A few years ago I bought my first valve power amplifier, my very first venture into tube audio. I had always had integrated mid-priced amps before, so the concepts of a separate power- and preamp were new to me. I knew I would need a preamp to drive the power amp but I never made an effort to buy one. I started off using a Bourns stereo pot to control the volume, a simple passive preamp. A small PVC casing with two sets of RCA inputs and an output to my power amp, that was my entire preamp at the time. That was about six years ago. Since then I've made some experimental preamps but none of them stayed for long. The last year or so I have been using a stepped attenuator as a passive preamp which was a nice improvement over the Bourns and the Alps that I had later on.
opamps and tubes :(
After all this time I got fed up with the ugly casings and makeshift circuits, so I decided to design a real preamp. A little bit more "attaque" in the music wouldn't hurt either. Passive sounds great, but it also misses out on some things like punch and openness. My first idea came out of some electronics magazines, and they seemed fine at the time. After reading article after article, the design slowly started changing. The circuit I have right now doesn't even resemble the circuit I started out with, four years of reading and experimenting has had it's influence on my projects. The first concept had a 6SN7 in a SRPP configuration, my final design a choke loaded 26. The power supply has changed from a diode rectified, solid state regulated supply to a valve rectified, CLC filter supply. We're talking two different worlds here. How much my views have changed over these past years, unbelievable...
All in all I've come up with a totally different design. A little like the difference between black and white, total opposites. I think this has to do with the fact that I've been working on this design for such a long time. During that time my taste in sound has changed considerably, and some of my equipment as well.
My original idea was for a SRPP circuit using the 6SN7 type tube, in my case a beautiful ST shaped Mullard ECC32. The ECC32 has different specs to that of the 6SN7, but they come close. The design came out of a local electronics magazine which I have been cursing ever since I got my first copy of Glass Audio and Sound Practices. Until I started reading these fabulous magazines I was living in the dark ages of tube audio. They opened my eyes to some great things in the audio world. I can honestly say that I owe most of my knowledge on different tube gear to these guys.
Schematic of my first ever (serious) preamp
I was reading several articles in GA, as well as Sound Practices, about DHT's. David O'Rourke's article "In search of the perfect tube" in Glass Audio vol. 8 #2 really sparked my imagination. He put me onto the idea to use a DHT in my preamp. Something that I hadn't seen or thought of before then and not many other people seem to appreciate.
THE tube ?!
I first went for the 1H4G, later the 30, because it seemed best suited for my demands. It had the desired plate voltage, was directly heated, had a mu of 9.3 and it's plate resistance was well within the limits at 10K. The 30 is electrically the same as the 1H4G except it has a UX4 base as opposed to the octal base of the 1H4G. I chose the 30 over the 1H4 because it was more nostalgic and simply just looked cooler ;)
A few weeks later Mattijs de Vries mentioned the specs of the 26 tube to me in one of our e-mail sprees, but I didn't really pay attention at the time. Only a few weeks after that did I begin considering using it in my preamp. It had a lower plate resistance and could "deliver" more current. It also had a lot more emission than the 30 which finally tipped the scales in favor of the 26. I looked for a dealer who had these in stock and quickly found one. Tube World had 4 tubes, 2 Radiotrons and 2 Cunninghams for a nice price. These were balloon shaped tubes, which I find more attractive, more nostalgic than the ST versions.
A few weeks later I finally started on my final design, then still RC loaded. A wooden board would be the foundation of my breadboard trial run. Some caps here, some resistors there... it was up and running within a week or so. I hooked it up to my power amp and turned it on.... Whoa, forgot to connect the grounds of the outputs, pretty hazardous for my power amp, the hum was deafening. The problem was solved and after some minor corrections and additions to the preamp I found that it sounded very nice already. I was using a makeshift transformer and some electrolytic caps for my breadboard, why waste the good stuff if you don't know it will sound good or not or just plain explode right in your face. Electrolytic fluid is better than PCB's in this situation :)
I listened to it for a few days and experimented with some other DHT's to see which sounded the best. I tried the 1G4 and 30, but the 26 beat them, hands down, no comparison. The 26 sounds more natural and open than the 30 or 1G4 could ever sound. I think it is because of the 26's higher heater current (=emission). The 1G4 and 30 are meant for battery operating, you don't even see the heater turn red unless your turning your variac up too much (...pling... shit!!!!)
My mind was set, this will be the tube that will stay in there, no more changes. So out went the breadboard while I ordered my parts, back came the passive preamp once more. It has been almost a year since then, and it is still not ready. I had a fabulous looking brass chassis made (1997) and some housings to pot my transformers and caps in. The chassis would rest in a hardwood frame and has brass volume and channel control-knobs. It's looks alone would have to guarantee a great sound.
Potting the trannies
A look inside
But alas, after building it into the chassis and hooking it up there was a loud hum that was there to stay. I tried all sorts of stuff, changed parts, tubes, modified the supply, added extra grounding... Nothing worked. Up to this day I still don't know what could have caused it. I found a spike on the B+ which was caused by my choke, but I managed to solve this little problem, but the loud hum was still there. Since I potted the transformers myself it is a safe bet that something went wrong during this procedure. Below is a picture of the spike I had on the B+ choke which caused a little frustration by itself.
Big spikes right after the choke
PS caps also potted (big mistake!)
Layout of the parts on the chassis
There have been several requests for me to post the RC-loaded version of my 26 preamp. With all the updates I never thought of leaving the old schematics on my site. So here they are...
Sound you say... well... Excellent, really musical and "easy" to listen too. You won't get tired of this tube! Choke loaded is even better, it opens up the image quite a bit and makes it even more natural. This amp will give you a LOT of enjoyment and is not expensive to brew up like the choke loaded version. You might want to reserve some space though while making the chassis, you never know ;)
I've compared the 26 with the 1H4 and 1G4 in similar setups. My favorites in order of best/least is 26/1G4/1H4.
- The 1G4 is a bit too warm and dark. Pretty microfonic as well. Filament is very delicate and snaps without mercy.
- The 1H4 is really laid-back but also very transparent and lively. No microphonics at all, a big contender for the 26.
- The 26 is the most open and sweet tube, very musical. Only slightly microphonic, not as with the 1G4 and some 1H4 at all.
The 22K load gives a gain of around 5 times. Raising the value of Ra also raises the gain; downside however is the increase in B+ needed to compensate for the drop over Ua and the narrowing of the frequency response (high range). 22K works well and gives more than enough gain for a line-level preamp. The 220nF output cap can be raised in value if really necessary, max should be 0.47uF. The most important parts in the circuit are the bypass capacitor and the filament supply. The 26 is absolutely ruthless if these are not top-notch. It is very sensitive to hum due to ripple on the supplies etc, so take care on these two points. Not much else to say about the circuit.
The 80 can be substituted with the 5Y3, this is an electrical equivalent but in a GT shape with octal base. It is a lot cheaper and easier to find. The bleeder is a must, the supply has to be drained when it is turned off, safety and everything! It is also a matter of getting the right voltage, you can experiment a bit with this resistor to fine tune the supply. If the voltage you get is not around 340-350V then you have the right voltage.
As for the filament supply... The output voltage to the 26 filament has to be flat DC. You can use regulators or a good passive supply to get 1.5V at 1A (just about). The regulator removes a bit of openness and musicality which the passive supply does not, it is easier to use though requiring less capacity etc.
Pictures of the first ever version of the 26 preamp that I made.
Front view of the last RC-loaded version I built.