Traynor YBA- Custom Sp.- convert 6ca7 (EL34) to 6550

[pbman1953]

Hi

Has anyone converted a YBA-3 from 6ca7 to 6550?

Here's the process-

Disconnected the bias voltage from pin 1 of the output tube sockets, shorted pin 1 to pin 8 (ground)

• Removed the two 47Ω screen resistors, installed four 1kΩ 5W screen resistors
• 
  My questions-
  
  Is this mod worth it? With this mod will the sound be closer to a Sunn type sound. Meaning smoother with less grunge. Before I do anything I will check the bias to be sure it's not running (hot) on the higher of the scale for the stock tube.
  
  
  I'm not a tech but have worked on many amps including cap upgrades. I am very careful. All I need is some direction and I know I can perform the mod. 
  
  Step one says to connect pin 1 to pin 8
  Above step two is to remove the 2- 47 ohm resistors. The install 4- 1k 5W , but does't explain what connection type.
  
  Thanks for any help
   
  
  

[sluckey]

I don't think this mod will make the amp sound any closer to a Sunn amp than it does with the 6CA7s. Bass may be a bit stronger. Sunn used a completely different preamp and power amp circuit. BTW, a lot of Sunn amps used 6CA7s, but the circuit was still the same as the 6550/KT88 amps.

Disconnected the bias voltage from pin 1 of the output tube sockets, shorted pin 1 to pin 8 (ground)

This step will result in running the suppressor grid of a 6CA7 at zero volts rather than -50V. I don't think you'll be able to hear a difference. But... The 6550 has no suppressor grid and pin 1 'may' be connected to the tube base if the base is metal. Connecting pin 1 to ground makes sense. Doing so will allow you to use either tube type.

Removed the two 47Ω screen resistors, installed four 1kΩ 5W screen resistors...
...Above step two is to remove the 2- 47 ohm resistors. The install 4- 1k 5W , but does't explain what connection type.

This is probably a good idea. If pin 6 is not used this is how I would do it...
First remove all wires and or resistors from pin 4 of all four tubes. Now connect all pin 6s together and also connect a wire from one pin 6 to R30 (If R30 had been connected directly to pin 4 of a socket just reconnect it to pin 6 of that same socket). Finally, mount a 1K/5W resistor between pin 4 and 6 of each power tube socket. (Leave the leads a little long so the resistor will sit high above the socket.)

[pbman1953]

My replies-


Disconnected the bias voltage from pin 1 of the output tube sockets, shorted pin 1 to pin 8 (ground)


So are you saying, don't bother with this step?


This is probably a good idea. If pin 6 is not used this is how I would do it...


All pin 6's are being used and are already tied together. On the left side the first pin 6 is connected to V3 preamp tube. On the right and last in the row is connected to the Standby switch and light

[sluckey]

So are you saying, don't bother with this step?

No. What I said was... "Doing so will allow you to use either tube type."

All pin 6's are being used and are already tied together. On the left side the first pin 6 is connected to V3 preamp tube. On the right and last in the row is connected to the Standby switch and light

WHAT? That's not what I see in the pics at all. There is nothing connected to pin 6 on the left or right socket. The two sockets in the middle have a 1500Ω connected between pin 6 and pin 5.

But, since two of the sockets are using pin 6 you'll need to just use a terminal strip or standoff turrets to mount the 1K/5W. Or you could still do it as I said above but you'll need to 'fly' the 1500Ω resistors to pin 5, thus freeing up pin 6 for the two middle sockets.

[eleventeen]

In your second pix above...the way you have the lead from your 5W screen resistor (connected to pin 4) crossing over tube socket pin 5.....

makes me rather nervous! I realize that often, posted pix are very lacking in their portrayal of depth. But, if that lead touches your pin 5 that output tube is probably destroyed in seconds.


If you threaded that lead through pin 6 and ran on to pin 4, then seriously bent pin 5 down towards the chassis...OK.

[pbman1953]

I looked closer at the sockets-


Starting from the left, socket closest to V3 preamp tube


1-  pin 6 is un-used
2 & 3 (middle sockets)-  you are correct -pin 6 & 5 are connected with a 1.5k resistor on each socket
4- pin 6 - un-used


While working the mod I will make sure nothing is close to pin 5 and any un-used pin 5's will be pushed away from any exposed wire leads- Thanks for noticing that




Ok, let's re-cap the steps-




1)  On all tubes jump pins 1 & 8


2) Remove both 5 watt resistors


3) Remove remaining wires from the other pin 4's, which is the 1k/20 watt feed 


4) tie all pin 6's


5) Take one wire from the 1k / 20 watt to only one pin 6


6) Install a 1k/5w between each pin 4 & 6


I have height in the  chassis to be able to " fly" each resistor


One question - does the resistor (1.5k) connection between pins 5 & 6 , on the middle sockets, change? Meaning, when I tie all pin 6's together that connection still stays?

Thanks

[sluckey]

...any un-used pin 5's will be pushed away (There won't be any unused pin 5.)

1)  On all tubes jump pins 1 & 8 (and be sure pin 8 is connected to ground)

One question - does the resistor (1.5k) connection between pins 5 & 6 , on the middle sockets, change?

Yes. Unsolder the wire and the end of the resistor connected to pin 6 to free it up. Leave the resistor connected to pin 5 and swing the resistor body around so it is pointing away from the socket. Now solder the dangling wire to the dangling end of the resistor. I would use heatshrink tubing to insulate the solder joint and the resistor body. Do this for both middle sockets.

[pbman1953]

Great, thanks for the help!

I'm not sure if this amp can handle them but I was wondering if I could use kt 88 in this amp. I already a Sunn 2000s with 6550 tubes.

Thanks

[sluckey]

KT88 and 6550 is pretty much the same thing.

[pbman1953]

"1)  On all tubes jump pins 1 & 8 (and be sure pin 8 is connected to ground)"


All pin 8's are tied together and with checking with my meter, pin 8 does ground out to chassis.

[pbman1953]

I just bought the 1k's and will probably do the work later. I do have a mached pair of 6550's and need another pair. While looking can I test the mod with 2- 6550 and 2 el34's?


I won't stress it .. Also, I'm going to remove the spring clips, because they are too small, and install spring tube retainers

[pbman1953]

I just noticed that all the 8's are jumped to each other  and the 1's are jumped to each other.


I suppose all I have to do is jump one set and all the 1's & 8's will be joined to ground?

[sluckey]

While looking can I test the mod with 2- 6550 and 2 el34's?

I would not mix them. Test it with 4 6CA7s or just two 6550s. You'll probably have to tweak the bias for each type tube.

I suppose all I have to do is jump one set and all the 1's & 8's will be joined to ground?

That's electrically correct. It would certainly look nicer to just jumper pin 1 to pin 8 on each socket and do away with half that duplicate wiring.

[pbman1953]

Hmmmm, I get a hum and the 6550's get real red

[Willabe]

Shut it down, their red-plating, it will kill them running them that hot!   

You have to adjust the -bias, increase it to cool them off.

[pbman1953]

with the bias control maxed it still hums and tube are red, but as soon as I see red I shut down

[pbman1953]

I just checked with my Bias Rite,


Voltage is just over 500v


Idle current is off the chart

[Willabe]

Leave the power tubes out until you are sure you have -dcv on all the power tube grid pins and have enough -dcv.

Take the power tubes out and measure the -dcv at each grid on the power tube sockets with the -bias set full up.

Post what you get.

If you have -dcv then;

Here's a hi-lighted drawing of you amps -bias. Blue circle shows 10K adjust pot with a 47K R below it. (You might have to make that R lager, maybe 68K or a 100K.)

The red circle shows a series R of 15K, you might need to make that 10K or a little less.

You might need to play with both a little.

         

[Willabe]

I just checked with my Bias Rite, Voltage is just over 500v. Idle current is off the chart.

Leave the power tubes out until you are sure you have -dcv on all the power tube grid pins and have enough -dcv.

You already should know that the idle current is 'off the chart' because the plates are red hot. No need to test them for current and cook them more.   

[Willabe]

Big typo, should be    -DCV    not -acv from the bias supply. Very sorry!   

[pbman1953]

Not to sound ignorant but want to be safe because I can do this but I'm not a full fledged tech.




Tubes are out I have a meter . Placing the DC on 1000?


You want me to test pins 4 &5 ?

[Willabe]

To be safe use your clip on gator meter probes. With amp off clip on pin 5 which is the power tubes control grid and the other clip to ground. (Pin 4 is the screen grid.)

Keep your hands out/off of the chassis, turn on the amp, take reading, write it down, turn off amp, wait a few seconds for B+dcv to drain off, like 30 or so, then repeat. 

Yes DCV on meter and it should be around 50 to 100-dcv.So you don't need to use the 1000 v setting.

[pbman1953]

Ok, meter is at 1000 volts DC setting




Black to ground




pin 4-  567 volts




pin 5 - 0 volts

[Willabe]

Ok, forget about pin 4, that's the screen grid and has nothing to do with the (negative) -dcv on pins 5.

So you have no -dcv on pins 5?

Was the amp working before with no red plating?

[pbman1953]

Yes, I did exactly the steps that I wrote above

[pbman1953]

Possibly jumping pins 1 & 8 changed things?

[pbman1953]

This may be it-   I didn't remove the bias wire from pin 1








QuoteDisconnected the bias voltage from pin 1 of the output tube sockets, shorted pin 1 to pin 8 (ground)

This step will result in running the suppressor grid of a 6CA7 at zero volts rather than -50V. I don't think you'll be able to hear a difference. But... The 6550 has no suppressor grid and pin 1 'may' be connected to the tube base if the base is metal. Connecting pin 1 to ground makes sense. Doing so will allow you to use either tube type.

[Willabe]

Possibly jumping pins 1 & 8 changed things?

It should not have and you have the tubes out so pins 5 should be isolated.

On the other hand if the amp was working before you changed some things then maybe you miswired something of broke off a wire?

Or your not hooking up your meter correctly?

Look at this drawing, measure for -dcv in the yellow and green circles and post -dcv's for them.

[Willabe]

This may be it-   I didn't remove the bias wire from pin 1

QuoteDisconnected the bias voltage from pin 1 of the output tube sockets, shorted pin 1 to pin 8 (ground)

That should be it alright. The -bias is being shorted to ground.

[pbman1953]

So the wire is purely removed and shrink , unused?

[Willabe]

I don't know how it's actually wired in the amp but looking at the schemo, you leave the -bias wire to feed the ends of the 220K grid leak R's where they come together.

Any other wire that goes from that junction over to pin 1 take it out or if you might want to go back to stock, then yes just tape/shrink wrap 1 end.   

[pbman1953]

Now pin 5 has - 50 volts DC

[pbman1953]

Actually -97.2

[Willabe]

Actually -97.2

OK, that's good. You need to check all 4 power tube sockets pin 4. Doh!!!!!  PIN 5

Here's another drawing but I think you have it fixed, get rid of red wire.

[pbman1953]

It's quiet now and I shrink off that wire




The only left over is that the bias is all the way down but the idle current high. In the past I've always got   30-40 miliamps, it's showing in the mid 70's




Pin 4's




1-  518


2-  526


3-  524


4- 516

[Willabe]

Once your sure you have -bias on all power tubes pin 5, then leave the -bias full up, -97.2-dcv so the tubes run the coolest and then use your bias right to set the bias for idol current.

[pbman1953]

pin 5




-106


 

[Willabe]

You have to measure EACH pin 5 on all 4 power tube sockets for -dcv.

[Willabe]

Pin 4's.....

My mistake again. Very sorry. Pin 5.   

[pbman1953]

how do i get the idle current down?

[Willabe]

The only left over is that the bias is all the way down but the idle current high. In the past I've always got   30-40 miliamps, it's showing in the mid 70's

More negative dcv will idol the tubes cooler, ie, -60dcv is more than -40dcv.

Are you sure you have -dcv on all 4 pin 5's?

[pbman1953]

yes, but niw has gone down to -55

[Willabe]

The -bias supply should be stable and should not be going up and down.

Power tubes draw no current from the -bias so they should be stable.

Did you move/bump the -bias adjust pot?

[pbman1953]

No, the pot is solid, still -55, when it's all the way down. If turned up it goes lower but not much, only to 53 or 52

[Willabe]

OK, somethings wrong, you had as much as -106dcv.

You might have a partial short, like a single strand of thin wire or a little blob of solder shorting somewhere.

Look at my reply #27 and with the power tubes out take -dcv readings for the yellow and green circles and post them.

[pbman1953]

Will post in the AM. I'm done for today.

Thanks!!!

[pbman1953]

Here's what measured




                                        On                          off




Yel Circle                          -59                       -108.9




Green Cir.                        -74                         -135.8

[sluckey]

What do you mean by "ON" and "OFF"? Is that the standby switch? Power switch? Tubes in versus tubes out?

Those negative voltages should not change. They should be pretty much rock solid.

BTW, did this amp work properly before you started this mod?

[pbman1953]

on  = power is on


off =    amp in standby mode


I think I have a bad socket . 


This amp did work before he mod , but possibly problems emerging 

[Willabe]

Measure for ACV in the red and blue circle in stand by mode and post. They should be 0 acv's.