Hi
I have published a project on Audioreview magazine in Italy about a push pull amp that use EL34 or KT120 with fixed bias, pentode configuration.
It uses a 6922 for input, ECC99 for splitter.
In fig. 1 there is the inut diagram
In fig. 2 the splitter
In fig. 3 the output
The voltage is around 430 vdc, the bias current is zout 30mA.
I used a one trafo for HT supply with 4 caps of 470uF -450 Vdc.
One toroidal trafo only for filaments and neg bias.
The Output trafo is made by F.I.A.T. of Roma with a C core; ratio is 27:1 the secondary is one at 5 ohm.
The global feedback is about 18 dB
In fig. 6 there is a very intersting test we normally use in the Lab of Audio, it is a particular test to check the power of an amp with continuous or impulsive signal
The AP1 is driven by a software written for this purpose, we have a resistive load with. 6-8- 5,3-4 -3,2- 2,7- 2,2- 2 ohm the there is an interface driven by AP1 that change in sequence the load, the software detect in automatic the distortion (written on a window where is possible to put the information of the amp as max power and the max level of THD we want to test).
When the THD defined previously is reached the software change the load and so on.
This test is continos and impulsive; in theory a perfect amp have a vertical line; more the line became orizontal less the amp can delivery currento on load.
In fig. 6 there is a result with EL34 at 35 mA of bias; the power on 8 ohms is 41.5 Watts , at 4 ohm are 26.4 (it is reasonable); if we look on red line (contineous input signals) we see that the line become horizontally for a lower value of load.
The blue line is impulsive and it is better, of course.
In fig. 7 there is the freq. response for three different power; it is very wide.
In fig. 8 there is a THD vs freq. test for four differetn power level; the results is good because the line is costant for frequency with an increment at low freq.
The we put the KT120 on this circuit and something is happened.
In fig. 9 the test with a bias current at 90 mA (in this situation we have class A until 32 Watts), we can see that the power is much better due to the internal lower resistence of the tube compare with the EL34.
We reach 57 Watts impulsive on 8 ohm and 76,9 watt on 4 ohms.
In fig. 8_90 there is a freq. response that is similar with EL34.
In fig. 6_Bias there are two results for a bias at 55 mA and 90 mA , we can see the difference. Also in this case a good results mainly for the higher current, of course.
The use of a KT120 or KT150, where is possible, give us a better performance because the capacity to drive more current due to a lower Rp ; mainly when the trafo, at low and high frequency, has the tendence to get a lower L.
A good performance are obtained with 6CA7 EH, very very good. It has a 30% less in the Rp of EL34B Tungsol I used.
Ciao
Walter
I have published a project on Audioreview magazine in Italy about a push pull amp that use EL34 or KT120 with fixed bias, pentode configuration.
It uses a 6922 for input, ECC99 for splitter.
In fig. 1 there is the inut diagram
In fig. 2 the splitter
In fig. 3 the output
The voltage is around 430 vdc, the bias current is zout 30mA.
I used a one trafo for HT supply with 4 caps of 470uF -450 Vdc.
One toroidal trafo only for filaments and neg bias.
The Output trafo is made by F.I.A.T. of Roma with a C core; ratio is 27:1 the secondary is one at 5 ohm.
The global feedback is about 18 dB
In fig. 6 there is a very intersting test we normally use in the Lab of Audio, it is a particular test to check the power of an amp with continuous or impulsive signal
The AP1 is driven by a software written for this purpose, we have a resistive load with. 6-8- 5,3-4 -3,2- 2,7- 2,2- 2 ohm the there is an interface driven by AP1 that change in sequence the load, the software detect in automatic the distortion (written on a window where is possible to put the information of the amp as max power and the max level of THD we want to test).
When the THD defined previously is reached the software change the load and so on.
This test is continos and impulsive; in theory a perfect amp have a vertical line; more the line became orizontal less the amp can delivery currento on load.
In fig. 6 there is a result with EL34 at 35 mA of bias; the power on 8 ohms is 41.5 Watts , at 4 ohm are 26.4 (it is reasonable); if we look on red line (contineous input signals) we see that the line become horizontally for a lower value of load.
The blue line is impulsive and it is better, of course.
In fig. 7 there is the freq. response for three different power; it is very wide.
In fig. 8 there is a THD vs freq. test for four differetn power level; the results is good because the line is costant for frequency with an increment at low freq.
The we put the KT120 on this circuit and something is happened.
In fig. 9 the test with a bias current at 90 mA (in this situation we have class A until 32 Watts), we can see that the power is much better due to the internal lower resistence of the tube compare with the EL34.
We reach 57 Watts impulsive on 8 ohm and 76,9 watt on 4 ohms.
In fig. 8_90 there is a freq. response that is similar with EL34.
In fig. 6_Bias there are two results for a bias at 55 mA and 90 mA , we can see the difference. Also in this case a good results mainly for the higher current, of course.
The use of a KT120 or KT150, where is possible, give us a better performance because the capacity to drive more current due to a lower Rp ; mainly when the trafo, at low and high frequency, has the tendence to get a lower L.
A good performance are obtained with 6CA7 EH, very very good. It has a 30% less in the Rp of EL34B Tungsol I used.
Ciao
Walter
Attachments
Last edited by a moderator:
Yes, of course
In reality I have test this circuit with: EL34B, 6CA7, 5881, 6L6GC, KT88 and KT120.
I decided to publish the two edge of the mesurements, in the middle the all tubes works in a right way.
In each configuration the bandwidth at open loop is around 60 kHz, this to have an idea also of the quality of the O.T.
I will post in a next future a mono amp with 4 x KT150 .
Walter
In reality I have test this circuit with: EL34B, 6CA7, 5881, 6L6GC, KT88 and KT120.
I decided to publish the two edge of the mesurements, in the middle the all tubes works in a right way.
In each configuration the bandwidth at open loop is around 60 kHz, this to have an idea also of the quality of the O.T.
I will post in a next future a mono amp with 4 x KT150 .
Walter
I agree, with KT150 you can reach better results but it is expensive.
I will show a 4xKT150 monos with a very good performances.
In this circuit a very good solution is the use of 6CA7.
If you look on freq. answer, with 18 dB of feedback ,I got -2db at 150 Khz at 1 watt. And 120 kHz at 10 watt.
Walter
Walter
I will show a 4xKT150 monos with a very good performances.
In this circuit a very good solution is the use of 6CA7.
If you look on freq. answer, with 18 dB of feedback ,I got -2db at 150 Khz at 1 watt. And 120 kHz at 10 watt.
Walter
Walter
A serious test published in the journal can be targeted not only at the audience of consumers but also manufacturers of tubes. Perhaps EL156 will not only clone the Chinese. And it is possible that relatively affordable 6P45S will suit many according to the test results. For example, in Russia KT150 is considered an expensive tube and it is difficult to buy even though it is produced in the country.
Please excuse me, but it's not my fault. I first type in a text editor in normal size. And I insert the finished text. Why the font is larger than I have not yet figured out.Kindly turn off bold and/or large size formatting, they are akin to shouting online.
EL509 is a curious tube with an extremely powerful cathode and a reduced load resistance of about 1 kΩ per tube. But it does not have a very powerful anode. Although the Russian analogues 6P45S stand for a short time up to 45W!
http://audioportal.hi-fi.ru/attachment.php?attachmentid=180828&d=1464587589
Last edited:
In the former USSR, these tubes were very many produced and they are still there in warehouses in East European countries and CMEA countries. Svetlana and now manufactures and sells but for audio. There is even a version in oktal base. But maybe they are harder to find. There are people from Lithuania refer to it on the forum. I personally don't know him. But the reviews seems positive. http://www.diyaudio.com/forums/tubes-valves/4691-russian-el509-6kg6-quality-question.html
There is also on this forum my friend from Latvia. You can write to him. diyAudio - View Profile: LinuksGuru
Last edited:
I decided to do this mainly as a challenge to reach a good results in one configuration that is considered not so easy.
The FB is 18 max, you can choose also for 12 dB with always a good results.
Then the main reason is to use a simple circuit with a very good power supply as capacity.
Then the use o double C trafo ( normal wire, but Litz will be used) that has non a very high L but the use of KT120 helps.
The output impedance with 18 dB of FB is around 1 ohm; the ratio of the trafo is 1:28, single secondary at 5 ohms
Bye
Walter
The FB is 18 max, you can choose also for 12 dB with always a good results.
Then the main reason is to use a simple circuit with a very good power supply as capacity.
Then the use o double C trafo ( normal wire, but Litz will be used) that has non a very high L but the use of KT120 helps.
The output impedance with 18 dB of FB is around 1 ohm; the ratio of the trafo is 1:28, single secondary at 5 ohms
Bye
Walter
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.