Less than 90 volts but no less than 60 volts. If you see 50 volts across R19, then the top valve will see approx. 100 volts of heater-cathode voltage difference (assuming B+ of 300 volts).
There's a reason why the volt divider is aiming at 1/4 of B+. This is because of each valve sees 1/2 B+.
So, with B+ of 300 volts, top valve's cathode will be floating at about 150v above ground. If you don't float the heater supply and instead ground it, the top valve's heater will be 150 volts below cathode voltage (which is clearly more than the allowed 90 volts spec), meanwhile the bottom valve is happily sitting at nearly zero heater-cathode volt difference.
If you aim at 1/4 of heater floating voltage (75 volts), the top valve will see 150-75 = 75 volts of heater-cathode difference while the bottom valve will also see 75-0 = 75 volts of difference. Both are 15 volts below the maximum allowable difference (i.e 90-75).
So you see that 1/4 gives the best headroom.
Ok thanks guys.
Measured across R19 and it reads 67.3v, so it's close to optimal.
The b+ is closer to 315v, but 315/4 = 79v, so 67v is probably a little on the low side for the best headroom, however given the 12AT7 limit is 90v, can we safely assume 67v is ok?
Measured across R19 and it reads 67.3v, so it's close to optimal.
The b+ is closer to 315v, but 315/4 = 79v, so 67v is probably a little on the low side for the best headroom, however given the 12AT7 limit is 90v, can we safely assume 67v is ok?
Oh I see how you are calculating it now. 90.5 then.
That's awfully close to the 90v threshold for the 12AT7.
6CG7 is 100v max so the input tubes are safe.
The 12AT7 at the output are slightly concerning.
Do you think it would be better to change the tubes out for a different type that has a higher heater-cathode rating? Or what would a more suitable value for R19 be?
That's awfully close to the 90v threshold for the 12AT7.
6CG7 is 100v max so the input tubes are safe.
The 12AT7 at the output are slightly concerning.
Do you think it would be better to change the tubes out for a different type that has a higher heater-cathode rating? Or what would a more suitable value for R19 be?
Last edited:
If you're on the phase of planning, yes i'd find better suited tubes, in terms of the gain you want and the maximum rating.. or find resistor values that would really put you spot on B+/4.
If you've already finished the build, i'd take the risk as its only slightly higher than the allowed max rating.. Or take the extra trouble of aiming B+/4. One thing about tube that i learned is that spec sheets are not razor sharp precise. Deviation of 5-10 percent is acceptable. Tubes are electronically robust and mechanically fragile. Transistors are the other way around.
If you've already finished the build, i'd take the risk as its only slightly higher than the allowed max rating.. Or take the extra trouble of aiming B+/4. One thing about tube that i learned is that spec sheets are not razor sharp precise. Deviation of 5-10 percent is acceptable. Tubes are electronically robust and mechanically fragile. Transistors are the other way around.
Determine R19 using simple voltage divider formula. To get B+/4, you will find that R18 should be 3 times R19 (I assume R19 is the one connected to ground, i forgot which one).
Ok I see what Broskie has done:
http://www.diyaudio.com/forums/tube...stor-calculations-software-2.html#post3771688
In the first schematic he shows R19 as 47k, and in the second it is 100k.
This must be because in the first schematic the supply feeds four tubes (the line gain stage plus the buffer).
So for the first schematic he has two R18, which are 300k in the first schematic. Hence 300/3 = 100/2 = 50k which is approx 47k as he shows.
Whereas in the second, there's only a single R19 of 300k, divided by three gives an R18 of 100k.
So it's probably (roughly) correct at the moment.
Other alternative would be to change out the 12AT7 for a 12BH7, which has a higher heather-cathode rating of 200v.
With a mu of 18.4 and an Rp of 4300 and with R3/R6 as 300R, the current will be approaching 15mA, which might in fact be better for driving a power amplifer (despite the fact that the transconductance of the 12BH7 is nowhere near as good as the 12AT7.
(the build is complete, so I thought changing the tube is easier than desoldering the resistor)
http://www.diyaudio.com/forums/tube...stor-calculations-software-2.html#post3771688
In the first schematic he shows R19 as 47k, and in the second it is 100k.
This must be because in the first schematic the supply feeds four tubes (the line gain stage plus the buffer).
So for the first schematic he has two R18, which are 300k in the first schematic. Hence 300/3 = 100/2 = 50k which is approx 47k as he shows.
Whereas in the second, there's only a single R19 of 300k, divided by three gives an R18 of 100k.
So it's probably (roughly) correct at the moment.
Other alternative would be to change out the 12AT7 for a 12BH7, which has a higher heather-cathode rating of 200v.
With a mu of 18.4 and an Rp of 4300 and with R3/R6 as 300R, the current will be approaching 15mA, which might in fact be better for driving a power amplifer (despite the fact that the transconductance of the 12BH7 is nowhere near as good as the 12AT7.
(the build is complete, so I thought changing the tube is easier than desoldering the resistor)
Broskie says to hold the heater supply at 1/4 of B+ in order to avoid exceedint the tube heater/cathode voltage ratings. He says to use a voltage divider circuit and decoupling capacitor (as I did too, in my earlier comment). Broskie gives two examples of a suitable voltage divider circuit. In the first he has: - R18 - two 300k 1W resistors in parallel (this is equivalent to one 150k 2W rating resistor) - R19 - one 47k resistor This will provide B+ * 47k / (47k + 150k) = 0.24 * B+ (assuming the tubes meet their specification) In the second example Broskie has: - (R18) - one 300k 1W resistor - (R19) - one 10 k 1W resistor This will provide B+ * 100k / (100k + 300k) = 0.25 * B+ (assuming the tubes meet their specification) Note that Broskie has specified 1W rating resistors. He seems to have thought of everything necessary for a reliably operating circuit. I would go with his recommendations for the tube type as well - you would need to change much more than one resistor if you start changing the tube characteristics.
It's quite a diverse circuit - you can see on the last page of the manual (attached) that he offers a number of solutions using different tubes, most of the parameters are very similar.
So with a b+ of 315 and R3/R6 or 300R, probably the safest bet is the 12BH7 in terms of a close match. The 12AT7 he suggests a B+ of 200...
The other thing to remember is that most tubes have a maximum heater to cathode resistance as well as a maximum voltage. For this reason I generally make my V+/4 divider using a 22K at the bottom and a pair of 33K resistors in series at the top and I use two watt types for all three. As well as keeping the heater cathode resistance comfortably low, it gives an exact B+/4 using standard value components.
Cheers
Ian
- Status
- Not open for further replies.