6.3v indirectly heater valve PCB (possible group buy?)
#46
OK, I took the files around to a friend who said he'd make the prototype of the PCB for us, he does this stuff for a living so I got some coaching, this is what I ended up with.
The most significant change is back to a variable reg but with 3 resistors to set the voltage, he thought this was a better than the diode chain idea, the rest is all PCB layout, the board is now the 1/4 of a standard eurocard so should be cost effective to build
Andrew
The most significant change is back to a variable reg but with 3 resistors to set the voltage, he thought this was a better than the diode chain idea, the rest is all PCB layout, the board is now the 1/4 of a standard eurocard so should be cost effective to build
Andrew
Analogue, the lost world that lies between 0 and 1.
- Paul Barker
- Social Sevices have been notified
- Posts: 8989
- Joined: Mon May 21, 2007 9:42 pm
#47
Thank you Andrew.
So what is the flexibility. If we change those three resistors to alter the voltage regulated portion and we alter the AC applied at the input can we go from 4v to 7.5v for instance?
What is the maximum current again?
OK stated as 1 amp.
Is that just to conservatively rate the 1.5 amp regulator and the 1.5 amp bridge?
Will the rest of the circuit stand an uprating in the bridge and rectifier to 5 amp?
So we could use it for the VT25/62 7.5v 1.25A, the 6b4G is covered, so is PX4 at 4v 1A but PX25 4v 2A.
If it is possible at our end to alter the Bridge and the Reg chip to a 338, is the rest of the circuit able to cope with the extra rating?
Thank you?
So what is the flexibility. If we change those three resistors to alter the voltage regulated portion and we alter the AC applied at the input can we go from 4v to 7.5v for instance?
What is the maximum current again?
OK stated as 1 amp.
Is that just to conservatively rate the 1.5 amp regulator and the 1.5 amp bridge?
Will the rest of the circuit stand an uprating in the bridge and rectifier to 5 amp?
So we could use it for the VT25/62 7.5v 1.25A, the 6b4G is covered, so is PX4 at 4v 1A but PX25 4v 2A.
If it is possible at our end to alter the Bridge and the Reg chip to a 338, is the rest of the circuit able to cope with the extra rating?
Thank you?
#48
Hi Paul,
The PCB is indeed more flexible, just change those three resistors and the input voltage et viola. 4v to 7.5v is well within range.
As you rightly say other parts could be changed or uprated providing they fit on the PCB. My only concern would be increasing the current will cause more heat, I will measure the temp of the heatsink when I get prototype back and try a higher current and see what happens, the current unit runs at about 45 deg above ambient at 1 amp, but I have found a bigger heatsink which should help. Its much cooler at ~350mA, which is what you'd expect.
The main problem is the extra heat will degrade the electrolytics, there's no doubt about that and, of course, current is a power law etc. The PCB is small to keep costs down but the downside is the caps are quite close to that big heatsink. You could usefully mount a bigger reg in a TO-3 package off-board and the rest would indeed cope with a much higher current. Equally, the Darlington could be changed for a higher spec model as well; its all very generic I even tried a jfet Nick sent me.
For directly heated triodes I am working on a circuit that uses a voltage controlled current source which I think will be a better solution to that specific problem, but its some way off being done.
Andrew
The PCB is indeed more flexible, just change those three resistors and the input voltage et viola. 4v to 7.5v is well within range.
As you rightly say other parts could be changed or uprated providing they fit on the PCB. My only concern would be increasing the current will cause more heat, I will measure the temp of the heatsink when I get prototype back and try a higher current and see what happens, the current unit runs at about 45 deg above ambient at 1 amp, but I have found a bigger heatsink which should help. Its much cooler at ~350mA, which is what you'd expect.
The main problem is the extra heat will degrade the electrolytics, there's no doubt about that and, of course, current is a power law etc. The PCB is small to keep costs down but the downside is the caps are quite close to that big heatsink. You could usefully mount a bigger reg in a TO-3 package off-board and the rest would indeed cope with a much higher current. Equally, the Darlington could be changed for a higher spec model as well; its all very generic I even tried a jfet Nick sent me.
For directly heated triodes I am working on a circuit that uses a voltage controlled current source which I think will be a better solution to that specific problem, but its some way off being done.
Andrew
Analogue, the lost world that lies between 0 and 1.
- Paul Barker
- Social Sevices have been notified
- Posts: 8989
- Joined: Mon May 21, 2007 9:42 pm
#49
That's about what i thought. Would probably take the T03 off board.
Yes the future dht project is a good idea.
Yes the future dht project is a good idea.
#50
Just a thought, I assume pad4 is to ground the centre of the heater supply. Is it worth splitting C4 into two, and taking the centre of the two to the virtual ground? Otherwise there is no low impedance path to ground to AC.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
#51
That's not a bad idea! The 'center tap' idea is just gilding a lilly really but if its worth doing, it's worth doing etc....Nick wrote:Just a thought, I assume pad4 is to ground the centre of the heater supply. Is it worth splitting C4 into two, and taking the centre of the two to the virtual ground? Otherwise there is no low impedance path to ground to AC.
Although you can always de-couple at the valve pins.
I was going to use the center tap but I wasn't sure who else might.
I'll add it to the list of refinements for the next rev, not sure if that will be the final thing or another prototype. I'll see how this one goes.
Andrew
Analogue, the lost world that lies between 0 and 1.
#52
Yep, always something after the cutoff time .
I find it amazing how pressing the "order now" button causes you to remember something 5 min later.
I find it amazing how pressing the "order now" button causes you to remember something 5 min later.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
- IslandPink
- Amstrad Tower of Power
- Posts: 10041
- Joined: Tue May 29, 2007 7:01 pm
- Location: Denbigh, N.Wales
#53
Cool !
In the end I think I concluded that the standard ( 3U ? ) eurocard is 100 x 160 , so you're talking 50 x 80 mm for your PCB, is this correct ?
I went looking for the dimensions on this and got thoroughly confused as none of the official sites can actually come out and say what they mean in plain language . What is wrong with these electronics people eh ?Andrew wrote:...the board is now the 1/4 of a standard eurocard....
In the end I think I concluded that the standard ( 3U ? ) eurocard is 100 x 160 , so you're talking 50 x 80 mm for your PCB, is this correct ?
"Once you find out ... the Circumstances ; then you can go out"
#54
Yep, 50 x 80, its only a standard photoresist test board so I can add the change before we go to PCB Express or whoever you use Nick.IslandPink wrote:
In the end I think I concluded that the standard ( 3U ? ) eurocard is 100 x 160 , so you're talking 50 x 80 mm for your PCB, is this correct ?
Andrew
Analogue, the lost world that lies between 0 and 1.
#55
For small runs and prototypes I use
http://www.pcbtrain.co.uk/
For larger quantities
http://www.quick-teck.co.uk/index.php
100 off 50 * 80, 1oz copper, silk screen, £140 + VAT + postage 12 day service.
http://www.pcbtrain.co.uk/
For larger quantities
http://www.quick-teck.co.uk/index.php
100 off 50 * 80, 1oz copper, silk screen, £140 + VAT + postage 12 day service.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
- jack
- Thermionic Monk Status
- Posts: 5503
- Joined: Wed Dec 29, 2010 8:58 pm
- Location: ɐılɐɹʇsnɐ oʇ ƃuıʌoɯ ƃuıɹǝpısuoɔ
- Contact:
#58
The point of D1 is to protect IC1 from reverse bias at power-off caused by electrolytics further down the chain - traditional monolithic regulators are typically extremely vulnerable to reverse bias... e.g. the 780x datasheet states that they break down at only 7V which is why you need D1 for those devices.
In the diagram above, D1 anode is connected to C4, therefore exposing IC1 to voltage from C5 at power off.
Granted, this would probably not be a problem if a tube was connected as C5 would discharge quickly, but D1's anode should go to C5 (IMHO) otherwise its pointless.
Edit: Just looked at the LT1086 datasheet - page 8 makes it clear that as the LT1086 is a "newer" device, D1 is not even necessary in this configuration - you only need it in exceptional circumstances e.g. when using a crowbar protection on the input side which could cause a large reverse surge current - the LT1086 internal protection will handle uS surges of 10A - 20A, so D1 is not needed...
...but if it was, its still in the wrong place. (IMHO!)
In the diagram above, D1 anode is connected to C4, therefore exposing IC1 to voltage from C5 at power off.
Granted, this would probably not be a problem if a tube was connected as C5 would discharge quickly, but D1's anode should go to C5 (IMHO) otherwise its pointless.
Edit: Just looked at the LT1086 datasheet - page 8 makes it clear that as the LT1086 is a "newer" device, D1 is not even necessary in this configuration - you only need it in exceptional circumstances e.g. when using a crowbar protection on the input side which could cause a large reverse surge current - the LT1086 internal protection will handle uS surges of 10A - 20A, so D1 is not needed...
...but if it was, its still in the wrong place. (IMHO!)
Last edited by jack on Mon Apr 16, 2012 3:25 pm, edited 5 times in total.
Vivitur ingenio, caetera mortis erunt
- jack
- Thermionic Monk Status
- Posts: 5503
- Joined: Wed Dec 29, 2010 8:58 pm
- Location: ɐılɐɹʇsnɐ oʇ ƃuıʌoɯ ƃuıɹǝpısuoɔ
- Contact:
#59
Small point: Note that the saw used to panellise boards is not zero-kerf - its typically about 1.5mm, so your panels should be 1mm smaller in each direction.Andrew wrote:Yep, 50 x 80, ...IslandPink wrote:In the end I think I concluded that the standard ( 3U ? ) eurocard is 100 x 160 , so you're talking 50 x 80 mm for your PCB, is this correct ?
Or for prototyping try Olimex who charge 30 euros for a 100x160mm double-sided, plated through, silk-screened & solder-masked panel with Au plating, panelised & cut at no extra charge assuming its a step-and-repeat panelisation, i.e. one file repeated.Mike H wrote:Or, FreePCB & P & M Services ... just another option ...
They take Eagle BRD files directly - no need for Gerber.
Used them many times (e.g. http://4hv.org/e107_plugins/forum/forum ... .php?72000 or http://www.desmith.net/NMdS/Data/SLA%20 ... %20Top.JPG). Not bad quality at all.
They may be based in Plovdiv but the guy that runs it, Tsvetan, speaks good English, responds to emails quickly, and is very helpful.
Vivitur ingenio, caetera mortis erunt
#60
I debated about the connection of D1 several times, I ended up with this since it seem sensible to protected the reg against over voltage on the load, say a slipped scope probe, but I think I may add two diodes on the final PCB, one as above and one in the traditional location back across the reg; as its hardly an expensive part.
The chap who helped me couldn't see how over voltage might happen on the small track between the reg and the darlington; he probably doesn't see many older type regs in his job. However, I concede your point that C5 might be the cause and be a problem with certain regs if it isn't able to discharge quickly enough.
This is only a first pass...
The PCB is slight under the 50*80, I forget how much but enough for a hacksaw blade to be run between.
Andrew
The chap who helped me couldn't see how over voltage might happen on the small track between the reg and the darlington; he probably doesn't see many older type regs in his job. However, I concede your point that C5 might be the cause and be a problem with certain regs if it isn't able to discharge quickly enough.
This is only a first pass...
The PCB is slight under the 50*80, I forget how much but enough for a hacksaw blade to be run between.
Andrew
Analogue, the lost world that lies between 0 and 1.