Class A 8417 Amp

[desperateaudio]

I have a pair of Bogen MO60A 8417 amps. The circuit is the forerunner of the Quicksilver and it matches the Bogen MO100A circuit almost Identically. The differences are one resistor change. The MO60A circuit is very similar to what Ned says you should do if you want to correct some of the issues of the Quicksilver amps.

I want to convert these amps into Class A.

I plan to run 300VDC on the 8417
I plan to also try a 100 ohm resistor between Pins 3 & 4 of the 8417 to see what I get with triode mode.
I plan to run 10 to 12VDC neg voltage of the grids
I plan to use a 6SL7 in replacement of the first section of the 7247 and a 12BH7 as a replacment of the second section of the 7247.
I plan to leave the front end voltage as originally designed even though I plan to lower the voltage on the 8417.

Question is do I need the feedback loop from the secondary of the OPT to the cathode of the 1st amplification section?

The feedback loop is basically a 25v winding on the OPT secondary with one side to ground and the other to the cathode of the 1st amplification stage via a 56K resistor in parallel with a 15pf cap. In class A why would I still need this feedback?

Also if I really need to keep it could I not change the resistor with a 50K pot so I could dial it in as I see fit?

Thanks guys!

 

[revintage]

I want to convert these amps into Class A.

I plan to run 300VDC on the 8417
I plan to also try a 100 ohm resistor between Pins 3 & 4 of the 8417 to see what I get with triode mode.

One must know the primary impedance of the output transformer to tell. A schematic would also be helpful.

Together with an UaIa-diagram of 8417 (think there are Us 250V and 330V available) and www.revintage.se/PPABAMP.xls you can do your own estimations. About triode I think you have to plot your own curves.

But I would go for 6550 or EL34 instead.

 

[desperateaudio]

Well I finished the Class A 8417 Triode Amp. It started life as a Bogen M60 8417 Amp. I have many 8417 and think when ran correctly they are great so it was not an issue for me to choose this tube. I started with the Quick Silver schematic and made the following changes. SS rectifier, input tube became half a 6SL7, driver tube became half a 7247, bia voltage is -12, output tube voltage is 300vdc, 8417 are in triode mode via 100 ohm resistor, the original feedback loop is removed, front tubes are ran with same voltages as original QS amp. New amp can also run 6550 in triode mode at 300vdc +/- 15 vdc. I use the Bogen M60 OPT's since I find them to be sleepers. The amps sound great, very full without any straining. If you got the time and the tubes give it a shot.

 

[Burnedfingers]

As a Quicksilver owner I am quite interested in your modifications and wondered if you would mind sharing a schematic with your changes. I have also considered picking up several Bogan commercial amplifiers to play with.

 

[desperateaudio]

New baby makes that tough right now but it is right on with the QS schematic except for the following.

1. 1/2 a 12ax7 becomes 1/2 a 6sl7 with a 1K cathode resistor no feedback connected to the cathode. Plate load resistor becomes 220K This is specific to the PS I built which is SS rectified

2. 1/2 a 12au7 becomes 1/2 a 7247. Cathode resisitor is reduced to 22K. Plate load resistor becomes 15K. Again specific to my PS but I am wondering if the plate resistor and the cathode resistor must match. Can not remember the rule??? Can you clarify?

3. Coupling caps become .47@600vdc from .18 and the 150K grid resistors become 51K each.

4. Bia voltage is -12vdc.

5. 8417 plates are seeing 325vdc and they are coupled to the sg2 via a 100 ohm resitor and ran in triode mode. This could come up a little if need be since it is well within tolerance I just do not think I need it as of yet.

If you could let me know about the need for matching the plate and cathode resistors on the 7247 that would be great. If they do need to match I am hoping 15K on each works as the PS is not as high as the original.

So far for me the OPT's on the big Bogens have been very good. I have cut a few in half and they are built very well!!!

Thanks in advance on the resistor question.

Also I am run Sylvania's rebranded as RCA's on the 8417.

 

[Burnedfingers]

Could you post the stock Bogan schematic?

 

[desperateaudio]

Here is the QS schematic I started with. I do not have the MO60 schematic but it is exactly like the Bogen MO100 amp except that amp uses 4 - 8417 tubes on the output and that schematic is out on the web somewhere I have seen it.

Any comments on the driver setup for the 7247 I spoke about having the 15K plate load resistor and a 22K cathode resistor?

Would it be better off if they both had the same value?

 

[Burnedfingers]

Why would you use a 7247 instead of a 12BH7 or better yet a 6sN7?

How many mA thru the 7247?

What is your voltage feeding the front end?

 

[Burnedfingers]

I could be clearly wrong here but I would think that a trimmer would be in order to match the waveform off the plate and cathode side of the driver tube. The 15K with a trimmer may be the way to go.

Maybe someone with more experience could jump in here and answer this.

 

[revintage]

Go for two matched 15k resistors for the PI to work best. Maybe you could add a trimmer to adjust other unbalances in your amp.

 

[desperateaudio]

Well one half of a 7247 is very similar to a 12AU7/12BH7 but yes I will most likely change to the 12BH7 shortly. I prefer the 12BH7 over the 6SN7 in this instance. Otherwise I love the tube.

I will change to matching 15K resistors this evening. I just could not remember the rule well enough so THANK YOU SO MUCH!!! for the help on that revintage.

Does anyone think that I should have kept the original feedback loop?

Thanks,

 

[desperateaudio]

Here is the schematic as I have already modified it and as I will modify it from the changes suggested today.

The real last question I have is if anyone thinks I should have kept the original feedback loop?

Thanks for everything!

 

[Boris_The_Blade]

You may want to put in some small value plate or cathode fusing resistors on those 8417s....just in came something bad happens

As for feedback, increase or decrease the resistor until your gain is suitable while keeping stability, then add a parallel capacitance until you have a nice square wave output.

nice work

edit: how much gain do you have now? (this is your open loop gain since you have no feedback)

 

[desperateaudio]

I have not officially measured the gain but it is driving pleanty currently. I really do not want the feedback if I can help it. I have no stability problems so I would like to do without. If anything it is pushing to much. I have swapped out the 8417 with 6550 and rebiased with no issues and the amp has a high ceiling. I will redue the PDf schematic tonight with actual voltage measurements and post tomorrow. From there I can calculate the gain.

Oh she sounds pretty good and with these final touches she should really sing.

Thanks

 

[Boris_The_Blade]

There is nothing wrong with feedback. Most if not all hi-fi amplifiers use it.

Some benefits:
gain reduction (if you have too much)
noise reduction
hum reduction
lower distortion

some cons:
gain reduction (if you don't have enough)
potential instability (if not done correctly or there is too much feedback)

I think you would enjoy how much cleaner the amps would sound if you did add feedback. As an option you could have a switch that turns the amp from open loop to closed loop with feedback...or even to what degree of feedback you want.

just a thought

 

[desperateaudio]

Ok so I will give it a try no harm no foul. With that said would you go with the original feedback setup as shown in the QS schematic I originally posted? The 100pf to the bottom output tube seem unnecessary. I am thinking a 10K pot replaces the 5K resistor so I can adjust and them jump that with a small cap of no more them 390pf.

What do you think?

Thanks as always.

 

[Boris_The_Blade]

Yup a 10K pot is a good place to start. May want to put a 3.3k resistor in series to avoid problems having the pot in the wrong place at startup, you'll probly not need to go below 3.3k anyways. Make sure it is not a wire wound pot. Instead of the 100pF from plate to feedback, you can just place the cap across the feedback resistance (pot+3.3k). It also doesn't need to have such a high voltage rating then. To verify if you have enough or too much feedback, use something like a 1KHz sine and measure the gain (around 1Vrms input for full scale output power, whatever it is..., is usually good).

Once you have the gain figured out, then play with different cap values until you have a nice square wave response. I'm assuming you have a function generator for this...if not you can pick them up cheap. You can usually try a 1KHz square or 10KHz square wave. The flatter the better. I'm curious to see the square waves then if you can get them on the computer somehow. I'm working on a bogen MO200 which probably has similar quality transformers.

good luck

 

[desperateaudio]

Thanks so much.

So no wire wound then I expect a good multi-turn precision unit is preferred?

So start with the 100pf cap across the pot/series resistor and go up from there got it.

 

[Boris_The_Blade]

So no wire wound then I expect a good multi-turn precision unit is preferred?

Sometimes those pots are wirewound. Here, a simple one-turn carbon will do. Rat shack ones are fine.

yup, 100pF sounds like a good place to start.

 

[Burnedfingers]

Did you modify the bias circuit to allow for separate bias adjustment for each output tube?

I don't know but if it were me I would surely try a 6sn7 for a driver. I would also try the feedback loop and see how it sounded.

Just my .02