Magnavox Flea Power: Getting More Out Of The 8600 Series - A Lot More!

dcgillespie

Fisher SA-100 Clone
Subscriber
So much has been written about this brand and these fabled little amplifiers that it could pose a serious threat to the amount written about Dynaco products! And just like Dynaco products, the modifications available are never ending for virtually any model you chose to discuss. And so it is with the Magnavox 8600 series as well. However, unlike the Dynaco products, the little 8600s rarely get any serious engineering attention to get the most out of them.

A search returned that the vast majority of circuit modifications are basically limited to removing the voicing components in the NFB loop (not always done correctly), or adjusting voltages to match the published schematics (also done with varying degrees of effectiveness). The rest of the efforts typically center around which tubes and components sound best, and dressing the little suckers up to disguise their humble beginnings.

I mean nothing against these efforts in the least. But after getting one of these cuties for myself (for a variety of reasons), the potential to get so much more out of it was immediately obvious. Ultimately, because these amplifiers represent so many different things to so many different folks (myself included), I decided to tear into mine and see how far a complete makeover could take it. Within the world of flea power and what the unit represents in stock in stock form, the results have been nothing short of spectacular.

Consider the wide attraction this amplifier has (in no particular order) to:

1. Those longing for the holy grail of single ended midrange sweetness.

2. Those using one in a biamped flea system for the mids or highs.

3. Those who are simply enamored with flea power.

4. Those who want to see what tube sound is all about in the cheap.

5. Those newbies to tube audio who want to take the plunge and get hands on service experience, but want to start out small and simple (very wise).

6. Those who were paid to haul away an old console; found one installed in it and wanted to see if it would still work (hey, we had one of those growing up!).

7. Those like myself, who have a young grandson that seems to have every attraction to vacuum tubes and music that I did at his age. That makes this a pilot/development project to set up the real one down the road in a couple of years.......

However or why ever you came into the possession of one of these units then, and regardless of what you use it for, with a working example, a moderate budget, and a few fun evenings, you can turn that (basically) stock unit into one that will cause folks to take notice -- of the sound.

Now first off, your ticket into this flea show starts with good, efficient speakers. Without those, you simply have to take your toys and go play elsewhere. In my case, I've owned Cornwalls from before it was popular to own Cornwalls, so at 101 db, my ticket is punched for entrance. So how is the stock amplifier with these speakers?

I got my unit (an 8604-00) without tubes from a fellow AKer battradio, in stock unmolested condition. A quick once over determined that all the parts that count were in good working order -- enough so that an accurate base line of performance could be established. A fresh set of tubes were installed, and the amp was then tested in the listening room. I gotta tell you, I was frankly underwhelmed. The unit displayed classic ground loop hum all by itself (I haaate hum), and just seemed asthmatic: It just seemed to run out of steam more than I am accustomed to. I know this is flea power. But I am accustomed to what good flea power can do, and to me, this wasn't it. Beyond that, the sound was not engaging, and didn't hold my attention. But it was in bone stock condition, so next up was putting it through the paces on the test bench to try and correlate what it was doing, to what I was hearing. Oh boy......

For perspective, I doubt there was ever any published specifications for these amplifiers. Be that as it may however, they are generally regarded as having anywhere from 3 to 5 watts per channel depending on who is offering the information, and that's about it. Beyond that and the tube complement, about the only other information known is that each channel is designed to operate into a 4 ohm load.

A quick glance at the tube manual shows that tubes of the 6BQ5 class are capable of developing 5.7 watts of power when operated in single ended mode. Of course that power level is never realized due to losses in the output transformer. Still, when that is accounted for, about 4.6 watts of power can be realized. Looking at the schematic for these amplifiers, they operate with similar or near similar voltage levels as those shown in the tube manual, so the claim of 3-5 watts would certainly seem reasonable. Except, that isn't what's happening..........

Using a a proper 4 ohm load and a mid band frequency of 1 kHz, each channel could develop just 2.3 watts of power before (unequal) clipping commenced. The power bandwidth (-3db or half power points) extended from 100 Hz to 20 kHz. But that's at 4 ohms. When an 8 ohm load is used -- as most speakers are today -- the available power is then cut in half. Now the unit is capable of just 1.2 watts per channel before clipping commences, with little change in the power bandwidth. That's not a very good head of steam, and correlates exactly with what I heard.

Next, some basic response and stability tests were run to finish out the base line testing. Now remember that the unit still includes the voicing components of the stock design. Still, the only time I've ever seen such contorted square waves is when something was drastically wrong -- except that nothing was in this case. Needless to say, the response was anything but flat......

I didn't get any pics of my unit in stock form before I got started, but needless to say, it was in very good physical shape, and pics of these units -- in all manner of physical form -- abound all over the net, as does the schematic. Search engines will bring it up on the first attempt. Still, a few pics will get the ball rolling:

1. Here channel 1 trying to develop 1.35 watts into an 8 ohm load at 1 kHz. Clipping is obvious, as is the fact that it is unequal as well. Channel 2 was no different.

2. This is a 2 kHz square wave applied to channel 1. This frequency is used because to accurately reproduce a square wave at this frequency requires a flat response within the audio bandwidth to 20 kHz. From this display, it is hard to even tell it's a square wave, with strong peaking and notable ringing across the top of the wave. The peaking causes early clipping within a narrow band for the power that's available, and the ringing can always be shown to affect clarity and detail. In short, it's a mess.....

3. With a 10 kHz square wave, you might as well take your toys and go home. It isn't square, it isn't.... heck I don't know what you'd call it...... But knowing what's going on in the supersonic realm of the amplifier is important to amplifier stability -- as will soon be shown..

This is all pretty bleak -- and likely why many single ended advocates simply avoid performance specifications and only talk about the "sound". The battle cry is that you don't listen to specifications. Uuuuhhh. I'm sorry, but you do -- and I heard these specifications in my listening room. However, it can all be dealt with quite effectively, with simple changes to turn Cinderella into a princess:

1. Frequency response can extend +0/-.4 db from 40 Hz to 40 kHz.

2. Power bandwidth can extend from 90 Hz to 25 kHz.

3. Mid-band power output into 8 ohms can be increased by over 300%.

4. High frequency transient response (square wave display) can take on an ideal wave shape, and

5. Stability can be maintained to a very good degree into capacitive only loading.

We'll start getting into all of this, next time.

Dave
 

Attachments

  • SAM_1447.jpg
    SAM_1447.jpg
    30.6 KB · Views: 598
  • SAM_1429.jpg
    SAM_1429.jpg
    49 KB · Views: 548
  • SAM_1431.jpg
    SAM_1431.jpg
    49.7 KB · Views: 519
Fascinating...

I'll be looking forward to your treatise on this topic, since I have an 86XX I use as a bench test amp. I've always been interested to see what performance could be squeezed out of a small chassis, and may have some slightly larger 6BQ5 console OPTs lying around as well. :thmbsp:
 
Hey, Dave,

Do you think some of your general findings will be applicable to the 196 / 197 series (similar amp, except for choke-loaded power supply and octal rectification, and for 196BA and 197, bigger OPTs)?

Thanks!

-D
 
Schematic please. There seems to be many different variants of this amp.

thanx
 
D -- very well could be. I'll try and rustle up a schematic and see.....

Dave

Here's a slightly marked up version:

18474819824_770f755440_o.jpg


-D
 
Impedance Mismatch: The Big Elephant In The Middle Of the Room

With flea power, every portion of a watt counts, so anything that works to waste power must be addressed. In the case of these amps (and many other Maggies I would assume), that starts with the impedance mis-match that exists between the 4 ohm output of the amplifier, and the (almost) universal 8 ohm speakers of today. It's a big hurt too, as whatever power the amplifier can make, is automatically cut in half right at the gate due to the mismatch created. There is simply no way around replacing the output transformers. However, the job is not hard, and can be done with little more than disconnecting the old ones, drilling out the four rivets mounting them in place, drilling three new holes, and then mounting the new transformers in place.

The biggest problem in replacing the existing transformers, was not so much in finding one that was electrically correct, but in one that was both electrically correct, and physically correct. When I modify any piece of equipment, I make every effort to maintain the original identity of what the unit was. When you see a modified unit of some type with an umbilical running to an outboard add on, or things strapped to the side that were never there originally, then the original identity of the unit gets lost quickly. Part of the attractiveness of these amplifiers is their small, self contained footprint, and I did not want to compromise that in my quest for improved performance.

Enter part# P-T31 at Antique Electronic Supply. These transformers fill the bill perfectly. They just fit, perform well, and are economical at (currently) just under $14 a piece. Electrically, they have a 5000 ohm primary winding, but have a single tap 8 ohm secondary winding -- exactly what is needed. Rated for 8 watts (a bit of a stretch, but still perfectly adequate for the needs at hand), the ones I received do not look like the one in the pic on their website. Rather, I have a feeling that the sales of the original (likely) surplus transformer (pictured) was so good that Antique had to have it cloned -- with that being what I received. While there are no markings on them to indicate who made them (just a stickered part number), they have all the trappings of a Hammond product. In any event, while they are hardly the sole reason for the success of the project, they are none the less a vital part, paving the way for other modifications that collectively along with the new transformers, produce the total improvement in performance achieved.

Once the old transformers are removed, the one next to the power supply can cap can be mounted in using the existing mount and wire holes. Unlike the old transformers, on the new ones, the leads come off both sides of the transformer -- primary (Blu and Rd) on one side, secondary (Grn and Blk) on the other. Therefore, unless you want to drill more holes, the leads for one winding need to wrap around the back side of the transformer, so that all four leads can go down through the original lead hole provided. However, while the new transformers have the same height and front to back dimensions, they have a thicker lam stack, so they must be installed as described, as the ultimate space between the power transformer and filter can cap is limited.

For the transformer next to the can cap, the transformer must be orientated so that the primary leads are closest to the can cap, with the transformer basically butted up next to the can. So situated, the original mount holes and lead hole can be used to mount this transformer, with the primary leads wrapping around the back side of the transformer as discussed so that all four leads go down through the original lead hole.

For the other transformer however, new mount and lead holes will need to be drilled. These will take on the same configuration as the original transformer holes, but will all be offset towards the power transformer. The new transformer should be orientated so that primary leads face the power transformer. Line this transformer up with the other new transformer already mounted, leaving the smallest of gaps between the two new transformers, and mark the holes for drilling. This produces max distance from the power transformer, and also allows the front mount tab for this transformer to just miss the tube socket for the 6BQ5 closest to the power transformer. As with the other transformer, the (secondary in this case) leads wrap around the rear of the transformer to ultimately enter the chassis through the new lead hole.

Once installed, the new transformers connect similarly as the old ones did:

1. Red to the B+ terminal on the T-Strip where the AC wiring is. The red lead from both transformers connect here.

2. Blue to pin #7 of the respective 6BQ5 power tube it serves.

3. Green to the center ground terminal of the T-strip containing the NFB voicing components. The Green lead from both transformers connects here.

4. Black to the respective output terminal for the amplifier it serves.

With the change in output transformers, the original 6.8K feedback resistor coming from the output terminals needs to be changed to a 10K resistor of the same wattage. This will then maintain the same feedback level as that of the original design.

At this point then, the amplifier is the same as it was, but with an 8 ohm output impedance. This change alone now allows the amplifier to develop and undistorted 2.3 watts per channel into an 8 ohm load -- which already means that an additional watt of power has been reclaimed -- nearly doubling what it used to develop into your 8 ohm speakers with the old transformers. But I encourage you to not stop here, as there is more power yet to be had, and much better quality as well.

Pics include:

1. The new transformers are beefier, using the same size lams, but more of them as this pic shows.

2. In an early stage of development, this pic shows how the new transformers nestle right in place, and do not draw attention to themselves as replacement units.

3. The small gap mentioned between the two transformers. It's not a precise measurement, just make sure they don't touch to allow for expansion with heat or general movement of the amplifier that might cause them to rub together. You can also see how the leads coming out of the right side of both transformers in this shot wrap around the rear of the transformer to enter the single lead hole provided for each transformer as discussed.

4. And the gap left between the new Channel 2 output transformer, and the power transformer. As I said, space is at a premium.

Since I use phone plugs for most of my speaker connections, you can also see that I took the opportunity to enlarge the right most rear chassis hole so that both holes on that end of the chassis could accommodate a 1/4 inch phone jack. Their center terminals connect to the leads exiting the rear of the chassis and connect to the original speaker connectors using appropriate slip-on terminals. The new jacks are color coded as: Red = Channel 2 or Right, Green = Channel 1 or Left.

Next time, we'll take on feedback and stability.

Dave
 

Attachments

  • SAM_1433.jpg
    SAM_1433.jpg
    83.2 KB · Views: 1,928
  • SAM_1438.jpg
    SAM_1438.jpg
    82 KB · Views: 2,061
  • SAM_1441.jpg
    SAM_1441.jpg
    64.8 KB · Views: 1,938
  • SAM_1442.jpg
    SAM_1442.jpg
    92.8 KB · Views: 2,050
Really looking forward to the rest of this, but I have to ask one REALLY stupid question - Why not just use 4 ohm speakers? Most of the really good speakers I have from 1985 onward are 4 ohms, and some older ones were as well.
 
D-- do you have any primary and secondary impedance information on OPTs used in this model?

Dave
 
I am impressed with the clarity of your writing and the thoroughness with which you have approached this project.

Well done!
 
Zira -- Many modern speakers are in fact 4 ohm units -- particularly in the auto industry -- but few of the modern speakers have the efficiency of the older Klipsch type speakers. I'm not saying they can't be found. But a lot of the folks using vintage tube amplifiers also use vintage speakers as well for a total package -- and there in lies the problem. The early (vintage) speakers that were 4 ohms (AR for example) were also highly inefficient. For those that have modern 4 ohm speakers of the required efficiency, then no change is needed. But most likely, most folks will welcome an 8 ohm output on this little guy.

Thanks for following along!

Dave
 
So just to be clear here, the change in transformer was purely to get an 8 ohm output, not to resolve any other issues?
 
That was surely a significant reason to change the transformer, but there were other reasons as well that will unfold along the way.

Dave
 
Back
Top Bottom