DHT preamplifier using the 1H4

I had some 1H4's laying around and thought I would build a quick and easy preamp. I had the frame and chassis plate made for me which saved me a lot of time (something that I do not have much of these days). The frame is made out of oak wood and the chassis plate out of 2mm copper.

The setup for the 1H4 is quite simple; RC coupling and cathode biasing. Easy as cake... well... so I thought...

I completed the amp and put it on the test bench, just to be sure I wouldn't blow out the rest of my system :)  All voltages were as I expected, pretty good for a first try. I used Duncan Munro's nifty little PSU Design program. If you're a tube DIY'er than you can't afford not to have this program. It allows you to simulate the power supply so that you need to build and rebuild it until it finally meets the expectations. Anyway...

The 1H4 sounds really really nice, and it's easy to operate. Low voltage, low current, easy filament, mu of 8 makes is great for a preamp. The only thing that surprised me was when I turned up the volume and accidentally hit the chassis with my foot. PLOOOONNGGGG. Microfonic-city !!! If I just whistle or clap my hands, I can hear the result through my speakers. Yikes, this tube is microfonic. When I turn the power amp (Sansui for testing and my subs) to 12-noon the noise from the speakers starts a nasty feedback loop on my tubes. For those who don't know: this is not a good thing!!!

I have since tried pasting the tubes with silicon in the attempt to dampen the glass envelope (hey, these tubes only cost $5 each, who cares if I blow a few). The result is only marginal. The largest part of the resonances are picked up by the tube socket by way of the chassis plate and internal wiring that resonates with the sound. I've already ordered some teflon tube sockets, but don't think this will help much.

Another freak thing I came across is something that puzzles me to this day. When running the amp I measured a funky 290-300mV hum on the output. With conventional amps having an input sensitivity of only 500mV, this is a problem to say the least. The first thing I tried was a battery heating of the filament. This solved the problem immediately, but why??? The filament supply had a ripple of only 6mV, even with a gain of 8 (which this preamp does not have!) it would never amount to 300mV. So it must be grounding I thought. I posted a message on JoeNet and immediately got some replies. Tom Ronan came up with the solution, or better; remedy. He had had some of the same problems in the past and found that a bypass cap over the cathode resistor solved the hum problem. My original design did not use this cap since I was not interested in the low end for my system. Once I hooked up a 470uF Black Gate, the hum was as good as gone... I tried using a few other values as well: 10uF, 100uF and 470uF. The experiment showed that the smaller cap (10uF) still left a ripple/hum on the output of 50mV, the 100uF however only left 16mV and the 470uF left 15mV.

The lowest value I could get on the output was 15mV, this was still too much for me. Still can't figure where it is coming from. I will most likely switch to battery heating in the future (when I have time). This is the only way to get zero noise out of the preamp. But that's another tale...

As to the sound... wow... pretty good... It is not the sound of the 26 that I like so much, but it will do very well until my ultimate 26 preamp is done (yes, battery heating as well).

After refitting my preamp for the 1G4 and finding better results than last time I figured I had to give the 1H4 another chance as well. The supplies are almost identical to those of the 1G4, just a few alterations to get more voltage on the plate etc. Gain is a bit higher than that of the 1G4, this is a good thing in my system. Just enough to get my 'power' amp playing a little louder.

The sound is absolutely gorgeous; very detailed, transparent and fresh. Better than I had heard the last time. Microphonics are also very tolerable now, I can't figure why this has changed since last time (same tubes, same chassis, etc.). Anyway, WOW!!! I was actually enjoying every minute it was playing.

I did hear some sharp sounding distortion during loud attacks/passages, I'll see what that is about this weekend. Conclusion, this tube has gone up in rank one step. What a few alterations in a power supply can accomplish ;)

I am not getting enough gain from the new setup, I can hardly get anything loud out of my system. I changed the load to 56K and adjusted the power supply accordingly.

The new signal section schematic.


B+ supplies for both channels

Filament supply. The switch connects the battery to the charger for loading the batteries or to the filament for playing.


The copper plate has not been treated. This setup was
just to see if everything would actually fit


rear view


a now blasted & lacquered chassis plate during the rebuild


finished version, now to find some knobs for the input and volume controls


internal wiring (old version)


inside of the new version

Jim de Kort