For example my 300B SE (d3a, CCS, source follower, 300B, 5K OPT) measuring at 1kHz.
PWR L_THD R_THD L_Ug THD R_Ug THD
1W 0.35% 0.31% 22V 0.11% 21V 0.12%
4W 0.86% 0.80% 47V 0.31% 45V 0.31%
6W 1.23% 1.08% 57V 0.39% 54V 0.38%
8W 1.52% 1.37% 65V 0.43% 63V 0.45%
Fonnesbek, most of soundcard distort at full level. Don't go over the -10dB.
Sample (1kHz, 4W/8R):
PWR L_THD R_THD L_Ug THD R_Ug THD
1W 0.35% 0.31% 22V 0.11% 21V 0.12%
4W 0.86% 0.80% 47V 0.31% 45V 0.31%
6W 1.23% 1.08% 57V 0.39% 54V 0.38%
8W 1.52% 1.37% 65V 0.43% 63V 0.45%
Fonnesbek, most of soundcard distort at full level. Don't go over the -10dB.
Sample (1kHz, 4W/8R):
Attachments
Last edited:
i guess, if. there is CCS, you are running d3a in triode mode, and i probable not at 2ma, correct ?
Yes, of course triode mode, 10mA.
LED bias (old type green LED), cascode DN2540 CCS, capacitor coupling to FET source follower (5mA), direct driving 300B (375V, 70mA, 5K:8 OPT).
The topology here is c3m in pentode, with current ~ 2ma, anode load of 110k Capacitor coupled to 300b, another words almost weco-91 clone.
Distortion figures of your driver and this one, should be different, that is given. However DHT driven by pentode is very interesting subjectively due to distortion cancelation and resulting distortion distribution.
Admittedly, the question remains, is we91 the best what can be done, or driving 300b with much more capable pentode, such as c3m, EL84, el34 etc. at 15+ma is the way to go. Personally i have made my choice, after trying 6sn7, 5687, 6em7, we810, c3m in triode and a few different operating points of c3m pentode, i have ended with similar to Thorsten "legacy" c3m-300b. However the only recommendation i can make is to try as much as possible. Everyone of the above amplifiers has it pluses and minuses. The same c3m pentode - 300b combo, configured a-la a09s plays small jazz combo wonderfully, but don't try to play Mahler 6th with it... running c3m at 15-16 ma brings Mahler to life, but unfortunately Ben Webster is not that intoxicating, like he was with a09 clone.
Again, my eyers, my speakers, my taste..
Independently of the driver topology, in my opinion Fonnesbek THD measuring too high.
Measuring many dozens of C3m (and C3g), all of them capable swinging 150Vpp at about 0.3-0.4% THD (most of them 200Vpp below 0.5% THD). Driving 300B with this low distortion signal, correct 300B SE circuit don't produce 1% THD at 1W.
Measuring many dozens of C3m (and C3g), all of them capable swinging 150Vpp at about 0.3-0.4% THD (most of them 200Vpp below 0.5% THD). Driving 300B with this low distortion signal, correct 300B SE circuit don't produce 1% THD at 1W.
Whatever you have mesured in triode-connected c3m has nothing to do with the discussed amplifier.
In operating mode of a09 (140 Ua, 66v G2, 350v B+, 110k anode R) at 20vrms
c3g has about 2 %THD . I did not measure many dozens, just about 6 or 8,
with 16 ma current and G2 =150v (Siemens recommended oper. points), c3m
has about 0.8 % THD at 20vrms and about 2.3% at 50vrms.
If you know about pentode caable of swinging 150vpp with 0.3-0.4 % THD please let me know.
I finally did the rebuild of the input stage, as I expected the distortion is much lower, I havn't had a time to listne to it yet.
I'm running the c3m at 18mA utilizing the "best pentode" described by Frank Blöhbaum in LinearAudio vol.1
2,73V RMS 1 W 0,3113%
5.65V RMS 4 W 0,7857%
8.00V 8 Watt 1,629%
I would like to lower the 50Hz noise, but that requires a rebuild of the power supply, I'm using a eletronic chocke which requiers bigge caps then I can use with GZ37, I think I'm pushing it how with the first cap being 60uF
I'm running the c3m at 18mA utilizing the "best pentode" described by Frank Blöhbaum in LinearAudio vol.1
2,73V RMS 1 W 0,3113%
5.65V RMS 4 W 0,7857%
8.00V 8 Watt 1,629%
I would like to lower the 50Hz noise, but that requires a rebuild of the power supply, I'm using a eletronic chocke which requiers bigge caps then I can use with GZ37, I think I'm pushing it how with the first cap being 60uF
Attachments
Are you using DC on the 300B filament?
If so, what is the ripple on the DC filament supply?
AC poured filaments can use a balance pot to get the 2 ends of the filament to balance to the "mid-point" of the filament. But unfortunately, that is only part of the problem. The magnetic field of 1.25A AC filament current can cause vibration of the filament wires if there is any difference of spacing from the filament to the 2 sides of the plate (magnetic field of filament, is attracted to a steel plate).
You will find some 300Bs have less hum than others when the filament pot is adjusted optimally for each tube.
Hmm. Stainless plates anybody?
Who is producing such DHTs?
If so, what is the ripple on the DC filament supply?
AC poured filaments can use a balance pot to get the 2 ends of the filament to balance to the "mid-point" of the filament. But unfortunately, that is only part of the problem. The magnetic field of 1.25A AC filament current can cause vibration of the filament wires if there is any difference of spacing from the filament to the 2 sides of the plate (magnetic field of filament, is attracted to a steel plate).
You will find some 300Bs have less hum than others when the filament pot is adjusted optimally for each tube.
Hmm. Stainless plates anybody?
Who is producing such DHTs?
The 1kHz carrier has +12dB level, the 50Hz is at about -57dB (the measurement software vertical subdivision is almos useless, the decade division is commonly used).
About 69dB distance for 50Hz hum is enough good.
If the 300B heater is not so good (poor AC cancelisation, DC filament with hum etc.) you have +/- 50Hz, 100Hz, etc. sidebands around the carrier.
About 69dB distance for 50Hz hum is enough good.
If the 300B heater is not so good (poor AC cancelisation, DC filament with hum etc.) you have +/- 50Hz, 100Hz, etc. sidebands around the carrier.
I'm using DC for both the 300B an C3m tentlabs filament supply and heater module, so I don't think the 50Hz originates from the heaters.
Plus if it was from the heaters it would be 100Hz as they have full wave briged rectifiers.
Hmm my B+ is also fullwave to halfs GZ37 with AC heater, so could the 50Hz be from the rectifier or a ground loop
Plus if it was from the heaters it would be 100Hz as they have full wave briged rectifiers.
Hmm my B+ is also fullwave to halfs GZ37 with AC heater, so could the 50Hz be from the rectifier or a ground loop
c3m in A06-3
Hi All,
I am reading the thread with great interest as I use Yamamoto's A-06-3 amplifier which is c3m driving AD1 as output. The circuit is similar (low current bias ~2mA) but with CCS and Zeners to stabilize the B+ and grid supply g2 to the c3m, plus the cathode is biased by diodes.
Recently I need to replace the c3m tubes and what I found is that different c3m tubes appear to bias very differently, resulting in the anode voltage of the c3m varying from 10s of volts to 150+ volts (nominal is 120V). Out of about 10 pairs of c3m I tested (Siemens, Telefunken) not one pair result in the intended anode voltage range.
I wonder if this is common in c3m tubes or in this type of topology? This makes it very difficult to find replacement tubes for the c3m.
Thanks.
Hi All,
I am reading the thread with great interest as I use Yamamoto's A-06-3 amplifier which is c3m driving AD1 as output. The circuit is similar (low current bias ~2mA) but with CCS and Zeners to stabilize the B+ and grid supply g2 to the c3m, plus the cathode is biased by diodes.
Recently I need to replace the c3m tubes and what I found is that different c3m tubes appear to bias very differently, resulting in the anode voltage of the c3m varying from 10s of volts to 150+ volts (nominal is 120V). Out of about 10 pairs of c3m I tested (Siemens, Telefunken) not one pair result in the intended anode voltage range.
I wonder if this is common in c3m tubes or in this type of topology? This makes it very difficult to find replacement tubes for the c3m.
Thanks.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.