Peavey Valvex 6x2 tube mixer - impedance matching...?

[PRR]

 

[okgb]

That's not the best / worst example of peavey stuff [ of which there are more than a few ]
and wouldn't stop me , but
I'd have to call that pepto bismol , which of course suggests other things

[ or do we blame it on the 80's ....................who ? how did they make design choices
and if they were worried about poor product perception , well ................. ]

 

[JohnRoberts]

That's not the best / worst example of peavey stuff [ of which there are more than a few ]
and wouldn't stop me , but
I'd have to call that pepto bismol , which of course suggests other things

[ or do we blame it on the 80's ....................who ? how did they make design choices
and if they were worried about poor product perception , well ................. ]

Unfortunately I was there during this so I know more than I would like to about the subject. It is easy to be critical, not as easy to do better, but even I found those knob colors unattractive.  My suspicion is they were channeling miami vice with those pastels (coral, etc).

Those knobs were tooled up for the console program(s). Peavey made a large sound reinforcement console (Mark VIII) and I did a large split recording console (AMR Production series). Both consoles shared common control components, etc.  We both used the same knob tooling but pushed different color plastic through the molds. Thankfully I got to use less extreme color schemes. As i recall there were many many design drawings and this  knob design was not thrown together but the result of many hours of work. I may have been in or near the room where decisions were made but I didn't really get a vote. These knobs were pretty cool with soft rubber outside, but I repeat I do not care for those colors.  The Mark VIII console was saddled with that same exact color scheme. :-(

There was a pretty significant internal conflict over the Peavey logo... that one on the tube mixer was referred to internally as the "lightning bolt" logo and designed by Hartley on his high school notebook cover (surprise surprise).  The alternate logo that I and others lobbied to use was called the block logo

We developed this modern improved logo initially to differentiate the recording and install products from the MI products. There is even a story about the 5150 getting the block logo. Eddie had artistic control over the look of the 5150 and he refused to use the lighting bolt logo after seeing the block logo. So even MI guys had better taste.  8)

The corporate politics were such that the same people who picked those colors, preferred the lightning bolt logo, and some working on the product side were actually threatened by #2 (RIP) to never even show #1 the block logo. This was enough to intimidate most.  At the time I left I had put the block logo on a number of MI products I was responsible for, but after I left the block logo has pretty much retreated to only remain on the fixed install product side. FWIW I left because I found myself arguing with the guy who's name was on all the buildings. Most employees decline to argue with the guy who signs their paychecks.

====
OK here's a related story about taste... It is common when doing artistic designs to show the boss man or boss lady at least three renderings, with a range of different looks. For a relatively routine business card design for the whole company the young designer from the art department tasked with the design, did his best effort, then added two throw away ugly designs, that he ASSumed nobody would ever pick, so his favorite design would get used...  Guess what, she picked the ugliest card design and we were saddled with that for years. An object lesson for designers everywhere, never submit a design you don't want to see used. I somehow managed to not have my business cards with me, and people thought I was snubbing them when I said sorry I don't have a card, but I really hated that card.       

JR

PS: I don't consider myself some super artistic judge of aesthetics and design, but I know what I like and don't, and most customers I spoke with seemed sympathetic to my preferences (like even Eddie).

 

[okgb]

And to be fair , there are certainly other companies that seem to make curious decisions
with potentially so much at stake for some design details .
I see it often from installers , where somebody claims to ask for the opinions of professional's and people
who will work there , and then of course you're stuck with quirky at best but possibly inadequate or
even unusable installs .
[ the early T series gtrs are another  , what were they thinking ? Look like no fun to play ]
I guess Quality has to shine through before it can do the talking .

 

[tonedude]

Thanx a for the great stories. I had the first version of the 5150 and that was my first high gain "insanely" loud amp, that loved to be played LOUD, and so I did...        ;D

For the record; main PCB

 

[tonedude]

Main PCB, front:

 

[JohnRoberts]

I notice some of the PCB attachment holes are broken out, you may want to be careful about that when reassembling so the daughter boards do not move around loosely inside.

For more than you may want to know about how it was manufactured, it looks like most of the smaller boards were originally part of a single larger board and was originally wave soldered as one big board with the several ribbon cables between subassemblies inserted before the wave solder step.  After wave soldering, the boards were broken apart, rotated, and attached. This is a lot less labor than performing point to point wiring or even using multi-pin cables and connectors between the separate PCB at final assembly.  I see from his initials that the PC board was designed by Keith (JKV).

JR

 

[ruffrecords]

I am still trying to work out how this was done with 8 triodes. ASSuming we have an input pot followed by a mute switch followed by a pan pot and FX send controls, all passive, that's three buses required. Cheapest we can do is to passive mix the lot right away - no buffers. With line levels in and 10dB lost in the pan and and about 16dB in passive mixing we end up at about -26dBu on the L&R buses and a bit less on the FX - not too bad but we probably had to use 470K or 1Meg bus resistor to maintain the 1Meg input impedance requirement. Actually its a bit worse than that because we have the two aux returns to add to the L&R buses but still basically OK. Looking at the tube PCB layout, it looks like all the stages are pretty much the same except for two which have bigger output caps - probably the main outs. Since the layouts all look the same they are probably straightforward common cathode stages. Main output stages probably with lower plate resistors and higher idle current and bigger output caps. Probably a CC stage for bus amplifying the 3 buses up to about 0dBu, thence to the master control and, for the mains at least, through the big cap CC stage to get up to the required +10dBU (3V). Possibly the same for the FX send, so that uses 6 triodes leaving two. FX returns are at -10dBu so to mix them on the main L&R bus but we really need to get them up to 0dBu so use two triodes for that and none left.

That's my guess, anyway.

Cheers

Ian

 

[tonedude]

I notice some of the PCB attachment holes are broken out, you may want to be careful about that when reassembling so the daughter boards do not move around loosely inside.

For more than you may want to know about how it was manufactured, it looks like most of the smaller boards were originally part of a single larger board and was originally wave soldered as one big board with the several ribbon cables between subassemblies inserted before the wave solder step.  After wave soldering, the boards were broken apart, rotated, and attached. This is a lot less labor than performing point to point wiring or even using multi-pin cables and connectors between the separate PCB at final assembly.  I see from his initials that the PC board was designed by Keith (JKV).

JR

Yes I saw that they where broken and that made me a little confused as I was very cautious taking it apart, but it's not really a problem as both boards are firmly attached to the chassis on their own.

Cool info on the manufacturing process! It's always nice to learn something new, and I knew nothing about that part. Quite an effective way to have the board soldered. It's a lot of thought put into designing the board itself to cut time and cost assembling it. Quite impressive!

 

[tonedude]

I am still trying to work out how this was done with 8 triodes. ASSuming we have an input pot followed by a mute switch followed by a pan pot and FX send controls, all passive, that's three buses required. Cheapest we can do is to passive mix the lot right away - no buffers. With line levels in and 10dB lost in the pan and and about 16dB in passive mixing we end up at about -26dBu on the L&R buses and a bit less on the FX - not too bad but we probably had to use 470K or 1Meg bus resistor to maintain the 1Meg input impedance requirement. Actually its a bit worse than that because we have the two aux returns to add to the L&R buses but still basically OK. Looking at the tube PCB layout, it looks like all the stages are pretty much the same except for two which have bigger output caps - probably the main outs. Since the layouts all look the same they are probably straightforward common cathode stages. Main output stages probably with lower plate resistors and higher idle current and bigger output caps. Probably a CC stage for bus amplifying the 3 buses up to about 0dBu, thence to the master control and, for the mains at least, through the big cap CC stage to get up to the required +10dBU (3V). Possibly the same for the FX send, so that uses 6 triodes leaving two. FX returns are at -10dBu so to mix them on the main L&R bus but we really need to get them up to 0dBu so use two triodes for that and none left.

That's my guess, anyway.

Cheers

Ian

I have some time tomorrow and I will start tracing the circuit to learn and share! I will post it here when it's done!

/ Mattias

 

[okgb]

I would assume at least 1/2 a triode per / channel 3 tubes , perhaps a whole tube , 6 Tubes
at least 1/2 a stage  maybe two stages per output , that's potentially 8 , depends whether any are used for the aux sends
What is the gain , ch , output , aux or overall for the unit ?

 

[tonedude]

Ok, here is my first attempt trying to trace a circuit... If you find something that looks weird and out of place, it's a good chance it is...  ;D

The things I think looks strange is the two 100 Ohm resistors drawn in the schematic (the color code brown, black, brown, and gold gives hundred ohms, but messuring the resistor in place gives something else... ). The other ting is the input and master volumes, i've been trying to understand how the arrangement works, but it beats me...  ???

 

[tonedude]

I would assume at least 1/2 a triode per / channel 3 tubes , perhaps a whole tube , 6 Tubes
at least 1/2 a stage  maybe two stages per output , that's potentially 8 , depends whether any are used for the aux sends
What is the gain , ch , output , aux or overall for the unit ?

- Gain out from one channel sent on to the master section is 10dB (Sending in -20 dB into channel 1 panned hard left out then measuring -10dB from L out (reference is +4dBu)). I think here the impedance mismatch comes in lowering possible gain out in a negative way (the load on the output is not optimal as it is 10kOhm)

- Gain out from channel 1 sent to aux out into 10kOhm is actually lower than the input, -3.5dB lower

 

[JohnRoberts]

I'm no tube guy but that looks straightforward.  Each half-tube gain stage makes an inverting amp with variable gain since the gain/level pot is across the feedback network.

These outputs go through lossy pan stage to sum busses, and final two tube halfs buffer the stereo sum bus outputs

The aux send is passive, but keep in mind this is targeted for guitars and stomp boxes so low impedance should not be necessary.

or not...  The channel fader kill with that kind of gain pot is not great but adequate for a KISS design.

The real tube guys may have better insights

JR

 

[okgb]

the tube stages themselves look standard , [ plate , grid cathode values ]  the dividing network between the two stages
seems curious , where does the first 100 k to cap go ?  [ the one  below the feedback from the second stage ]
fixed gain with attentuation on the input  before  the first stage and gain through feedback on the output stage
nice work , thanks for taking the effort

 

[abbey road d enfer]

Added larger/more caps in PSU which made it less prone to distort bass frequencies.

The original complement of 100uF for B+ is more than enough for 8 sections of 12AX7. Many power amps have not more than 100uF on B+, check out Twin Reverb, with only 35uF on B+. If your unit distorts on bass, adding zillions of caps won't change a thing, it comes from somewhere else, unless the original caps are dry...
First thing I would have done is replacing these 20 year old caps.

If I hit it really hard I can make it distort, a little, not much really...

You can make distort anything if you hit it hard enough.

I plan to upgrade caps in the audiopath, any suggestions? Wima MPK?

Without knowing positively if the ones in place are not good enough and what type of improvement you are expecting, it would be preposterous to give you a definite answer. I know it is quite common in audiophool circles to claim that X brand capacitors or Y brand resistors sound "better", but here we like to have a more pragmatic approach.

 

[ruffrecords]

Well my guess was completely wrong. The actual circuit has some gain per input and I has ASSumed this was not necessary as the inputs were to be line level.

The output stage is a simple virtual earth mixer stage with 100K bus resistors for the channels and 47K bus feed resistors for the FX return.

The FX send circuit is dreadful.

Cheers

Ian

 

[okgb]

Sure, gain to overcome the passive mixing loss , another section on the output to give it a
buffered low imp drive would have been reasonable ,  3 stereo input or 6 mono
I still think it was a good idea

yeah , not ragging on it John , what was the original price point ?

btw I noticed a variation on the lightening bolt logo , a oval line around it that softens
the goofyness while adding a slight retro feel , reasonable compromise who ever came up with that

 

[JohnRoberts]

Keep in mind this mixer was for a guitar rig, to drive some amps and pedals. All high Z stuff...

JR

 

[mlange]

Hey All-

Old topic, but as I have the actual Valvex schematic, I'd be happy to email it to anyone who needs her.

Cheers,

Mark