Bartel Interstage Transformer Circuit

[W5FH]

Hello-
    Out of curiousity I have been viewing abovechassis and underchassis photos of the Bartel Sugarland, Starwood and Roseland all of which use an interstage transformer. A 12AU7 drives the interstage and from what I can deduce the 12AU7 is in push-pull mode driving the interstage transformer. In the Starwood a photo (attached) shows the 12AU7 with the I/S transformer yellow wire going to one plate (pin #1) and the orange wire of I/S going to other plate (pin #6). The cathodes (pins 3 and 8) are tied together with a wire going to a turret board. Each grid of the 12AU7 (pins 2 and 7) have a 1.6K resistor (it appears) going to turrets on a board to the right of the socket. It appears that there is an active solid state device on this board, appearing as an integrated circuit DIP profile with 8 pins. It appears voltage is being supplied to this board (R+ and R- terminals). My thoughts are that the IC is serving the function as the "single ended to push-pull" conversion or better said unbalanced to balanced function. It appears to me that the unbalanced dry signal is going to the second from top turret. I say this as a green wire leaves this turret and feeds a 470K resistor feeding the grids of the 12AT7 reverb transformer driver tube (this is tube to left side of picture). The two turrets that the two 1.6K resistors are attached to must be carrying a push-pull signal. If that IC on the board is not doing the single ended to push-pull conversion then the orange and green wires attached to these two turrets are bringing a push-pull signal from elsewhere in the amp. If that is the case, then possibly the IC is serving as an attenuator for the post phase inverter master volume.
   My question is anyone familiar with an IC cased like this that would serve this function?
Possibly I am all wet behind the ears here, or maybe I am nosing into something that I should not, and if so I will respectfully cease being nosy. I am just trying to further my knowledge. Sincerely and respectfully.

[PRR]

The "DIP16 with only 8 legs" is probably a relay?

[W5FH]

Thank you for the reply PRR-
     You are correct in that there are DIP-cased relays that look like this subject device. And to further support your statement is the supplied voltage turrets are labelled "R+ and R-". Not sure what switching functions would be going on at this board unless it is associated with the "standby" switching arrangement (possibly grounding signal path/disconnecting).
     If my thinking is correct, an I/S transformer would be used between the output of the preamp and the grids of the power amp tubes. A benefit of using transformer coupling is that blocking distortion on power amp grids cannot occur. I was expecting to see a single ended stage driving the I/S transformer primary like in other amps. It is stated in technical info that a 12AU7 is the "phase inverter". I was not sure how to interpret this; thinking the 12AU7 was actually a driver tube to I/S trans (and trans was serving P/I role). The 1.6K grid resistors look pretty low for the 12AU7 to be a phase inverter, plus there would need to be other components on turret boards. The 12AU7 plates appear to be connected to the I/S trans primary in a balanced drive fashion; very much like output tube plates connected to output trans primary. I guess it did not sink through my thick head that the 12AU7 is a P/I just as stated (and not just a driver); I envisioned the I/S transf serving that role as P/I.   

[PRR]

I never heard of this brand (I don't go out much), it does look interesting, but he's not spilling many beans (why would he?). Until you buy one and dissect it, it may remain a mystery.

[echuta13]

He's a pretty interesting builder. Casts his own knobs (+ some other components), does his own silk-screens, and cabinet work.  Very nice stuff!

[sluckey]

The "DIP16 with only 8 legs" is probably a relay?

My initial thoughts too. However, I don't remember ever seeing a sensitive electronic circuit without a diode across the coil.

Another possibility is a dual optocoupler.

[shooter]

the turrets at the tail of the IC are labeled R+ R-, even appears to be a hole drilled where I would expect a diode, although could be resin splat.
the 4 pins on the left - looks like a wire shorting all 4 legs, behind the pins, could be a light reflection, can't really tell

[W5FH]

Hey Shooter-
   Thanks for comments. Concerning what appears as a wire going across left side of DIP relay: Possibly this may be the edge of the "backing" material that I have seen on the underside of these DIP relays. I have noticed on some DIP relays a white plastic looking material applied to underside of relay body.

[NewYorker]

Bartel founded Tone King. Based in Baltimore. Tone King marched to the beat of a different drummer, too.

[dbishopbliss]

I never heard of this brand (I don't go out much), it does look interesting, but he's not spilling many beans (why would he?). Until you buy one and dissect it, it may remain a mystery.

You could try sending an email to Mark. He came to a local "tone fest" once. He drove 2 hours from Baltimore to hang out with a bunch of us in the upstairs of a community pool house playing guitars and listening to amps. I showed him my Sluckey AC15 and he was very complimentary of the layout. Nice guy.

[PRR]

....I don't remember ever seeing a sensitive electronic circuit without a diode across the coil....

They make relays with the diode built-in. When they do that, the coil (usually non-directional) needs + and - marks. We see this on the PCB silkscreen. OTOH there are latching relays where one way is Set and the other is Reset... I donno how they mark these.

[MWaldorf]

I was working a friend's booth at NAMM a few years back and Bartel was the next booth over.  Not only is his work impeccable, he's a really nice guy.  Plus his amps sound great.

[HotBluePlates]

... A 12AU7 drives the interstage and from what I can deduce the 12AU7 is in push-pull mode driving the interstage transformer. In the Starwood a photo (attached) shows the 12AU7 with the I/S transformer yellow wire going to one plate (pin #1) and the orange wire of I/S going to other plate (pin #6). The cathodes (pins 3 and 8) are tied together with a wire going to a turret board. Each grid of the 12AU7 (pins 2 and 7) have a 1.6K resistor (it appears) going to turrets on a board to the right of the socket. ...

I recommend re-counting those pins.

What I think I see is Pin 2 & 7 (grids) receiving the same signal via a wire-jumper.  Meanwhile, Pins 3 and 8 each have an individual 1.5kΩ (Brown Green Red) cathode resistor.

I don't know what the Yellow/Orange plate wires go to.

Overall, any "interstage transformer" is very very likely to be a phase inverter, receiving a single-ended input and delivering a push-pull output.  Support for this comes from Bartel's website:  "... An interstage transformer retains symmetry in the phase inverter even in hard clipping. ..."  I believe "phase inverter" is being used here to denote a function, not a specific tube (because the transformer is the inverter).

[W5FH]

Hey Hot Blue Plates -
    Thank you for the help here. I went back and carefully reviewed the picture of that tube socket. Respectfully, I still believe the wiring is as I originally said. It sure appears the orange and yellow wires (going to plates) are slightly larger than the other wiring being used in amp (suggesting they are transformer leads). The 1.6K resistors are going to pins 2 and 7 (grids). Pins 3 and 8 (cathodes) are tied together with a wire leaving socket.
    When I get a moment I will post a schematic here, it is the MD-129 modulator from a 1950's era Air Force UHF AM transmitter made by Collins. I my Ham radio hobby of many years I have scrapped out a lot of military electronics for parts. There is an interstage audio transformer in there, made by Chicago, that is very wide banded as the modulator handled an early form of data transmission requiring an unusually large bandwidth for an AM transmitter. The modulator used a 12AX7 phase inverter that drove a pair of 6BA6's driving a pair of 6AQ5's in push-pull. The 6AQ5's are driving the primary of this I/S transformer. The secondary of I/S transformer drives a pair of 811 modulator tubes. So in my thinking of Bartel calling the 12AU7 a phase inverter this MD-129 modulator came to mind since a 12AX7 phase inverter is driving an I/S transformer (through two amplifying stages). I guess the whole crux of my interpretation of the picture is if the two wires (orange and yellow going to plates) are actually transformer leads. But I am all ears here and open to any and all corrections.

[PRR]

> MD-129 modulator

Like this? (attached)

The actual phase inversion is in the first stage. An odd choice which adds several cents. From there it runs R-C coupled through to the plates of the 6AQ5s (small 6V6es) and a transformer to the monster 811 bottles.

I suspect the point is "redundancy". After the first 12AX7 stage, any single tube could crap-out and signal would still pass. You see this some in commercial broadcast gear, where off-the-air may be hundreds of dollars of unbillable commercial time. Cold-war radio failure could have been even more costly.

The 12AX7 stage has another weirdness. It is all on NEGative 65 volts! At 65V and 220k load, it is not ever going to wear-out the plate section or cathode. It may be working 0.15-0.2mA and 32V on the tube, so 5mW in a 1Watt plate and 5mA cathode.

[W5FH]

Yessir PRR-
    That is the MD-129. There were two USAF transmitters, one VHF and the other UHF. They had very similar power supply/modulators. The other supply was the MD-141 if I recall correctly; I have parted out a couple of each over the last few decades. Thank you for going to the trouble here to download the schematic, post the relavant portions and the interesting commentary. Recently I dug out T-302 and reviewed the specs on it. Frequency response is 200-15,000 CPS +/- 0.5 dB, W.V. 535 vdc. Not sure if enough low end frequency range?? The use of an I/S transformer in a guitar amp has piqued my interest, mainly for reducing/eliminating blocking distortion on grids of output stage. I have been researching circuits for driving the primary of an I/S transformer but have found none (except this modulator) that use a tube phase inverter to drive transformer primary in push-pull. PRR what is your take on viewing that 12AU7 tube socket wiring. What type of circuit would have the cathodes tied together, separate 1.6K "grid stoppers", and (if I am interpreting picture correctly) each plate going to a transformer lead? If the two plate leads are a transformer winding (center tap supplying B+) then this 12AU7 tube would have to be a phase inverter, correct?

[HotBluePlates]

... I went back and carefully reviewed the picture of that tube socket. Respectfully, I still believe the wiring is as I originally said. ...

Yep, you are correct!  I counted the pins counter-clockwise from Pin 9 (thinking it was Pin 1), so I got everything backwards.

... So in my thinking of Bartel calling the 12AU7 a phase inverter this MD-129 modulator came to mind since a 12AX7 phase inverter is driving an I/S transformer (through two amplifying stages). ...

Ken Fischer called a phase inversion transformer in old Gibson amps an "interstage transformer."  I suppose it is, but the point is that it performs phase inversion, while later designs allocated a triode or 2 for that function.  See the Gibson BR-9 for an example.


Why does Bartel care?

On another forum, we debated someone's observation when using a scope who claimed the phase inverter of an amp distorts first.  They made this claim because they turned up an amp until it distorted, then scoped each stage, and saw "distortion" at the output of the phase inverter.

    - The phase inverter outputs are hooked to the output tube grids.

    - When peak drive signal equals/exceeds the output tube bias voltage, the output tubes draw grid current.

    - The grid current is a low impedance:  the grids go from looking like ∞Ω to ~800Ω.

    - The phase inverter was able to drive a load of 60-80kΩ; when the load drops down near 1kΩ the output of the phase inverter is clamped/clipped.

    - My contention is to perceive this as "output tube distortion" because the output tubes are indeed distorting (they're pushed to/beyond their clean power output), but also because they are inducing the clipped output of the phase inverter.  It would not be clipped if the grid current weren't happening.

    - The coupling caps between the phase inverter tube & the output tubes are getting charged by the grid current, and shifting the actual bias of the tubes.  This leads to blocking distortion, which the zener diode approach at the link seeks to resolve.


I believe Bartel sought to sidestep the output tube grid current/coupling cap/blocking distortion issue altogether by using a "single-ended to push-pull transformer" to generate the equal & opposite signals for driving the power tubes.

[PRR]

...See the Gibson BR-9 for an example.

Or Fender Musicmaster Bass.

[W5FH]

Thanks comments HotBlue Plates and PRR -
      I agree with what you are saying regarding the differences in driving output tubes with transformer coupling as compared to driving them with the triode phase inverter outputs using coupling capacitors.
      Up until I read the Bartel manual, which lists the 12AU7 as a phase inverter, I have always envisioned a single ended stage driving the I/S transformer primary and the I/S transformer performing the phase inversion, driving output tube grids in push-pull. And I agree with a previous statement explaining that possibly Bartel called the 12AU7 a phase inverter because it was all-inclusive (12AU7 driver tube and I/S transformer with phase inversion performed by I/S transformer). What bothers me is this -- based on the tube location on the chassis as shown on layout in the manual I feel confident I am viewing correct tube socket (You can see the paralleled triodes of reverb driver tube to the left of 12AU7 phase inverter). When studying the wiring of this 12AU7 socket I question the tube being a single ended driver to transformer. Those two heavier appearing wires (orange and yellow) each go to a plate pin, this suggests to me that the 12AU7 plates are connected to a transformer winding. If that is the case that means the plates are receiving B+ from (a center tap on) that transformer winding, and that the plates are push-pull driving the I/S transformer primary. You can see that the two cathode pins are tied together. Each grid has a 1.6K resistor going to separate terminals on the terminal board. So what can be inferred from studying this tube socket wiring, do you feel the orange and yellow wires, being a little larger size than other wiring, are transformer leads? I have studied pictures of chassis topside and you can see an orange and brown lead coming off of I/S transformer. Only conjecture here and all for the sake of learning and curiousity about the circuitry. 

[HotBluePlates]

... until I read the Bartel manual, which lists the 12AU7 as a phase inverter ...

I am stuffed from eating all this crow!!     Thank you at least for being gentle in pointing out my mistakes!  I should read/research more before speaking.

You're right, Bartel's manual does say a 12AU7 is the phase inverter.  In which case the only smart approach is center-tapped primary receiving a push-pull signal (from the 12AU7) to a center-tapped secondary delivering a push-pull output to the output tube grids.

In which case it would be a lot like the modulator you mentioned, and (sorta) for the same reason:  output tube grid current won't upset the biasing.

[W5FH]

No problem at all HotBluePlates -
     You give a ton of help and assistance to folks here, I am glad that you have taken enough interest to help me here. Maybe we can figure it out. It really gnaws on me when I cannot figure something out. My goal here is learning and thank you for all you do.

[W5FH]

Some interesting info about the Bartel I/S transformer circuit:
     In viewing the Sugarland chassis photos the topside view shows the I/S transformer #TX20P/40-17662 with six leads (Orange, Yellow, Red, Blue, Brown, Black). In viewing the chassis underside photo from a dealer's advertisement (could not copy here, wrong file type), it appears to me that the I/S transformer primary orange lead is going to (U-4) 12AU7 pin #1 (plate) and the primary yellow lead is going to pin #6 (plate). The red lead (primary center tap) of I/S goes to B+ (not visible this picture). U-4 is listed as phase inverter. Each grid has a 1.6K resistor each attached to a turret. There is a 220K resistor across these two turrets and a 1 watt resistor (can't read value) connected to the pin #2 grid resistor turret. What is interesting is the circuitry involving the secondary leads of I/S transformer, the brown and blue lead. It is my opinion the brown and blue leads (shown coming up from under the turret board near the AC line fuse) are from the I/S transformer secondary. These two transformer leads go to a pair of turrets. It appears a rectifier diode, a 6.9K resistor and a (zener?) diode are all 3 in series, with a set of these 3 coming off of these two turrets. They combine at the anodes of the (zener?) diodes and a black wire goes to the 1 watt resistor mentioned previously. Also a black/green twisted pair of wires come off of the I/S transformer secondary lead (brown, blue) turrets; this twisted pair go up to the master volume pot. It is mentioned in descriptive literature that the master volume is a post-phase inverter type and circuit operates on the saturation properties of the I/S transformer. So I am assuming the diode circuitry is involved in this; has anyone studied a circuit like this? I cannot find the I/S transformer secondary black lead, I believe it to be the center tap and would need to be grounded or AC coupled to ground. The 6V6's are cathode biased.   

[Matt R.]

Been intrigued about this circuit myself. The diode/resistor network is a feedback network from the secondary, going back to the 12au7 input via 820 ohm R to grids. MV is directly off the secondary of IS xfmr, as are the diode/resistor network. Question is: where are the 12au7 cathodes attached?

[W5FH]

Hey Matt-
     Very nice to see your reply here. I spent quite a bit of time viewing photos on the Bartel Amplifiers Facebook pages. There are quite a number of views of both assembled units and units in various stages of assembly. I have plans to go back and look some more to try and possibly get clues as to what goes where. I cannot see where the lead from cathodes goes.  I am curious about the comments made, in an advertised technical description, in regard to using core saturation of the I/S transformer as a means to both adjust master volume and add tonal effects. To me, when increasing core saturation you would have to be careful to not lose too much of the frequency response of transformer. I would think this would be done by increasing the DC current flow (would have to be in primary). The master volume is a dual pot and it is stated it is a post phase inverter master volume. It would appear this may involve the diode/resistor circuitry connected to the secondary of I/S transformer (brown and blue leads). The red lead of I/S transformer is the center tap of secondary winding; you can see it soldered to ground rail near 12AU7 phase inverter in some of the views. The black lead is the center tap of the primary winding, but I could never see where it tied in (to B+ I assume). The 1.6K resistor coming off of each grid of 12AU7 P/I perplexes me, normally these resistor values would be much higher for P/I. However, these grid resistors of high value (1 meg) are for cathode biasing the P/I and possibly this P/I is not cathode biased but instead externally biased from the resistor/diode network. These grid resistors go to turrets on the DIP relay turret board and multiple wires from the twisted wire bundles attach. I am thinking the DIP relay is for opening up/grounding out the reverb signal. Due to the added circuitry of relay I focused on studying the Sugarland as it has no reverb or tremolo, but uses this same intriguing I/S transformer circuitry. I am very curious just how that 12AU7 P/I is configured. It is called the phase inverter in tube function listing. It's plates are in push-pull as you can see the I/S transformer primary leads connected to pin #1 and #6. If the 12AU7 is a LTP P/I, as you know, it is a differential amplifier. Normally there is no signal input to the grid of the in-phase triode (non-inverting triode of P/I). If the same signal that is applied to the grid of the inverting triode then there will be no output from plate circuits. Possibly this property is being used in a fashion, involving a bias voltage from that rectifying circuit off of I/S transformer secondary. I surely look forward to more of your observations and comments! 

[Matt R.]

Haven’t looked at the other models only the Sugarland, but this is something I’ve never seen. I believe some kind of bias is coming from that secondary network, fed through 820R and cathodes tied to junction of 820R/220k/1k6. Still, a lot of pieces missing from this puzzle…  I’m about 45 mins from Mark in Md. Talked to him online about his design briefly but no secrets revealed. Awesome amp/design for sure, I probably wouldn’t reveal anything either.

[W5FH]

OK Matt-
   Great on visiting with Mark. I can understand why he would protect his work product and I have great respect for that. I am curious in the technical sense about innovative circuitry and admit that I would like to learn and understand such. From viewing the photos of his shop and work I have the utmost respect for his meticulous approach to all aspects of amp construction. Very, very impressive and in a league of his own.
   I did a search on core saturation of audio transformers and found some interesting info. Folks are experimenting with it to replicate distortion properties of vintage amps that used undersized output transformers. Jensen published a 31 page treatice on audio transformers that shows up on the search; this is an informative paper.
   At this point it appears to me Mark uses the fairly much the same topology in preamp sections of all 3 versions of amps. The first two stages (tubes) are triode stage/cathode follower. Based on description of tone circuitry, and using a dual pot for preamp volume, I am wondering how these two stages (tubes) are arranged. The third stage (tube) appears to be all parallel connected like a reverb driver tube. Very interesting for sure.   

[W5FH]

Here is a hand drawn schematic of the 12AU7 phase inverter I/S transformer master volume circuit. Not sure if correct, the best that I could do from studying pictures.

[W5FH]

Interstage transformer driven by phase inverter:
     For reference take a look at the Magnatone MP-1 schematic, link included below. A 12AU7A phase inverter is driving a center-tapped reverb tank driver transformer,  shown as having a 25K center-tapped primary. A similar application.
 https://www.magnatoneamps.com/schematics/magnatone_mp1.gif