A modified Magnavox 8802 in the Gillespie style

[RUmad]

Aahhhhh, the sweet sounds of success!
When you can, maybe you can tell me about why you are using solid state diodes over the tube rectifier? Do you feel it's necessary? I added a 47uf before the 100R resistor and a 9 henri choke in place of the 470R between 40's. Gonna check voltages again after she warms up a bit. If you guys think I should change out for diodes I will. Also, I know you replaced the otx's with Edcor, but for now at least I'm staying with stock. Advice on what to do with the original feedback loop? I was thinking, even left a question on Dave's 9300 thread, that I would use the same as the 9300 I did. But you did it differently. Because of the Edcors? Love some input on that. Thanks for answering so quickly this afternoon, I was able to get it done because of that.

 

[thorpej]

I went with SS diodes because I wanted to use different OTs and change the operating point when I did so, requiring higher voltages.

As for the original feedback loop, if your open loop gain changed, you may need to revisit the feedback loop, yes. I estimated the gain of the circuit, and calculated for -12dB feedback. I ended up being very close on the gain estimation, and the feedback measures at -11.1dB... close enough! The 330pF feedback cap was then added to help tame down the square wave overshoot I was getting from Edcors. (I still haven't done anymore HF tuning -- my kid wanted his amp!)

 

[RUmad]

I hear that. This one's for my daughter. She knows I'm building it, but doesn't know it's for her yet. I have a feeling I'm gonna pull the rectifier and add the diodes. I like 6v6's at higher voltages. I've got 290 at the first cap now, down to 265 on the screens. Thanks for your help! I'm glad I could incorporate Dave's AWESOME EFB into this, and someone ELSE did the math for me!
Thanks brother.

 

[RUmad]

Pin 1 on a 6v6 has no internal connection correct? I already used pin six to add the screen stabilizer resistor. Is it safe to tie the (3) plate connections to pin 1, eliminating those wires from the rear terminal strip?
These;

 

[thorpej]

I used pin 1 on my 6V6 sockets as a tie point for the cathode connections, and the GE data sheet explicitly says "no connection". The pin is even omitted on some tube samples.

 

[thorpej]

If you're using the original OTs, I'm not sure there's a lot of benefit to higher voltages... You might consider drawing up a load line to see what the plate current ends up at.

 

[RUmad]

Good point. I'll have to see if I can even remember how to do that. Thanks. I can't wait to convert to floating paraphase now, it sounds ridiculously good already.

 

[thorpej]

Happy to help out with that. Movie-watching with kiddos right now, but I can draw up a load line tomorrow.

 

[RUmad]

Yes sir. Busy weekend here too. I'll be around. Thanks man!

 

[thorpej]

Well, I quickly threw together the class B load line for 285V plate / 250V screen with the 6K-CT stock OT, and it looks like you can get about 10W max output, but you spend a fair bit of time above the max dissipation curve. And increasing the screen voltage won't help you squeeze much more output power out of it because the line is so steep. I don't think increasing the plate voltage with the stock OTs is worth it.

 

[RUmad]

That's ok. I like tube rec better anyway, I just figured I'd do it if you guys said it'd make a difference in sound. I appreciate you taking the time to do that for me. I think I'm going to just rework the existence ng setup. What I have now is; rec-47uf-100R-47uf-choke-47uf-10k-22uf. I think I can neaten it up a bit, while lowering the screen voltage a bit by doing;
Rec-47uf-choke-47uf-470R-47uf-10K-22uf.
Essentially moving the choke into the first slot, removing the 100 resistor and putting the 470 back in. Make sense?
Also have to figure the right bias with the voltage lower than yours, but that I can figure out.

 

[thorpej]

My EFB configuration has considerable range. You probably are looking for something in the 15-17V range, I guess.

 

[dcgillespie]

If the OPTs are in fact 6K primary pieces, then the very best thing you could do for the amplifier is to install EFB(tm) by way of using a cathode bias regulator. Besides adding EFB action (the "Enhanced" portion of the concept), the cathode bias regulator also converts the output stage from straight resistive cathode bias operation, to true fixed bias operation (modified by the enhanced element), as surely as if a conversion to conventional fixed bias operation with a negative bias voltage source had been made. A 6K load is much too low of a load for use with 6V6 tubes and resistive cathode bias. Adding EFB then would/will/did produce a major improvement in performance.

Dave

 

[RUmad]

You're darn right it did Dave. The highs and lows are so beautiful I almost don't miss the mids that the 6bq5's provide. I'm also getting jealous that I have to give this away. And all this before I've even converted to your floating paraphase circuit! (Although I DID replace every single resistor and cap in the thing). Can't wait to finish! And my bias test point jacks just showed up. Thanks for the input gentlemen, pleasure as always

 

[RUmad]

So, as family functions wouldn't allow it, I didn't get much done over the weekend. Spent the whole morning converting to the floating paraphase. Two things. One, it's a-ma-zing. Two, I don't think I'm going to do the whole 'square one' 5"thing anymore. I didn't realize how many brand new resistors I'd be throwing out. Oh well, you live and learn. Hopefully. Anyway, just a few bits a pieces left to do. Sounds fantastic. Thank you guys.

 

[dallen9494]

Hi,

After the truly amazing results I have achieved from rebuilding my 9300 (thanks Dave, and all who have contributed!), I've decided to rebuild an old 8802 that I'm going to give to my son, following this rebuild that Thorpej created. If it turns out as nice as my 9300, it will be spectacular, I'm sure. My question is in regards to the output transformers. I have a couple of options and wanted to get some input on your thoughts. I've got the original 9300 transformers (I used the Z-565 48's), I've also got a pair of Bogen RP 230 transformers (7408 transformers), and I also have a pair of Harman Kardon A500 (actually the receiver t5000) transformers (7355 tubes). The Bogen is complete, so I'm debating whether I should pull them off it, I'm thinking I might eventually get to it!

If I use the 9300 transformers, should I use the feedback that Dave has on the 9300 build, or is it entirely different since I'm using 6V6's?

Thanks for all the amazing work everyone does here, I'm slowly learning!
Dennis

 

[dcgillespie]

What type of output tube biasing system will you be using?

 

[gadget73]

7408 is a lot closer to a 6V6 than a 7355 is. A 7355 is pretty nearly a 5881, which is in the 6L6 family. You could probably re-work it to run a 6L6 family tube but at that point you'd basically be gutting it to nothing and using the chassis. If you wanted to do that, I'd probably just take the A500 and copy it's amp circuit directly, subbing a 5881 or 6L6GB for the 7355 tubes. Possible the power transformer won't be able to provide enough B+ current though.

 

[dallen9494]

Thanks for the quick response. I was planning on following the build Thorpej did with EFB. I'm leaning towards the 9300 transformers, since I already have them leftover from the amp I upgraded. I plan on using a pair of 6 ohm ADS speakers with the 8802.

I wasn't sure about the 7355 transformers, I've got two quads of nos tubes, so maybe when I get a little more experience I'll try to build a copy of the A500 amp circuit.

Thanks for the help!
Dennis

 

[dcgillespie]

Dennis -- If you are planning on implementing the EFB circuit with 6V6s, you will want to ensure adequate heatsinking of the 337 regulator. With an estimated 100 mA flowing through it, and the higher bias requirements of the 6V6 family, it will be dissipating 3 watts of power -- which is fine as long as the heat sinking is adequate. As you know, I simply used the chassis, which should be fine, but not knowing your build, it's a point to be aware of.

Also, you can use the same exact NFB arrangement, but you will want to increase the value of the 100Ω resistor running between the two cathodes of the driver tube to 180Ω to make up for the loss of output stage gain. Input sensitivity will require just over 1 volt to develop full power output.

Good luck with it!

Dave