian for for testing that all out and posting !   
     
      I would need 240vdc out , not sure on the circuit you used ,  would the 150R you suggest replace R35 and R35 series resistors ( for a total resistance of 527R) and the 33k 5W you suggest for the R34 ( 100k 2W resistor ) listed in the Gyraf G9  circuit ?    ...  just reduce R34 in the Gyraf circuit to the 33K ?

  Thanks so much,
                          Chip

audiophreak said:

Thank you Ian for for testing that all out and posting !   
     
      I would need 240vdc out , not sure on the circuit you used ,  would the 150R you suggest replace R35 and R35 series resistors ( for a total resistance of 527R) and the 33k 5W you suggest for the R34 ( 100k 2W resistor ) listed in the Gyraf G9  circuit ?    ...  just reduce R34 in the Gyraf circuit to the 33K ?

  Thanks so much,
                          Chip

Click to expand...

Yes, replace R25 and R36 with a single 150 ohm resistor and the 33K replaces R34. That will give you 275 volts. For 240V dc you need to change the 33K to 28800 ohms. You could use a 27K 5W resistor in series with a 1.8K 1W resistor.

Cheers

Ian

Thanks so much Ian !!!   
   
  I got a last minute gig , so I wont be able to try this for a few days  ... and Thank you All for the experienced perspective,  help and schooling 

audiophreak said:

Thanks so much Ian !!!   
   
  I got a last minute gig , so I wont be able to try this for a few days  ... and Thank you All for the experienced perspective,  help and schooling 

Click to expand...

One tip. Whatever you do, do NOT omit the zeners across the TL783 as shown on the G9 schematic. They are essential to protect the TL783 from accidental output short circuits. I learnt this the hard way - it is incredibly easy to destroy them.

Cheers

Ian

Noted  ;D        .. hopefully I'll get the chance to try this out tomorrow 

    Thanks again !!

Hi Ian ,
          Got a chance to do this today , I could only find a  26k 6.5w resistor on hand, but it worked beautifully   
      un loaded  it went up to 251vdc , loaded with four 12av7's  it dropped to 224vdc, I think I might have a 33k 6.5w left over from my Dual G-Pultecs I did with the same PT, I'll try that tomorrow , but a 25vdc  difference from un loaded to loaded is far better than the 75vdc spread I was getting.

  .. just so I can learn from this , do these resistor values load the regulator better / more efficiently ??

  Thanks again for taking the time,
                                                      Chip