| sonata149 |
I built an SE amplifier using 807s with the screen grid 250 volts fed from two valve regulators. On crescendo passages, the VRs' glow flickers - brightens and dims and even blocks out. At first I thought it was the bass but it also happens with rich female vocals (Mina - this evening).
Has any member had a similar experience with VRs? Any info and guidance would be greatly appreciated. Thanks - Joe A |
|
|
| EC8010 |
| You need more quiescent current through your VR valve. |
|
|
| kevinkr |
Had no idea that Mina was still around. My sister's preferred artist for a while..
Reduce the value of the dropping resistor feeding the regulator tubes, depending on type they can handle up to about 30mA or so quiescent. A slight amount of flicker is acceptable at high levels, but they should never extinguish. |
|
|
| sonata149 |
| Thanks EC8010. As I'm not very technical, does that mean I should lower the value of the dropper resistor? |
|
|
| EC8010 |
| Yes, as kevinkr says. If you can post a diagram, we can even tell you what value to reduce the resistor to. |
|
|
| Miles Prower |
| quote: | Originally posted by sonata149
I built an SE amplifier using 807s with the screen grid 250 volts fed from two valve regulators. On crescendo passages, the VRs' glow flickers - brightens and dims and even blocks out. At first I thought it was the bass but it also happens with rich female vocals (Mina - this evening).
|
I would not rule out the possibility that your regulator isn't completely stable. You might have a damped oscillation in there. You'll need to o'scope to see if this is happening. |
|
|
| Johan Potgieter |
I agree with Miles.
In days of yore we had several occasions of less than perfect tube regulators. Also, take into account that with beam tubes the g2 current can vary appreciably between low-max. output. You do not mention whether the VR tubes are used as primary regulators or whether they form part of a higher current tube regulator. Whatever topology, the problem is the same. Make sure the regulator can cope. |
|
|
| jlsem |
| quote: | | You need more quiescent current through your VR valve. |
I agree with EC8010. In a Class A 807 amp the screen grid (@250V) will probably not draw the 5ma minimum current requirement of most gas regulator tubes.
John |
|
|
| sonata149 |
Hi Folks,
Thanks a lot for your feedback. Although it's getting a bit confusing for me (with damped oscillations), I'm starting to think seriously about it. I have dropped the resistance by about 500 ohms and I'm still getting the same results.
I'll post a diagram of that section later on as I'll be out. But I wanted to say thanks to you all for your help. |
|
|
| kevinkr |
Depending on the value of that resistor you may need to drop to something like half its current value. As a minimum you would probably need >25mA flowing through these vr tubes if they are supplying the screens of both output tubes.
Got your email, am writing a long winded reply.. :D |
|
|
| Steven-H |
| Interesting topic. While I don't have anything to add (my knowledge is limited...), I would like to see a schematic. I've wanted to make an 807 sort of amp for a while now... |
|
|
| sonata149 |
To all the members who are trying to help me sort out this problem I hope I can manage to post these 2 partial scans showing the screen grid VR section.
As can be seen, from a B+ of around 450V there is a 10K/10W dropper resistor feeding 2 VRs for 250V - an OC3 and OD3 - which is then taken to the 807 screen grid via a smaller 1K0. (Is this a grid leak dropper similar to that used for the signal grid in audio amplifiers?) There is then a 47uF/0.22uF capacitor combo from the 807 screen grids to earth (ground). I'm not sure if Kevin is referring to this capacitance.
I now realise that by halving the 1K0 resistor I didn't change the current draw of the VRs. Maybe, I need to lower the 10K resistor. Perhaps, you guys can guide me as to the value to use. The real B+ voltage is rather lower than that specified due to the slightly lower secondary AC voltage rating of the transformer I used.
I hope this gives you a better idea what I'm talking about.
Thanks and regards to ALL.
Joe A |
|
|
| kevinkr |
| HMMM, I'm not sure about that 47uF cap and the combination of VR tubes through a 1K resistor, but presumably other amps based on this design do work.. The 10 K resistor would result in a total VR tube + screen current of 20mA which may be rather marginal at full power - I would decrease that resistor to say 6.8K and see what happens. |
|
|
| sonata149 |
| Hopefully I get the image posted this time. |
|
|
| sonata149 |
| This is the second part showing how the VR supply is connected to the screen grid. The member who asked me for the 807 circuit can email me directly as I'm still being 'moderated'... |
|
|
|
