• Disclaimer: This Vendor's Forum is a paid-for commercial area. Unlike the rest of diyAudio, the Vendor has complete control of what may or may not be posted in this forum. If you wish to discuss technical matters outside the bounds of what is permitted by the Vendor, please use the non-commercial areas of diyAudio to do so.

TSE 2A3 C3

It wasn’t supposed to make a difference but, my monoblocks vastly improved when I replaced the recommended Elna 4700uf with something fancier. I’m not saying what I used so as to keep any discussion theoretical. I’m not shilling for anyone.

I understand that ac has a direct path to ground at the other (-) end of the filament but this cap sure looks like a cathode bypass cap across the 1 ohm filament of the 2A3.

The reason I became even curious about this is because I also recently built a 45 version TSE that utilised (fancy) filament regs (that have no cap at all there, in addition to other “thought-provoking” properties) and it sounded so good I thought my 2A3 amps were broken.

I’m curious what the ee’s and George think. Am I trippin?

While I’m here, I’d like to say thank you for all of it. I learned a lot from building these circuits and have enjoyed their sounds immensely over the years.

This is the Tubelab forum, and this post is about the TSE, but there are actually TWO TSEs and the difference is in the filament regulator. The OP did not specify which one he is referring to. The "original TSE" was the first PC board offered by Tubelab. It is a 20 year old design that used a Sharp regulator chip that went extinct when ROHS made the lead free mandate. It appeared that nobody was buying the parts except for TSE builders so Sharp just killed the part rather than making a no lead version. Fortunately Mouser, Digikey, and RS all had plenty in stock, until someone bought up all of them in the same week. This led to a redesign of the board in 2018. The new board is called the TSE-II. It uses a Microchip part while also incorporates a list of requests from this forum, most notable being a larger board to better handle some "300Bs" that are quite large. The audio path in both designs are identical.

The TSE-II redesign details are here. The schematic and BOM are in post #1:


The original TSE schematic is included.

My opinion about the capacitor across the regulator?

No capacitor is perfect. All have some associated internal resistance and inductance. The filament regulator is a voltage reference feeding a current amplifier stage which is stabilized with a huge amount of negative feedback. The output voltage is compared against the reference and the difference steers a feedback loop and current amplifier that tries to make this error voltage zero. Ideally this works much like a solid state amp that compares the speaker voltage to the input voltage and tries to make the error zero while keeping the output impedance as low as possible. As with any high feedback solid state amp, this loop can be confused, especially when running at or near its maximum load current with an inductive load. This means that it is theoretically possible to hear differences with different caps across the filament in your TSE or TSE-II. The 2A3 is the worse case as two tubes are in parallel so there is a 5 amp load on the chip. Some "2A3s" are a bit greedy in the current department too. It is also possible that your regulator is on the edge of dropout if you have low line voltage and some greedy tubes. See if the same differences exist with one output tube removed and only one channel running.

Not sure whose regulators you have but some builders seem to prefer Coleman's boards over the Microchip part, but I have never tried them.


  • TSE_board.pdf
    70.7 KB · Views: 63
Ok. Sorry. It’s crude but I think this is it. The regulator is right out of George’s schemo.


  • 36EABC93-57C0-4373-9046-2D9A2D331B33.jpeg
    652.7 KB · Views: 58