Yule Tide Update -- and a solving a 47 year mystery!
Work on the big Bogen has certainly slowed with all the holiday activities, but not stopped. The power supply area of this particular unit was nothing short of a complete mess (see the last three pics of post #1), with wiring that was simply scary, given the over 700 volt operation this unit routinely sees. So, time has been taken to tidy up that house.
Also, with EFB the next batter up, consideration needed to be given to the somewhat unique bias supply these amplifiers have. For what they are, the bias supply in these units are really quite dependable, and given the close tolerance components used, quite accurate as well. For the needs of the original design then (where the quiescent output stage current is intentionally set quite low), they are quite appropriate. However, they are also a rather high impedance supply, meaning that if you vary the load on their output (as adjusting the new DC Bias and Balance controls does), the output of the supply varies right along with it. In addition to the manual controls then, the EFB circuit will also vary the load somewhat on the output of this supply as well. Having the output of the bias supply vary as the manual and/or EFB circuits work to achieve the ideal bias voltage applied to the output stage means that the range of these controls and the EFB circuits need to be extended greater than normal. This can be done, but it's usually easier to look at stabilizing the bias supply itself.
Within its topology, the existing bias supply could be beefed up to address this concern, but the original design already consumes much energy for what it does, and trying to beef it up would only aggravate this issue. In the end, given the re-purposing goals of this project, it was decided to scrap the old bias supply all together, and go with something new.
For this project, the junk box yielded a small 18 volt power transformer, that if you can believe it, came out of one of those "100 watt" junk computer speaker systems. The silly thing is half the size of the big ol' 600 volt coupling cap used to tap the original bias supply into the HV winding of the power transformer! In any event, pairing this up with a full wave voltage doubler circuit will provide all the stable bias voltage needed, not only to power the new EFB circuits, but also to remain quite stable as the various controls of both channels are adjusted, or the EFB circuit works to dynamically hold the output stage at the point of lowest distortion with varying B+ power supply conditions.
So, progress has still been made during the slow down, with the power supply cleanup nearly done, and work on the new bias supply begun. More info as that work progresses.
As for the 47 year old mystery:
Christmas time always brings a time of reflection on my life, as I'm sure it does for many of you. Last night (well, in the wee hours of today), my thoughts turned to my mid-teenage years in the mid 60's. This was a time when multiple garage bands could be found practicing on any given Saturday in many of the neighborhoods in the city I grew up in. High Schools and even the State Fair sponsored large 'Battle of the Bands" that most groups strived to just win a spot to be allowed to compete in, let alone win. The band I was in was no different. I was fortunate enough that our band eventually rose up to be a well sought after band, winning a number of the competitions, with about a two year stretch where we were booked every Friday and Saturday night with good paying gigs.
At the time, the idea of using large sound reinforcement systems that are so common today, were unheard of. It was every man for himself with his own equipment, so you tried to balance the sound of all the individual pieces, and just hoped the vocals could be heard over the guitars, organ, and drums. Mics were for singing into, and that was it. Fender amps ruled the day, and so did Bogen PA equipment.
None of the members in our band grew up with any silver spoons, so all of our equipment was homemade. My brother could produce leatherette covered speaker cabinets with Fender Bandmaster speakers in them (sold by RS no less!) that looked every bit as professional as the Fender gear. I built all the amps for the group from blank Bud chassis -- bass, lead, rhythm, Leslie, and PA, which were all basically Fender Twin Reverb knockoffs. We were the only band that had reverb on our voices! What a time. This kind of approach produced a very clean, very loud, but raw sound that is completely absent today. It was so neat. If you've ever seen the movie "That Thing You do", you can get a sense of it -- particularly in the scene where the band plays the state fair with all the period correct Fender equipment.
So back to the mystery. The biggest bands (with seed money) always used a Bogen MX-60, driving a pair of University ID-60 horn drivers, screwed on to University CLH horns. We used the same drivers and horns, which was the only real equipment that could raise the vocals over the band amps -- if the PA amp was willing.
Now the PA amp was always taxed pretty hard to do its job, but the horns were so efficient that an honest 60 watts through those horns could drive people out of the area all by themselves. Our Twin based PA system always sounded truly great, never having any trouble elevating our vocals over the band with clean, clear, sound. The number of other band members that would ask about our "no name" equipment was endless. But the Bogen MX-60? It never failed to UN-impress.
Without exception, at the time, whenever I heard an MX-60 in a band, it always had this pulsing, overloaded sound about it when pushed to its upper limits. Not an output stage clipping type of sound, but clearly an overloaded sound about it, that would pulsate in sound level produced. But it was a Bogen -- and an MX-60 at that, in a time when a name was everything in terms of "band" equipment -- so it was the best you could get (except for that pulsation........)
So as I came back to reality last night and started thinking about this current Bogen project, it hit me like a ton of bricks: With the distortion recently discovered in the driver design of the MO-100 and MO-200 units (see post #158), could this have been the source of the overload heard so frequently, so long ago in those MX-60s? After all, they use 8417s in their output stage as well!
A quick check of the MX-60's schematic solved the mystery: The phase inverter/driver design used in the MX-60 is a (virtual) carbon copy of that used in the MO-100/200 amplifiers. If the driver overloads in them, it would overload in the MX-60 as well. And, it would overload in the exact way I heard it to overload as well: As the input grid to the AF amplifier was driven positive, its plate voltage would fall pretty dramatically. This would cause the rather large R/C time constant between the AF Amplifier stage and the inverter stage to seriously off-bias the inverter stage negatively, causing the sound to fade. As the overload went the other direction on the AF stage, it's plate voltage returned through normal, but now, moving to a very high plate voltage, off-biasing the inverter stage in the other direction. This would bring the sound back with authority, but only to have it go the other way again, continuing on until things settled down and the overload was eliminated. There is no doubt that this was the source of the pulsing sound I heard so many times from some many MX-60s..
I know there are some MX-60 aficionados out there, so you folks as well might take heed of the driver modifications made the the MO-100 and 200 amps being discussed here. Without any of the other changes also made, the NFB loop components would require some alteration in the MX-60s, but they would be very easy to determine. Otherwise the basic driver change would directly apply to the MX-60 as well, and eliminate high power level pulsing problems so evident in these units. Just thought you interested folks might like to know. If anybody wants to know the new value NFB values that would be appropriate when installing the modified driver circuit in their MX-60, just let me know.
Just a couple of pics to show the latest progress:
1. Basically what I started with, shortly after starting to tear it all out.
2. The HV supply all neatly reinstalled, with no worries about where that 700+ volts might be going now. The only changes made to the HV supply (besides removing the old Bias Supply design that stole power from it), was to include a 470K 1W bleeder resistor. The original design is self discharging -- but only on the bottom half of the voltage doubler design. The top half is only discharged by the plate current of the output tubes -- but if they're not in, then the charge remains. The addition of the bleeder resistor means that the complete HV supply will now self discharge itself whether the tubes are in or not, in less than 1 minute.
Also, you will see the new "100 watt" bias supply power transformer now mounted to the side of the chassis. Work will begin on the new bias supply next, but for now, with everything cleaned up, the evidence of the original ground zero destruction can be seen with all the black residue -- and this is after a basic clean up too!
Once the new bias supply is installed, then the original and new designs can be sonically compared, with the EFB installation beginning after that.
Dave
