ampeg b15 build, now trouble shooting

[nateflanigan]

I've wanted to do a bass amp build for a while now and I just picked up a solid state B15 for fairly cheap.  The plan is of course to build a tube B15, and I think it would be cool to do a tweed bassman channel instead of the two identical channels.  What I'm wondering about though is the price of Ampeg replacement transformers vs say bassman transformers, the difference is ridiculous.  The internet claims that the ampeg ot primaries are 6.6k.  Any reason I shouldn't use a hammond or edcor with appropriate specs, or any high wattage 6l6 OT?

Thanks friends!

[ernie_jr]

I have built one ampeg  (b26) with an old stancor transformer and I currently building another (B15) with the same type o/t. Sound is great. Hammonds will work fine. They may sound a bit different, but they are great trannies. Get the 60 watt, 4300 ohm primary and play around with the ultra linear taps, you may like that sound even better.
Good luck,
Ernie

[nateflanigan]

Thanks, I was thinking it'd be a nice amp to play around with UL (having never done that).  Why 4.3k instead of 6.6k on the primaries?  I understand that either would work but since it's very unlikely I'll try out both to see which I prefer what are the theoretical differences?

[ernie_jr]

4300 ohm will give you a bit more power. Sunn uses 4300 phm on their UL o/t with great results
Ernie

[PRR]

One of Ampeg's designers favored high voltage with high impedances. 60W in 6.6K load requires nearly 600 Volts! Such an amp can be modestly more efficient (more output from the same tubes). IMHO, fat tubes with less-insane B+ voltage and lower load to suck about as much power is a wiser path.

Since it isn't really a tube Ampeg, and there's little magic in the hot Ampeg output section, just steal the whole power system (PT, tubes, driver, OT) from one of the Usual Suspects (Fender Bassman/Twin, Marshall) and put Ampeg preamps in front.

[nateflanigan]

Hi PRR, as always nice to hear from you.
On the B15n schematic I don't see any mention of 600v.  It shows 450v on the plates and the wiring diagram for mojo's replacement PT shows secondaries at 375-0-375.  I'm sure there's something in there I'm not understanding, regardless I like your idea. 
http://www.el34world.com/charts/Schematics/files/ampeg/Ampeg_B15N.pdf

I don't know if I could squeeze a twin output section into the existing chassis, maybe if I used a SS recto and cut out one of the channels. 

[Willabe]

One of Ampeg's designers favored high voltage with high impedances. 60W in 6.6K load requires nearly 600 Volts!

I think he was referring to design choice used on -- some -- of the amps at Ampeg over all, in reference to the question about the OT primary impedance. The higher Z load needs a higher dcv to get X output power. (Read it again, might be clearer to you now?)  

Take a look at schemo's of SVT, V4, V22, ect. 500vdc to 650vdc.


                     Brad      

[ernie_jr]

Later models that used 7027a's (B18, B 25) had 500+ volts on the plates.
Ernie

[PRR]

That plan with 450V and 6,600 load will make about 45W and suck about 200mA.

Same but with 4,300 load may exceed 60W output and needs 300mA.

The 6,600 load plan is less stress on the tubes; at this point, Ampeg still was not beating tubes to death. The 4300 plan is hard work, but thousands of Fenders and copycats run these values with acceptable reliability.

[VMS]

B15 schematic with two different channels is in the owner's manual of the Heritage B-15:

http://www.ampeg.com/products/heritage/b-15/index.html

[nateflanigan]

Thanks guys.  Those numbers are really helpful prr, I'll make sure my PT can supply the 300ma if I go that way. 

That heritage b15 looks great, Chip helped me sort out a brown vibroverb channel on an AB763 build a while back, what I like best about it is that there are two different voices but they are not at odds with each other.  The 64 and 66 channels seem like they'd be different but complimentary as well.

[ernie_jr]

Sunn ran  kt 88's with a 4300 ohm load with P/T's supplying 200ma. Unless you plan on running the amp full out all the time, you probably wont need  a 300ma P/T. 200ma should be plenty and much more cost effective.
I am currently building an ampeg with one channel B15 and one channel B25.  P/T and bias is left over from a Fender 5f6a build (P/T can supply 200ma)  B15's used 6l6's, I will be using 807's. We should keep in touch as we build our toys and see what results we get.
Ernie

[pullshocks]

Nate, would you be interested in building this project in an actual Ampeg B-18 chassis?  I have one in decent shape that I am not going to use. 

[xm52]

The early model B-15 amps were rated at 25W RMS. This was a cathode biased amp. The later revisions moved to fixed bias and the output power was rated at 30W.

I have some specs for the transformers that came from a service manual. For the fixed bias models, the PT108 power transformer was 750V CT, 150 ma with 117VAC on the primary, the OT-214 output transformer, had a 6.5K primary with a 290R DCR, the secondary had 8 and 16 ohm taps with a DCR of 1 and 1.4 ohms.

The load lines reveal that the B15 was a conservative design. The high B+ voltage allowed them to position the operating points to achieve this.

[nateflanigan]

Insert Quote
Nate, would you be interested in building this project in an actual Ampeg B-18 chassis?  I have one in decent shape that I am not going to use.

PM sent.

Ernie and '52, excellent info, thank you.

[nateflanigan]

Since we last spoke I've settled on building the "heritage" B-15.  I've built the board and am mostly just waiting on the transformers from edcor, I don't want to get to far into wiring before I have the iron installed.  I do have a few questions/ideas though.

-I've attached a screen shot of the preamp section, it's pretty obvious which pot is the volume but I can't tell what's bass and what's treble in the tone stack.
-I bought pullshocks b-18 chassis, it has three holes for input jacks instead of four.  I'm thinking I could do one input for each channel, then in the middle hole install a dpdt switch to toggle between the two resistor values.  Does that make sense?

Thanks!

[tdvt]

I only know this because the Duncan Tone Stack calculator has this style (James) of tone controls & I've been fooling with the calculator for a small amp build I'm planning. No experience otherwise, but the pot towards the second stage grid is the treble.

[nateflanigan]

Cool, thanks! 

[ernie_jr]

The pot on the left is the bass, the pot on the right (the one that feeds the PI) is the treble
Ernie

[sluckey]

A good clue for this type question is to look at the size of the caps connected to each pot. Big caps usually means bass. Little caps usually means treble.

[pullshocks]

Thanks for the update Nate.  Can't wait to see that chassis come back to life.  It had been sitting in my basement for over 30 years.

[nateflanigan]

A good clue for this type question is to look at the size of the caps connected to each pot. Big caps usually means bass. Little caps usually means treble.

It's funny I didn't think about that but since you've pointed it out it actually seems kind of obvious.

It had been sitting in my basement for over 30 years.

No kidding?!  That's cool, I usually sort of half ass it when it comes to documenting things and sharing with the board but I'll make sure I post some decent pics.

[sluckey]

The James circuit is a passive tone stack that is based on the Baxandall circuit. The Bax is an active tone circuit that is placed in a negative feedback circuit for a gain stage. I've never seen a true Bax in a guitar amp, but I'm sure there probably is one somewhere. Probably better used for acoustic guitar though. Bax is usually found in hifi equipment.

James and Baxandall are many times used interchangeably in the guitar amp community. Probably because of the circuit similarities.

[xm52]

It is interesting to note that James published the article "Simple Tone Control Circuit, Bass and Treble, Cut and Lift", E. J. James, Wireless World, February 1949. Baxandall followed up with "Negative Feedback Tone Control", P. J. Baxandall, Wireless World, October 1952. It seems that Baxandall had better public relations people and got his name cited more often.

By the way in the power supply, they use a Heyboer HY027707, 9Hy@120ma choke. It isn't indicated on the schematic.

Good luck with your build.

[nateflanigan]

By the way in the power supply, they use a Heyboer HY027707, 9Hy@120ma choke. It isn't indicated on the schematic.

Thanks for the tip, I have the standard "big" fender-ish choke.  Next time I'm ordering that type of stuff I'll have to remember that.  It'll be interesting to check out the difference.

[nateflanigan]

Question about the OT secondaries, on the schematic there's a 250r 10w resistor from the 8 ohm tap to ground.  I've never seen that before, what's it there for? 

[darryl]

Question about the OT secondaries, on the schematic there's a 250r 10w resistor from the 8 ohm tap to ground.  I've never seen that before, what's it there for? 

It is to place some load on the output transformer if the amplifier is accidentally operated with no speaker connected. If there is no load at all, dangerously high voltages may be generated. These can destroy the output transformer, and/or cause flashovers on output valve bases or sockets.

[tubeswell]

Question about the OT secondaries, on the schematic there's a 250r 10w resistor from the 8 ohm tap to ground.  I've never seen that before, what's it there for? 

It is to place some load on the output transformer if the amplifier is accidentally operated with no speaker connected. If there is no load at all, dangerously high voltages may be generated. These can destroy the output transformer, and/or cause flashovers on output valve bases or sockets.

It can also help snub HF oscillation from the NFb loop.

[nateflanigan]

Thanks guys.  I have my secondaries going to an impedance selector switch, where as the schematic has individual jacks for the different taps would hanging that power resistor and the NFB off the tip terminal of of one output jack cause any problems? (as opposed to having them on separate jacks like in the schematic)

[darryl]

To maintain consistency with the original circuit, the NFB should come off the 4 ohm transformer tap, while the 250 ohm resistor should be from the 8 ohm tap to ground. Both these connections are made between the output transformer and the impedance selector, not after the impedance selector.

[sluckey]

would hanging that power resistor and the NFB off the tip terminal of of one output jack cause any problems?

The amount of NFB will change depending on which secondary tap is selected. I would hard wire the NFB to the 4 ohm tap just like the original. (Just connect to the proper selector switch terminal.)

The power resistor is used for failsafe protection. I'd leave the power resistor hardwired to the 8 ohm tap just like the original. Having it hardwired will provide protection against open speaker load or selector switch failure. (Just connect to the proper selector switch terminal.)

Wiring the amp as close to the original as you can will result in an amp that sounds as close to the original as you can get. 'course you may be looking for a different sound.

[nateflanigan]

The amount of NFB will change depending on which secondary tap is selected

I thought this might be the case, but really had no idea what the theory behind those parts was, thanks for sorting it out guys.

[nateflanigan]

I fired up the b15 tonight, and I got problems...

-no sound, none what so ever.

-Generally the voltages are on the high side, not sound the alarm high, but on the PI the plates are over 400v.  

-There's a switch to switch from cathode to fixed bias, I'm keeping it in cathode biased mode for now as something is wrong with the little bias supply board I built.  At the moment the bias supply is disconnected from everything.  I've got appropriate-ish voltages on the power tube plates but no voltage at all on the cathodes.  

-If it seems I'm being a bit vague with the voltages it's because the rectifier seems to have blown before I had a chance to note the voltages.  I thought I'd check to see if the pre-amps we're working by using the external amp jack and running it into another amplifier.  As soon as I kicked on the stand-by the rec flashed an I blew the fuse, I disconnected the ext amp set up, and another flash and blown fuse.

Weird stuff...

Where shall we start?

[Willabe]

Are you using a lamp limiter to check for power supply shorts at start up? This is a must have/must use amp building/reparing tool.

It can save your PT on start up if you have a short.

http://home.comcast.net/~seluckey/amps/misc/Amp_Scrapbook.pdf


                            Brad                                      

[nateflanigan]

I was aware of the light bulb test existing but have never seen it explained so succinctly, leave it sluckey. 

I have another 5ar4 at my practice space I can pick up over the weekend, I'll post back with the results from the current limiter and a voltage chart. 

Thanks

[Willabe]

I was aware of the light bulb test existing but have never seen it explained so succinctly, leave it sluckey.

Yep.



                  Brad     

[nateflanigan]

Alrighty, here's some data that will hopefully help you help me.  Firstly the amp passed the light bulb test, I went slowly and did it first with no tubes, then just the rectifier, then added the rest.  I got the initial flash followed by a dim glow. 

Voltages are...

B+ at the standby switch before kicking it on 402v, then it jumps to 512v once the switch is on.

V1
Pin 2: 180v
Pin 3: 2v
Pin 5: 249v
Pin 6: 3.5v

V2
Pin 2: 183v
Pin 3: 2.3v
Pin 5: 185v
Pin 6: 2v

V3
Pin 2: 445v
Pin 3: 0v
Pin 5: 445v
Pin 6: 0v

V4/V5
Pin 1: 0v
Pin 3: 515v
Pin 4: 515v

It's obvious that those are problematic, and of course I'll start quadruple checking component values and wiring, but any one have any direction to point me in?

I've attached a screen shot of the schematic from the heritage B-15 manual, I can't seem to extract just that page from the document.

Thanks

[Willabe]

V3 somethings wrong with the cathode. V4/5 same thing.

High volatage on the plates of those tubes with no voltage on their cathodes means the tubes not drawing any current.

Check the cathode R's/C's and their connections, right pin on tubes, cold solder joint, open R.

V4/5 check the bias switch wiring. Measure from pin 1 to ground, it should read what ever you used for the cathode R, 250R?  


                         Brad    

[PRR]

Are pins 1 and 4 missing on your tubes?

But my bet is that V3 isn't really getting heater voltage.

[LooseChange]

I know I'm late with this one but just finished one of these. Did only one channel in a Champ size chassis. Sounds great.

How can I help?

[nateflanigan]

Hi guys,
I made some progress.  

-I found a cold joint in my bias supply board, that's working now so I'm running it in fixed bias mode.  On the output tubes I still have 515v on the plates and zero volts on the cathodes (should be normal now right?) and of course the -44v.  I've attached a drawing of my implementation of the bias switch, seemed pretty straight forward but maybe I missed something.

-It seems one of my 6sl7's is bad.  That would be the one that was in the PI position.  Swapping it out with one of the others I'm getting 238v on the plates and about 3v on the cathodes.  And similarly using the bad one in a preamp position screws up those voltages.  Unfortunately I don't have any spares.

However, still no sound.  Totally quiet, I'd guess the output is shorting to ground, but I can't find it.  How do you guys look for that sort of thing?  All of the OT secondary taps show continuity to ground and each other using the "beep" function on my MM, reading resistance they all have a few ohms between a given tap and ground.

I'm tempted to fuss with the external jack and try sending signals in and out, but I really don't want to blow another rectifier, I have no clue why that happened.

Are pins 1 and 4 missing on your tubes?

But my bet is that V3 isn't really getting heater voltage.

Nope, the 6sl7 are all NOS and have all the pins, the out put tubes are JJ's.

Measure from pin 1 to ground, it should read what ever you used for the cathode R, 250R?

That checks out.

I know I'm late with this one but just finished one of these. Did only one channel in a Champ size chassis. Sounds great.

Oh, I think we're just getting started.  Which schematic did you use?  

[nateflanigan]

Wait a sec, The output tubes have no pin1 (or 6).  On my other builds I used Doug's AB763 layout for those connections which jumpers pins 1 and 8,  Looking at the 6L6 data sheet looks like the cathodes are indeed connected to my unused pin 8. 

I don't want to jump to conclusions, so I'll wait on connecting pins 1 and 8.  There's probably something to this I'm unaware of.

[sluckey]

Pin 8 is the cathode on a 6L6. The tube won't work if it's not connected to anything.

[nateflanigan]

Ok, more, err, um, progress.  I don't know why but the schematic in the ampeg manual shows the cathode connections on pin 1.  I didn't question this.  So I've jummpered pins 1 and 8.  I do have output now, extremely loud wild oscillation.  If I pull the PI it cuts the noise.  

I remember reading something about swapping the OT primaries and oscillation.

[Willabe]

I'm sorry Nate I saw the schemo and it showed pin 1 for the cathode, I should know better by now and have that memorised. Amp companies/builders use pin 1/6 on 6L6, 6V6 and others in this family, as solder posts. Look at most any Fender schemo and it's there. I'm not sure why when they don't use pin 1 as a solder post they tie it to pin 8, maybe just to ground it for some reason?

Try disconnecting the feed back loop, just 1 wire. If the oscillation stops, then swap the OT primary wires and reconnect the feed back loop.


                 Brad    

[sluckey]

I'm not sure why when they don't use pin 1 as a solder post they tie it to pin 8, maybe just to ground it for some reason?

That's an EL34 thing. Pin 8 is still cathode. Pin 1 is suppresor grid. Whenever you see a 6L6 amp with pin 1 and 8 tied together, it's usually so you can also plug in an EL34.

A long time ago, when 6V6s were put in a metal shell, the shell was connected to pin 1 and it was customary to connect pin 1 to ground for shielding.

[nateflanigan]

Flipping the OT primaries OR flipping the PI plate leads OR flipping the power tube grid leads all achieves the same purpose - it changes the phase of the signal that is used for the NFB.  When you have them flipped backwards it actually performs a positive feedback (which you noticed), and very quickly starts to howl or oscillate.

I found this in an old post, I'll start with disconnecting the nfb tomorrow if that's the cause of the oscillation it'll be nice to have a few spots to swap the phase.  

I should know better by now and have that memorised.

Eh, me too, we'll probably both remember it now though.

[sluckey]

I just looked at your schematic again. The bias circuit is wired such that the pot can adjust the bias voltage down to zero. That's not good! I'm thinking the original is not really wired per the drawing and I suspect a drawing error. That would make at least two errors on that schematic. Might be worthwhile to look for other errors, maybe compare to an original drawing.

Here's a pic showing how I would wire the bias circuit. A simple change...

[nateflanigan]

Great, thanks for pointing that out! 

I just went through the "heritage" schematic comparing it to the "Joe Piazza"schematics in the library.  There's plenty of slightly different component values but the circuits appear to be the same. 

However, the "heritage" schematic also shows the wrong pins for the heater connections on the 6sl7's.  I noticed this while initially wiring the heaters.   Go figure.

[PRR]

> the 6sl7 are all NOS and have all the pins

My point was: report all 3 voltages on a triode. Grid voltage is important. Even if it is supposed to be zero. The odd numbers on V3 could have been a missing grid connection.

> when 6V6s were put in a metal shell

AFAIK, metal 6V6 is almost unknown, but metal 6L6 used to be THE bee's-knees. And the shell was supposed to be tied to chassis.