An intriguing 6S45P direct coulpled 300B amplifier
I’d like to submit and discuss the schematic of an integrated amplifier (direct connected to a CD-player … without pre).
We (no-profit association – Agorà) would like to realize this project together with the child that take part to the “Corso di Elettronica” (electronic course, promoted by our association).
Since I’m not an expert in Electronic, I’d be very grateful if somebody could help us in this adventure.
Points to be defined:
Current setting of 6S45P and bias of 300B:
1) 15-20 mA …with “combined” anode load (H + R; i.e. H=50H/400DCR+ R=3.000 ohm).
2) 35 mA … with “pure” choke anode load (i.e. 170H/2000DCR).
6S45P-Bias:
1) vacuum tube diode on the cathode (which one?).
2) battery on the cathode.
3) RC on the cathode.
4) RC on the cathode (with R being a potentiometer … in order to match the two 6S45P).
Input transformer:
1) Sowther 3575 in configuration 1:1
2) Sowther 8920 in configuration 1:1 having the two secondary in parallel.
3) Sowther 8920 in configuration 1:2 … with a less power input-driver tube (i.e. ECC99).
Power supply:
1) L2 and L4 --> 30H/100mA/190DCR
2) V ?
3) C1=160 uF, C3=80 uF, C5=1 uF
I’d like to submit and discuss the schematic of an integrated amplifier (direct connected to a CD-player … without pre).
We (no-profit association – Agorà) would like to realize this project together with the child that take part to the “Corso di Elettronica” (electronic course, promoted by our association).
Since I’m not an expert in Electronic, I’d be very grateful if somebody could help us in this adventure.
Points to be defined:
Current setting of 6S45P and bias of 300B:
1) 15-20 mA …with “combined” anode load (H + R; i.e. H=50H/400DCR+ R=3.000 ohm).
2) 35 mA … with “pure” choke anode load (i.e. 170H/2000DCR).
6S45P-Bias:
1) vacuum tube diode on the cathode (which one?).
2) battery on the cathode.
3) RC on the cathode.
4) RC on the cathode (with R being a potentiometer … in order to match the two 6S45P).
Input transformer:
1) Sowther 3575 in configuration 1:1
2) Sowther 8920 in configuration 1:1 having the two secondary in parallel.
3) Sowther 8920 in configuration 1:2 … with a less power input-driver tube (i.e. ECC99).
Power supply:
1) L2 and L4 --> 30H/100mA/190DCR
2) V ?
3) C1=160 uF, C3=80 uF, C5=1 uF
Attachments
Nice looking project!
6S45
I don't quite understand the need for a combined choke. If it's purely dc considerations it is possible to adjust the anode voltage as you have a separate PS. For 6S45 even 50H would suffice IMO.
So my choice would be 50H @15mA.
6S45 bias
Vacuum tube? Don't think any will give you such a low bias at 15mA, let alone 35mA. If you still think you must use it try several 80 in parallel (2-4?)
IME rechargable batteries work well with the 6S45. Just don't ask what type as they all sound annoyingly different (as capacitors) in this position. As you have an input transformer it may be a much better sonic solution to connect the battery in series with the transformer windings.
regards
6S45
I don't quite understand the need for a combined choke. If it's purely dc considerations it is possible to adjust the anode voltage as you have a separate PS. For 6S45 even 50H would suffice IMO.
So my choice would be 50H @15mA.
6S45 bias
Vacuum tube? Don't think any will give you such a low bias at 15mA, let alone 35mA. If you still think you must use it try several 80 in parallel (2-4?)
IME rechargable batteries work well with the 6S45. Just don't ask what type as they all sound annoyingly different (as capacitors) in this position. As you have an input transformer it may be a much better sonic solution to connect the battery in series with the transformer windings.
regards
Re: An intriguing 6S45P direct coulpled 300B amplifier
Konnichiwa,
From experience with the 6S45-PE, you may find neither chosen operating point too good sounding. I would recommend 170V/10mA as a starting point for optimisation.
In my experience and if the desire is not to exploit the maximum power of the 300B, try 350V/60mA/5K load with 74V Bias. Then consider the use of one or two LNR Diodes, carefully matched to the circuit (that requires access to serious Audio Analyser and the skills to use it, plus a box full of different diodes). If you are not certain how select the correct LNR diode, leave them out as the wrong diode will make things worse rather than better.
May I suggest a better solution entierly? Battery on Grid. As you already have intended an input transformer (for an integrated amplifier unit a Volume Control Transformer may be a better choice) just place the battery in the ground lead of the Transformer making sure to bypass the battery with a capacitor.
The 6S45 choke loaded will allow the required 74V Peak at around 0.5% 2nd harmonics, so the driver stage will be very linear, using a LNR Diode will not likely make things significantly better.
Then to correctly Bias the 300B add a small adjustable resistor in series with the 6S45-PE Anode Load choke which allows the 300B Bias to be trimmed.
The suggested operating points will allow around 6 - 7 Watt RMS @ 5% THD (mostly 2nd Harmonics) without using LNR Diodes and will require around 1.3V RMS for full output with a 1:1 Input connection.
Sayonara
Konnichiwa,
Antonio Tucci said:
From experience with the 6S45-PE, you may find neither chosen operating point too good sounding. I would recommend 170V/10mA as a starting point for optimisation.
Antonio Tucci said:
In my experience and if the desire is not to exploit the maximum power of the 300B, try 350V/60mA/5K load with 74V Bias. Then consider the use of one or two LNR Diodes, carefully matched to the circuit (that requires access to serious Audio Analyser and the skills to use it, plus a box full of different diodes). If you are not certain how select the correct LNR diode, leave them out as the wrong diode will make things worse rather than better.
Antonio Tucci said:
May I suggest a better solution entierly? Battery on Grid. As you already have intended an input transformer (for an integrated amplifier unit a Volume Control Transformer may be a better choice) just place the battery in the ground lead of the Transformer making sure to bypass the battery with a capacitor.
The 6S45 choke loaded will allow the required 74V Peak at around 0.5% 2nd harmonics, so the driver stage will be very linear, using a LNR Diode will not likely make things significantly better.
Then to correctly Bias the 300B add a small adjustable resistor in series with the 6S45-PE Anode Load choke which allows the 300B Bias to be trimmed.
The suggested operating points will allow around 6 - 7 Watt RMS @ 5% THD (mostly 2nd Harmonics) without using LNR Diodes and will require around 1.3V RMS for full output with a 1:1 Input connection.
Sayonara
Some specifications...
KYW,
thank you for your replay.
Specific questions:
1) 6S45P Operating point.
OK for your suggestion. However ... how to obtain the bias for the 300B with only 10 mA ? We should have to add a 17Kohm R (better ... 17.000-400 ohm of the choke) in series with the anode choke.
Can the high value of R represent a problem ... or make worse the operating condition of 6S45P ?
2) Battery bypassed by a capacitor on the ground lag of the input Transformer.
OK...but which V value and C value ? ... 2.4 V ? ... uF ?
3) LNR diode on the cathode of 300B. (what does LNR mean ?)
Why do you believe that the use of 6CG3 may represent a problem ? Since it has been largely experimentated and tested by Dennis Boyle ... why not to use it ?
Dennis Boyle told me that a mixed bias is better than a fixed bias. I do not know the reasons of such assertion. However, I believe that it must be important ... if he utilizes, in addition to the stacked supply, a LNR diode on the cathode.
Furthermore, eliminating the LNR diode ... what we need on the cathode of 300B ? ... nothing ? ...or a Capacitor ?
KYW,
thank you for your replay.
Specific questions:
1) 6S45P Operating point.
OK for your suggestion. However ... how to obtain the bias for the 300B with only 10 mA ? We should have to add a 17Kohm R (better ... 17.000-400 ohm of the choke) in series with the anode choke.
Can the high value of R represent a problem ... or make worse the operating condition of 6S45P ?
2) Battery bypassed by a capacitor on the ground lag of the input Transformer.
OK...but which V value and C value ? ... 2.4 V ? ... uF ?
3) LNR diode on the cathode of 300B. (what does LNR mean ?)
Why do you believe that the use of 6CG3 may represent a problem ? Since it has been largely experimentated and tested by Dennis Boyle ... why not to use it ?
Dennis Boyle told me that a mixed bias is better than a fixed bias. I do not know the reasons of such assertion. However, I believe that it must be important ... if he utilizes, in addition to the stacked supply, a LNR diode on the cathode.
Furthermore, eliminating the LNR diode ... what we need on the cathode of 300B ? ... nothing ? ...or a Capacitor ?
For analog...
...The combined load (choke+R) is necessary to obtain the bias for the 300B.
Alternatively, with 15 mA ... how to obtain 70-75 for the bias ?
...The combined load (choke+R) is necessary to obtain the bias for the 300B.
Alternatively, with 15 mA ... how to obtain 70-75 for the bias ?
Are you sure that a 5AR4 is the correct rectifier for the 6C45 B+? I don't see how you can draw the B+ from its center-tapped filament.
John
I believe the 6CG3 will work in this position. Dennis Boyle's Chimera Labs website shows lower distortion figures when using (low internal resistance) damper diodes for setting bias.
John
Jlsem,
I do not understand the problem with 5AR4 ...
Please, may you better explain the problem ?
I do not understand the problem with 5AR4 ...
Please, may you better explain the problem ?
B+
jlsem,
you can take B+ ad libitum from either side of the DH rctf. eliminating the pot (or two matched R's), it does not change anything except you get better and more equall and linear exploatation of the rctf. cathode, compairing with the case when it's taken from one side.Or better try a separate,good, with e- static shield,bifillary wound heater T/X!!
Regards,
Yugovitz
jlsem said:
jlsem,
you can take B+ ad libitum from either side of the DH rctf. eliminating the pot (or two matched R's), it does not change anything except you get better and more equall and linear exploatation of the rctf. cathode, compairing with the case when it's taken from one side.Or better try a separate,good, with e- static shield,bifillary wound heater T/X!!
Regards,
Yugovitz
Re: Some specifications...
Konnichiwa,
Well, first, to get 74V Bias @ 10mA you need 7400 Ohm (including Choke DCR). Now you do need some form of interlock against the 300B Turning on without Bias anyway, how about a nice 48V Relais with a 4800 Ohm Coil? This will turn on the HT for the 300B once the 6S45 draws enough current.
Not as far as I am concerend, as the resistance value is still rather low.
I'd say 2.4V looks good enough, the Capacitor - it's a question of quality mainly. I would use a 0.1uF Silverleaf/Mica type I guess, others have used a few uF PIO.
Also, you can go for higher inductance, lower current Chokes in the PSU for the 6S45. If you intend to build more than one amplifier or two, getting things made EXACTLY TO SPEC is cheaper than buying off the shelf. If you get Chokes made you can specify 4 to 8 section layered disk windings which will reduce stray capacitance no end.
Also, with custom chokes consider having them made in two sections with a deliberatly very high leakage inductance between the halves. If you then loop the choke into the PSU so that one section is in the positive and one section is in the negative line the leakage inductance between the halves in effect operates as common mode filter, you get a lot of added common mode mains born RF noise rejection that way, essentially for free.
LNR = Lineariser. This is a common HF design feature here applied to Audio as per the stuff from Hoekstra.
The point is that the 6CG3 works for a very specific circuit, with a specific driver valve and a specific operating point. Unless you copy the Axiom Circuit exactly it is likely that the result will not neccesarily as intended. Of course it MAY be, hard to say, hence my point that you must be able to compare and qualify perfromance if you wish to implement this technique, otherwise the LNR can easily turn into a NLNR (Nonlineariser).
I think the mixed bias is of little relevance as the diode contributes very little Bias. What is important is that the bias varies nonlineary with current. The result is, that if the nonlinearity is chosen well the nonlinearity of the system is reduced, as both nonlinearitier have opposing signs.
We need nothing, straight to PSU Return (ground).
I will also say that having AC on the heaters of the 300B WITHOUT significant added hum cancellation strikes me as adventerous. When laying out the chassis and circuit provide for enough room and heat rejection capacity to allow a DC supply (perhaps Guido Tent's module).
I have attached a Circuit I did some partial testing for and which should work pretty well, but which I never quite got around to complete which might be of interrest to you.
Sayonara
Konnichiwa,
Antonio Tucci said:
Well, first, to get 74V Bias @ 10mA you need 7400 Ohm (including Choke DCR). Now you do need some form of interlock against the 300B Turning on without Bias anyway, how about a nice 48V Relais with a 4800 Ohm Coil? This will turn on the HT for the 300B once the 6S45 draws enough current.
Antonio Tucci said:
Not as far as I am concerend, as the resistance value is still rather low.
Antonio Tucci said:
I'd say 2.4V looks good enough, the Capacitor - it's a question of quality mainly. I would use a 0.1uF Silverleaf/Mica type I guess, others have used a few uF PIO.
Also, you can go for higher inductance, lower current Chokes in the PSU for the 6S45. If you intend to build more than one amplifier or two, getting things made EXACTLY TO SPEC is cheaper than buying off the shelf. If you get Chokes made you can specify 4 to 8 section layered disk windings which will reduce stray capacitance no end.
Also, with custom chokes consider having them made in two sections with a deliberatly very high leakage inductance between the halves. If you then loop the choke into the PSU so that one section is in the positive and one section is in the negative line the leakage inductance between the halves in effect operates as common mode filter, you get a lot of added common mode mains born RF noise rejection that way, essentially for free.
Antonio Tucci said:
LNR = Lineariser. This is a common HF design feature here applied to Audio as per the stuff from Hoekstra.
Antonio Tucci said:
The point is that the 6CG3 works for a very specific circuit, with a specific driver valve and a specific operating point. Unless you copy the Axiom Circuit exactly it is likely that the result will not neccesarily as intended. Of course it MAY be, hard to say, hence my point that you must be able to compare and qualify perfromance if you wish to implement this technique, otherwise the LNR can easily turn into a NLNR (Nonlineariser).
Antonio Tucci said:
I think the mixed bias is of little relevance as the diode contributes very little Bias. What is important is that the bias varies nonlineary with current. The result is, that if the nonlinearity is chosen well the nonlinearity of the system is reduced, as both nonlinearitier have opposing signs.
Antonio Tucci said:
We need nothing, straight to PSU Return (ground).
I will also say that having AC on the heaters of the 300B WITHOUT significant added hum cancellation strikes me as adventerous. When laying out the chassis and circuit provide for enough room and heat rejection capacity to allow a DC supply (perhaps Guido Tent's module).
I have attached a Circuit I did some partial testing for and which should work pretty well, but which I never quite got around to complete which might be of interrest to you.
Sayonara
Attachments
Antonio, I think the circuit you are investigating has the advantage---and perhaps a significant advantage at that---of providing a degree of PSU noise cancellation. Any non-signal related PSU disturbances will show up on the grid and cathode of the 300B, cancelling to the degree those disturbances are not reduced by your driver anode choke. Elegant indeed.
My mistake. I have been playing with 6V rectifiers and damper diodes so much lately that I'd forgotten the cathode is tied to the filament on 5V ones.
John
KYW,
… very interesting suggestions ! However, I’d like to submit some specific questions about them.
If so … is your proposal substantially different from the split chokes in a “hum bucking” configuration (providing phase cancellation of hum and noise) proposed by others ?
… very interesting suggestions ! However, I’d like to submit some specific questions about them.
You are right, my stupid wrong calculations gave me a R value of 17.000 ohm … rather high for a correct load. With 7.400 ohm all the things are OK.
I already have a DIY line conditioner with separate 220 Vac outputs (2 for the anodic and 2 for the filament lines) and with a variable delay between filaments and anodic lines. So, turning on the anodic lines (both 6S45P and 300B) with a delay of 30-40 sec … is still necessary the relay you suggest ?
It is no clear the use and the configuration of the custom chokes you suggested. What I have understood is a two-choke component … with one section (choke 1) connected, in series, in the negative line and one section (choke 2) connected, in series, in the positive line (as reported below).
If so … is your proposal substantially different from the split chokes in a “hum bucking” configuration (providing phase cancellation of hum and noise) proposed by others ?
AC on the heaters of the 300B. I have considered the eventuality of an excessive hum. In this case I’ll add a rectifier+CLC section (I do not love the regulated supply on the heaters of DHT). In this case, however, I have a doubt … which characteristics should have the secondary of the filament transformer ? Ok for the current (I think that 4 A is enough) … but how many volts ? In some schematics it is reported a secondary with 5V … while in others …6.3 V.
Attachments
Konnichiwa,
Well, it would make me feel better to have it. Using the relais interlock I suggest makes sure the +B for the 300B is available only if the bias is present. This way switching the Amp on without driver valve or having a driver valve fail or having something in the Amp fail that causes the driver valve to stop conducting will not allow the 300B to be fried.
As I have build a few DC coupled Amp's where the Bias relied on the driver circuit I would not want to be withoput this type of interlock.
Not? Look at this schematic:
This shows one of the two +B Chokes (the one marked FC-6 which is a 6H "Wideband" Choke) connected in that fashion.
I am not sure what "hum bucking" confiuration you refer to. Any reference?
Fine.
In my Schematic above the Winding is basically 7V (lengthy and difficult to explain, so don't ask).
In a new implementation I am using 8V tapped at 5V & 2.5V AC to be Schottky rectified and to allow CRC/CLC Filtering with adjustable Heater Voltage for 4V/5V/7.5V DC (PX-4/PX-25; 300B/50; 10/10Y) with also options for 2.5 & 5V AC.
Your rectifier and filtering leads to losses in voltage. If these would lead to a DC below 5V you need to boost the Voltage to compensate.
As I find it hard to lengthen something when you have already cut it short I would suggest to start from a higher voltage than theory suggests and use series resistors to drop the excess DC voltage. This also helps to slow down and reduce the turn on currenet surge which is more kind to the valve filament.
Sayonara
Antonio Tucci said:
Well, it would make me feel better to have it. Using the relais interlock I suggest makes sure the +B for the 300B is available only if the bias is present. This way switching the Amp on without driver valve or having a driver valve fail or having something in the Amp fail that causes the driver valve to stop conducting will not allow the 300B to be fried.
As I have build a few DC coupled Amp's where the Bias relied on the driver circuit I would not want to be withoput this type of interlock.
Antonio Tucci said:
Not? Look at this schematic:
This shows one of the two +B Chokes (the one marked FC-6 which is a 6H "Wideband" Choke) connected in that fashion.
Antonio Tucci said:
I am not sure what "hum bucking" confiuration you refer to. Any reference?
Antonio Tucci said:
Fine.
Antonio Tucci said:
In my Schematic above the Winding is basically 7V (lengthy and difficult to explain, so don't ask).
In a new implementation I am using 8V tapped at 5V & 2.5V AC to be Schottky rectified and to allow CRC/CLC Filtering with adjustable Heater Voltage for 4V/5V/7.5V DC (PX-4/PX-25; 300B/50; 10/10Y) with also options for 2.5 & 5V AC.
Antonio Tucci said:
Your rectifier and filtering leads to losses in voltage. If these would lead to a DC below 5V you need to boost the Voltage to compensate.
As I find it hard to lengthen something when you have already cut it short I would suggest to start from a higher voltage than theory suggests and use series resistors to drop the excess DC voltage. This also helps to slow down and reduce the turn on currenet surge which is more kind to the valve filament.
Sayonara
5AR4 rectifier cathode is indirectly heated and the cathode is connected to one of the filament pins, pin 8 I believe. The shown topology would be ok with a 5U4 which has a direct heated filamentary cathode. If use of the 5AR4 is contemplated the pot should be removed at the B+ connection made to pin 8. Did I miss something here?😕
No, you do not miss anything. You are probably right !
...but why with indirected heated diodes the pot does not work ?
Now all is clear ! Many thanks.
...and now I am again confused.
Which characteristics must have the secondary for the 300B heaters ?
I'd have contemplated to request a secondary with the following charcteristics: 4A and 2.5V, 5.0V, 7.0V center tapped.
Is it correct ? .. or do you suggest 4A and 2.5V, 5.0V, 8.0V center tapped ?
Here is a reference:
http://home.att.net/~chimeraone/axiom300bdesigninfo.html
Konnichiwa,
I would recommedn 0-2.5V-5V-8V @ 4A
This gives 2.5V, 5V, 5.5V & 8V AC, to allow a lot of different Valves with AC and DC Heaters.
I don't see any special in the use of Chokes in the Axiom. Looks to me just like a fancy name for using an LCLCL filter instead of CLC.
Sayonara
Antonio Tucci said:
I would recommedn 0-2.5V-5V-8V @ 4A
This gives 2.5V, 5V, 5.5V & 8V AC, to allow a lot of different Valves with AC and DC Heaters.
Antonio Tucci said:
I don't see any special in the use of Chokes in the Axiom. Looks to me just like a fancy name for using an LCLCL filter instead of CLC.
Sayonara
Something is wrong in my calculations !
KYW,
Something is wrong in my calculations !
considering that the current for the heaters of 300B is 1.2 A (as reported in the data sheet) …
and suppose that I want to have the filament of 300B supplied with 5Vcc in a CLC combination …
with 8Vac on the secondary of the transformer for the heaters, I have,
- After the bridge rectifier and the first capacitor: 8x1.4-1.4 = 9.8
- Reduction produced by the power choke (let say 0.1H, 4A, 0.47 DCR): V=RxI = 0.47x1.2 = 0.56V
- After the choke: 9.9–0.56 = 9.34 Vcc …????
with 5Vac on the secondary of the transformer for the heaters, I have,
- After the bridge rectifier and the first capacitor: 5x1.4-1.4 = 5.6
- Reduction produced by the power choke (let say 0.1H, 4A, 0.47 DCR): V=RxI = 0.47x1.2 = 0.56V
- After the choke: 5.6–0.56 = 5.04 !!!!
How can I supply the heaters of 300B with a Vcc voltage (with a CLC filter) having 8 Vac on the secondary of heaters transformer ?
... Where is the error ?
KYW,
Something is wrong in my calculations !
considering that the current for the heaters of 300B is 1.2 A (as reported in the data sheet) …
and suppose that I want to have the filament of 300B supplied with 5Vcc in a CLC combination …
with 8Vac on the secondary of the transformer for the heaters, I have,
- After the bridge rectifier and the first capacitor: 8x1.4-1.4 = 9.8
- Reduction produced by the power choke (let say 0.1H, 4A, 0.47 DCR): V=RxI = 0.47x1.2 = 0.56V
- After the choke: 9.9–0.56 = 9.34 Vcc …????
with 5Vac on the secondary of the transformer for the heaters, I have,
- After the bridge rectifier and the first capacitor: 5x1.4-1.4 = 5.6
- Reduction produced by the power choke (let say 0.1H, 4A, 0.47 DCR): V=RxI = 0.47x1.2 = 0.56V
- After the choke: 5.6–0.56 = 5.04 !!!!
How can I supply the heaters of 300B with a Vcc voltage (with a CLC filter) having 8 Vac on the secondary of heaters transformer ?
... Where is the error ?
- Status
- Not open for further replies.