Seems like you forgot English grammar too. " ' "

The SG505 and AA501 are the ones to have. Mine have been flawless for the years I have had them (knock-on-wood...) I always wanted that DM501 also but ended up getting other meters. I am pleased you enjoy my videos. I will try to make a few more. Stay safe.

Thermisters are good to use in the AC power line. I have used them many times and they will reduce inrush current. As for preventing over-voltage on electrolytic capacitors when using solid state rectifiers , they may not help so much as there is no load on the power supply in a vacuum tube amplifier until the filaments warm up but overall they are often a good idea to use.

I agree with you and they come in higher voltage ratings also. Thanks for your comment.

Hi Lou, That is a big question. First off, we must know the filament voltage and current of each tube. In virtually all cases, the filament voltage is 6.3 but the required current is higher in the higher power tubes. Start with a 6V6 - fil current is 0.45A, 6L6GC fil current 0.9A, EL34 fil current 1.5A and KT88 fil current 1.6A. So, from this perspective, the KT88 although it is 100mA more per tube than the EL34, it will probably be OK. Now we have a bias issue. Back to the four tubes above - the 6V6 needs about -15V bias, the 6L6 about -37, the EL34 around -37 similar to the 6L6 and the KT88 more like -55 V bias. These are ball park bias values and must be adjusted to not exceed plate dissipation for each tube type. The point is, if you put a KT88 in the socket of a EL34 and can not adjust the bias to a safe plate current, the KT88 can get way too hot by drawing too much anode current. The funny thing about this is, you may be able to get away by putting a 6V6 in a KT88 amplifier (because the bias is so negative it will keep the 6V6 current low) but you can not put a KT88 in a 6V6 socket for the same reasons, i.e the bias voltage is too low and, and as can be seen from filament current requirements, the filament of the KT88 is almost four (4) times higher. If you try the KT88 in the Dynaco ST70, you should monitor the plate current with something like a bias-probe to make sure anode (plate) current is not exceeded. Hope this helps.

2500 VDC is about the highest plate voltage you want for a pair of 813. A 3200 volt CT transformer into a full wave center tap rectifier (FWCT) and then into a capacitor filter should give you about 2000 VDC. A large sized microwave oven transformer (MOT) is between 1800 and 2000 volts and into a bridge (four diodes) will end up between 2250 and 2500 VDC with a capacitor filter. However... microwave transformers have one end of the HV winding connected to the frame (ground) of the transformer so that wire will have to be carefully removed and properly insulated - probably not the best idea to use MOT's. If the 3200 you are mentioning above is DC out of a rectifier, that is a bit high even into a fully loaded choke input supply which will drop it to about 90% or 2900 volts. Be sure to use a bleeder resistor on your HV PS and be safe... High Voltages at this level should be considered, Death On Contact. You can find my email address on QRZ.com 73 WA4QGA

I think that happened to me twice as the voltage doubler capacitors failed twice. After the last replacements, a few years ago, they have been stable and reliable and run several hours per day.

This is the third time I have had issues with the electrolytic capacitors in this unit. In this case, I think it was a little over-voltaged at start up but on two occasions a few years back, the capacitors in the power supply went bad - I believe opened. The HV power supply is a voltage doubler and nothing is over-voltage in that area - the two larger brown ones in the video. Maybe I changed brands and that is why they have lasted the last few years. Like you say - just junk capacitors. I have electrolytic capacitors in McIntosh amplifiers 50+ years old and they work perfectly. This little amplifier can run 12-15 hours a day at times as I have it on with background music. Thanks for your comments.