OK need assistance or clarification of output transformer values

[AdrianH]

Chris, thanks for the comments.

Never thought of that. I know that the response rolled off towards the top, but never considered as I go up it would also a.c. ground the cathode and could up the gain.

What I could possibly do is remove the 270 pF cap and stick a 20pF cap across R9, the 470K, the reactance would be around 470K at 17KHz and should start to shunt the signal to ground.

C3 is large just to provide feedback at bass frequencies, I hope to stop bass boost? Feeding back about 1/3 of the audio voltage to the speaker, there seems to be enough gain in the second stage and it certainly helped on the distortion front. Most of the feedback was done by trail and error trying different value resistors and seeing what happened to distortion. I am still trying to understand it all really.

Anode Volts for U1A/B are 102/104 Volts the cathodes at 3.0 and 2.8 Volts with 0.84 and 1.5mA Ia respectively.
The output stage is 24mA total and I am relying on the valves being reasonably balanced to share correctly.

I am enjoying this too much, but my Ebay source has closed shop during the pandemic so will get stuck for parts to build anything bigger then with pencil valves.

Adrian

[Diabolical Artificer]

Well done, mucking about with valves is fun, gets the old noggin ticking over too. A few observations on your amp, instead of the 270p on the cathode you could try a 2200p and 1k or similar from each anode to ground to roll off HF gain. Also why not apply NFB to the first IP valve, split Rk either by adding a 390r on the bottom or using a 3k3 and 390r, apply NFB between the two, bypass cap to the same junction. Also sub a 5k pot for the 3k FB resistor, 10 turn is good if you have one to see what level of NFB is best. this will improve F response. Other thoughts, move the 470k grid leak from the 2nd stage to the 1st, and try loosing the 33k grid stopper, just to see how it affects F response. To check F response display the OP on your scope, then switch ranges on your sig gen (if this is how yours works) to give a quick sweep, EG 40hz, 400hz, 4khz or 100hz, 1khz, 10khz.

I've loads of bits, as have many folk here, more than enough to see me out, if you want a few valves, [ have bags full of EF80's you can have : ) ] etc drop us a PM, if you pay the courier they're free.

Andy.

[AdrianH]

Hello Andy, again thanks for the comments
A few responses, it started with a circuit from the net called Powerman, it looked simple so I built it, it uses the valves I have to hand the 6021's, and for me it has grown from there.

The 470K grid leak on the second stage, I put there to make sure that if a wire broke from the pot, or if the pot was intermittent, the grid was always grounded stopping higher anode current, it probably would not be needed but I was unsure.
Yes I could add NFB to the first valve, changing the sense from the output transformer as there would be another 180 phase shift, move the Pot to the first stage. At the moment the first stage is just acting like a pre-amp. So I will try and work out options.

OK on the EF80's something I will make a note of, the issue is the hardware as well valve's. Holders, turret board, chokes, caps etc. Things that I am sure most of you have from years of collecting and disassembly. I was hoping to order old MOD units from a vendor on Ebay allowing me to strip them down and re-use what I could, or go to the BVWS sales and buy parts, but I will just have to wait as non of it is essential.

I have managed to download the Radio Designers handbook, in two forms before I realised that Radiotron Designers Handbook was the same. But there is a lot of reading there on screen which I find hard to do but it gives me another reference as well as Scroggie.

Last night and a few others, I have been either with test gear or on the PC until 3am, much to the disgust of the wife. I get something in my mind and sort of forget the time. Last night was working on Kicad and slowly progressing from circuit to PCB. It is fun and the only way I learn is by having a target and doing.

No surface mount devices here!

Adrian

[Herald1360]

C6 could be if you cut the wires off!


If you're doing hand assembled PCBs, the larger formats of SMDs are easy enough to work with and still use less board area than those "Tokyo skyscrapers" I see in the picture

[AdrianH]

Think you mean the resistors, I used them as I managed to get a complete E6 1/2Watt resistor set with about 100 pieces in each value, so may as well use them.

Was aiming to use them with turret board which I have since ordered and SMD are not very good on turret connections.

Adrian

[Al (astral highway)]

Neat little project, Adrian. It seems like a good fit way to spend lockdown hours

One query, what's the deal with the valves being mechanically attached to a big heatsink, and a fan for good measure?

[AdrianH]

Hello Al;

When I got the old MOD guard receivers they came in a small metal box with all the circuits and every valve has a metal shield that goes around it and also the clip that takes the heat into the case, so I reused them and put them on a heat-sink to act as a cooler. The fan I added as I still could not put my hand onto the heat-sink after a while, it is a 12 Volt fan running from 6.3 Volts DC so a lot slower then it should be.

The specs do have a maximum temperature and I was not sure if without the shields and clips if this would get exceeded?

The heaters are still 300mA at 6.3 Volts so there is ~1.9 watts just from the heaters. The output pair have 12ma at 145 Volts in them so that is a further ~1.7 Watts, so ~ 3.6 a valve max into a small pencil valve which I was thinking could not dissipate the heat as well as a larger valve.
I am guessing that if I remove the heat I will have less issues with the strip board/pcb material. It may not be a problem without the metal work, I am just not sure.

Just being rather safe then getting it wrong I guess?

I have ordered 5 boards from China, which will be a first for m to draw up and layout.on a PC. I just want to see what they are like.

Think my next project after reading one of the threads on here will be a Top Band micro power a.m. transmitter with the 5840 Pentodes and 6021 Triodes, that will be interesting for me, then..... I will probably run out of valves.

Adrian

[AdrianH]

Just to bring this little project to an end. The boards turned up from China. Very pleased with the results, everything seems to be in order, over the next few days I hope to replace the strip board amp with the finished unit, even have it playing 50/60's music from the PC headphone socket.

Adrian