Hi Kevin,
I was actually waiting for your suggestions and response!
Like Pieter said, I might change to fixed biasing in a later stage. Will keep that option open for now.
Im quite concern when you mention "bypass capacitor which could impart its own unnecessary flavor".
Does it means that I'll have to consider a better cap for better sonics?
I was planning to use some NOS russian electrolytic capacitors that I have previously collected for that bypass cap...
All the caps in the PS are NOS russian PIO too.
- Louis
Hi Louis,
Unfortunately the cathode bypass cap is directly in the signal path and therefore you need to use a very high quality unit in that location, that also goes for the capacitor after the choke which the output transformers will see as the return loop for the audio signal currents. You might also want to look at hum canceling topology if you are using autobias.
Note that you can also use batteries for bias, in this case you could borrow a page from my book and use 6 x A23 or A27 12V camera batteries in series for fixed bias - since this is a class A amplifier no significant grid current will flow and they will last a long time. They're also tiny and won't take up much space - I recommend separate batteries to limit damage to the amp in the event of a battery failure. Some will suggest a mix of autobias and fixed bias - it is definitely "safer" however with sufficient cathode resistance a bypass cap is required. In any event making provisions to quickly and safely check the output stage bias when the amp is operating is a good idea. (Tip sockets on amp chassis)
Something important to mention is that conventional fuses are not safe at these voltages. There are special high voltage fuses available if you feel you need to use HV fuses. Keep the HV plate transformer separate from the rest of the supplies and you can fuse its primary independently of everything else and size the fuse to limit damage in the event of tube or bias failure. (Note that I have not had a bias failure in any amp I have ever built, but there is a first time for everything..)
Please excuse this newbie, but I don't really understand "filament CT midpoint grounded".
A wild guess will be having filament power transformer with Center-Tap for grounding.
- Louis
Direct heated rectifiers may be heated with center taped heater windings the idea being that there is no heater AC superimposed on the rectified dc. In practice I have not really been able to determine a significant difference with standard 5V rectifiers - I think it is a real issue though with very high filament current rectifiers like the 866, 872, etc..
maybe it was this one I read
Effects of AC Heating Power Applied to Directly Heated Triodes
a different matter, and you may know this, but I have found that it sometimes help to search on google images
Effects of AC Heating Power Applied to Directly Heated Triodes
a different matter, and you may know this, but I have found that it sometimes help to search on google images
Both of the above schematics can be used when the filament supply is floating. In the first one, you can leave out the hum potmeter. Just join all three resistors together at that point. The function of the two upper resistors is to provide a path for the tube's current without shorting out the filament supply, if you use 100ohm for these they have a minimal effect on bias (a few volts) and the filament supply will only need to supply an extra 20V/200ohm = 100mA.
BTW, I agree with Arnulf and Kevin that fixed bias is probably better here. The GM70 sure likes a high B+ and it's a pity to waste it on cathode biasing.
Still not correct Louis. As far as I can read 830V is your supply voltage at 125 mA so the final anode voltage that you get is 830V-70V=760V.
However at 760V with -70V the tube will draw clearly less than 125 mA, more or less 85-90mA.
But, because the GM70 is working at lower current the actual supply voltage will be somewhat higher than 830V (depending on the rectification efficiency of the diode, supply transformer etc...). In the end, having the bias set at -70V, most probably you can estimate 850V supply voltage at approx. 95-100 mA....
In summary, as a line of guide, I think you need something in the range of 680-750 ohm.
Practically you can start using a resistor + a pot in series and once you find the desired bias you measure the total resistance and use one o more fixed resistors that add up to that value.
This is good practice also because your actual tubes could have significantly different curves from those in the picture above.
I am for fixed bias anyway.....
45
Last edited:
Of course, but this is as trivial as using a teeny-weeny transformer (even with voltage multiplier if necessary to obtain higher voltage) and comparetively large filter compared to the current draw behind it and/or active regulation.
Current draw is negligible (tens of microamperes) so we're talking about $10 worrth of components versus large power resistors, relatively high voltage capacitors (50+ uF @ 100+V isn't as cheap anymore) and wasted B+ (= higher rating filtering caps required, again driving the price up) and dissipated power in general.
Cathode bias is fine for tubes where relatively little power is dissipated (pre-amplification stages, smaller output stages).
Alright folks, looks like fixed bias is the way to go.
To complicate the design a little, the C3g and IT is going to be in another chassis that can run on its own.
The C3g is actually my current headphone amp using 10uf cap coupling.
I listen to headphones more often than speakers. So would like to keep it separated from the power-hungry monster.
The idea is to replace the coupling cap with IT with two secondary taps. One for headphones and the other to drive the GM70.
Correct me if my calculations are wrong:
preferred load for C3g = 8k~10k
headphones impedence = 64ohm
10k:64 = 12:1
The ratio to drive the headphones should be 12:1
Back to my question in the first post:
What IT ratio to drive the GM70?
How should I implement fixed bias for the GM70 without the IT in the same chassis?
Will this work?
- Louis
To complicate the design a little, the C3g and IT is going to be in another chassis that can run on its own.
The C3g is actually my current headphone amp using 10uf cap coupling.
I listen to headphones more often than speakers. So would like to keep it separated from the power-hungry monster.
The idea is to replace the coupling cap with IT with two secondary taps. One for headphones and the other to drive the GM70.
Correct me if my calculations are wrong:
preferred load for C3g = 8k~10k
headphones impedence = 64ohm
10k:64 = 12:1
The ratio to drive the headphones should be 12:1
Back to my question in the first post:
What IT ratio to drive the GM70?
How should I implement fixed bias for the GM70 without the IT in the same chassis?
Will this work?
An externally hosted image should be here but it was not working when we last tested it.
- Louis
You calculated impedance ratio. Transformer turns ratio is square root of impedance ratio (sqrt(10000/64)).
Yes on bias, but bias supply also needs to be connected to a common reference somewhere (usually this will be ground).
You don't need cathode resistors in this case (unless you want to reduce gain).
Regarding IT ratio to use: whatever suits you and sets big enough a load for the driver. If you need extra voltage swing, you can use IT to your advantage. If you don't, you can use IT to get more current (but since you're not going to leave class A1 this shouldn't be a concern for you).
Yes on bias, but bias supply also needs to be connected to a common reference somewhere (usually this will be ground).
You don't need cathode resistors in this case (unless you want to reduce gain).
Regarding IT ratio to use: whatever suits you and sets big enough a load for the driver. If you need extra voltage swing, you can use IT to your advantage. If you don't, you can use IT to get more current (but since you're not going to leave class A1 this shouldn't be a concern for you).
You could put the bias supply in the preamp chassis, and connect it in series with the IT secondary winding.
Or you could put it in the GM70 chassis but then you would need an interlock of some sorts, to make sure that the plate voltage of the GM70 cannot be switched on if the cable to the preamp is disconnected.
oh, that means turns ratio for headphones should be 3.5:1?
The bias supply will be in the GM70 chassis. I intend to use schottky diodes for full wave bridging with C-R-C. Is a "clean" power important here?
Please forgive this newbie here, am I right that I should isolate the bias supply first, ground the "positive" and connect "negative" to GM70 grid?
That means I just ground the GM70 filament supply?
Still not sure how to find the IT ratio for the GM70. Do you mean that it doesn't really matter in my design? If 3.5:1 is the ratio for my headphones, can I just use it to drive the GM70?
- Louis
Come on Arnulf!
Louis has trouble enough to find the right direction.
His 12:1 winding ratio IT for the C3g to drive his 64 ohms headphones is correct; the 12:1 turns ratio means a 9k2 impedance for the tube and that is just fine. Gain of this stage will be a bit over 2 and that is enough to drive his headphone.
For the C3g/IT to drive the GM70 it is best to make the IT 1:1 because all gain of the tube is
necessary (and 1:1 gives the best IT in terms of bandwidth). With a step-up IT for higher voltage gain the current will be reduced, and that is not good to drive the GM70 grid (miller capacitance).
Pieter
Louis has trouble enough to find the right direction.
His 12:1 winding ratio IT for the C3g to drive his 64 ohms headphones is correct; the 12:1 turns ratio means a 9k2 impedance for the tube and that is just fine. Gain of this stage will be a bit over 2 and that is enough to drive his headphone.
For the C3g/IT to drive the GM70 it is best to make the IT 1:1 because all gain of the tube is
necessary (and 1:1 gives the best IT in terms of bandwidth). With a step-up IT for higher voltage gain the current will be reduced, and that is not good to drive the GM70 grid (miller capacitance).
Pieter
This way, the GM70 biasing voltage will always be there even if I only use headphones!
Im thinking on this part too. Based on my last posted schematic, what will happen if B+ comes on without bias supply? Which should turn on first? I might need to add a delay power-on circuit of some sort.
What if the GM70 is switched on but C3g not connected? Will it kill the GM70?
There will be voltage on the input connector too... Is that ok?
- Louis
Hahahaaa! u r the man Pieter!
Alright! we shall stick to 12:1 for headphones and 1:1 to drive GM70!
What do you think about fixed biasing?
- Louis
No, your calculation looked wrong but your end result was okay: divide 10000 by 64 gives you 156:1, take the square root and you get 12:1.
IF you happen to have 'too much' output power for the headphones, you can always increase this ratio a little. That will give you less power, less distortion, and a higher damping factor.
Yes, but the current drawn from this supply is VERY small, so it is easy to make a clean supply. Just throw 100uF at it and it will be okay.
Yes, connect bias supply positive to anode supply negative. Bias supply negative goes to IT secondary and the other IT secondary end goes to GM70 grid.
You should make sure that the bias supply can never die without taking down the entire amplifier. For example, if the bias supply has a separate transformer, choose a fuse with a somewhat higher current rating than necessary for that transformer.
Yes.
You need to check if the input signal times the amplification of the c3g stage is enough to provide the voltage swing needed for the GM70.
If you use high level sources (CD player) 1:1 should be fine. With older sources (tape deck, ...) maybe you won't get full output power with 1:1.
Yes, you are right. It's not a problem, though.
ALWAYS make sure that bias is switched on BEFORE B+!! Without bias supply, the tube current will runaway and hopefully the fuse blows before the output transformer blows
In amps with little tubes, it's okay to switch everything (B+, bias, heaters) at the same time. But with big tubes like this, it's better to first turn on the heater and the bias, wait 30 ... 60 seconds, only then turn on B+.
If the bias supply is on (and reaches the grid!!!), no problem.
If the bias supply is off, big problem.
Not sure what you mean?
Louis,
A multi secondary IT (1:1/0,08) is very unpractical when applying fixed bias, apart from the fact that an IT like that is not optimal. Forget it!
I think there are two possible solutions:
1. Build the headphone amp with it's dedicated 12:1 transformer.
Mount also a coupling capacitor from the anode of the C3g to an output RCA.
Input of the GM70 amp starts with a 100k resistor (from input to negative bias supply).
In this case we have full gain of the C3g because the primary of the IT acts like an anode
choke. But hmmm we have a capacitor.
Best solution is:
2. Separate headphone amp and power amp by building the C3g stage into the power amp.
When applying fixed bias a good quality 1 ohm resistor at the GM70 anode or cathode is very practical to measure the anode current.
Pieter
A multi secondary IT (1:1/0,08) is very unpractical when applying fixed bias, apart from the fact that an IT like that is not optimal. Forget it!
I think there are two possible solutions:
1. Build the headphone amp with it's dedicated 12:1 transformer.
Mount also a coupling capacitor from the anode of the C3g to an output RCA.
Input of the GM70 amp starts with a 100k resistor (from input to negative bias supply).
In this case we have full gain of the C3g because the primary of the IT acts like an anode
choke. But hmmm we have a capacitor.
Best solution is:
2. Separate headphone amp and power amp by building the C3g stage into the power amp.
When applying fixed bias a good quality 1 ohm resistor at the GM70 anode or cathode is very practical to measure the anode current.
Pieter
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.