Vox Buckingham V112 rebuild (was Vox Buckingham questions)

[Stevem]

I just zoomed in on your photo and I replied wrong to you.

Your red circles are on the collector terminal of the transistors, not the Base.

[seveneves]

I just zoomed in on your photo and I replied wrong to you.

Your red circles are on the collector terminal of the transistors, not the Base.

I appreciate the follow-up but I do think I have it right. I even metered the socket terminals to verify. Here's a shot with a new socket placed above, oriented the same:



And a close-up with a flash. You can see the terminal initials:



The 'C'ollector is the middle terminal (long strip with the screws going through it).

[Stevem]

Yes, you are correct, I could not zoom in enough in your original photo I looked at.

Here’s another detail.
Try not to use the mica type insulators under the TO3s.
They insulate great from voltage but also heat transfer, this is why many of the windows in wood stoves is actually mica, not glass or Pyrex .

Try to locate the blue or red type plastic ones.

I bought a ton of them when this type first came out decades ago, and I forget from who, but most supply houses stock them.

The pain in the Butt is that the shipping is 5 times the cost of the parts!

[Stevem]

Here’s some possible good info and a question for RG.

Sunn 100 watts rms SS amps like the Concert bass and the lead version use a PI driver transformer and 4 2n3055 outputs.
This is a good size transformer, but it has a separate winding for each transistor.

These Sunn amps can be had dirt cheap when blown up so I wonder if this driver / splitter trans can be used in a 4 output Vox amp that needs one?

Could this same transformer and only two of the secondary windings be used in a two output Vox amp?

RG, here is the schematic for the output section of the Sunn amp I am speaking of.

Interestingly these Sunn amp have the outputs lined up length wise along the rear of the amp and the driver / splitter as seen in my photo is on one end of the amp.

This makes for really long wires going to the far side outputs and a real propensity for oscillation issues at high volume levels.
It’s no wonder so many of these have blown up !

[seveneves]

Yes, you are correct, I could not zoom in enough in your original photo I looked at.

No worries, I had to triple check myself...

Try not to use the mica type insulators under the TO3s.
They insulate great from voltage but also heat transfer, this is why many of the windows in wood stoves is actually mica, not glass or Pyrex .

Try to locate the blue or red type plastic ones.

Re: the insulators, I am using mica with thermal joint compound (i.e., "Thermalcote").

Are you talking about using those red/amber colored silicone insulators instead?

My understanding (and based on some quick googling of what others think) is that mica with a *thin* layer of grease is not terribly less effective than silicone pads, if at all. Have you observed differently? I'm not sure if your experiments with mica included the thermal grease.

[R.G.]

I believe very high frequency oscillation may be why and the reason for this is the crazy long wires going to the output transistors and also there unequal length.

Maybe. I've scoped quite a few of these, and never see it, but then I wouldn't have because they might only oscillate just before they die.

I spent a lot of time researching this output stage. They make a big deal out of both the low impedance bias from the resistor string and the low impedance of the secondary resistance of the secondary base windings. The windings themselves have some leakage inductance, but surprisingly little because the secondaries are bifilar wound. The wires have parasitic inductance too, and that might profit from a base stopper. The stopper is worth trying experimentally, though. The output devices in the Buckingham are germanium, and have funnier characteristics than the silicon ones in the Guardsman and Beatle. I've not heard of reliability issues in the 30w power amps.

In the Beatle and Guardsman, the problems are much bigger. These things are right at the edge of what you get with 1960s vintage 2N3055's safe operating area and second breakdown.

Sunn 100 watts rms SS amps like the Concert bass and the lead version use a PI driver transformer and 4 2n3055 outputs.
This is a good size transformer, but it has a separate winding for each transistor.
These Sunn amps can be had dirt cheap when blown up so I wonder if this driver / splitter trans can be used in a 4 output Vox amp that needs one?

Could this same transformer and only two of the secondary windings be used in a two output Vox amp?

Hmmm. Thanks for the schematic. I'm not all that familiar with Sunn.
It's possible that the driver could be used for a Guardsman or Beatle. I'll dig in a bit more. I think I'd just parallel two of the secondary windings for each pair of output transistors if the primary-secondary voltage ratio is right.

If I had one of the Sunn transformers to measure I might be able to get closer. The schemo looks like Sunn went for higher voltage across the speakers and higher impedance speakers than Vox did for 100Wrms. The bridge configuration gives you double the voltage and half the current capability on the speaker outputs.

I dug pretty deeply into the Beatle driver transformer. Here's an article on analysis of the driver trans and a blow-by-blow on how to make a works-alike from a small filament transformer.
http://www.geofex.com/Article_Folders/T ... former.pdf

[LOUDthud]

Those Sunn driver transformers tend to develop an internal short between the Primary and one of the secondaries. I think I have one that I can send you R.G. The earliest Sunn Concert amps (150W at 4 Ohms) had a bridge output stage with a floating power supply and one side of the bridge output grounded. Quite a clever get-around for the low breakdown Voltage of the 2N3055's at the time. The later version of the amp used the same driver transformer but a more conventional parallel output configuration (like the Vox Beatle) with a split rail power supply for 200W at 2 Ohms.

Mercury Magnetics sells a replacement for the Sunn driver transformer.

[seveneves]

Update on the preamp recapping.

I had some spare radial electrolytics in the proper values so I jumpered them in. I have axials on order.

Upon fire-up, the first thing that jumps out at me is the hiss... yes, I read this in R.G.'s Repair Supplement so it is no surprise... there's quite a bit of it, to the point where it's hard to tolerate, personally.

So, I read the section on Hiss:

"[Hiss] is by far worst for parts in the first/input stages. Thomas made this issue much worse by the design of their tremolo and reverb stages and to some extent the mixer/limiter stage."

"The solution [for hissy input transistors] is to replace the input transistors with new, low noise types."

"The best choice is to replace all the transistors with low noise types like the 2N5088 or 2SC1815 and replace all the resistors in [the reverb and tremolo] stages with metal film types for the tremolo, reverb recovery input and mixer-limiter."

With that, I set out to find out what parts I'll specifically need to do the replacement, once. The wiring in these seems pretty fragile. The less I have to flex the wiring, the better.

Of course all electrolytics throughout the amp will be replaced.

Each preamp will have their transistors replaced. That's pretty straightforward. For the other sections, I am a little hazy. I took a shot at identifying the transistors and resistors that should be replaced to help minimize hiss. I'm just not sure if I'm missing anything else in these sections.

In the tremolo section, is there any other component that should be replaced? All I could surmise from R.G.'s publication is Q205 and R221:

https://i.imgur.com/lpY1Sqx.png

Similarly in the mixer/limiter section, but there's quite a bit I've identified. Those in red as for sure, green is unknown if there's any benefit in replacement. Did I go overboard here? Overlook anything?

https://i.imgur.com/zpsn49k.png

In the reverb, I understand most of the noise is in recovery. Just not sure about those resistors in green:

https://i.imgur.com/oBG8D3C.png

Thoughts?

[R.G.]

Upon fire-up, the first thing that jumps out at me is the hiss... yes, I read this in R.G.'s Repair Supplement so it is no surprise... there's quite a bit of it, to the point where it's hard to tolerate, personally.

It's like the old bromide about peeling onions - you peel off a layer, then you cry, and then you go peel some more.

With that, I set out to find out what parts I'll specifically need to do the replacement, once. The wiring in these seems pretty fragile. The less I have to flex the wiring, the better.

I cannot fathom what Thomas was thinking with the solid core hookup wire, other than they simply had no idea of the hard life guitar amplifiers live.

Each preamp will have their transistors replaced. That's pretty straightforward. For the other sections, I am a little hazy. I took a shot at identifying the transistors and resistors that should be replaced to help minimize hiss. I'm just not sure if I'm missing anything else in these sections.

You can get by with replacing only the input transistors of the two-transistor input pair. Those are the ones that get their base-emitters zapped.

In the tremolo section, is there any other component that should be replaced? All I could surmise from R.G.'s publication is Q205 and R221:
Similarly in the mixer/limiter section, but there's quite a bit I've identified. Those in red as for sure, green is unknown if there's any benefit in replacement. Did I go overboard here? Overlook anything?

Although the tremo and mixer/limiter do work at low signal levels, their inputs don't go off-board to the outside world. The actual syndrome for excess hiss is that inputs that go outside the enclosure get zapped by static electricity and reverse-break the base emitter junction. This is not fatal, but each breakover increases the transistor's noise a little. Over time, the hiss builds up. This is the reasoning behind the suggestion to replace the input transistors first. Other internal transistors may go hissy, but that would be just a random failure, not part of a typical aging process.

If it were me, I would rig the amp up on the work bench so I could do a listening test after replacing parts. This would let you hear which parts did the best for reducing hiss. Notice carefully: that preamp section has live AC wires on it any time it's plugged in, so doing any replace-and-test work means you need to always, always, always unplug the line cord any time you are touching anything except the knobs. Seriously, you could die.

Then, I would replace all three input transistors for the three channels, and the reverb recovery input transistor. This single step could be huge in cutting hiss. It has been for me.

As a side note, you will likely never get the hiss down to un-noticeble levels. The internal "architecture" runs the signal levels inside the amp down at about 25-50mV, not up at line-level 1Vrms like most modern practice. So any hiss added by anything up through the input to the mixer will get glommed onto what is passed to the power amp. It's a poor design for getting low noise. But you can fix quite a bit of the excess noise from the inputs. You'll have to make your peace with getting it down to a level you can live with.

In the reverb, I understand most of the noise is in recovery. Just not sure about those resistors in green:

My advice would be to not arbitrarily replace any of the resistors unless replacing transistors didn't get the hiss to be tolerable. It is possible that a resistor going bad could introduce a lot of hiss, but that isn't the general case. I did have one resistor making the hiss unbearable in my first Beatle. It had simply gone bad and hissed. It happened to be in the second reverb mix transistor's circuit, so it got amplified and passed on. That one was a b----h to find. I had to resort to "shorting" signals to ground with a 47uF capacitor at the output of section by section to even find the offending section. But once I located and replaced the resistor, magic happened. Advice: don't go looking for where magic has to happen unless you just have to. Shotgunning resistors is sometimes a viable last resort, but wait until nothing else works. The reasoning behind that is that replacing lots of resistors involves lots of flexing the PCB up and down, and can get you into the scenario where fixing one wire breaks another wire.

[LOUDthud]

Can you protect those input transistors with a diode ?

I always thought it was the "death cap" that made the input transistors go hissy. Do the JFETs go hissy or do they just die ?

[R.G.]

Can you protect those input transistors with a diode ?

Yes. A silicon diode hooked up anode to emitter, cathode to base, that is, always reverse biased in normal operation, is a near-complete protection. The diode adds only its nano amps of leakage current to the base current, and its reverse biased capacitance to the base-emitter junction operation. This is nearly always benign at audio frequencies. The diode clamps the B-E junction to one diode drop, far below the 5-8V breakdown of most NPN bipolars. Works for PNPs too, just reverse the protection diode.

I always thought it was the "death cap" that made the input transistors go hissy. Do the JFETs go hissy or do they just die ?

Well, it could be one cause. The death cap can certainly cause the signal ground to be wandering up and down at line frequency, more than the BVbeo of the input transistors, and has the potential (yuk, yuk, yuk) to cause transients when flipped. The received wisdom from the guy who taught me this one was that un-terminated guitar cords that got zapped on the tip contact were the real disasters. But anything that can make the base-emitter go backwards by more than maybe 10V could do it.

It's a slow death. Any one zap adds an undetectable amount of hiss, but over the years, it adds up. I suspect that the idea that old transistors just go hissy may be grounded in this process.

[seveneves]

It's been a while since my last post. Had a question as I start to go through the preamp and replace the channel input transistors.

I have about twenty (20) new 2N5088s that I am checking gain/hfe, mostly out of curiosity and using this method (but adjusted for NPNs). They are all having gain in the 420-440 range.

I checked one of the original NPNs in the normal channel (Q101) and it has a gain of about half that, 220.

Just curious what happens (if anything) when subbing NPNs originally having lower gain with modern NPNs having higher gain?

[Stevem]

I don’t think I would be concerned about it unless the amp goes into oscillation .

One thing I have noticed at times with a higher signal level going into a tone stack of some amps is that it can cut back on how much the controls can effect the frequency range they are working with.

Here’s what BC184 test out like which is another sub for the 5088.

You should check out one of these meters, the things it can test and report back to you about is insane for the price!
Just be careful if you do get one and always remember to discharge electro caps before testing them.

[R.G.]

What happens (if anything) when subbing NPNs originally having lower gain with modern NPNs having higher gain?

You've hit on another of those questions that are reasonable and insightful to ask that seem like there should be a simple, straightforward answer - but the answer is complicated.

The shortest answer is that in the Vox preamps, nothing much happens. That is because the Vox preamp circuits mostly have both local feedback for the NPNs (usually emitter resistors) and feedback around a pair of transistors that keep things steady.

The effect of higher or lower gain from a transistor was the subject of a number of lectures in my circuits courses in university. Engineers in the tube era could count on tubes having a well defined gain. Not so with transistors - the early transistors had wildly varying gains. So tube engineers had a number of product disasters with switching to transistors. A good deal of circuit design work in the 1960s and 1970s was oriented towards using local and global feedback to make a given circuit perform depending on resistor values, not on the transistor gains. This discipline turned out to be very successful, as local feedback both lowers their effective gain and extends their frequency response to limit parasitic oscillations.

Stevem's comment applies. In every circuit, if you turn up the gain enough, it will oscillate eventually. But in this case, doubling the current gain does not get there. There are on the order of 30 Vox preamp boards built with 2N5088s, and all work fine. I have also subbed in the 5088 in repairing a number of original Vox preamps, and had no issues.

@stevem: Cool meter. I gotta get one of those!

[seveneves]

R.G., I suspected the higher gain devices would be fine in this application based on your other writings about it. I just wanted to know how/why. I really enjoy the explanation and the science behind the technology. I suspect I am not the only one and am continuing to learn things, which ultimately is what it's all about IMO.

Thanks again!