Warning LED when valve stops conducting - a little SS help needed

What people are working on at the moment
User avatar
Nick
Site Admin
Posts: 15711
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#106 Re: Warning LED when valve stops conducting - a little SS help needed

Post by Nick »

1) with a DC heated DHT using Andrews DC board, is there a benefit to using virtual cathode resistors
When this question was asked, my answer was "as there is no disadvantage to doing so, and some theoretical justification, then may as well"

Now there is a downside, that justification is removed.
2) with a pair of DHT with a shared cathode, is there any reason not to wire the heaters in series
Is there any reason to expect two heaters to require the exact same current to run at the ideal temperature?
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
RhythMick
Old Hand
Posts: 747
Joined: Fri Jun 23, 2017 12:24 pm

#107 Re: Warning LED when valve stops conducting - a little SS help needed

Post by RhythMick »

Nick wrote: Sat Feb 16, 2019 12:07 pm
1) with a DC heated DHT using Andrews DC board, is there a benefit to using virtual cathode resistors
When this question was asked, my answer was "as there is no disadvantage to doing so, and some theoretical justification, then may as well"

Now there is a downside, that justification is removed.
2) with a pair of DHT with a shared cathode, is there any reason not to wire the heaters in series
Is there any reason to expect two heaters to require the exact same current to run at the ideal temperature?
Thanks Nick. Let's explore both the downside and the upside to using VC resistors when a DHT is DC heated.

Downside is that they appear in parallel to the filament to the DC heater board.

When the filament is working they lower the overall impedance and take DC current. Nick was a great help to me in understanding the different currents and I put a spreadsheet together to calculate it. I settled on using 200R VC because it minimised the heater current through them.

When the filament fails, the VC resistors provide a current path for the DC heater board, which sees a big increase in impedance and reduces the output current accordingly to maintain the voltage across the onboard sense resistor. Andrew may want to correct my description here. The existence of the VCs makes filament failure detection more complex, though it can still be done (ala Mike).

Upside of the VC resistors is something I don't understand well enough to document.
I recall reading articles saying that DC heating meant that one side of the filament is at a higher voltage than the other and thus that the filament coating was likely to become depleted, implying shortened tube life. I don't know if that's true or not, or whether the charge space surrounding the cathode negates or minimises that problem. Nor do I know if as one end of the filament becomes depleted whether the other end naturally compensates, nor if the problem can be reduced by occasionally switching the polarity of the DC voltage.
I also don't understand how the Virtual Cathode resistors would help in this regard. Even with them in place they don't change the voltages at either end of the filament.

I'm glad you re-opened the question Andrew. I suspect there is little evidence one way or the other. I'll do a bit more digging and see if I can collect references to anything useful, I'm sure others can weigh in also. At the end of the day though, one of the big advantages of DIY is that I can prove it to myself by trying it.

I may well try changing the 26 valve pairs (which are easily accessed at the front) to series heaters, something like this.
155032356326513-1.png
There will be a voltage difference and thus a bias difference between the two valves, which could be minimised by taking the cathode connection to the joined filaments from the centre pins.
RhythMick
Old Hand
Posts: 747
Joined: Fri Jun 23, 2017 12:24 pm

#108 Re: Warning LED when valve stops conducting - a little SS help needed

Post by RhythMick »

Andrew you asked for a refresh of the current circuit, see below. The same pattern is used for all valve pairs, with the same Virtual Cathode resistors (200R) and the same measurement resistors (10R).
1550333053838421-1.png
RhythMick
Old Hand
Posts: 747
Joined: Fri Jun 23, 2017 12:24 pm

#109 Re: Warning LED when valve stops conducting - a little SS help needed

Post by RhythMick »

Ok I've converted the right channel 26 pair to series, removed the virtual cathodes, moved the 10R measure resistors to the anodes, disconnected the unused board from the transformer. I'll trim the board for 3v. Will report back after listening probably tomorrow, wife not well poor love so off to soothe fevered brow etc.
RhythMick
Old Hand
Posts: 747
Joined: Fri Jun 23, 2017 12:24 pm

#110 Re: Warning LED when valve stops conducting - a little SS help needed

Post by RhythMick »

RhythMick wrote: Sat Feb 16, 2019 7:11 pm Ok I've converted the right channel 26 pair to series, removed the virtual cathodes, moved the 10R measure resistors to the anodes, disconnected the unused board from the transformer. I'll trim the board for 3v. Will report back after listening probably tomorrow, wife not well poor love so off to soothe fevered brow etc.
Trimmed to 3v across the pair. As expected slight difference between the valves. Trimmed the higher 1 down to 1% high, the other is 1.5% low.

Obviously in series the regulator only sees the voltage across the pair, so 1 disadvantage is that if 1 of the valve filaments changes impedance (over time or with different valve pairs) the regulator can only regulate the end to end 3v.

I'll have a listen tomorrow.
RhythMick
Old Hand
Posts: 747
Joined: Fri Jun 23, 2017 12:24 pm

#111 Re: Warning LED when valve stops conducting - a little SS help needed

Post by RhythMick »

Left channel now done. Both valves within 0.02v (one high, one low), that's 1.3%.
RhythMick
Old Hand
Posts: 747
Joined: Fri Jun 23, 2017 12:24 pm

#112 Re: Warning LED when valve stops conducting - a little SS help needed

Post by RhythMick »

Switched over to my favourite valves (old Cunningham globes) and the variation is a little greater. One channel they are identical, the other one is 5.5% high the other 3.5% low.

The sound is every bit as good, first impressions are the bass is tighter, more tuneful, follows the micro sounds better. Not hearing any edginess, nor loss of air, space or detail. Amp is cold, I'll play radio through it for an hour then listen properly again. First impressions good.

So, assuming nothing changes significantly, it's been a good experiment. 2 things have changed here. 1 board driving 2 filaments in series, but also removal of the virtual cathode resistors. If there are improvements to the sound I'm of the opinion that will come from the removal of the virtual cathodes, not from running the filaments in series. The difference in heater voltage between the 2 valves in the pair will affect the sound, I'll try and work out if one channel sounds better than the other.

The effect on valve life of that variation is a concern. I think many would say anything within 10% is good enough but I'm inclined to think the further away from spec you are the damage will increase exponentially. I don't want to be left continually wondering whether the filament voltages are right. Glad I experimented, but I'm fairly sure I'll be going back to individual boards.

I'll listen for a week like this then try individual DC boards but no virtual cathode resistors.
RhythMick
Old Hand
Posts: 747
Joined: Fri Jun 23, 2017 12:24 pm

#113 Re: Warning LED when valve stops conducting - a little SS help needed

Post by RhythMick »

Hmmm as always there is a little more to it.

The outer valves are at the -ve end of the heater voltage. Each valve has a voltage gradient on the filament, but the midpoints of the filaments are different. The outer valves have less bias and unsurprisingly are running slightly higher DC current (6mA vs 5mA). As I was methodical in writing the amp, the outer valves take the same phase.

I'm still of the view that there's been an improvement, but it's from the removal of the virtual cathodes and it may improve further when I watch back to separate DC boards. Fascinating.
User avatar
pre65
Amstrad Tower of Power
Posts: 21373
Joined: Wed Aug 22, 2007 11:13 pm
Location: North Essex/Suffolk border.

#114 Re: Warning LED when valve stops conducting - a little SS help needed

Post by pre65 »

In my most humble opinion, getting the filament heating "right" is the key to directly heated valve performance.

To that end, each directly heated valve should have it's own separate filament supply.
The only thing necessary for the triumph of evil is for good men to do nothing.

Edmund Burke

G-Popz THE easy listening connoisseur. (Philip)
RhythMick
Old Hand
Posts: 747
Joined: Fri Jun 23, 2017 12:24 pm

#115 Re: Warning LED when valve stops conducting - a little SS help needed

Post by RhythMick »

pre65 wrote: Sun Feb 17, 2019 1:40 pm In my most humble opinion, getting the filament heating "right" is the key to directly heated valve performance.

To that end, each directly heated valve should have it's own separate filament supply.
Makes sense. Nick tried to tell me that several times. I was listening honest, but nice to be able to experiment.
RhythMick
Old Hand
Posts: 747
Joined: Fri Jun 23, 2017 12:24 pm

#116 Re: Warning LED when valve stops conducting - a little SS help needed

Post by RhythMick »

Well I'm convinced something has changed for the better. Unless I like the sound of one valve at a higher bias than the other (consequence of series heating) those virtual cathode resistors were holding the amp back. Hard to believe quite honestly, given where it was, but that was an unexpected result for sure.

I'm definitely going back to individual dc boards, but without the virtual cathodes. I'll do that to the 26s first. Unless that takes me backwards I'll repeat for all the valve pairs.

Andrew, your DC boards are truly stunning I think.
User avatar
Mike H
Amstrad Tower of Power
Posts: 20157
Joined: Sat Oct 04, 2008 5:38 pm
Location: The Fens
Contact:

#117 Re: Warning LED when valve stops conducting - a little SS help needed

Post by Mike H »

Watch out for imbalanced anode currents, If the two valves are not an exactly matched pair. DC offset will reduce transformer inductance. Unless the transformers have gapped cores. Ideally needs presets to indiviually adjust grid bias and equalise the anode currents.

I had this aggro with a GEC Williamson clone, common cathode resistor on KT66 pair, and a bias trim preset pot, which needed resetting every couple of weeks. :shock:
 
"No matter how fast light travels it finds that the darkness has always got there first, and is waiting for it."
RhythMick
Old Hand
Posts: 747
Joined: Fri Jun 23, 2017 12:24 pm

#118 Re: Warning LED when valve stops conducting - a little SS help needed

Post by RhythMick »

Mike H wrote: Sun Feb 17, 2019 5:36 pm Watch out for imbalanced anode currents, If the two valves are not an exactly matched pair. DC offset will reduce transformer inductance. Unless the transformers have gapped cores. Ideally needs presets to indiviually adjust grid bias and equalise the anode currents.

I had this aggro with a GEC Williamson clone, common cathode resistor on KT66 pair, and a bias trim preset pot, which needed resetting every couple of weeks. :shock:
Thanks Mike. The transformers are Lundahl and have an air gap for that purpose. I do agree it's worth having the valves in balance though, minimise the inductance lost.
User avatar
Mike H
Amstrad Tower of Power
Posts: 20157
Joined: Sat Oct 04, 2008 5:38 pm
Location: The Fens
Contact:

#119 Re: Warning LED when valve stops conducting - a little SS help needed

Post by Mike H »

That would be a great help then. (gapped cores.) :D
 
"No matter how fast light travels it finds that the darkness has always got there first, and is waiting for it."
RhythMick
Old Hand
Posts: 747
Joined: Fri Jun 23, 2017 12:24 pm

#120 Re: Warning LED when valve stops conducting - a little SS help needed

Post by RhythMick »

So, back to the original topic for a second. Once the virtual cathodes are out I'll implement the led warning and report back.
Post Reply