I am trying to get a 300Vpp voltage swing in my driver stage for a 845 amp. I guess that if I use a choke load, and have a plate voltage about 300V, I should be able to do that.
Question :
1. How do I plot the loadline for a choke load? Is it a straight horizontal line with constant current?
2. How do I calculate the gain of the stage and the voltage swing?
3. The max voltage at the plate of the driver stage will be Vp + 1/2 of voltage swing. If Vp is already at the design centre max limit, does this instantaneous higher Vp kills the tube? I am thinking of using a 6pi9 ( 6ag7 ) as the driver tube but the datasheet says the maximum plate voltage is only 300V. If I use the B+ at 300V, and get a 300Vpp swing, then the maximum plate voltage will be 450V.
Am I taking sense of is my ignorance of tubes showing? I am a silicon era guy, with no formal tube training. My tube bible, Morgan Jones, does not seem to have the theory I need.
Thanks for any help
CM
Question :
1. How do I plot the loadline for a choke load? Is it a straight horizontal line with constant current?
2. How do I calculate the gain of the stage and the voltage swing?
3. The max voltage at the plate of the driver stage will be Vp + 1/2 of voltage swing. If Vp is already at the design centre max limit, does this instantaneous higher Vp kills the tube? I am thinking of using a 6pi9 ( 6ag7 ) as the driver tube but the datasheet says the maximum plate voltage is only 300V. If I use the B+ at 300V, and get a 300Vpp swing, then the maximum plate voltage will be 450V.
Am I taking sense of is my ignorance of tubes showing? I am a silicon era guy, with no formal tube training. My tube bible, Morgan Jones, does not seem to have the theory I need.
Thanks for any help
CM
Here’s the circuit of the amp, a pair of 845 SE monoblock. I have found several problems with the design, and trying to fix all of them.
1. Bias of the 845 seems wrong. With 125 volts at the cathode of the current with 1150 B+ from the curves should be much higher than the 83mA calculated from 125/1k5. The actual measured voltage is 130V, close enough. I am changing the cathode resistor to 2k3. That should give a bias of 65mA, 150V at cathode, and B1+ at 1000V.
2. The 330V +B2 actually measures 380V. At turn on, the voltage of +B2 goes all the way to 480V, stressing the two 450V ecap. Once current is drawn from +B2 when the heaters warmup, it is not a problem. Changing to 500V ecaps solves the problem. However, applying +B2 before heaters warm up is not good and I will add a delay relay circuit for +B2..
3. The actual OPT is only 4K impedance, not 6K as shown in the schematic. Looking at the 845 curves, I cannot imagine how it can work. I will use the 4 ohm output to my 8 ohm speakers so that the load is effectively 8K. If this does not work, then I will replace the OPT with Audionote 10K 50W SE OPT. These cost about USD400 per pair in Singapore.
4. The max anode voltage of the 6AG7 from the datasheet is 300V. 320V were used, and actual measured voltage is 350V. I guess the 6AG7 will not last that long with such a design. If 300 Vpp swing is possible, then the anode of the 6AG7 can be 150+ 350 = 500V. Will this kill the tube? I am thinking of a simple rewiring to use a EL34 in place of the 6AG7. Bias is perhaps at 40mA. I am struggling to calculate the value of the cathode resistor of the EL34 at this setting as I still do not understand the curves.
Anyway, first listening before any mods – not remarkable. In fact the worst tube amp I have ever heard. My Ella EL34 amp sounds a lot better. But for USD600, you cannot go wrong. All the modifications I want to do above will only cost me a small amount of money. Even if I replace the OPTs with Audionotes, it will cost me only slightly more than one grand. And I get a vehicle to learn about SE amp designs. BTW, all the components used are good quality, PIOs for coupling, GE PP caps for PS filtering. PIOs and GE PP cathode resistor decoupling. Stainless steel housings. Military grade NOS tubes ( except for the 845 ). The Russian 6AG7 metal tube is a piece of art.
Any help to assist in my learning is much appreciated.
http://www.bezdz.com/Dl/T10A.htm
1. Bias of the 845 seems wrong. With 125 volts at the cathode of the current with 1150 B+ from the curves should be much higher than the 83mA calculated from 125/1k5. The actual measured voltage is 130V, close enough. I am changing the cathode resistor to 2k3. That should give a bias of 65mA, 150V at cathode, and B1+ at 1000V.
2. The 330V +B2 actually measures 380V. At turn on, the voltage of +B2 goes all the way to 480V, stressing the two 450V ecap. Once current is drawn from +B2 when the heaters warmup, it is not a problem. Changing to 500V ecaps solves the problem. However, applying +B2 before heaters warm up is not good and I will add a delay relay circuit for +B2..
3. The actual OPT is only 4K impedance, not 6K as shown in the schematic. Looking at the 845 curves, I cannot imagine how it can work. I will use the 4 ohm output to my 8 ohm speakers so that the load is effectively 8K. If this does not work, then I will replace the OPT with Audionote 10K 50W SE OPT. These cost about USD400 per pair in Singapore.
4. The max anode voltage of the 6AG7 from the datasheet is 300V. 320V were used, and actual measured voltage is 350V. I guess the 6AG7 will not last that long with such a design. If 300 Vpp swing is possible, then the anode of the 6AG7 can be 150+ 350 = 500V. Will this kill the tube? I am thinking of a simple rewiring to use a EL34 in place of the 6AG7. Bias is perhaps at 40mA. I am struggling to calculate the value of the cathode resistor of the EL34 at this setting as I still do not understand the curves.
Anyway, first listening before any mods – not remarkable. In fact the worst tube amp I have ever heard. My Ella EL34 amp sounds a lot better. But for USD600, you cannot go wrong. All the modifications I want to do above will only cost me a small amount of money. Even if I replace the OPTs with Audionotes, it will cost me only slightly more than one grand. And I get a vehicle to learn about SE amp designs. BTW, all the components used are good quality, PIOs for coupling, GE PP caps for PS filtering. PIOs and GE PP cathode resistor decoupling. Stainless steel housings. Military grade NOS tubes ( except for the 845 ). The Russian 6AG7 metal tube is a piece of art.
Any help to assist in my learning is much appreciated.
http://www.bezdz.com/Dl/T10A.htm
First is the amp working, if it is working why try to fix it.
1. Bias of the 845 seems wrong. With 125 volts at the cathode of the current with 1150 B+ from the curves should be much higher than the 83mA calculated from 125/1k5. The actual measured voltage is 130V, close enough. I am changing the cathode resistor to 2k3. That should give a bias of 65mA, 150V at cathode, and B1+ at 1000V.
Even if you change the resistor B1+ would be still 1150. Why should the B1+ drop. UNless your trans has higher resistance. On the pdf and actual measurements sometimes do not tally. Are using the exact valve and brand to the pdf you compared too?
2. The 330V +B2 actually measures 380V. At turn on, the voltage of +B2 goes all the way to 480V, stressing the two 450V ecap. Once current is drawn from +B2 when the heaters warmup, it is not a problem. Changing to 500V ecaps solves the problem. However, applying +B2 before heaters warm up is not good and I will add a delay relay circuit for +B2..
Most caps have a turn on surge but delays would be great
3. The actual OPT is only 4K impedance, not 6K as shown in the schematic. Looking at the 845 curves, I cannot imagine how it can work. I will use the 4 ohm output to my 8 ohm speakers so that the load is effectively 8K. If this does not work, then I will replace the OPT with Audionote 10K 50W SE OPT. These cost about USD400 per pair in Singapore.
Why not? something against using 4k? Have you plotted the load line for it? it works but not ideally to what is design too. How do you know it is 4K anyway?
4. The max anode voltage of the 6AG7 from the datasheet is 300V. 320V were used, and actual measured voltage is 350V. I guess the 6AG7 will not last that long with such a design. If 300 Vpp swing is possible, then the anode of the 6AG7 can be 150+ 350 = 500V. Will this kill the tube? I am thinking of a simple rewiring to use a EL34 in place of the 6AG7. Bias is perhaps at 40mA. I am struggling to calculate the value of the cathode resistor of the EL34 at this setting as I still do not understand the curves.
at 40ma at 320volts Vgk should be around -24 to -28. THen just used a 700 ohm resistor then assuming -28 Vgk.
It probably sounds bad due to the opt being to low in value. By the way voltage doublers never sound good compared to normal votlage. To much spikes or the charging capacitance.
At 1150 Volts, 10k would be the proper value of impedance which is commonly used not 4k,6k,8k but they design with that values.
1. Bias of the 845 seems wrong. With 125 volts at the cathode of the current with 1150 B+ from the curves should be much higher than the 83mA calculated from 125/1k5. The actual measured voltage is 130V, close enough. I am changing the cathode resistor to 2k3. That should give a bias of 65mA, 150V at cathode, and B1+ at 1000V.
Even if you change the resistor B1+ would be still 1150. Why should the B1+ drop. UNless your trans has higher resistance. On the pdf and actual measurements sometimes do not tally. Are using the exact valve and brand to the pdf you compared too?
2. The 330V +B2 actually measures 380V. At turn on, the voltage of +B2 goes all the way to 480V, stressing the two 450V ecap. Once current is drawn from +B2 when the heaters warmup, it is not a problem. Changing to 500V ecaps solves the problem. However, applying +B2 before heaters warm up is not good and I will add a delay relay circuit for +B2..
Most caps have a turn on surge but delays would be great
3. The actual OPT is only 4K impedance, not 6K as shown in the schematic. Looking at the 845 curves, I cannot imagine how it can work. I will use the 4 ohm output to my 8 ohm speakers so that the load is effectively 8K. If this does not work, then I will replace the OPT with Audionote 10K 50W SE OPT. These cost about USD400 per pair in Singapore.
Why not? something against using 4k? Have you plotted the load line for it? it works but not ideally to what is design too. How do you know it is 4K anyway?
4. The max anode voltage of the 6AG7 from the datasheet is 300V. 320V were used, and actual measured voltage is 350V. I guess the 6AG7 will not last that long with such a design. If 300 Vpp swing is possible, then the anode of the 6AG7 can be 150+ 350 = 500V. Will this kill the tube? I am thinking of a simple rewiring to use a EL34 in place of the 6AG7. Bias is perhaps at 40mA. I am struggling to calculate the value of the cathode resistor of the EL34 at this setting as I still do not understand the curves.
at 40ma at 320volts Vgk should be around -24 to -28. THen just used a 700 ohm resistor then assuming -28 Vgk.
It probably sounds bad due to the opt being to low in value. By the way voltage doublers never sound good compared to normal votlage. To much spikes or the charging capacitance.
At 1150 Volts, 10k would be the proper value of impedance which is commonly used not 4k,6k,8k but they design with that values.
First is the amp working, if it is working why try to fix it.
Yes it works, but the sound ? Adequate for vocal and jazz, but with one bass drum, the sound that emerges is totally undamped. My Philips music station sounds better!
Even if you change the resistor B1+ would be still 1150. Why should the B1+ drop. UNless your trans has higher resistance. On the pdf and actual measurements sometimes do not tally. Are using the exact valve and brand to the pdf you compared too?
Forgot to mention that I willl change the transformer tap to the 380V from 420V to lower +B1. I am not using the datasheet of the Chinese 845, just the RCA ones.
Why not? something against using 4k? Have you plotted the load line for it? it works but not ideally to what is design too. How do you know it is 4K anyway?
4K was printed on the transformer. I will take it out, pump in a 6V AC on the output, and measure the ouput voltage. The sq of the voltage will give me the impedance ratio. That will confrim if the schematic or the printing is correct. At 1150V and 4K, the loadline intersect the Y axis of the curve at 287 mA. If the 83mA bias voltage is correct, the voltage gain for positive swing and negative swing far from symetrical. Not good at all. Anyway, is the idea of using the 4 ohm tap to 8ohm a goo idea to get a 8K load for the 845? Will this work or will there be a side effect. I think that the low inductance of the primary will give me a bad bass response. Changing to a proper 10K OPT is perhaps the best way foward.
Thanks for your help.
Yes it works, but the sound ? Adequate for vocal and jazz, but with one bass drum, the sound that emerges is totally undamped. My Philips music station sounds better!
Even if you change the resistor B1+ would be still 1150. Why should the B1+ drop. UNless your trans has higher resistance. On the pdf and actual measurements sometimes do not tally. Are using the exact valve and brand to the pdf you compared too?
Forgot to mention that I willl change the transformer tap to the 380V from 420V to lower +B1. I am not using the datasheet of the Chinese 845, just the RCA ones.
Why not? something against using 4k? Have you plotted the load line for it? it works but not ideally to what is design too. How do you know it is 4K anyway?
4K was printed on the transformer. I will take it out, pump in a 6V AC on the output, and measure the ouput voltage. The sq of the voltage will give me the impedance ratio. That will confrim if the schematic or the printing is correct. At 1150V and 4K, the loadline intersect the Y axis of the curve at 287 mA. If the 83mA bias voltage is correct, the voltage gain for positive swing and negative swing far from symetrical. Not good at all. Anyway, is the idea of using the 4 ohm tap to 8ohm a goo idea to get a 8K load for the 845? Will this work or will there be a side effect. I think that the low inductance of the primary will give me a bad bass response. Changing to a proper 10K OPT is perhaps the best way foward.
Thanks for your help.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.