Hello-
I am researching the HiWatt fixed bias phase inverter. Have done much searching, read the Amp Books discussion and studied Mark Huss site. I see there were two versions - the early version where the signal was passed to P/I through a coupling cap into the DC bias/cathode follower triode and the later version where the signal was passed to P/I through a coupling cap and it did not pass through this DC bias triode/cathode follower.
I have read good, and bad, things about both schemes. My question is which is the most preferred? I have read where some liked the sound better when it passed through the DC bias triode, others prefer it the other way around. I would very much appreciate hearing opinions and discussion about the (early method) sonic effects of that cathode follower acting as a buffer.
I have used the later version fixed bias Hiwatt P/I in some of my non-Hiwatt builds and as far as I can tell it is doing as it should. I used the more involved presence circuit such as shown on the late 70's preamp. I never have tried the early method where the bias triode/cathode follower is acting as a buffer. It has really piqued my curiosity as it seems about half the DR-103 builds I see documented are using the early pass-through method.
Also, does choice of 12AX7 or 12AT7 as P/I play a role in any of this?
Sincere thanks,
Byron W5FH
HiWatt DR-103 Phase Inverter
Moderators: pompeiisneaks, Colossal
Re: HiWatt DR-103 Phase Inverter
I did an experimenting last year to find an answer to the same question: https://ampgarage.com/forum/viewtopic.php?f=10&t=36459
My experience was that the amp was a lot less stiff without feeling loose, i.e., without losing what makes Hiwatts sound and feel like Hiwatts. A lot of preferences will come down to players, though, and I'm used to playing as the only guitarist in a band and have mostly played Marshall style amps. So the feel difference and the difference in how the sustained notes develop associated with the early style PI better fit me. It's totally possible that someone who is used to playing silver panel Twins would prefer the later style PI. As far as I know, the unique behavior of the circuit stems chiefly from the relationship of V3 to V4, not necessarily what's in V4. I do know a 12AX7 will work safely as a PI in these amps and that it will significantly increase overdrive, but without a scope I'm not sure if the substitution breaks down the operating parameters of the PI that keep it sounding clean, if it's a decrease in input headroom at the PI itself, or if it's driving the output section harder.
My experience was that the amp was a lot less stiff without feeling loose, i.e., without losing what makes Hiwatts sound and feel like Hiwatts. A lot of preferences will come down to players, though, and I'm used to playing as the only guitarist in a band and have mostly played Marshall style amps. So the feel difference and the difference in how the sustained notes develop associated with the early style PI better fit me. It's totally possible that someone who is used to playing silver panel Twins would prefer the later style PI. As far as I know, the unique behavior of the circuit stems chiefly from the relationship of V3 to V4, not necessarily what's in V4. I do know a 12AX7 will work safely as a PI in these amps and that it will significantly increase overdrive, but without a scope I'm not sure if the substitution breaks down the operating parameters of the PI that keep it sounding clean, if it's a decrease in input headroom at the PI itself, or if it's driving the output section harder.
Re: HiWatt DR-103 Phase Inverter
Thank you cdemike!
I have studied that two-input Jimmy Page Hiwatt preamp and your excellent write-up. It bothered me looking at the 100K resistor coming off of B+ line that feeds the plate circuits of both triodes of V3. The first section of V3 is the post master volume gain stage and has its own 100K plate load resistor. This reminds me in a way of the split load style arrangement of series plate resistors used in some amps to reduce signal levels. At the junction of the two 100K resistors the 8 ufd filter cap would be shunting a portion of audio signal to ground from plate of V3 first section, due to action of the other 100K resistor supplying B+. In studying the early 6G16 Vibroverb split load series plate resistors (82K/22K series pair with 22K nearest B+ node)) are used in second gain stage to reduce the dry signal to 25% (taken at junction of 82K/22K) and reverb drive to 75% roughly (taken off plate). So in this Jimmy Page preamp it looks like to me the audio signal is split in some fashion between the two 100K series resistors, with 8 ufd taking that to ground. The other thing I wonder is the amount of DC voltage swing between these 2 100K resistors during periods of heavy drive. This voltage is supplying the bias voltage divider/cathode follower. Possibly the phase inverter DC grid bias is swinging along with it? Maybe the Page preamp was designed to get a little extra effects to the tone from this circuit?
I have measured the DC bias coming off of this cathode follower in several of my builds using it (same as the one in early 80's SAP preamp) and it is always in 70 volt region. I don't believe I would want its source voltage to be swung like in the Page circuit.
Just my take on it. I really appreciate the info. I hope others that have experimented with this will chime in.
Thanks!
I have studied that two-input Jimmy Page Hiwatt preamp and your excellent write-up. It bothered me looking at the 100K resistor coming off of B+ line that feeds the plate circuits of both triodes of V3. The first section of V3 is the post master volume gain stage and has its own 100K plate load resistor. This reminds me in a way of the split load style arrangement of series plate resistors used in some amps to reduce signal levels. At the junction of the two 100K resistors the 8 ufd filter cap would be shunting a portion of audio signal to ground from plate of V3 first section, due to action of the other 100K resistor supplying B+. In studying the early 6G16 Vibroverb split load series plate resistors (82K/22K series pair with 22K nearest B+ node)) are used in second gain stage to reduce the dry signal to 25% (taken at junction of 82K/22K) and reverb drive to 75% roughly (taken off plate). So in this Jimmy Page preamp it looks like to me the audio signal is split in some fashion between the two 100K series resistors, with 8 ufd taking that to ground. The other thing I wonder is the amount of DC voltage swing between these 2 100K resistors during periods of heavy drive. This voltage is supplying the bias voltage divider/cathode follower. Possibly the phase inverter DC grid bias is swinging along with it? Maybe the Page preamp was designed to get a little extra effects to the tone from this circuit?
I have measured the DC bias coming off of this cathode follower in several of my builds using it (same as the one in early 80's SAP preamp) and it is always in 70 volt region. I don't believe I would want its source voltage to be swung like in the Page circuit.
Just my take on it. I really appreciate the info. I hope others that have experimented with this will chime in.
Thanks!
Re: HiWatt DR-103 Phase Inverter
I'd also be very eager to hear other's thoughts on this. There doesn't seem to be a lot written about how the two circuits impact the sound and why they do. I know Greg Fryette did a series on Hiwatts on the Fryette youtube channel a while back and he expressed a strong preference for the version with the 580k resistor in the bottom leg of the voltage divider feeding the grid of the cathode follower. I'll need to rewatch it in case he gets into some more detail on why it sounds the way it does.
The similarities to a split load resistor hadn't occurred to me until your post -- interesting! FWIW, though, I think I've only seen schematics with 100K plate resistors there (never a 220K), and in versions with the more common B+ rail where V1 and V3 share B+ feeds, I think there would be the same effect whereby signal gets shunted to ground, but in the V1 filter cap. I also suspect the effect would be limited in any event since it seems that the primary effect of 100K/8uF filter network in the early style B+ rail (i.e., the one in the Jimmy Page preamp schematic) would be to effectively split the B+ rail into parallel lines.
Cool to see someone else is messing with these. What was you experience on your builds?
The similarities to a split load resistor hadn't occurred to me until your post -- interesting! FWIW, though, I think I've only seen schematics with 100K plate resistors there (never a 220K), and in versions with the more common B+ rail where V1 and V3 share B+ feeds, I think there would be the same effect whereby signal gets shunted to ground, but in the V1 filter cap. I also suspect the effect would be limited in any event since it seems that the primary effect of 100K/8uF filter network in the early style B+ rail (i.e., the one in the Jimmy Page preamp schematic) would be to effectively split the B+ rail into parallel lines.
Also curious to know if someone has put a scope on one of these amps to see how those resistors behave. I suspect the reason this circuit was borne our of the broader Reeves design philosophy, and having a separate node might provide better isolation and control of the PI's bias feed. I also suspect the reason the B+ rail changed in its later form with V1 and V3 sharing a node was a diminishing returns situation where, between the phase cancellation between V1 and V3 (read: no possibility of oscillation/positive feedback) and the extra cost of a 8uF capacitor, it made more sense to simplify the circuit.W5FH wrote: ↑Thu Jan 25, 2024 1:12 am The other thing I wonder is the amount of DC voltage swing between these 2 100K resistors during periods of heavy drive. This voltage is supplying the bias voltage divider/cathode follower. Possibly the phase inverter DC grid bias is swinging along with it? Maybe the Page preamp was designed to get a little extra effects to the tone from this circuit?
I have measured the DC bias coming off of this cathode follower in several of my builds using it (same as the one in early 80's SAP preamp) and it is always in 70 volt region. I don't believe I would want its source voltage to be swung like in the Page circuit.
Cool to see someone else is messing with these. What was you experience on your builds?
Re: HiWatt DR-103 Phase Inverter
Hey there cdemike-
To answer your question about my builds using the Hiwatt fixed bias P/I I used the separate triode (V3-B) of 12AX7 (no audio passing through it) to supply grid bias to phase inverter. The circuit I used (for fixed bias and P/I only) is identical to the one shown on the Hiwatt early 80's two input SAP preamp. I did not have the V3-A triode gain stage. I came off of master volume wiper of my build and tied the Hiwatt circuit to it at junction of 47n and 1n capacitors of early 80's preamp.
I have noticed differences in the voltage divider resistors for the bias regulator tube, depending on where the voltage source for plate comes from. I believe this is most important when integrating this circuit into a build. Have noticed at least 3 different arrangements of voltage divider and cathode resistors. In the early 70's linked input SAP preamp the voltage divider for grid is supplied from previous stage tube plate (having to pass through its 100K plate resistor). On many of the other Hiwatt preamps the grid voltage divider is supplied from same source voltage as plate of regulator tube. I believe you must be very careful when using this circuit to match up your grid divider resistor values, and cathode resistor, of regulator triode to source voltages.This is relating to use of regulator triode as a bias voltage source. If the circuit is not holding the P/I bias voltage steady like it should I would imagine you would hear the effects on tone.
If you are going to pass audio through the bias regulator triode you not only have to supply the correct bias voltage to P/I but you must also have bias triode biased correctly. I can see where this could be even more critical.
I am sure you have read the treatise on the Hiwatt fixed bias phase inverter at the Amp Books.Com site. I based my circuits on what was shown there. Of course I am not passing the audio through the bias triode. So my several builds using this are just "conventional" in regards to feeding the phase inverter with audio signal. I just wanted to try the fixed bias Hiwatt P/I due to its low distortion being touted, plus try the unique presence circuit associated with it.
I want to try a build of the earlier design where the audio is passed through the bias regulator triode. A portion of an early 70's Hiwatt preamp, showing driving stage, bias triode and P/I, is detailed with voltages in the Amp Books.Com treatise. This was from personal communication between author and Mark Huss it is noted in biblography. This would probably be the circuit I next try. I am not technically qualified to evaluate the effects of passing audio through the bias triode of which it is mostly acting as a low output impedance buffer. The bias triode grid circuit can only "steal" a very tiny current from driver triode plate, in acting as a DC coulped cathode follower, to generate compression. I respect the observations of others who report the preference of the tone when passing audio through bias triode. My only comments to this is that in most of the cases a swap was made from one style bias triode circuit to another. I am curious if all voltage divider and cathode resistors were the correct values for the particular circuit (involving bias triode voltage sources).
Just as a matter of convienence I stumbled across a schematic of a 2004 Hiwatt preamp (Main Preamp PCB Schematic, Hiwatt 2004 Range, Drawing HIW-CD001, Issue 1.2, Sept. 5, 2004) showing the use of a high voltage TO-92 cased MPSA44 transistor, used strictly as the bias regulator and not in audio path. This would surely simplify a build! I am going to try that to free up a triode being it is just for a voltage reference.
Thank you for commenting on this!
Byron W5FH
To answer your question about my builds using the Hiwatt fixed bias P/I I used the separate triode (V3-B) of 12AX7 (no audio passing through it) to supply grid bias to phase inverter. The circuit I used (for fixed bias and P/I only) is identical to the one shown on the Hiwatt early 80's two input SAP preamp. I did not have the V3-A triode gain stage. I came off of master volume wiper of my build and tied the Hiwatt circuit to it at junction of 47n and 1n capacitors of early 80's preamp.
I have noticed differences in the voltage divider resistors for the bias regulator tube, depending on where the voltage source for plate comes from. I believe this is most important when integrating this circuit into a build. Have noticed at least 3 different arrangements of voltage divider and cathode resistors. In the early 70's linked input SAP preamp the voltage divider for grid is supplied from previous stage tube plate (having to pass through its 100K plate resistor). On many of the other Hiwatt preamps the grid voltage divider is supplied from same source voltage as plate of regulator tube. I believe you must be very careful when using this circuit to match up your grid divider resistor values, and cathode resistor, of regulator triode to source voltages.This is relating to use of regulator triode as a bias voltage source. If the circuit is not holding the P/I bias voltage steady like it should I would imagine you would hear the effects on tone.
If you are going to pass audio through the bias regulator triode you not only have to supply the correct bias voltage to P/I but you must also have bias triode biased correctly. I can see where this could be even more critical.
I am sure you have read the treatise on the Hiwatt fixed bias phase inverter at the Amp Books.Com site. I based my circuits on what was shown there. Of course I am not passing the audio through the bias triode. So my several builds using this are just "conventional" in regards to feeding the phase inverter with audio signal. I just wanted to try the fixed bias Hiwatt P/I due to its low distortion being touted, plus try the unique presence circuit associated with it.
I want to try a build of the earlier design where the audio is passed through the bias regulator triode. A portion of an early 70's Hiwatt preamp, showing driving stage, bias triode and P/I, is detailed with voltages in the Amp Books.Com treatise. This was from personal communication between author and Mark Huss it is noted in biblography. This would probably be the circuit I next try. I am not technically qualified to evaluate the effects of passing audio through the bias triode of which it is mostly acting as a low output impedance buffer. The bias triode grid circuit can only "steal" a very tiny current from driver triode plate, in acting as a DC coulped cathode follower, to generate compression. I respect the observations of others who report the preference of the tone when passing audio through bias triode. My only comments to this is that in most of the cases a swap was made from one style bias triode circuit to another. I am curious if all voltage divider and cathode resistors were the correct values for the particular circuit (involving bias triode voltage sources).
Just as a matter of convienence I stumbled across a schematic of a 2004 Hiwatt preamp (Main Preamp PCB Schematic, Hiwatt 2004 Range, Drawing HIW-CD001, Issue 1.2, Sept. 5, 2004) showing the use of a high voltage TO-92 cased MPSA44 transistor, used strictly as the bias regulator and not in audio path. This would surely simplify a build! I am going to try that to free up a triode being it is just for a voltage reference.
Thank you for commenting on this!
Byron W5FH
Re: HiWatt DR-103 Phase Inverter
There's actually very little practical difference between the two.
In the earlier case, the signal travels through the bias-setting tube (which will add a smidgen of harmonic distortion) and in the later case the signal bypasses the tube and goes directly to the PI input.
The loading on the previous stage is about the same either way, and the medium high impedance of the previous stage driving the PI vs. the low impedance of the cathode follower driving it shouldn't make a lot of difference either.
In the earlier case, the signal travels through the bias-setting tube (which will add a smidgen of harmonic distortion) and in the later case the signal bypasses the tube and goes directly to the PI input.
The loading on the previous stage is about the same either way, and the medium high impedance of the previous stage driving the PI vs. the low impedance of the cathode follower driving it shouldn't make a lot of difference either.
Re: HiWatt DR-103 Phase Inverter
Thanks for comments Mark and thank you for the great Hiwatt site and all of the info, much appreciated.
Byron W5FH
Byron W5FH
Re: HiWatt DR-103 Phase Inverter
If someone with more understanding can explain to me please-
I have attached two Hiwatt preamp schematics to refer to. What I cannot get my head wrapped around is how the bias "regulator" triode can output the same bias voltage to phase inverter grids (in vicinity of 73 vdc) when it has its grid voltage divider resistors supplied by a varying DC voltage? This is in reference to the voltage divider pair of resistors supplying grid voltage to the bias "regulator" triode. On one schematic this voltage divider pair is supplied from the main B+ rail that is also supplying its plate, a fairly steady dc voltage. On the other schematic the voltage divider pair is supplied from the plate-side (under 100K plate load resistor) of gain stage. As I understand it the dc voltage on plate of this gain stage could vary a total of around 100 volts during drive.
I have spent a lot of time reviewing the theory of cathode followers and that light has yet to go off in my head to understand the operation of this bias voltage "regulator". If someone has the time to explain this I would surely be grateful. I apologize for my lack of understanding at this point.
Byron W5FH
I have attached two Hiwatt preamp schematics to refer to. What I cannot get my head wrapped around is how the bias "regulator" triode can output the same bias voltage to phase inverter grids (in vicinity of 73 vdc) when it has its grid voltage divider resistors supplied by a varying DC voltage? This is in reference to the voltage divider pair of resistors supplying grid voltage to the bias "regulator" triode. On one schematic this voltage divider pair is supplied from the main B+ rail that is also supplying its plate, a fairly steady dc voltage. On the other schematic the voltage divider pair is supplied from the plate-side (under 100K plate load resistor) of gain stage. As I understand it the dc voltage on plate of this gain stage could vary a total of around 100 volts during drive.
I have spent a lot of time reviewing the theory of cathode followers and that light has yet to go off in my head to understand the operation of this bias voltage "regulator". If someone has the time to explain this I would surely be grateful. I apologize for my lack of understanding at this point.
Byron W5FH
You do not have the required permissions to view the files attached to this post.
-
martin manning
- Posts: 13209
- Joined: Sun Jul 06, 2008 12:43 am
- Location: 39°06' N 84°30' W
Re: HiWatt DR-103 Phase Inverter
In V2, the triode with the 100k cathode resistor is strictly providing a DC reference voltage. In V1a, it is providing a DC reference voltage and driving the PI. The signal from the preceding tube's plate is bypassed around the upper (1.8 Meg) divider resistor by the 22n cap, and into the grid of the PI. That causes the voltage across the upper resistor to remain essentially constant, while the ends rise and fall together with the signal.
Re: HiWatt DR-103 Phase Inverter
Thanks Martin-
If you will bear with me, I still am confused in regards to preamp circuit V1a. The gain stage is supplying the DC voltage to the 1.8M and 1 M voltage divider pair from plate side of the 100K plate load resistor. Won't this DC voltage be varying up/down at an audio rate as this gain stage is driven? Is this what you are trying to tell me -- the coupling capacitor that is bypassing the 1.8M divider resistor (and passing audio signal into cathode follower) is "smoothing" out this swinging DC voltage, similar to a power supply filter cap? Is it smoothing it out to the point that the DC voltage to divider would be nearly the same as at idle conditions with no signal drive to gain stage? Maybe that light is starting to glow a bit now...
Thanks,
Byron W5FH
If you will bear with me, I still am confused in regards to preamp circuit V1a. The gain stage is supplying the DC voltage to the 1.8M and 1 M voltage divider pair from plate side of the 100K plate load resistor. Won't this DC voltage be varying up/down at an audio rate as this gain stage is driven? Is this what you are trying to tell me -- the coupling capacitor that is bypassing the 1.8M divider resistor (and passing audio signal into cathode follower) is "smoothing" out this swinging DC voltage, similar to a power supply filter cap? Is it smoothing it out to the point that the DC voltage to divider would be nearly the same as at idle conditions with no signal drive to gain stage? Maybe that light is starting to glow a bit now...
Thanks,
Byron W5FH
-
martin manning
- Posts: 13209
- Joined: Sun Jul 06, 2008 12:43 am
- Location: 39°06' N 84°30' W
1 others liked this
Re: HiWatt DR-103 Phase Inverter
Yes.W5FH wrote: ↑Fri Jan 26, 2024 6:36 pm Thanks Martin-
If you will bear with me, I still am confused in regards to preamp circuit V1a. The gain stage is supplying the DC voltage to the 1.8M and 1 M voltage divider pair from plate side of the 100K plate load resistor. Won't this DC voltage be varying up/down at an audio rate as this gain stage is driven?
No, the cap passes the AC on to the other end of the 1.8 Meg so it also goes up and down like the voltage at the input end (and the CF's grid follows the plates), but the average (DC) level is shifted down by the resistive divider, same as if there were no signal present. In other words, the divider only works on the DC, because to AC the upper resistor is shorted. The CF's cathode follows its grid, so the DC level and average AC at the PI grid are also the same.
Re: HiWatt DR-103 Phase Inverter
Thanks Martin-
Been reviewing Blencowe book and some other references, I believe I am grasping your explanation. Thank you for taking the time to help.
Byron W5FH
Been reviewing Blencowe book and some other references, I believe I am grasping your explanation. Thank you for taking the time to help.
Byron W5FH