Voltage regulated Grid Bias supplies.

[IslandPink]

Anyway the Onkaku & Ankora (?) both have caps to ground for the AC path.
So what Thomas said

If I look eg. at the Ankoru, there's a small power supply creating smoothed negative volts of -100 to -200 ( adjustable ) that connects to the bottom of the secondary of the interstage transformer and via that , to the grid of the 845's .

Where's the (AC) cap to ground ? - I guess the final cap on the power supply ? Yes, Ok, I'm with you now , sorry .


Should I have read the other thread that's going, on Dave's amp ?

[IslandPink]

Caps aren't much good for holding a DC bias voltage !

Ummm I meant from the interstage secondary to ground to decouple ac from the bias dc / regulator do dah...

Sorry I got my wires crossed a bit trying to visualise this. I had the lower end of the interstage coupled to signal ground with cap and trying to apply the bias to the signal end ...blah blah

I'll get me coat .

( But if the +30V reg. supply can supply 0 to 20mA or whatever, then surely it itself represents a low-impedance to ground ?
So the cap isn't needed ? )

[Dave the bass]

In the last A2 amp I built I made up 2 x variable grid bias supplies. Simple BR/CRC pot divider plus cap decoupling the wiper of the pot feeding the earthy side of the IT. So in effect I did have a C in the position Thomas suggested and yes it worked OK. But I had a bit of a task setting up and balancing the +ve grid bias as both came from the same TX but via separate CRC networks and pot dividers. Adjusting one channel had a small bot notciable affect on the other grid, buty I could adjust them to get (say) 70mA anode current per channel after a bit of AB fettling.

I wondered if using Volt reg IC's was a better way, hence OP. But then the AC developed across the IT sec got me thinking again again... and hey Tesco, we're back to square one again

I had to walk to the pub y'know, I was doing that much thinking!!!

DTB

[Nick]

Yep, I was going to answer this after Dave's initial question, but its gone all over the place now. As Mark said the cap is to provide a AC link to ground, a regulator would not need this in theory, but in practice, it will, first most regs are inductive in nature and have rising impedance with frequency, secondly if its a series reg, then its not equally good at sinking current as sourcing it (or the other way for -ve regs), so again the cap will help out with tha.

Regulated bias supplies are not always a good thing. As the mains voltage changes the anode voltage on the valve will change (unless its also regulated), if the grid voltage is unregulated it will vary at the same time and help to keep current constant (ish), regulate the bias and not the anode voltage and that won't happen. Not a big deal with low gm valves, but makes a big difference with something like the 6c33c.

[Dave the bass]

Thanks for all the advice and replies everyone. 'preciate it.

As an aside...Did the word ''Euterpe'' in one of Marks replies light up any other Gong-heads brains by chance?


Maybe its just me then.

DTB

[IslandPink]

Dave, I'm ashamed to say I'm not enough of a Gong-head - but I approve of any track that is over 9 minutes long .

Let's just hope you are able to use this A2 bias supply for something over a foot tall with a heavy graphite anode.

[pre65]

I'm going to get a cheap 0 - 30v dual bench supply for my CV2799 in A2 experiments because my current PSU only goes up to 12v.

[Dave the bass]

....first most regs are inductive in nature and have rising impedance with frequency, secondly if its a series reg, then its not equally good at sinking current as sourcing it (or the other way for -ve regs), so again the cap will help out with tha.

Ah, got it (I think), the chip will be seen as an inductive reactance XL by the waveform across the secondary of the IT, the cap gives a low Zpath to ground and gets it out the way of the reg chip function yes?

Regulated bias supplies are not always a good thing. As the mains voltage changes the anode voltage on the valve will change (unless its also regulated), if the grid voltage is unregulated it will vary at the same time and help to keep current constant (ish), regulate the bias and not the anode voltage and that won't happen. Not a big deal with low gm valves, but makes a big difference with something like the 6c33c.

Wow, yeah, I hadn't thought of that. So, combined tracking of both grid and anode supplies keeps the amp bias points more stable as they follow the mains supply fluctuations together as they're not locked to ref voltage like a voltage reg.... but.... if I was to regulate the HT in addition to the grid bias it might not be a problem even with a high gm valve? Both supplies would be locked in regulation and be oblivious (hopefully) to changes in mains voltage.

Dave, I'm ashamed to say I'm not enough of a Gong-head .

This I need to work on!

I'm going to get a cheap 0 - 30v dual bench supply for my CV2799 in A2 experiments because my current PSU only goes up to 12v.

Nice one Phil, bench supplies are very very handy. I should maybe treat myself to a dual LT supply @ a good few amps. Instant heater supplies!

DTB

[Nick]

but.... if I was to regulate the HT in addition to the grid bias it might not be a problem even with a high gm valve? Both supplies would be locked in regulation and be oblivious (hopefully) to changes in mains voltage.

Yep, just make sure the bias is there before the B+

[Dave the bass]

Yep, just make sure the bias is there before the B+

Possible proviso, unless the valve is 'manly'/designed/happy with zero bias for a while <pushes nerdy glasses up nose and blinks awkwardly> eg TZ40?

DTB

[Nick]

Yep, just make sure the bias is there before the B+

Possible proviso, unless the valve is 'manly'/designed/happy with zero bias for a while <pushes nerdy glasses up nose and blinks awkwardly> eg TZ40?

DTB

Yep, that right. But I will rephrase that, make sure the bias is stable before B+ is applied. Zero bias is a special simple case.

[Dave the bass]

Noted, will do.

DTB

[Mike H]

Wot Nick said (earlier), that's how I was thinking about it re the series pass reg. Plus the HF impedance aspect did also occur but I didn't mention.

As for the reg supply being constant whereas the HT isn't, here's a radical idea, you could just set a resistor divider for its reference, ergo it will also vary as its mains supply varies?

Or even more radical, reference it with a divider derived off the actual HT.