Debugging ideas for B-15n - distorted sound

[dmp]

I recently bought a broken Ampeg B-15n to fix up and use for recording. I'm stuck with an unpleasant clipping type distortion on the preamp stage (6SL7), independent of volume. I'm stuck for the moment and was looking for some ideas to debug the problem. Any ideas?
What I've done:
- Installed new filter caps.
- added a trimmer and biased new power tubes
- checked resistances to be within 10%
- Checked and reflowed solder joints that looked suspect at all.
- checked input jacks have good connection to ground

What I know:
- with the preamp tubes removed, the amp sounds great with an active ss preamp plugged into the Ext Amp jack. So, the phase invertor and power stage seem to work fine.
- Bypassed the tone stack by connecting the coupling cap from the first stage to the 270k mix resistor and still heard the distortion. So the problem seems to be at the first tube stage
- Measured voltages and found ~170v at plate and ~2.2v at base for the 6sl7 preamp tube, which is correct according to the schematic
- Swapped out several preamp tubes and heard no change in the distortion
- Changed all coupling caps for mallory 150s (I didn't want to do this, but it seemed likely to be the problem) and still had the same crackly distortion.
- Carefully used a chopstick with the amp on to try to find a bad joint or component and found nothing.
- Added a 1meg resistor from the preamp grid to ground (in parallel w/ the 5.6meg) to see if there was high grid leakage causing the tube to not have a proper bias - but still had the same distortion.

So I'm really stumped. Thanks in advance!
 

[sluckey]

There are two preamps. Do both sound the same?
What are the voltages for all pins of both preamp tubes?

[xm52]

First of all, let me say that these amps are dead quiet when they are working properly. I should add that the onset of distortion is typically at around 11 or 12 o'clock with the tone controls set flat at noon.

It sounds like there is some sort of bonehead problem with the wiring. Check everything carefully against the schematic. Don't assume that the wiring, as is, is correct. Check under the eyelet board with a dental mirror to ensure that there are no shorts between eyelets. All it takes is one errant strand of wire to create a problem.

I would remove the second tube (channel 2) and concentrate on fixing channel 1.

I have encountered a similar problem with these amps. There are several gray shielded wires in the pre-amp that run from the input jacks and tone stages to the board and tubes. The braided shield is grounded at one end to keep noise down. These wires tend to rust and can cause all sorts of problems.  It seems that the ones that Ampeg used were not intended to withstand over 40 years of service. I would change them all and see if that helps.

[dmp]

Both channels behave the same way. I've concentrated on channel one, and pulled the tube for the second channel, as xm52 suggested.
I didn't write down the voltages for all the pins when I checked them. I'll check them again and post the values.
The amp does have the gray shielded cable so I'll look at that as well. There isn't much else that could be causing the problem.
The amp is a '68 and has a through hole pcb. I had assumed it was 1 sided and that there isn't anything on the bottom to cause problems, but I'll try to get a peek under as well. it is soldered down to the tube sockets, but looks like it was lifted at some time.

[xm52]

Those PCB models are a lot harder to get under to inspect. You don't want to have to de-solder the tube sockets or drill out the socket rivets. On the plus side, there is less of a chance of a short under the board with the non eyelet versions.

The "ext amp" jack on the back of the amp is intended as a pre-amp out/power-amp input.  You can run an instrument cable from this jack to the input of another amp and remove the PI and/or power tubes to avoid sound coming out the speaker. This will allow you to isolate the pre-amp and might help with the debugging. You need to keep the volume level on the B-15 down as you can overload the input of your second amp which would result in the type of distortion that you are trying to fix.

Check your power supply grounds in the pre-amp carefully. They have an odd grounding scheme on these amps. The power supply cap returns (-) are isolated from the chassis and connected to a buss that is grounded at a single point at the input jacks. Although this seems odd, it works very well in this amp. Having your ground depending on the tightness of the nut on a jack makes me cringe. I've tried different ground topologies but have never been able to improve upon it. The SVT uses the same scheme.

[dmp]

Great suggestions xm52!
I replaced the gray cables and saw no change. But... I ran the preamp from the 'ext amp' to a different power amp. I tried this and the preamp alone sounds great (I pulled the power tubes for this test). So I'm thinking it is a problem with the grounding, causing the power tubes to affect the preamp tubes. I'll post a picture, but the ground from the power supply cap returns are not going to the input jacks. They connect near the 22k power supply resistors on the pcb.

[dmp]

Here is a picture of my b-15n.  The two black wires on the right are the power supply grounds (from the caps) to the pcb ground. The input jacks have a wire connecting them to the pcb ground over on the left side.
So the original ground scheme should have the power supply grounds go over to the input jacks?
Where does the pcb ground connect to the chassis?   

[jjasilli]

May I digress?  What are those 2 brown rectangular things atop the small vertical boards -- multi-section caps?  I've got a sh**load of them from an old accordion amp.

[Willabe]

May I digress?  What are those 2 brown rectangular things atop the small vertical boards -- multi-section caps?  

I think those are the tone circuit's. I remember Ken Fisher talking about them.

If you look at some Ampeg schematics it shows them with a dotted line around them.


         Brad        

[xm52]

Yes, those are tone stage circuits composed of networks of resistors and ceramic capacitors encased. They called them PEC for printed electronic circuit. You used to see them in radio's and TV's. They saved on assembly time since they were often mounted in close proximity to the pots.

Google "centralab PEC modules". You'll find some info with part numbers and schematics.

[xm52]

The two black wires on the right are the power supply grounds (from the caps) to the pcb ground. The input jacks have a wire connecting them to the pcb ground over on the left side.
So the original ground scheme should have the power supply grounds go over to the input jacks?
Where does the pcb ground connect to the chassis?  

It is difficult to see the details in your image. From your description, it sounds like the connections are correct.

The PCB buss ground is on the left side, at the pre-amp end, and connects to the chassis at the input jack. The copper circuit board buss runs to the other side where the power amp is components are located. The (-) side of the power supply caps connect to the buss at the power amp end. The idea is, the high current components from the power supply are connected at one end (but not grounded to the chassis -- if you have an FP type cap can check that it has an insulating washer installed that isolates the case from the chassis). The ground buss extends to the PI, then the pre-amp ground, then the buss is grounded at the input jack.

Check that the tabs from the tube sockets are in contact with the copper pads of the circuit board. Sometimes you find them lifted off the surface and there is only a blob of solder making the connection between the tab and the copper pad. A solder bridge makes a lousy connection.

There should be an isolation washer on the main speaker jack. The speaker return is best connected at point where the PI ground connects to the ground buss. There should also be an isolation washer on the ext speaker jack. Otherwise you will get a ground loop and hum.

[jjasilli]

May I digress?  What are those 2 brown rectangular things atop the small vertical boards -- multi-section caps?  I've got a sh**load of them from an old accordion amp.

Thanks guys!

[dmp]

The main speaker jack was not isolated - but after doing that, I confirmed that the only connection from the chassis to the pcb is the wire from the input jack to the pcb on the left. Made no difference in the problem.
The tabs on from the tube sockets to the pcb don't look perfect, but I've checked everyone and I don't think any have a problem.
And I just confirmed again that operated separately with the 'ext amp' jack, the PI/power amp works great, and the preamp works great. It is only when using both that I get a sizzly distortion at even low volumes. One funny thing is that the preamp works with the standby switch off. I'm not sure why that is??? Could the PT center tap be leaking to ground?
Here's better pics:

[dmp]

And preamp tube voltages (ch 1):
pin2: 181v
pin3: 2.17v
pin5: 162v
pin6: 2.6v
pin1: 0v
pin4: 0v
pin7,8: 6.2vAC

With power on, standby off (coming from rest):
pin2: 63v
pin3: .87v
pin5: 55v
pin6: 1.04v

[xm52]

And I just confirmed again that operated separately with the 'ext amp' jack, the PI/power amp works great, and the preamp works great. It is only when using both that I get a sizzly distortion at even low volumes. One funny thing is that the preamp works with the standby switch off. I'm not sure why that is??? Could the PT center tap be leaking to ground?

The standby switch should be connected to the HV secondary center tap. When standby is on the switch is open, when off the switch closed. It is not normal that you should hear the preamp operating when the standby is on so there is a problem there. One possibility is that the switch is not opening fully when you are turning the standby on. This would mean that the switch is faulty and needs to be replaced. It is also dangerous to leave it as is. Another possibility is that someone modified your amp by adding a resistor across the switch somewhere  in the circuit. This would be intended to put the amp in a semi standby state, allowing a lower B+ to be maintained. I would try a jumper, or temporarily solder a wire across the standby switch and see if it helps resolve the standby issue.

You first set of voltages when the amp is in operating mode are good. I would check all the other voltages that are indicated on the schematic. There is a copy of the schematic for your model (with the ultra hi and lo switches) in the archive. The second set of readings are a result of the "problem" with the standby circuit. There shouldn't be any voltage with the standby on.

Since the pre and power amps operate without any issues on their own, there is the possibility that there is a partial short or an oscillation that is causing a problem when they are connected. On a hunch you can try jumping in a small cap, maybe 50-100pf, in parallel with the 10K feedback resistor (R40). It is mounted on the "ext speaker" jack sleeve. In the event that there is a partial short, I would first carefully check out the ext amp jack. Try inserting a 1/4" jack with nothing connected to it and see.

[dmp]

The standby switch is odd - I disconnected the center tap (rec/yel) and verified that the switch is good (inf resistance off, 0 ohms on). The resistance from the center tap to ground is 165k ohms (probably due to the resistors in the bias circuit). Note that the amp does not work until the standby is switched on unless the power tubes are removed - then the preamp tubes have a low voltage allowing them to operate. I need to think through it more, but I think a standby switch on the ct would allow some voltage to be rectified, since the bias circuit is providing a path to ground.
I think the problem I am having may be an oscillation issue. I tried replacing the gray shielded with from the mix resistors to the PI input and the amp squealed loudly. I put the gray wire back in and it went back to the way it was behaving before. Possibly the blu/brn wires from the output primary got flipped at some time. I may try disconnecting the feedback and see if the distortion changes.

[dmp]

There's definitely a problem with oscillations. I switched the blue/Brn wires to the OT, and the amp was very noisy. Then I switched them back, and it was still noisy. So perhaps I disturbed something that was a problem connection. I'd guess something between the output and the PI. I measured the voltages (while it is oscillating / noisy) like this and the voltages on the power tubes were OK (486v plates, 455v screens), but the voltages on the PI tube were off. pin2 was 275v, when the schematic says 235v, and pin 5 was 156v, when the schematic says 225v.

[xm52]

Looking at your pics, it looks like an offset voltage is being applied to the center tap of the hum balance pot. Normally this helps reduce noise. You might want to try removing this and connecting it to ground until you sort out the problem. Earlier NF models had it like this. It simplifies things a little.

Those PI voltages are interesting. Even more reason to try the parallel cap on the feedback resistor.

[dmp]

Those PI voltages are interesting. Even more reason to try the parallel cap on the feedback resistor.

This will increase negative feedback of high frequencies? I'm not sure what this test would help identify? I think there must be a bad connection or bad resistor somewhere that I am missing because now it won't stop squealing. The lower half of the PI is conducting heavily like the bias is wrong.

[xm52]

It is difficult to know what type of oscillation you have. The cap can tame high frequency oscillations and it is found in a number of Ampeg designs. The test could help provide some insight as to the nature of the problem.

It does look like the problem is due to a bad solder joint, shielded wire, component, or lead dress. Even a bad tube can cause similar problems. I recall having an HF oscillation and swapping out three different 6SL7's before the problem was resolved. They were all bad tubes. Probably all from the same batch.

I noticed in you pic that there is a green and yellow twisted pair near the PI. I assume that these are the OPT secondaries. These wires cross the gray shielded wire twice. Normally the speaker wires, which carry a relatively high current, are under the circuit board and run along the chassis allowing some space between them and the sensitive signals in the gray wire. This could be a problem. If you can't easily move the transformer wires, it might help if you could replace the gray wire with a shorter length that avoids the speaker wires.

[dmp]

First off, the gray wire / OT leads were causing the noisy oscillations. I changed how these were routed and the amp is quiet at idle again. It still has the original problem of distorted sound at all volumes. The voltages around the PI are now normal (pin2:255v, pin5:249v,pin3:2.7v,pin6:2.49v).
I'm going to try to describe the distortion a little bit more. It sounds like a buzzing fret or a rattle in a speaker cab. Just an annoying clipping type of distortion on all notes.
And it is definitely happening in the amp. I just checked again that the preamp sounds great alone with the power tubes removed and the power amp sounds great alone when run from a separate preamp (tested using the 'ext in' jack).

On a hunch you can try jumping in a small cap, maybe 50-100pf, in parallel with the 10K feedback resistor (R40)

Didn't make any difference. Changed the tone slightly of the amp but didn't affect the distortion.

offset voltage is being applied to the center tap of the hum balance pot

I removed this and the hum at idle increased, but the distortion problem wasn't affected

[dmp]

Ahh haa! With the 'Ext amp' connected to another power amp, the distortion goes away!
I had the B-15n running as normal playing through it connected to the speaker cab, and as before, the notes sounded distorted. But when I connected a cable from the 'ext in' to a different power head, the B15n immediately stops the distortion (and sounds great!). Presumably it is providing a pull down resistance? Could it be anything else? The resistance measured from pin1 of the PI (also the ext amp jack) without a connection to another amp is  233kohms. I would think this would be enough?

[sluckey]

What is the voltage reading on V3 pin 1? Should be dead zero.

Try this... Tack solder a 1M resistor across the EXT. IN jack. Does it get happy?

[xm52]

I love those Ahh haa! moments.

Just to be sure, is this the schematic (http://www.el34world.com/charts/Schematics/files/ampeg/Ampeg_B15N.pdf) for your amp? 

So loading the pre-amp out with another amp resolves the problem. Impedance matching between the tone stages, the PI, and whatever is connected via the ext amp jack is important here. I would start by measuring the resistance of the mixing resistors, R18 and R22, to ensure that they are both 270K. Make sure that the bass and treble pots are 1M. Then make sure that the other components in the tone stage are up to spec.

With the topology of this model, there isn't a low impedance driver like a cathode follower just before the PI. When you connect another amp via the ext amp jack, the load reduces the signal going into the power amp. It is noticeable when you try it. Not the best design but they saved on a tube. The input impedance of the PI 6SL7 is high. The tone stage output impedance is higher than it should be. It is supposed to be low enough to drive the PI.

[xm52]

Here is a voltage reference chart for the 68 B-15 compliments of Joe Piazza.

[dmp]

That is the schematic for this amp. I tried putting in a 1meg pulldown resistor like slucky suggested, but when I tried the amp out again the loud oscillations were back. I think there must be an intermittent short somewhere that rears it's ugly head now and again. I wish I could inspect the transformer wires under the board. I'm worried it is a problem in one of the transformers, because I have been over everything else several times. But I need to find this problem before figuring out issue at the PI input. Thanks for all the help - and I'll report back once I make some more progress.

[PRR]

What is the voltage reading on V3 pin 1? Should be dead zero.

[dmp]

What is the voltage reading on V3 pin 1? Should be dead zero.

It is 0.0v when I measure it, but I'm not sure I trust that measurement. I had thought a special DMM (very high impedance) was necessary for this so the act of measuring wouldn't pull down the grid voltage.
The latest test I did was to ground V3 pin 1 (P.I. grid). This stops the oscillations.  I'm going to check the tone stack pcbs next.

[dmp]

So after a long, tedious process of elimination, I found why the feedback/oscillation/distortion was happening and have finally fixed this amp. 
I started by grounding the input to the P.I. and worked back to the 270k mix resistors, and found that with them in the circuit (grounded on the preamp side) the amp had the problem. Inspecting things with a mirror, the wires running under the board to the output transformer were RIGHT UNDER the through holes for the 270k resistors. With the mix resistors lifted off and soldered together with the wire running to the P.I., the amp is dead quiet and plays perfectly. Another lesson to beware vintage amps that have been messed with! I knew the board had been lifted at one time and had not been put together with good wire routing...
Thanks for all the help, especially xm52!

[xm52]

Great news! Congratulations for sticking with it and sorting out the problem.

These problems can be discouraging so it is sometimes good to remind yourself to take the time to enjoy the process.

B-15's have a fantastic tone. Enjoy your amp!

For those interested, Ampeg has recently reissued the B-15N. They combined features from two of the most popular early models and added a choke in the power supply. The schematic is contained in the owner's manual (http://www.ampeg.com/pdf/Heritage_B-15_OM.PDF). The speaker cabinet offered with the Heritage B-15 is what they call a double baffle design. There has been a few different cabinets offered over the years. I like the double baffle one the best. Here is an interesting site with some details of the cabinet and amp's designer, Jess Oliver (http://www.vintage-blue.us/).  

[dmp]

It was definitely rewarding when I started reamping bass recordings through it last night!
I had definitely gotten tired of troubleshooting it though.
The link to the reissue is really interesting. Although they are only making 50 total apparently.
I have a SVT reissue that sounds amazing through the big 8x10" ('fridge) cab.  They did a great job with that one, IMO, except that it is too noisy at idle. Unbeatable for rock, but for a quiet country song it isn't quiet enough. Or else mine has a problem.

[xm52]

The stock pre-amp tubes that come with the reissues have been known to cause noise. I had an SVT-VR that had noise issues. Installing a set of good NOS tubes made a big difference.

There is a ground loop problem in some of the early re-issue (around 2006) models. The details are here http://music-electronics-forum.com/t15670/ .

[HotBluePlates]

May I digress?  What are those 2 brown rectangular things atop the small vertical boards -- multi-section caps?  I've got a sh**load of them from an old accordion amp.

Thought this might be of historical interest...

The attached are from a 1960 catalog of Sprague capacitors.  There are components printed on rectangular ceramic called "Bulplates".  The circuit package could be selected by circuit function, and then picked out by the specific values needed in each position.

The one for the Ampeg B-15N seems to be TC-13 (Tone Compensation), at $1.25 for four caps and two resistors ready to solder straight to your tone controls...

[pompeiisneaks]

Oh interesting, that's the one I'd been looking at that they use in the VT-22/V4, I've been looking over that circuit a lot, and that's definitely the Baxandall Circuit... Very cool.

Not for sale anymore I'm sure.

~Phil

[kagliostro]

That is VERY Cool HotBluePlates 

Thanks for sharing

---

Some time ago I've find, and put on my archive, this







Source
http://ctgelectronics.weebly.com/custom-amplifier---fcn-51-b.html
http://ctgelectronics.weebly.com/custom-amplifier---fcbn-51b.html

Franco

[sluckey]

Oh interesting, that's the one I'd been looking at that they use in the VT-22/V4, I've been looking over that circuit a lot, and that's definitely the Baxandall Circuit... Very cool.

Not for sale anymore I'm sure.

You can still get some for the old Ampegs...

     https://www.fliptops.net/catalog/p-100797/pec-board-tone-module-for-vintage-ampegs-used

[pompeiisneaks]

Ouch 29 bucks, I can build it myself for a lot less :) I saw it here: http://www.vintageampeg.com/ampeg-tone-stack-modification/

But I don't know that he has them for sale anymore, but again, my curiosity is the 'next project' type thing.  I've been working on the layout from the schematic.  (I think I'll do a 2 tube instead of 4 and not use the 7027A's due to them costing a bit more) but anyway, derailing this thread too much :)

~Phil

[sluckey]

All the B15-Ns use 6L6s. Earlier models were cathode bias, later models were fixed bias. All octal tubes.

BTW, I would not be concerned with derailing a thread that died over 5 years ago.

Here's that tone stack layout I used on my AO-63 conversion. Everything mounts on the pots. You can see it on my website. May give you some ideas...

[PRR]

> 7027A's due to them costing a bit more

7027 was always a 6L6GC but with an extra G2 pin so if you put 6L6GC in a fully wired 7027 socket it would blow-up the power supply.

Clear-out the socket, wire as for 6L6, don't use socket lugs as parts supports, and run any modern "6L6 type".

[pompeiisneaks]

Sluckey, oh excellent thanks!

PRR, yeah I'm doing a scratch build so I don't need to mod it, I'll just plan for 6L6's from the start.

~Phil