Yes the input stage is a diff amp with a CCS tail. The balanced shunt feedback has a forcing balance function in addition it's "Schade" feedback action. The pot in the cathodes of the diff amp is not absolutely required, it just allows you to set perfect balance in the diffamp itself by degenerating the gm of the triodes. Set the pot to get equal DC voltages at the 2 diffamp anodes. To understand the circuit visualize a virtual earth point ( a zero AC signal point) in the centre of that 12K cross connect resistor which moves about a bit according to the balance of the output tube anode voltage balance. This changes the amount of feedback to each side to enforce balance. The 220K loads on the diff amp are effectively "anti boot strapped" by that feedback and will appear to be a lower value.
I've built 8 HiFi Amps to this scheme using evrything from EL34 to EL84 to 6V6 to 6GW8 and all of them were lovely. I've also used 6SL7 in lieu of the 12AX7 for the diff amp. All of them have incredible pace, rhythm and attack which made me think I should try it for a guitar amp. Well also I'm of the Dumble school of guitar power amps, balance it as best you can.
Cheers,
Ian
I've built 8 HiFi Amps to this scheme using evrything from EL34 to EL84 to 6V6 to 6GW8 and all of them were lovely. I've also used 6SL7 in lieu of the 12AX7 for the diff amp. All of them have incredible pace, rhythm and attack which made me think I should try it for a guitar amp. Well also I'm of the Dumble school of guitar power amps, balance it as best you can.
Cheers,
Ian
Loudthud,
This point is exactly why I wanted to try this circuit for a guitar amp. The setup procedures for the Dumble suggest that the balance adjustment is set for perfect balance and that is certainly less usual than having some degree of intentional imbalance to emphasize 2nd harmonic distortion. My experience with the HiFI versions suggest that it has enough "natural" 2nd harmonic and also suggests that the overload characteristic is "nicer" than traditional Ultralinear alone. Certainly the HIFi Amps "grunge up" quite nicely when pushed hard and are great Rock Amps. All this is speculative of-course as I haven't yet tried it as a guitar amp. It was just something that I thought I'd flag to Nazaroo as something he may wish to try, particularly as he doesn't seeem to be locked into building the traditional stacked schmitt PI plus pentode mode output tubes you see in 90+% of guitar power amps.
Cheers,
Ian
This point is exactly why I wanted to try this circuit for a guitar amp. The setup procedures for the Dumble suggest that the balance adjustment is set for perfect balance and that is certainly less usual than having some degree of intentional imbalance to emphasize 2nd harmonic distortion. My experience with the HiFI versions suggest that it has enough "natural" 2nd harmonic and also suggests that the overload characteristic is "nicer" than traditional Ultralinear alone. Certainly the HIFi Amps "grunge up" quite nicely when pushed hard and are great Rock Amps. All this is speculative of-course as I haven't yet tried it as a guitar amp. It was just something that I thought I'd flag to Nazaroo as something he may wish to try, particularly as he doesn't seeem to be locked into building the traditional stacked schmitt PI plus pentode mode output tubes you see in 90+% of guitar power amps.
Cheers,
Ian
something that may or may not be useful
I own a Dean Markley signature T120R valve head for guitar. Im no valve expert but from what i do know it appears to use and LTP input driving 2pair of 6L6GC in class B. It claims 120 watts output, and it easily play intolerably loud whilst staying clean, IF you want it to
maybe the schematic is worth a gander? From memory the outputs are running around 40-50mA per tube, and that is hot, in my limited experience. I have had to choose my replacement 6L6s very carefully due to this. I know little of ultra linear designs so this info may be of zero use to you. Technicals aside, it is the most glorious sounding guitar amp ive had the pleasure of hearing, and using. My childhood best friend was somewhat depressed whemw he brought his Twin round to compare with it!
I own a Dean Markley signature T120R valve head for guitar. Im no valve expert but from what i do know it appears to use and LTP input driving 2pair of 6L6GC in class B. It claims 120 watts output, and it easily play intolerably loud whilst staying clean, IF you want it to
maybe the schematic is worth a gander? From memory the outputs are running around 40-50mA per tube, and that is hot, in my limited experience. I have had to choose my replacement 6L6s very carefully due to this. I know little of ultra linear designs so this info may be of zero use to you. Technicals aside, it is the most glorious sounding guitar amp ive had the pleasure of hearing, and using. My childhood best friend was somewhat depressed whemw he brought his Twin round to compare with it!
Last edited:
This is a great discussion. My cousin experimented with different tubes in the LTP position in one of his amps, and the result was a bit surprising. It was a somewhat blind test because he doesn't know much about tubes, and my Dad had given him a few to try. Typical AX, AT, AU, and then a 12DW7 (one hi-mu and one med-mu triode) and that was the one he liked the best, by far. He said it had a much richer tone than the rest.
It doesn't seem to be the best in a HiFi amp either. "Perfect balance" assumes that the OPT and output tubes are matched over all operating conditions and frequencies. Good luck with that requirement. I have found that an intentional imbalance in a HiFi amp can null different harmonics at different power levels.
I have found two unconventional ways to set the balance.
One is to apply a tone at about 800 Hz and about 1/4 to 1 watt (normal listening level vocal range) with the speaker and an FFT analyzer connected and adjust the balance control for minimum higher order harmonics.
The other (I use for guitar amps) is to connect a dummy load to the amp, apply an 800 Hz tone, crank the amp to the edge of clipping, and adjust the balance pot for minimum sound emission from the OPT. Don't know why this sounds good, but it does seem to work well with cheap OPT's designed for guitar amp use (zero interleaving).
Hi Guys
Ian, that circuit is pretty common in hifi and is referred to as "low-z Drive" in DSN-1.
As far as the 100W amp project goes, if the 300-0300V PT is used in the lethargic half-bridge mode, B+ will only be 430V. Power into a 5k-aa OT is limited to about 36Wrms. Why 5k? Is it because Hammond lists such a unit? The 1650N at 4k3 is more like what production 100W amps use with four tubes driving. This is rated at 60W hifi but supports 100W easily for guitar which does not require full bass extension.
The 1650R 5k is designed to match a single pair of 6550s or KT88s. The latter can use the UL taps without question. For either tube, use at least a 1k-5W screen-stop regardless of output wiring (UL, triode, tetrode). This same detail applies to all power tubes in guitar amps.
The typical 1650R app calls for 600Va and usually 300-350Vs with no UL connection. The low screen voltage allows you to beat the plate quite liberally but also requires a cathode follower or TX driver to push the grids positive for full output.
Have fun
Kevin O'Connor
Ian, that circuit is pretty common in hifi and is referred to as "low-z Drive" in DSN-1.
As far as the 100W amp project goes, if the 300-0300V PT is used in the lethargic half-bridge mode, B+ will only be 430V. Power into a 5k-aa OT is limited to about 36Wrms. Why 5k? Is it because Hammond lists such a unit? The 1650N at 4k3 is more like what production 100W amps use with four tubes driving. This is rated at 60W hifi but supports 100W easily for guitar which does not require full bass extension.
The 1650R 5k is designed to match a single pair of 6550s or KT88s. The latter can use the UL taps without question. For either tube, use at least a 1k-5W screen-stop regardless of output wiring (UL, triode, tetrode). This same detail applies to all power tubes in guitar amps.
The typical 1650R app calls for 600Va and usually 300-350Vs with no UL connection. The low screen voltage allows you to beat the plate quite liberally but also requires a cathode follower or TX driver to push the grids positive for full output.
Have fun
Kevin O'Connor
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.