04 September 2008

Ultra-Linear 6T9 DIY Tube Amplifier

It was always my intention to build two 6T9 vacuum tube amplifiers. For the first one I used a baking pan for the chassis and I was interested to hear just how good these simple single-ended (SE) tube amps could be with just basic components. From the first hour of play I was very impressed with just how musical they are when matched with the right speakers. My second build of this amplifier incorporated a lot of the original parts but this time more attention was paid to the construction and small upgrades to capacitors and the power supply. I also found that the sound could be further improved by converting the kit to operate from Pentode mode to Ultra-Linear (UL) mode. Both amplifiers are Class A.

The schematic below shows the tube amplifier circuit (click to enlarge). The items in red are the changes that are required for the conversion to UL mode.

6T9 Ultra-Linear Tube Amp Schematic
To change the output section to UL the screen grid resistor (R107, 1k) was removed and the UL tap (tap 3) from the Edcor audio output transformer (OT) is wired directly to the screen grid (pin 10). Other changes I made to the amplifier circuit were to replace the input capacitor C101 (and C201) and the coupling capacitor C102 (and C202) with 0.22 uF metalized polypropylene types. This helps improve the bass response slightly.

The power supply schematic is shown below (click to enlarge). On the power supply I made the following changes which are shown in blue on the schematic. 0.01 uF polyester capacitors across the rectifier diodes. 0.1 uF polypropylene capacitors across all the electrolytic capacitors. One additional 100 uF electrolytic capacitor in the power supply.

6T9 Tube Amp Power Supply Schematic
6T9 Tube Amplifier Edcor Transformers
Tube Amp Construction Notes
I wanted the amp to look cool (very black) and minimalist with its gold connections hidden and all hardware neatly tucked under a black enclosure. Two coats of Epoxy Enamel spray paint was used on the case once all drilling had been completed.


DIY 6T9 Valve Amplifier Kit
The amp is designed to deliver music well, to be unassuming in appearance and have a high wife acceptance factor (WAF). The OTs are slung under the top plate where the power transformer is mounted. Once again the power transformer and the OTs are rubber mounted. Multi-core shielded cable is used to connect the rear gold insulated RCA connectors to the 50k potentiometer and to the printed circuit board (PCB). The power switch was conveniently placed at the front of the amp and a captive power cord was used.

The potentiometer should be connected to the signal ground. Also keep the chassis earth and signal earth separate. Use the shielding on the input wiring to distribute the earth to the input RCA connections, potentiometer and PCB.

DIY 6T9 Vacuum Tube Amplifier Kit
6T9 Tube Amp Measurements

The photographs below show the sine and square wave response of the amplifier. There is little to no change in the response compared to pentode operation.

6T9 Tube Amp Square Wave Response

6T9 Tube Amp Sine Wave Response
Sine Wave Response - 1kHz @ 6.2V P-P, 8-ohm load

6T9 Tube Amp Sound
I played the new amp without NFB and with to determine the best sound. I found the amp sounded much better with NFB. The UL version of the 6T9 tube amplifier sounds better than the Pentode version. Because so many of the components are the same in both amps the improvement in sound is likely due to the UL operation mode.

Mark Houston




Additional 6T9 Tube Amplifier Links

9 comments:

  1. You show a 6.2Vp-p output into an 8ohm load. This is approximately 600mW into the load. I am assuming that this isn't full power. Do you have any estimates of maximum clean output power and input sensitivity for this UL version?

    ReplyDelete
  2. 4.5 Watts you mean, right? Power = Voltage x Voltage / Resistance.

    ReplyDelete
  3. Power calculation must use RMS voltage...(6.2Vp-p / 2) * .707

    ReplyDelete
  4. What make/model xformer was used in the power supply?

    ReplyDelete
  5. It looks like the same power transformer from the other 6T9 project build: http://diyaudioprojects.com/Tubes/6T9-Tube-Amp-Kit/

    "I used a Hammond 369JX power transformer and Edcor XSE15-8-5K for the output transformers (OT)"

    ReplyDelete
  6. I built this project and it sounds much better than I expected... very musical and crisp, yet mellow.

    I do have one comment and two questions:

    Comment: Using diodes to make the rectifier, the turn-on current is high enough to blow 0.75A slo fuses... I moved to 1.0A and all is fine.

    Question1: The metal chassis near the tubes get hot... like real hot. Hot enough to melt the epoxy I used to glue some spacers inside the chassis. Is that expected? Is there better glue to use?

    Question2: After an hour or so of use (first time) the gain seemed to decrease and a hissing noise seemed to increase over time. I assume related to the heat... as I turned it off, and later it sounded fine.

    ReplyDelete
  7. Hi Jim,

    Based on your description, you may have a heat build up problem. As the temp rises, the resistors will change in value and that could be the source of your problems.

    I would suggest by improving ventilation through the chassis. Drill holes in the top plate to exhaust the heat and holes in the bottom to take in fresh are.

    Hope this helps.

    ReplyDelete
  8. I built this amp, and when it works it sounds very nice. However, I do have an issue. It eats fuses like candy. The only deviation i made from the schematic was for the power transformer. I used an Edcor XPWR065-120 for the power transformer. It has the same voltage IO, but higher current capability. I assumed (maybe incorrectly) that since the output xfrmr voltages were the same, then the current would be dictated by the circuit connected and not related to the xfrmr itself. Anyway, I've used even 5A fuses and even they blow (usually at turn on).

    Any help would be appreciated... info or even the best way to debug the issue.

    Thanks,
    Jim

    ReplyDelete
  9. Hi Jim, try using a slow blow fuse.

    Cheers

    ReplyDelete

GET WEBSITE UPDATES VIA EMAIL