Electrolyticless Tubelab Simple SE

[nic6paul]

Well I decided with all the extra parts left from my first Simple SE build I have enough left to do a second amp. I wanted to make this one a bit more unique and go with all capacitors to be non-electrolytics. I have the power supply covered and I am using all motor runs. For C1 I will be using a 50uf motor run, C2 100uf and an additional 150uf to be added. There was another member of the board that built a similar Simple SE so I have basically copied his power supply.

The area I am not too sure on is 6.3v 1500uf Cathode bypass cap. How low of a value can I go here. Finding large value film caps is hard/expensive. Also the the two caps at C12,C22 how much lower of a value can be used in place the 1500uf caps specified on the BOM. I figure this is going to get kind of expensive so it may not be feasible but I thought why not give it a try.

[SpreadSpectrum]

It sounds like it might be cheaper/better(if you can't stand electrolytic caps) to modify for fixed bias. Remember, there is no cap like no cap.

[arnoldc]

Remove the cathode bypass caps and you lose some gain. So if you can live with that, then you're fine.

[rock4016]

These are $23 each, but would work fine as a cathode bypass cap.

http://www.parts-express.com/pe/psho...Number=027-447

[tubelab.com]

OK, I'll tell you how to make a non electrolytic Simple for very little money, but you will need to do some serious experimenting.

You already have the power supply figured out. Motor run caps. So build your board, wire in some motor run caps, leave the cathode bypass caps out, and test the board. The gain will be low without the caps, but you should still get sound. You can tack some caps on the back of the board to test if you want. It is always a good idea to start with a working board otherwize you won't know if your experiment was successful.

The area I am not too sure on is 6.3v 1500uf Cathode bypass cap. How low of a value can I go here.

I use oversize capacitor values in my amplifier designs. There is a real reason for this despite what some "experts" say in other forums. I took a bunch of capacitors into work and tested them on a fancy HP component analyzer. Real capacitors are not perfect, they have a resistance and an inductance associated with them. I simply tested a bunch of caps and picked the ones that had the lowest ESR and ESL. I guess that is the engineer in me. I doubt that you han hear the difference with any good cap over 200uF.

Remember, there is no cap like no cap.

OK, lets run with that. If we want NO CAP, how do we get NO CAP? Ask yourself, what would SY do? You guessed it LED's. Using an LED for the cathode of the 12AT7 works good. I have done it. No cap and no resistor, just an LED. The catch......you need the right LED. LED's are quite variable as to their forward voltage drop for different part numbers of the same color. 12AT7's are quite variable as to the cathode voltage required to get a reasonable plate voltage. The current is fixed by the CCS chip. It is possible that the two sections of a given tube are different enough to require two different colored LED's.

Finding the right LED requires some experimenting and a bunch of LED's. I soldered a pair of alligator clips on short wires into my board and started trying a bunch of LED's. You want an LED that puts the plate voltage in the 175 to 250 volt range. A more technical approach is to put your chosen 12AT7 in your working Simple SE and measure the cathode voltage. Then look for an LED that has that much forward voltage at 10 mA.

OK now what do you use for the cathode bypass in the output stage? I would be tempted to wire a few motor run caps together and hide them under the deck, but there may be other alternatives. LED's? Well yes this is possible, but it will take a BUNCH of LED's. Yes, I have tried it. Yes it works, and yes the amp lights up the whole room! I have a friend who builds all sorts of Gizmos using PIC chips and LED's and he buys LED's by the hundreds from a vendor in Hong Kong. There are some big LEDs available now that operate at 100+ mA and drop 3 or 4 volts. They are blue or white and are often sold as "1 watt" LED's even though they run at about 1/2 watt. He brought over a bunch and we wired some into a Simple SE. I think we had 12 in series in each channel to get the current right.

Another possibility is to use a big zener diode. It would take a 10 watt zener diode. The voltage would need to be the same as the cathode voltage measured on the same tubes in the board with cathode resistors. I have not tried this yet though.

It has been suggested to use fixed bias. I have done this, but a negative voltage source is required. This implies another power supply.

[Sheldon]

Some cathode loads proposed in this thread: http://www.diyaudio.com/forums/showt...65#post1867965

Also, you can use an LED or other similar load with a series resistor to linearize it further. So, something in between a bypassed resistor and a constant voltage load. Of course, gain is in between too.

Sheldon

[arnoldc]

Sheldon, thanks for that link. I've been away from DIY Audio for a while and I feel like I'm a newbie

[2bz]

I built mine with Elna Silmic II electrolytics. Sounds almost as good as films but alot cheaper.

[huskydawg9]

Originally Posted by tubelab.com

I doubt that you han hear the difference with any good cap over 200uF.

You can still find the red blackgate N series non-polarized 470uF 6.3V, for less then 10 bucks...

[nic6paul]

I did finish this project. It sounds great. I used four 600uf asc caps for the cathode bypass capacitors and the power supply is all motor runs. The downside....the amp is 2 feet wide.