Issue with 1971 Sunn Sceptre PI

[TheKT88KilledJFK]

Using Luckey's schematic located here: http://sluckeyamps.com/sunn/sunn_sceptre_1971.pdf,  I have scoped the signal on my sceptre to the outputs of the triode of the phase inverter.  It just stops on both sides of the waveform.  The signal input into the PI on the pentode side is about 70 mV while the output of the pentode on pin 6 and 2 is about half that.  After the signal hits the triode side, it just dies.  The tube is tested good on my BK 667 tube tester. 

Additionally, "node" B is setting at about 438 VDC, about 40-50v higher than most schematics indicate it should be.  Is it possible that the capacitor stack for the voltage divider is compromised? 

[sluckey]

The signal input into the PI on the pentode side is about 70 mV...

That's way low. Should be more like 1V_rms. A common cause for this low voltage would be the tremolo optoisolator. Put a jumper across points X and Y to see if the opto is killing your signal.

...while the output of the pentode on pin 6 and 2 is about half that.

That pentode should boost a 1V input up to about 50V or more. You may have issues with the tube circuit also. But for sure, you gotta get the input up to normal.

Is it possible that the capacitor stack for the voltage divider is compromised?

I don't understand. What cap stack? What voltage divider?

[PRR]

Welcome.

> 70 mV while the output of the pentode on pin 6 and 2 is about half that.

As Sluckey says, the pentode should have GAIN. Actually close to 300 alone, though less with NFB enabled. Certainly not "half".

DC voltages please. Something may be shorted or open. High volt on "B" suggests the triode side is not sucking as expected. Or could just be normal tolerances. DC volts at P G2 G1 K and P G K might show a problem.

[TheKT88KilledJFK]

I have already removed the reverb circuit. I had already removed and bypassed the entire reverb and tremolo circuit as the opto-isolator was no doubt dead.  I am still having the same issue except now a the 50mv signal going into pin 8 of the PI has jumped to .5 V.  Progress is progress I suppose but the output of the PI is still around 30 mV. 

The capacitors I was referencing before were just the voltage filter capacitors in the metal can that sits on top of the amp.  Specifically, the 20uf 525v capacitors..

As for as the phase inverter DC voltages:
P_triode = 444 VDC
G_triode = 406 VDC
C_triode = 0 VDC
P_pentode = 406 VDC
G1_pentode = 363 VDC
G2_pentode = 0 VDC
C_pentode = 0 VDC

[sluckey]

Your voltages indicate that neither the pentode nor triode are conducting. The tube could be dead. Are the filaments for this tube lighting up in the amp? Do you have a 6AN8 plugged in?

[TheKT88KilledJFK]

Your voltages indicate that neither the pentode nor triode are conducting. The tube could be dead. Are the filaments for this tube lighting up in the amp? Do you have a 6AN8 plugged in?

I have 4 tested good 6an8s on hand that I have been swapping.  The voltage is dropping entirely across the tube as you say so it has to be some sort of open.  I do have the tubes plugged in. I somehow got that far.

Really appreciate the help. 

[sluckey]

The voltage is dropping entirely across the tube as you say so it has to be some sort of open.

There is no voltage drop because there is no current flowing through the tubes. Your zero volt readings on the cathodes indicate the tubes are not conducting. Are the filaments lit? What AC voltage do you measure across pins 4 and 5?

[TheKT88KilledJFK]

The voltage is dropping entirely across the tube as you say so it has to be some sort of open.

There is no voltage drop because there is no current flowing through the tubes. Your zero volt readings on the cathodes indicate the tubes are not conducting. Are the filaments lit? What AC voltage do you measure across pins 4 and 5?

The heater pins are getting 6.4 VAC combined. 

Is it possible this is some sort of socked issue? I cannot visibly see lit filaments on the 6an8. 

[eleventeen]

Could definitely be something with the socket. Cathodes should not be zero unless the cathodes are directly connected to ground. If they *are* the tube is just not conducting; which is pretty much the same thing as full B+ appearing on the plates. On *both* 6AN8 sections?


You'll find it, you're getting close.

[sluckey]

It's very easy to see that my 6AN8 is lit up. It's "in yo face" bright. You say the tube is good (I don't happen to trust tube testers) and you measure correct voltage at the socket pins 4 and 5. That only leaves the socket. Just for kicks, put your tube back on the tester. You should easily see it lit up.

[TheKT88KilledJFK]

It's very easy to see that my 6AN8 is lit up. It's "in yo face" bright. You say the tube is good (I don't happen to trust tube testers) and you measure correct voltage at the socket pins 4 and 5. That only leaves the socket. Just for kicks, put your tube back on the tester. You should easily see it lit up.

I put it in the tube tester and it got bright.  I put it in the amp and it doesn't light up at all.  I measured the heater pins and it's still 6.4vac.  It has to be the socket. 

[TheKT88KilledJFK]

Sorry for the double post.  I re-soldered the heater pins and now the tube actually lights up. 

Still get 0 VDC across pin 9.  The amp is also not very loud at all.  The Power tubes are biased properly. 

[sluckey]

What are the other voltages on the 6AN8?

[TheKT88KilledJFK]

Phase inverter DC voltages:
P_triode = 341 VDC
G_triode = 106 VDC
C_triode = 119 VDC
P_pentode = 106 VDC
G2_pentode = 36 VDC
G1_pentode = 0 VDC
C_pentode = 1.01 VDC

Thanks again!

edit: the pentode's cathode reading showed as 0 due to a setting on my multimeter.  I changed the range and it came in as 1.01 VDC.  I suppose everything looks ok with the PI now. 

[sluckey]

Those voltages look much better. Is the amp still not loud enough?

BTW, I measured the voltages on my Sceptre this morning and put the voltages on my schematic and updated the pdf on my web site. Take a look. You may have to refresh your browser to see the updated file.

[TheKT88KilledJFK]

Those voltages look much better. Is the amp still not loud enough?

BTW, I measured the voltages on my Sceptre this morning and put the voltages on my schematic and updated the pdf on my web site. Take a look. You may have to refresh your browser to see the updated file.

Thanks again Steven.  The amp is only outputting around a watt of power.  The power tubes are certainly dissipating much more.  No matter the bias, the output is still really low.  I suppose the next logical step would be to directly measure the power off the power tubes so if that's okay then I'll assume it's an output transformer issue which I was really hoping was not a problem.  Oh well, Edcor has a 4500k UL 60w Tx for $75 so I'll probably purchase on of those if I can confirm it's actually the issue.

Also, waveform amplitude also reduces between V1a and V1b by about half.  It reduces from 30v p-p to ~12v p-p which makes no sense.   When I do the gain calculation, it should be 11 for each side, for a total of 121 at max volume.  I know the stages load each other but I didn't account for that. 

Here is my v1 voltages (VDC):
P2 = 297
G2 = 0
C2 = 1.2
P1 = 270
G1 = 0
C1 = 1.3

Pins 4&5 were 6.4 VAC

This strangely looks like the same issue as the PI. :(  Since the load resistor is 100k and plate resistance is around 80k, shouldn't I see less voltage across the tube?  It would have to be not conduction to have a high enough impedance to have the majority of the voltage across it.


EDIT: lol it was another loose heater pin.  This is what happens when you buy an amp off craigslist for $100 for parts. 

[sluckey]

What kind of signal are you injecting into the amp? What level and where? I suggest a 500Hz 200mV pp signal injected into one of the input jacks. That is a realistic input level. That input should give you at least 2v pp at V3 pin 8.

Since you are having tube socket issues I suggest cleaning all sockets with Deoxit or similar. Use a jeweler's screwdriver to retension the pins.

[TheKT88KilledJFK]

What kind of signal are you injecting into the amp? What level and where? I suggest a 500Hz 200mV pp signal injected into one of the input jacks. That is a realistic input level. That input should give you at least 2v pp at V3 pin 8.

Since you are having tube socket issues I suggest cleaning all sockets with Deoxit or similar. Use a jeweler's screwdriver to retension the pins.

I was doing 2vpp because I use actives sometimes but I have lowered it to 500mV.  I cleaned the sockets now.  Still having issues with the heaters not firing up at 6.3 VAC on the 12ax7.  It's bizarre.

EDIT:  The heater voltage on my tester is about twice the voltage in the amp so that's probably why it's not so bright? 

Anyways, at pin 8 of V3 it's ~.8 VAC p-p with a 500mVAC input.  It just seems like the second portion of V1 has too low of gain no matter what tube I swap in there.  Even when I unload the output from the tone stack it's still less than the input side. 

Also, I removed the .22 mF cap off the cathode of pin 8 because it had about .24 ohms of esr. 


Now I have removed the tone stack and am just feeding pin 8 of v3 from pin 6 of v1 coupled with a cap and 270k resistor in series. 

[sluckey]

EDIT:  The heater voltage on my tester is about twice the voltage in the amp so that's probably why it's not so bright?

Don't you have a knob on the tester to set the filament voltage? It can be difficult to see a 12AX7 filaments glowing. May need to dim the room lights. It for sure will not be as obvious as that 6AN8.

Also, I removed the .22 mF cap off the cathode of pin 8 because it had about .24 ohms of esr.

Don't just remove it. Replace it. That cap will increase the gain of V1B. If you don't have a low voltage .22µF, use anything you have, up to about 50µF. Bigger values will give a bigger bass response. Too big may even sound too boomy for lot's of folks.

Let's resolve the gain issues. Insert a 500mV signal into the amp. Now use your scope to measure the signal on pin 2 and pin 1. Gain will be equal to the voltage on pin 1 divided by the voltage on pin 2. Repeat for the second triode but measure voltage levels on pin 6 and 7. Gain will be voltage on pin 6 divided by voltage on pin 6. Furthermore, the gain for each triode should be the same, even though they are dealing with different levels. So, give me some real world numbers for gain. While you're at it, do this for the 6AN8 pentode also.

I'm still curious if you have properly bypassed the reverb and tremolo circuits. Explain what you did. And how did you determine the trem optocoupler was bad? Can you post some hi rez pics?

[TheKT88KilledJFK]

I'm still curious if you have properly bypassed the reverb and tremolo circuits. Explain what you did. And how did you determine the trem optocoupler was bad? Can you post some hi rez pics?

I bypassed the tremolo and reverb circuit by feeding pin 6 of the 12ax7 into the start of the tone stack, at the 56k resistor and 270pF node. 

I will get all the voltage measurements shortly and replace that cap. 

[sluckey]

I bypassed the tremolo and reverb circuit by feeding pin 6 of the 12ax7 into the start of the tone stack, at the 56k resistor and 270pF node.

OK. You should also disconnect the 470K at the reverb input and the 750pF at the reverb output. This will remove any loading that the reverb circuit may be putting on the signal chain. May as well pull V2.

Just curious, why don't you want reverb and tremolo?

[TheKT88KilledJFK]

Don't just remove it. Replace it. That cap will increase the gain of V1B. If you don't have a low voltage .22µF, use anything you have, up to about 50µF. Bigger values will give a bigger bass response. Too big may even sound too boomy for lot's of folks.

Let's resolve the gain issues. Insert a 500mV signal into the amp. Now use your scope to measure the signal on pin 2 and pin 1. Gain will be equal to the voltage on pin 1 divided by the voltage on pin 2. Repeat for the second triode but measure voltage levels on pin 6 and 7. Gain will be voltage on pin 6 divided by voltage on pin 6. Furthermore, the gain for each triode should be the same, even though they are dealing with different levels. So, give me some real world numbers for gain. While you're at it, do this for the 6AN8 pentode also.

For the 12ax7:
pin 2 = 600mV p-p
pin 1 = 18 V p-p

So that's a gain around 30.

For the the other side of the 12ax7:

pin 7 = 18v p-p
pin 6 = 7.5v p-p

So that gain is at .4 which is a problem.  Even if I detach it from any loading it only climbs slightly. 

I also added another .22 uf cap on the cathode.

[sluckey]

You do have a speaker or a dummy load connected to this amp, right?

pin 7 = 18v p-p

You actually measured 18V at pin 7, right? Not just assuming that since you had 18v on the previous plate, there must be 18v on the grid? If so, then there's a problem with V1B. Not much to go wrong. The first thing to do is put another 12AX7 in the socket. Then measure the voltages again for pins 6, 7, and 8. Also measure the voltage at the top of the 100K plate resistor (node D).

Now power down and measure the resistance from pin 8 to chassis ground (1.5K). Also measure the resistance from pin 6 to node D (100K). Make these checks directly from the tube pins, not just across the actual resistors.

What voltage and resistance readings do you get?

I still want to see some pics?

EDIT... BTW, the mid boost switch must be on for that .22µF cap to be in the circuit.

[TheKT88KilledJFK]

Swapped another 12ax7 in, same voltages for the most part.  I have a lot of 12ax7s floating around and all exhibit the same issue.  It totally makes sense to blame it, thats for sure.  BUT the .01 mF cap had shorted itself.  I am going to replace it and see what happens. 

[sluckey]

I bypassed the tremolo and reverb circuit by feeding pin 6 of the 12ax7 into the start of the tone stack

Ding Dong! I know what the problem is. I have the triodes for V1 labeled wrong. Pins 6, 7, 8 are the input triode (first triode). Pins 1, 2, 3 drive the tone stack (second triode). I've fixed the pin labels so get the revised schematic.

Since you have pin 6 tied directly to the tone stack, you have totally bypassed the second triode. You need to feed pin 1 into the tone stack. Use the updated schematic to correct any other changes you may have made based on my wrong pin labels. After you untangle this you should have plenty of gain.

Be sure to connect a speaker or dummy load. Sorry for the confusion. Let me know how it turns out.

[TheKT88KilledJFK]

I bypassed the tremolo and reverb circuit by feeding pin 6 of the 12ax7 into the start of the tone stack

Ding Dong! I know what the problem is. I have the triodes for V1 labeled wrong. Pins 6, 7, 8 are the input triode (first triode). Pins 1, 2, 3 drive the tone stack (second triode). I've fixed the pin labels so get the revised schematic.

Since you have pin 6 tied directly to the tone stack, you have totally bypassed the second triode. You need to feed pin 1 into the tone stack. Use the updated schematic to correct any other changes you may have made based on my wrong pin labels. After you untangle this you should have plenty of gain.

Be sure to connect a speaker or dummy load. Sorry for the confusion. Let me know how it turns out.

I realized that before I joined this forum and was merely communicating based off your interpretation since it really doesn't matter which side does what.  I didn't know if my amp had just been wired differently in the past or not, so I didn't see a point in "correcting" you. 

I have a dummy load I always use. 

edit:

At this point it's taking me longer to trouble shoot the amp than just to rebuild the entire preamp that takes 5 minutes to do.  I think I am just going to order all new parts although I would like to know what was actually wrong.

Two output transformer feedback resistors are off by about 20-25%.  The 680 ohm tests at around 900 ohm and the 1500 ohm tests around 1200 ohms.  They're original resistors so I thought that was interesting.   

[sluckey]

At this point it's taking me longer to trouble shoot the amp than just to rebuild the entire preamp that takes 5 minutes to do.  I think I am just going to order all new parts although I would like to know what was actually wrong.

I think you are only 5 minutes away from fixing the preamp issue. You know the first triode is working fine. The second triode (pins 1, 2, 3) is the problem. Measure the voltage on pins 1, 2, and 3.

Also measure the voltage at the top of the 100K plate resistor (node D).

Now power down and measure the resistance from pin 8 to chassis ground (1.5K). Also measure the resistance from pin 6 to node D (100K). Make these checks directly from the tube pins, not just across the actual resistors.

What voltage and resistance readings do you get? We should be able to determine what the problem is from these readings. Oh, and clean the tube socket. If you don't have deoxit or another electronics spray cleaner, use alcohol. Just let it dry before powering up.

[TheKT88KilledJFK]

At this point it's taking me longer to trouble shoot the amp than just to rebuild the entire preamp that takes 5 minutes to do.  I think I am just going to order all new parts although I would like to know what was actually wrong.

I think you are only 5 minutes away from fixing the preamp issue. You know the first triode is working fine. The second triode (pins 1, 2, 3) is the problem. Measure the voltage on pins 1, 2, and 3.

Also measure the voltage at the top of the 100K plate resistor (node D).

Now power down and measure the resistance from pin 8 to chassis ground (1.5K). Also measure the resistance from pin 6 to node D (100K). Make these checks directly from the tube pins, not just across the actual resistors.

What voltage and resistance readings do you get? We should be able to determine what the problem is from these readings. Oh, and clean the tube socket. If you don't have deoxit or another electronics spray cleaner, use alcohol. Just let it dry before powering up.

First off, I am following the updated scematic.  Everything is wired in accordance with that, except the reverb and tremolo sections are removed.  Additionally, the first 525v 20uf cap on the b rail stands alone versus combined like in your schematic.  I will probably do the same mod but for now here are my b rail measurements:

point A: 466 VDC
point B: 419 VDC
point C: 392 VDC
point D: 365 VDC

These are the voltages I am getting at v1:

pin	AC	DC
1	.24	225
2	2.8	0
3	0	2.7
4	3.13	0
5	3.12	0
6	2.8	228
7	.07	0
8	0	1.02
9	0	0

I also noticed that the output transformer outputs had been grounded in all states by inappropriate wiring.  I corrected that and the amp is loud and very clean which was anticipated because of the power measured being dissipated across the power tubes.   Still having an issue with the 2nd triode of the preamp tube no matter what.  Thanks a lot for the help so far.  I have no idea why this issue persists.  I think all of my preamp tubes have the same issue that my tester cannot detect. 

[sluckey]

Still having an issue with the 2nd triode of the preamp tube no matter what.

You listed the voltage for pin 2 as 145. Pin 2 should read zero. Did you possibly mean 145mV?

[TheKT88KilledJFK]

Still having an issue with the 2nd triode of the preamp tube no matter what.

You listed the voltage for pin 2 as 145. Pin 2 should read zero. Did you possibly mean 145mV?

Yes it reads 0 now. 

[Willabe]

I also noticed that the output transformer outputs had been grounded in all states by inappropriate wiring.  I corrected that and the amp is loud and very clean which was anticipated because of the power measured being dissipated across the power tubes.   

Can you fill us in?

[TheKT88KilledJFK]

I also noticed that the output transformer outputs had been grounded in all states by inappropriate wiring.  I corrected that and the amp is loud and very clean which was anticipated because of the power measured being dissipated across the power tubes.   

Can you fill us in?

With regards to which detail?  After fixing the phase inverter issue, I believe the power stage was receiving about 35 Vsin on each side from the PI, so the power tubes were getting driven fine and the bias was set right, so there was no reason for the amp to be as quiet as it was regardless of the issues with V1.   

To summarize the issue with the output, the output transformer's 8 ohm coil was actually grounded by the output jack's switching action.  The last owner had switched the ground and the signal side of the ohm jack he had replaced. 

I received this amp with a blown 6an8, bad fuse, bad opto-isolator, and various out of spec components.  However, the choke on the ps is good, the Power transformer is brand new, the output transformer tests good.  The power tubes and rectifier tube are brand new as well as all the pots.  Some of the caps were replaced with ATOMS, some were replaced with off the shelf parts, some just weren't replaced (eww).  Everything in the amp tests good now.  I took every part off and tested it yesterday.  There were some offenders but nothing that would cause the same issue I am seeing today with V1.   

[Willabe]

I also noticed that the output transformer outputs had been grounded in all states by inappropriate wiring. 

With regards to which detail?

This part, thanks;
     

To summarize the issue with the output, the output transformer's 8 ohm coil was actually grounded by the output jack's switching action.

[sluckey]

Sunn used an unusual switching scheme (Switchcraft 13A) on the speaker jacks. The circuit can be confusing to a casual observer.

If you plug into the speaker jack you will be connected to the 8Ω tap thru the ext. spkr jack normally closed switch.

If you plug into the ext. speaker jack you will be connected to the 4Ω tap thru the normally open switch. The speaker jack will also be connected to the 4Ω tap.

So, you can plug in an 8Ω cab to the speaker jack. Or you can plug in a 4Ω cab to the ext. speaker jack. Or you can plug in two 8Ω cabs, one to the speaker jack, the other to the ext. speaker jack. The jack labels do not make this obvious to the user or inexperienced tech.

I never can remember the switching operation without looking at the schematic so I rewired my speaker jacks using simple Switchcraft jacks. One jack is 8Ω. The other is 4Ω. No switching. No confusion.

[TheKT88KilledJFK]

Sunn used an unusual switching scheme (Switchcraft 13A) on the speaker jacks. The circuit can be confusing to a casual observer.

If you plug into the speaker jack you will be connected to the 8Ω tap thru the ext. spkr jack normally closed switch.

If you plug into the ext. speaker jack you will be connected to the 4Ω tap thru the normally open switch. The speaker jack will also be connected to the 4Ω tap.

So, you can plug in an 8Ω cab to the speaker jack. Or you can plug in a 4Ω cab to the ext. speaker jack. Or you can plug in two 8Ω cabs, one to the speaker jack, the other to the ext. speaker jack. The jack labels do not make this obvious to the user or inexperienced tech.

I never can remember the switching operation without looking at the schematic so I rewired my speaker jacks using simple Switchcraft jacks. One jack is 8Ω. The other is 4Ω. No switching. No confusion.

Definitely see how it could be confusing.  It was confusing enough for me to not notice it at first.

The ground wire to pin 9 of the 12ax7 was oxidized or burnt.  I am not sure which.  Once I replaced that, the triode began functioning normally.  Thanks again for the help!  It looks like all is well now. 

[sluckey]

Good news! Now that you have the basic amp working, you may want to consider fixing the reverb and tremolo circuits. Tackle them one at a time and it will go pretty fast. I've got everything worked out as far as replacing the trem optocouple with a Vactrol VTL5C1. Cheap and easy to find. I'm assuming you didn't rip out the circuit board.

I'd love to see some high rez pics of the inside of the amp.

[TheKT88KilledJFK]

Good news! Now that you have the basic amp working, you may want to consider fixing the reverb and tremolo circuits. Tackle them one at a time and it will go pretty fast. I've got everything worked out as far as replacing the trem optocouple with a Vactrol VTL5C1. Cheap and easy to find. I'm assuming you didn't rip out the circuit board.

I'd love to see some high rez pics of the inside of the amp.

I can get you some pics within the hour but it's not pretty!  I am going to redo everything heh. 

I did remove everything related to the other circuits, but I took a lot of pics since I plan on putting the board and reverb back in.  I still need to find a reverb tank as this amp didn't come with one.  Shouldn't be too hard. 

[sluckey]

The common 4AB3C1B tank used in most Fender amps works fine in the Sceptre. Hoffman sells that tank. The 9AB3C1B tank also works and gives a much fuller sounding reverb. And if you really want a Gibbs unit, you can find one in old Hammond organs that use the AO-39 or AO-44 reverb amps. DON'T GET A TANK USED WITH AO-43 AMPS! eBay is a good source.

Still waiting for pics.