Questions.
Hi all,
I've begun to order parts for this pre and I have a couple of questions. I will have two switched inputs on the pre. Do I need to use a DPDT and switch both the + and - of the input signal or should I just switch the + of the signal and use a common earth on all of the RCA jacks?
I also want to use Thermistors in the PSU to limit the current rush at turn on. Can anyone suggest a model and part# that you are familiar with for this task? Thanks.
Hi all,
I've begun to order parts for this pre and I have a couple of questions. I will have two switched inputs on the pre. Do I need to use a DPDT and switch both the + and - of the input signal or should I just switch the + of the signal and use a common earth on all of the RCA jacks?
I also want to use Thermistors in the PSU to limit the current rush at turn on. Can anyone suggest a model and part# that you are familiar with for this task? Thanks.
EC8010 said:
I'm not quite sure how to calulate the load. The total current will be 30mA but the supply is 300v rail to rail and 150v per leg to ground (bipolar). How do you calculate that load. Do I use the 300v value or the 150v value? Also do the thermistors go on the primary of the power transformer or the secondary before the bridge?
The thermistor is best placed on the mains side. You draw 150V x 30mA = 4.5VA. You must have some heaters, so if we allow 4VA for them, and a bit of loss in the core, means 10VA at the primary side. You have 120V mains, so 10VA / 120V = 83mA. You need a very small thermistor. 0.5A thermistors are available, often known as NTCs (negative temperature coefficient). If the thermistor doesn't get hot enough, its resistance will still be high, and that will upset things. However, there is a trick you can use. the thermistor loses heat mostly by convection. If you wrap it in rockwool loft insulation, that convection current will be restricted, so even if it is passing a small current, it will still get hot. This should enable a 0.5A thermistor to do the job if you can't find a lower value.
Both Allied and Mouser carry the Thermometrics brand of NTC thermistors.
If a separate filament trafo is used, placing the NTC device(s) on the primary of the B+ trafo is OK. If a trafo with secondaries for both filament and B+ is used, the NTC device(s) should be associated with the B+ winding.
If a separate filament trafo is used, placing the NTC device(s) on the primary of the B+ trafo is OK. If a trafo with secondaries for both filament and B+ is used, the NTC device(s) should be associated with the B+ winding.
Hmm. I don't see why you shouldn't use an NTC with a filament supply (on either mains or secondary side), so long as it's carefully selected for suitably low resistence when hot. Limiting inrush to the cold (PTC) filaments is a good idea, and most tubes are quite happy to run with a slightly low filament voltage.
Hi,
NTCs? Have been using those in tube equipment for over 15 years now.
Other useful goodies are snubber caps across secondary windings of transformers.
A MOV across a power on switch, with or without snubber caps...
Snubber networks across primary windings of OPTs can even work well when needed...
No, you won't see that too often in commercial equipment but they're certainly useful thingies.
Cheers,
NTCs? Have been using those in tube equipment for over 15 years now.
Other useful goodies are snubber caps across secondary windings of transformers.
A MOV across a power on switch, with or without snubber caps...
Snubber networks across primary windings of OPTs can even work well when needed...
No, you won't see that too often in commercial equipment but they're certainly useful thingies.
Cheers,
It has to do with cathode stripping. Associating the NTC device with the B+ only allows the filaments to warm up before the B+ rises.
IMO, SS rectified, current regulated DC is the "correct" way to protect filaments from a turn on surge. BTW, the "famous" Philips/Mullard turn on flash is an intentional design feature that uses the turn on surge to protect the cathodes.
Hi,
Current regulating heater supplies can be useful too...
Which reminds me of those old amperite iron/hydrogen ballasttubes as mentioned in this thread:
When I first stumbled across the Amperite site about a year or so ago I was really surprised they actually still make this stuff.
Does anyone see derived applications useful for audio other than those stated on the Amperite site already?
As CCSs they're way too slow in response time (about 1s depending) but there must be more...
Cheers,
Current regulating heater supplies can be useful too...
Which reminds me of those old amperite iron/hydrogen ballasttubes as mentioned in this thread:
When I first stumbled across the Amperite site about a year or so ago I was really surprised they actually still make this stuff.
Does anyone see derived applications useful for audio other than those stated on the Amperite site already?
As CCSs they're way too slow in response time (about 1s depending) but there must be more...
Cheers,
Eli Duttman said:
I like many others use AC to power the filaments. In my new amp when I flip the switch the filaments get a surge from the filament winding I guess but the B+ comes on slowly because of the 5AR4. I have decided to use two separate switches for the filament power and B+ on the preamp. Is this a bad thing? I guessed that as long as the filaments where warmed up for 20-30 seconds before I turned on the B+ that all would be well. Am I wrong?
Hi,
Cathode stripping occurs under the following conditions:
1/ Too low heater voltage will not heat up the cathode enough and whenever something is trying to pull of electrons from the cathodes these are then stripped of their emissive coating, leaving "dead" surfaces.
2/ Trying to pull current from a still not sufficiently heated cathode.
While 1/ is a repeat event till the tube reaches end of life, 2/ Can kill the tube in one go depending on the amount of current drawn.
In practice, if all voltages come up more or less simultaneously, current will start flowing while the cathode is still warming up.
No big deal for as long as there's no current demand. E.g. you're not blasting music through the amp.
Most of this kind of desaster scenarios happen in the guitar amp world as you can imagine.
To avoid that kind of expensive experience you use NTC, timers, whatever necessary to prevent it.
In audio gear I've seen at least as much damage done by stand by switches getting sequenced the wrong way around or left in the wrong position as by people switching on the amp and trying to draw power.
When you try to do that you'll hear the amp distort and anyone with a little common sense will soon realise that, ...well tubes just need to warm up a minute or two first.
Given a choice, the NTCs in the B+ are the best solution overall.
Cheers,
P.S. As for flashing heaters, they still flash, BTW. Only the flash is hidden from the users' sight by coating the heater ends and sticking it further up the cathode sleeve.
Cathode stripping occurs under the following conditions:
1/ Too low heater voltage will not heat up the cathode enough and whenever something is trying to pull of electrons from the cathodes these are then stripped of their emissive coating, leaving "dead" surfaces.
2/ Trying to pull current from a still not sufficiently heated cathode.
While 1/ is a repeat event till the tube reaches end of life, 2/ Can kill the tube in one go depending on the amount of current drawn.
In practice, if all voltages come up more or less simultaneously, current will start flowing while the cathode is still warming up.
No big deal for as long as there's no current demand. E.g. you're not blasting music through the amp.
Most of this kind of desaster scenarios happen in the guitar amp world as you can imagine.
To avoid that kind of expensive experience you use NTC, timers, whatever necessary to prevent it.
In audio gear I've seen at least as much damage done by stand by switches getting sequenced the wrong way around or left in the wrong position as by people switching on the amp and trying to draw power.
When you try to do that you'll hear the amp distort and anyone with a little common sense will soon realise that, ...well tubes just need to warm up a minute or two first.
Given a choice, the NTCs in the B+ are the best solution overall.
Cheers,
P.S. As for flashing heaters, they still flash, BTW. Only the flash is hidden from the users' sight by coating the heater ends and sticking it further up the cathode sleeve.
So is this flash what is meant when datasheets list 'controlled heater warm-up' as a feature? It always seemed to me that the thermal shock of such a rapid turn-on might outweigh cathode-stripping effects in terms of tube life, especially where tube rectifiers are used, or some other slow-start / delay mechanism for the B+.
No, "controlled warm-up time" meant that the valve was designed to warm up at the same rate as all others with "controlled warm-up time" enabling series heater strings without current regulation. (If the current is regulated, it doesn't matter if one valve warms at a different rate to another, but if the string is connected directly across the mains, differing warm-up times could temporarily leave one valve with excessive heater power, which would damage its cathode.)
Controlled heater warmup time was introduced for use in TV receivers with series connected heater strings. The idea is to equalize the turn on surge across the entire string and avoid overvoltage on any given heater. The "A" at the end of a 7 or 9 pin miniature type designation is an indication of controlled heater warmup time, eg. 6AU6A.
A 12AX7A will work any time a 12AX7 is called for. In a series heater string, a 12AX7 may or may not work correctly.
A 12AX7A will work any time a 12AX7 is called for. In a series heater string, a 12AX7 may or may not work correctly.
Hi,
Not 100% sure but the effect is the same anyway.
The term "controlled heater warm" you'll only find in U.S. datamanuals.
I've never encountered it in a European one but I've not seen all of them either.
Quite often you'll find this "controlled heater..." term in conjunction with small signal tubes carrying the A suffix (for improved version) as for example the 12AX7A.
On those tubes, carrying this A suffix, I've never seen the heater flash when powered up so I assume the effect was hidden from sight as well and that change procedure was part of the "things to do to acquire that A suffix".
Nowadays you won't find much of these small tubes with flashing heaters, at least over here they seem all gone.
As far as I kow the last ones came from the former East-German RFT factories or occasionally you'd see that on Ei tubes as well.
If that were the case, lightbulbs wouldn't last that long.
Once more: cold cathode, no current flow, no stripping of the cathode possible.
The danger is the intermediate status: cathode not yet hot enough and already too much current drawn.
When idling at warm up, nothing, absolutely nothing detrimental can happen to a tube.
For the anecdote:
One of my amps, a PP 2A3 has it's B+ charged for months in a row every time I use it as it doesn't have any bleeders whatsoever.
The tubes haven't changed since the day I bought them five years ago, in fact the tubes seem to prefer it this way.
As an aside my caps stay formed that way too.
I know that amp as I designed and built it myself so security isn't much of an issue for me but I wouldn't recommend this to anyone as forgetting to bleed of the caps (something you should always check first anyway) can lead to lethal shock.
Cheers,
Not 100% sure but the effect is the same anyway.
The term "controlled heater warm" you'll only find in U.S. datamanuals.
I've never encountered it in a European one but I've not seen all of them either.
Quite often you'll find this "controlled heater..." term in conjunction with small signal tubes carrying the A suffix (for improved version) as for example the 12AX7A.
On those tubes, carrying this A suffix, I've never seen the heater flash when powered up so I assume the effect was hidden from sight as well and that change procedure was part of the "things to do to acquire that A suffix".
Nowadays you won't find much of these small tubes with flashing heaters, at least over here they seem all gone.
As far as I kow the last ones came from the former East-German RFT factories or occasionally you'd see that on Ei tubes as well.
If that were the case, lightbulbs wouldn't last that long.
Once more: cold cathode, no current flow, no stripping of the cathode possible.
The danger is the intermediate status: cathode not yet hot enough and already too much current drawn.
When idling at warm up, nothing, absolutely nothing detrimental can happen to a tube.
For the anecdote:
One of my amps, a PP 2A3 has it's B+ charged for months in a row every time I use it as it doesn't have any bleeders whatsoever.
The tubes haven't changed since the day I bought them five years ago, in fact the tubes seem to prefer it this way.
As an aside my caps stay formed that way too.
I know that amp as I designed and built it myself so security isn't much of an issue for me but I wouldn't recommend this to anyone as forgetting to bleed of the caps (something you should always check first anyway) can lead to lethal shock.
Cheers,
fdegrove said:
I've got some Mullards (made in britain) that flash. Always a fun sight.
Hmm... then class AB has yet another thing going for it: fixed bias prevents the tubes from drawing much current during warmup. As opposed to auto-bias, especially with absurdly, stupifyingly large bypass caps, where bias is initially zero.
Hmm... so if that's what causes stripping... why do rectifiers work at all? They see full voltage across them during the first few cycles, and with no grid in the way of the electrons...
as forgetting to bleed of the caps (something you should always check first anyway) can lead to lethal shock.
Cheers,
I've never had a problem, not just because most of my amps now include some sort of divider somewhere (be it for tube bias or just heater bias), but because the tubes always bleed the PROPERLY SIZED PS caps down to under 10V. You must have some disturbingly large caps, or a seperate heater power switch, to do that.
Tim
Hi,
No, class AB hasn't got anything to do with it, tubes can deliver a certain amount of current even when not fully warmed up.
It's only when you exceed that amount that cathode stripping will occur.
Maybe I wasn't expressing myself clearly enough but I'd think you'd know better than that?
Why can't you stick a big cap behind a vacuum diode rectifier?
Tubes don't discharge capacitors when they don't have a heated cathode/filament as they'd present an open circuit.
What's a properly sized PS anyway?
In the example I gave the cap would probably be disturbingly large from your POV not mine.
For kicks: it's a 1000µF/400V cap.
There isn't any.
FWIW, my OTLs_ a commercial design_ use 660µF/700V of capacitance for the voltage amplifier and phasesplitter tubes alone and you know what? It's regulated too with more caps behind the reg....Crazy, huh?
Not surprisingly slew rates are well into the best ever seen category,the amp has very wide bandwidth and I still have to hear one that can beat it.
Oh, it even has a couple of pentHodes in it too.
Cheers,
No, class AB hasn't got anything to do with it, tubes can deliver a certain amount of current even when not fully warmed up.
It's only when you exceed that amount that cathode stripping will occur.
Maybe I wasn't expressing myself clearly enough but I'd think you'd know better than that?
Why can't you stick a big cap behind a vacuum diode rectifier?
Tubes don't discharge capacitors when they don't have a heated cathode/filament as they'd present an open circuit.
What's a properly sized PS anyway?
In the example I gave the cap would probably be disturbingly large from your POV not mine.
For kicks: it's a 1000µF/400V cap.
There isn't any.
FWIW, my OTLs_ a commercial design_ use 660µF/700V of capacitance for the voltage amplifier and phasesplitter tubes alone and you know what? It's regulated too with more caps behind the reg....Crazy, huh?
Not surprisingly slew rates are well into the best ever seen category,the amp has very wide bandwidth and I still have to hear one that can beat it.
Oh, it even has a couple of pentHodes in it too.
Cheers,
fdegrove said:
My point being, a tube will easily exceed current if the cathode is slow to react to the tube heating up. Example, 470 ohm Rk plus a few thousand uF bypass. It starts out zero biased.
On the other hand, class AB is fixed and almost never uses tubes for the bias supply, as such the tubes will ALWAYS be biased off such that no current can ever flow during heatup. Even if a signal is applied -- the preamp tubes aren't hot either!
Exactly. So why should one be worried about say, stripping a 2A3 or 300B? They have cathodes comparable to 5Y3 and 5U4 (yeah, I'll have to check to be sure, but they're in the ballpark as I recall), which are rated for peaks up to at least 2A.
Then how are my amps discharging? Again, I can remove any voltage dividers across the PSU leaving just tubes. And it'll still drop to safe levels (<40V) within 20 seconds.
Around 2mF/ampere. ISTR that's for 1V ripple, much more is tolerable in tube circuits (at least before a filter choke) so as little as 500uF/A is acceptable.
For instance, Revision 3 uses 2x50uF with a (rather small) choke inbetween, plus two RC's for the driver and preamp. The 6L6 draws 60mA.
Fine with me, but you haven't told me if it's a 12AX7 (er... wait, those suck.. I mean, uh, what's that called... 6SL7) preamp or a quad KT88 power amp.
Yep, utterly.
Now really Frank, what does slew rate have to do with excessive capacitance? Even you know it doesn't. I could get that kind of slew with a .1uF filter cap. Big whoop.
Oh, it even has a couple of pentHodes in it too.
Cheers,
Now there's a point of interest, as being CCS's they tend to have slow rise or fall time. If used in the signal path that is.
Tim
MORE OT....
Hi,
I already had the impression you had completely misread my argument from your first reply but nevermind.
It's good to see we both agree that all that paranoia about cathode stripping and artificial softstarts where it isn't needed is just in the heads of some people not knowing how a tube works.
Which doesn't imply that that amount of peak current can be pulled from the cathode before it is fully warmed up as it is precisely under those intermediate conditions that cathode stripping will occur.
Unless the warm up sequence is abnormal and someone is trying to blast music at loud levels through the amp demanding current that that cathode can't deliver at that point in time nothing bad will happen to the cathode coating.
I think I have figures for that for at least a few tube types based on research by the Philips Eindhoven Development Labs but I'll need to dig it up in case some people are interested in that.
You must have some discharge path somewhere down the line (bleeder resistor from connected preamp perhaps, load in // with tubes?) or else something's not quite right with those caps.
The test is easy enough to do, don't heat the filaments of the tubes and measure the B+, it should go up considerably.
Make sure the caps are up to the extra voltage, which they normally should be.
Fair enough, that's textbook theory. In practice more can but won't necessarily hurt depending on where and how you use the extra capacitance.
Either way we both have our ideas about this so it's no big deal.
Actually, I did say already it was a PP 2A3 amp and holy golly it does (or rather it can) use a 12AX7A as well....
Admittedly I changed it to a 6N1-P later on for sensitivity reasons.
Certainly. It's not just the capacitance of course, the circuit itself has to be up to it as well...Guess even you knew that too. [kidding]
They are in the signal path but who said anything about a CCS?
Cheers,
Hi,
I already had the impression you had completely misread my argument from your first reply but nevermind.
It's good to see we both agree that all that paranoia about cathode stripping and artificial softstarts where it isn't needed is just in the heads of some people not knowing how a tube works.
Which doesn't imply that that amount of peak current can be pulled from the cathode before it is fully warmed up as it is precisely under those intermediate conditions that cathode stripping will occur.
Unless the warm up sequence is abnormal and someone is trying to blast music at loud levels through the amp demanding current that that cathode can't deliver at that point in time nothing bad will happen to the cathode coating.
I think I have figures for that for at least a few tube types based on research by the Philips Eindhoven Development Labs but I'll need to dig it up in case some people are interested in that.
You must have some discharge path somewhere down the line (bleeder resistor from connected preamp perhaps, load in // with tubes?) or else something's not quite right with those caps.
The test is easy enough to do, don't heat the filaments of the tubes and measure the B+, it should go up considerably.
Make sure the caps are up to the extra voltage, which they normally should be.
Fair enough, that's textbook theory. In practice more can but won't necessarily hurt depending on where and how you use the extra capacitance.
Either way we both have our ideas about this so it's no big deal.
Actually, I did say already it was a PP 2A3 amp and holy golly it does (or rather it can) use a 12AX7A as well....
Admittedly I changed it to a 6N1-P later on for sensitivity reasons.
Certainly. It's not just the capacitance of course, the circuit itself has to be up to it as well...Guess even you knew that too. [kidding]
They are in the signal path but who said anything about a CCS?
Cheers,
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.