Exodus II

© VT52.com


This is the continuation of the original Exodus amplifier. In conjunction with my new AER drivers in the Oris 150 horns I needed to have my Exodus output transformers rewound. The previous PM4A units were 8 ohm speakers, the AER are now 14 ohms. Another reason to have them rewound was the switch in output tube. The UX210 has been moved up from driver to output tube. It delivers a whopping 1.2W of output power at the settings I am using. A bonus is that the VT25, VT25A, 10Y and 801A can be dropped in without modifying anything in the circuit. This will allow me to spare those ancient UX210/CX310 beauties when playing music just for background noise.


© VT52.com
a 12K (yellow) and 30K (blue) load line for the UX210 at 400V/18mA


The driver is the 71A, a very nice power tube by itself. It has a low gain meaning the amp will have a rather low input sensitivity. Seeing that the 10 is running at around -30V this means I don't really need a lot of swing anyway. The preamp delivers about 10-15V of swing, that leaves me with more than enough input signal. Give or take a volt or two... The miller capacity on the 10 is at it's worst at 30kHz and 25V swing. It requires 1.8mA of drive here. The 71 will have no problems at all with this load, this is one of the reasons I chose the 71A as driver.

I was thinking about using the 12A here as it has more gain (8), but at 6mA bias I don't like this as a driver that much for the 10 in my case.  The 10 can go into A2 a little requiring 2-3mA into the grid, so a bit more current is a nice thing to have. The driver is IT coupled to the 10 by a 1:1 ratio, this is the nicest ratio for getting a broad bandwidth (specially in the highs). All iron was made with a low frequency cut-off of 100Hz in mind.


© VT52.com

Why change from the 45 people will wonder. I love the old 45 tubes, but the tests with the RS241 and 10Y (not much in coming between these two tubes, granted) left a very nice impression lingering in the back of my mind. The 10 sounds very crisp and delicate in the highs, voices are very focused. The 45 beats the 10 in regards to bass and mid-lows. It is a compromise. Maybe I am just getting older?

The 71 as power tube was a bit muddy in the highs and lacked the extension that the 45 had. But that was as output tube where it had to do some work. As driver the 71 has very little work to do as the 10 will only ask around 2-3mA at a worst case situation swinging 30V into the grid at highs frequencies (30kHz). A piece of cake, a walk in the park. The 10 can even go into A2 a bit, giving around 1.5W of output before running out of current. The 71 is a good tube here, it can swing the needed voltage and also handle the change in grid impedance of the 10.


I've rewired the entire amplifier last week, the old wiring was a mess and I had to change some components anyway. Below are some pictures of that job. I now use silver wire all over the place (seen in the red cotton sleeves). The intention is that this will be the final amp I build for this purpose... at least for the next few years. I need to start finishing my system if I am ever to listen to it for more than a month per year :)


© VT52.com
The new layout


© VT52.com
All the parts needed for a single stage


© VT52.com
Hook-up to the OPT and the two current sources (heatsinks)


© VT52.com© VT52.com
Before and after...


© VT52.com
The power connectors (2x B+ and 1x filament)


© VT52.com
The top, nice and clean.


Looking at the load lines I wondered why not to use an even higher load on the 10. Right now I have a 12K output transformer delivering my sweet music, already a nice load. Thoughts are now floating towards having a pair of 30K OPT's made for this amp. Due to the even more linear response I will be able to drive it even further into A2 before it starts to distort as much as with the 12K load. I figure it will now handle about 50V drive into the grid giving around 2W of output power (34% efficiency). Not for the faint hearted drivers though!


The new OPT idea will be carried out, it will be a 30K:14R made for the 10/VT25 up to 30mA. Output will be around 0.5W staying within A1. Besides the higher ratio I will also pot these OPTs in a wood enclosure using bee wax. As the OPTs will not get hot, the bee wax is not a problem (melting). I can always stick a wick into it and use it as a candle should they not prove to be what I wanted ;)). Right now I'm looking for a source for silver wire as well, one that is affordable. The new inter-stage between 71 and 10 will get the same treatment as the OPT; new core/windings and then potted in wood/wax. The filament supply for the 71A will be changed to a 6V battery and no regulators, just a small resistor to drop the voltage a bit. This should simplify things even further. I'm also looking into giving the 10 a fixed bias instead of the auto bias, more on that later.


I just got back from the Triode Festival in Langenargen (DL). Inspirational juices are flowing again, must do some DIY!!! The 26 preamp was ripped apart more than a year ago to add a phono stage, but no work has been done on this since due to some other time consuming jobs. As a result I have been listening to a small backup amp all this time (prologue/epilogue amps). It is about time this changed. The general setup and tube line-up will not change, hence I put the modification under the Exodus 2 project and didn't start a new page for this amp. The power amp with it's 71A driver and 10 output tube will remain almost the same with exception to a few new parts; more on that later. 

The preamp will be expanded by a cross-over section to filter the signal going to my bass amps. In most of my projects I have built amps solely for the mid/high horn and used a filter on the output to drive the bass cabinets through a separate power amplifier. This time I will integrate the filter into the preamp section. This means two sections per channel will be added (input and output buffered). Having said this, the actual frequency range of the 26 (preamp), 71A and 10 (poweramp) will be limited to what the horn can handle; from around 80Hz and up. This relieves all three section of the bass reproduction since the signal is tapped from the input and handled by the cross-over now. What this means is that the loads used on the three tubes can be chosen lower (induction) resulting in few windings and possibly smaller cores for the transformers. Less is more is also true with transformers. I choose inductions so that the roll-off should start at around 50Hz.

The 10 will be loaded by a 30K load. "What, 're you nuts?!"... I can already hear people saying this :))  After plotting some load lines for the 10, taking into account the power I actually require and what the 10 can do into positive grid swing, I find that this load will give excellent linearity plus all the swing I could desire. Only prerequisite is that the driver is up to the task of delivering grid current to the 10. Using the 71A with IT loading will allow me to do just that. The curves point out that with the -30V bias I can swing the grid all the way to -80V and +20V before running into "problems". That's at least 50V of grid swing!!! Entering that into some calculations shows I get about 1.8W of output. That is with grid current of course... Normal A1 operations will deliver around 0.7W.


A note worth mentioning to those interested in the tube I use; not only can you use the hard to find UX210/CX310 but you can actually stick in the VT25/10Y or even VT25A and VT62/801A as well. These tubes all have more or less the same specs at the settings I use. The only real odd-ball is the VT25A which has an oxide filament instead of a thoriated tungsten filament like the UX210 / 10 / 10Y / VT25 / 801 / 801A / VT62. Another option would be the TJ 205D. Oddly they have put a UX4 base underneath this tube and filament specs are a bit off; this would mean it really isn't a 205D then wouldn't it?! :)   The 205D has a mu that is a little lower than that of the 10, so Rp is just a bit lower (around 4K-4K5), and of course filament voltage is lower (4.5V !!!!). If you use a current regulator set to 1.2A however, this will not matter as this TJ 205D pulls the same current. I might try this tube when the new amp is finished, although you could find about four pieces of NOS VT25 for the same price of a single TJ 205D tube (5 pieces if you take the mesh version). As the supply of 10Y equivalents is plentiful at this moment, there is no reason to substitute with an expensive current production tube. The VT25 family has been overlooked by the general DIY populous as it has a high Rp, oooooh, that is a bad thing people said, "give me a 2A3 or 300B with <1K. Thus the stocks of 10Y/VT25/801A has remained about the same as it was 10-15 years ago. Only recently have I seen more people start to buy this tube. Maybe the 2A3's are becoming a bore and the mono-plates harder and harder to find in NOS shape.



© VT52.com
Signal section


© VT52.com
B+ for the 10


© VT52.com
B+ for the 71


© VT52.com
Filament supplies.