I've read a lot of DIYaudio posts but haven't posted much myself:well here goes. I got into vintage tube radios about three years ago and that lead me to a Harman-Kardon tube AM-FM receiver. This dude weighs like 40 pounds and has twenty tubes; it will heat a small house but I love look and sound of this vintage of gear. I have gone through this receiver and brought it up to specs but I m having trouble finding power tubes (7355s). I have been able to find un-matched 7355s but it really needs a matched quad to perform as designed, and they are not available. I did some research on a replacement and the closest pentode to a 7355 currently in production is the 7591. 7355s and 7591s operate in push-pull at pretty similar voltages except the Grid #1 voltage, which are -23.5 vs -16, respectively. So how do you change the grid #1 voltage? H-K brings the cathode ground up with a voltage divider circuit that also drives the filaments of three pre-amp tubes instead of having true negative dc. And to boot, on my unit, the aforementioned filament voltages are low (8.9, 8.5, 7.6). Instead of trying to make a new divider circuit, I decided adding a new power supply would be the answer. The new power supply has an adjustable negative dc for grid #1 and powers the pre-amp tubes at 12.6 volts (±10%). The bonus to having an adjustable -dc is I can properly bias either 7355s or 7591s. I have completed step one, which is to install the new power supply, re-power the pre-amp filaments, connect grid #1 circuit to the neg dc supply, and connect the cathodes to ground. I haven't modified the tube sockets yet so I brought it up today with 7355s and it biased very easily and performed very well. I wasn't sure what effects the increased filament voltage in the pre-amp and increased plate to cathode voltage in the power tubes would have, but i'm happy to report no new hums and a slight increase output, particularly in the phono pre-amp. I attached the schematic of the mods. The only modification left is to decouple pin 4 on the power tube sockets, and put a jumper from pins 8 to 4. After the sockets are modified, I'll try try to bias a new matched quad of re-production Tung-Sol 7591s power tubes and report back. Thanks for reading.
Bill
Bill
My HK A300 kit had a complicated circuit, that made any modification hard to do.
The 4 output tube cathode currents ran the phono preamp tube filaments.
If I had that A300 amplifier again, I would completely change the output tube circuitry (but hard to do because of the lack of room).
I would take out the phono preamp stage; and use the space.
Either 4 individual self bias resistors with individual bypass capacitors; or an adjustable fixed bias supply with 4 potentiometers to have an adjustment for each of the output tube currents.
For your receiver model, that is a lot more than I would even attempt.
Look harder for matched 7355 output tubes; or really bite the bullet and completely transform the output circuit.
Changing to 7591 tubes seems like a major lot of work, especially because you not only have the phono preamp but also the tuner portion.
You do not want voltages to change, or you may have to re-calibrate the tuner portion.
Perhaps someone else has some good and re-assuring advice.
The 4 output tube cathode currents ran the phono preamp tube filaments.
If I had that A300 amplifier again, I would completely change the output tube circuitry (but hard to do because of the lack of room).
I would take out the phono preamp stage; and use the space.
Either 4 individual self bias resistors with individual bypass capacitors; or an adjustable fixed bias supply with 4 potentiometers to have an adjustment for each of the output tube currents.
For your receiver model, that is a lot more than I would even attempt.
Look harder for matched 7355 output tubes; or really bite the bullet and completely transform the output circuit.
Changing to 7591 tubes seems like a major lot of work, especially because you not only have the phono preamp but also the tuner portion.
You do not want voltages to change, or you may have to re-calibrate the tuner portion.
Perhaps someone else has some good and re-assuring advice.
This modification does completely change the bias control of these power tubes. Previously there was no bias control for the power tubes. The 50 ohm pots on the cathode circuit allow balancing of the pairs but is very limited and won't make up for poorly matched tubes. 7355 aren't getting easier to find and I want to have matched tubes like the other kids. As far as this mod being a lot of work, note that I listen to this receiver daily and swapping the power tubes has been on my todo list for awhile. To formulate this plan, I worked with a tech that knows way more than me. I designed and he critiqued this design on paper until I was comfortable I wasn't going to kill my receiver. I bench tested this power supply and adjusted the resistor valves to get the voltages I wanted. I now have installed a toroid transformer (about the size of a hockey puck) to the a 2"x3"x1.5" aluminum project box, which is mounted to the back of the chassis. It sticks out about an inch further than the AM antenna and when I have the wooden case on it, is only visible from the back; I can live with this, many couldn't. The new power supply circuit is in this device box. The voltages that are changed by the modifications are the power output tube cathodes are lowered to +3 vdc from +27 vdc and the grid voltages are changed from +5 vdc to -18 vdc (now adjustable from -6 to -37); this provides the specified -23vdc differential grid #1 voltage for 7355s or -15.5vdc for 7591s. Also, the phono pre-amp filament voltages are corrected by the mod from 8.6, 8.3, and 7.2 to 12.6 vdc ±10%. The all tube plate voltages remain powered from original transformer/power supply and are unchanged by the modification (I have checked). I have done this mod and am literally listening to the FM radio on it right now through 7355 power tubes. Also, with the phono pre-amp voltages corrected, I was able to use the short plate low-noise repro Tung Sol 12AX7s that didn't sound right before. I hope to find time to repin the sockets and try a new matched set of repro Tung Sol 7591s in the next couple of days. I let you know how the bias controls work and if there is adequate adjust. Thanks for your input.
Last edited:
