ferninando
Lunatic Member
who made the xfmrs?
Magnavox Flea Power: Getting More Out Of The 8600 Series - A Lot More!
- Thread starter dcgillespie
- Start date
No clue. Like I said, I think the ones I got are clones since they don't look like the one pictured. But they certainly measure as advertised. Besides the Antique part number on the label, the only other identifying information on the label is "MCD 15 - 02". The labels on both transformers are identical.
Dave
Dave
Klipsch KG1s would probably work well here.
Yeah, although I wouldn't want any less efficiency. They are only rated for 90 db/1 watt, and of course are also an 8 ohm speaker.
Dave
Dave
Whoops, I meant KG2. 90.5db efficiency, but 4 ohm. My Sansui 8080 is not particularly fond of them.
Hmmm. Maybe I leave my AMP-196BA alone and build a second SE amp using these transformers...
-D
Great thread Dave. Now I'm thinking of dragging out my little Mag. 8301 and following along. Thanks so much. Curt
Sadly I don't. Not sure how to measure, either. However, I can tell you that it was made by Midwest Coil and Transformer and the part number is C320083-1. There is a similar part number used on Motorola SE amps that is 5000 ohms / 3.2 ohms. Hopefully this helps.
-D
Last edited:
jazbo8
Well-Known Member
Some questions, why is the output low to begin with? Is the original OPT so lossy? And I wonder why the output would be cut in half when using 8 Ohm load? The datasheet suggests that higher RL increases the Po (well with somewhat different operating points)... Also since most "8 Ohm" speakers have a lower nonimal impedance around 6 Ohms, would that still too much a mis-match?
Last edited:
whoaru99
Epic Member
Yeah, from my personal messing around with some different OPTs to try to optimize the load I believe last part of the bolded bit is the key.
Jaz -- As a basic fact, when transferring power from a generator to a load, it can be shown that maximum power transfer occurs when the impedance of the generator matches the impedance of the load. As that relationship varies either way, energy is lost either in the generator (as heat) when the value of the load dips numerically below that of the generator, or is simply lost due to to lack of current flow when the value of the load rises numerically above that of the generator.
Applying this concept to the scenario at hand, the basic equation for determining power output developed in a load is P = E squared / R, where E is the voltage developed into the load in volts, and R is the value of the load in Ohms. Therefore, if an amplifier develops 20 volts across an 8 ohm load, it is developing 50 watts of power, as (20X20)/8=50.
Now an amplifier can only produce so much output voltage, and particularly so when only one output tap is available. Further, in a properly designed reproduction amplifier, the value of the output voltage in fact increases very little relatively, as the load attached to it increases numerically in value from optimum. But the value of the load in ohms (R) has a very direct impact on the amount of power the amplifier actually produces for the voltage it is capable of producing. As R numerically increases in the equation above (and above that of the generator itself), power being developed in the load is reduced: The same amplifier with a 16 ohm load in place of the 8 ohm load now produces just 25 watts, all else being equal. If the load were to become infinite, then no power would be produced at all. This is the basic concept at play here: The amplifier is only capable of so much output voltage from the single tap provided on the original OPT. But the load presented to the amplifier by vintage speakers (8 ohms) is basically double that which it represents as a generator (4 ohms). As a result, power produced is reduced -- and it is power (a product of voltage AND current) that moves the speaker cone. Solution? Install an output transformer with a lower turns ratio that supplies a greater output voltage. This in turn restores the current flow otherwise lost by using a numerically higher than optimum load for the original OPT -- so that's exactly what was done.
Now relating all of this to your questions:
1. No, the original OPT is not lossy for the load its rated for. But as I stated earlier, minimizing power loss is everything in flea amplifiers, so eliminating the impedance mismatch that often occurs with these particular amplifiers is a crucial first step towards that goal -- but it is not the only reason that power is reduced in this amplifier. The other reasons will be discussed as the thread unfolds.
2. The reduced power due to the mismatch as explained above.
3. I'm not sure what data sheets you are referring to or what you are looking at that is suggesting that Po increases with an increasing RL, but if it is for that of the tube itself, then operating points are of course everything. However, the often cited parameters for tubes of the 6BQ5 family show that with fixed bias, a plate and screen voltage of 250 volts, and a load of 4500 ohms, then the plate power developed will be 5.7 watts. As a practical matter, the revised output stage operates with cathode bias and a plate load of 5000 ohms and of course an output transformer, all of which will be discussed in more detail as the thread unfolds.
4. The impedance of any complete speaker system is always considered to be a "nominal" value, as in reality, the actual impedance of a speaker varies considerably with frequency. Therefore, its a misnomer to consider that most 8 ohm speakers have a nominal value of 6 ohms. The 8 ohm rating is already a nominal value. Individual 8 ohm bass drivers may display an open air value closer to 6 ohms, but once it is placed in a cabinet and electrical crossover networks are added, the 8 ohm rating becomes pretty accurate. No doubt however, there are some speakers that are designed from the get-go to have a 6 ohm nominal impedance value.
I hope this helps!
Dave
Applying this concept to the scenario at hand, the basic equation for determining power output developed in a load is P = E squared / R, where E is the voltage developed into the load in volts, and R is the value of the load in Ohms. Therefore, if an amplifier develops 20 volts across an 8 ohm load, it is developing 50 watts of power, as (20X20)/8=50.
Now an amplifier can only produce so much output voltage, and particularly so when only one output tap is available. Further, in a properly designed reproduction amplifier, the value of the output voltage in fact increases very little relatively, as the load attached to it increases numerically in value from optimum. But the value of the load in ohms (R) has a very direct impact on the amount of power the amplifier actually produces for the voltage it is capable of producing. As R numerically increases in the equation above (and above that of the generator itself), power being developed in the load is reduced: The same amplifier with a 16 ohm load in place of the 8 ohm load now produces just 25 watts, all else being equal. If the load were to become infinite, then no power would be produced at all. This is the basic concept at play here: The amplifier is only capable of so much output voltage from the single tap provided on the original OPT. But the load presented to the amplifier by vintage speakers (8 ohms) is basically double that which it represents as a generator (4 ohms). As a result, power produced is reduced -- and it is power (a product of voltage AND current) that moves the speaker cone. Solution? Install an output transformer with a lower turns ratio that supplies a greater output voltage. This in turn restores the current flow otherwise lost by using a numerically higher than optimum load for the original OPT -- so that's exactly what was done.
Now relating all of this to your questions:
1. No, the original OPT is not lossy for the load its rated for. But as I stated earlier, minimizing power loss is everything in flea amplifiers, so eliminating the impedance mismatch that often occurs with these particular amplifiers is a crucial first step towards that goal -- but it is not the only reason that power is reduced in this amplifier. The other reasons will be discussed as the thread unfolds.
2. The reduced power due to the mismatch as explained above.
3. I'm not sure what data sheets you are referring to or what you are looking at that is suggesting that Po increases with an increasing RL, but if it is for that of the tube itself, then operating points are of course everything. However, the often cited parameters for tubes of the 6BQ5 family show that with fixed bias, a plate and screen voltage of 250 volts, and a load of 4500 ohms, then the plate power developed will be 5.7 watts. As a practical matter, the revised output stage operates with cathode bias and a plate load of 5000 ohms and of course an output transformer, all of which will be discussed in more detail as the thread unfolds.
4. The impedance of any complete speaker system is always considered to be a "nominal" value, as in reality, the actual impedance of a speaker varies considerably with frequency. Therefore, its a misnomer to consider that most 8 ohm speakers have a nominal value of 6 ohms. The 8 ohm rating is already a nominal value. Individual 8 ohm bass drivers may display an open air value closer to 6 ohms, but once it is placed in a cabinet and electrical crossover networks are added, the 8 ohm rating becomes pretty accurate. No doubt however, there are some speakers that are designed from the get-go to have a 6 ohm nominal impedance value.
I hope this helps!
Dave
jazbo8
Well-Known Member
Thanks for the above replies, I really should not distract you from the task at hand with my follow-on questions, so I will leave them for later once you have gone through your presentation.
If not, you can search for Magnavox AMP196BA / AMP197 schematic or Magnavox 8601 schematic as the starting point and then follow along. The primary difference is the former used an octal rectifier and had a choke-loaded PS, but both used a 12AX7 driver (or 6EU7 - which is a 12AX7 with a different pinout) and two EL84 power tubes. Schematic for the former is on the first page of the thread. It's a very simple amp, thankfully.
-D
jazbo8
Well-Known Member
The 86-04's chassis does not seem to match that of the 196, are the schematics for them still the same? With the voltages shown in the 196, it seems capable of higher output than what Dave have measured in the 86-04.
analog addict
Glory or Death!
Here's the 8601-20 with eq section already deleted...If you wanted to get tricky, you could swap out the 6EU7 for the 12AX7 after changing around the pins....
Hey Dave,
I too will join the 'great thread!' commentary. I have a couple of these little amps, as well as a few 'clones' that I fabricated along the way.....and some time back I secured the parts to build Ned's (Redboy) clone of a SE amp that he had fabricated some time back. His amp used some OTs from Edcore....which are significantly larger than the Maggie OTs..... but alas, I've yet to complete that project in order to do any listening comparisons. But, I definitely fall into the 'enamored with flea amps' category, and having a set of Cornwalls myself, I too can relate to the idea of making a small amp sound great.
To that end, I will be watching and reading this thread with great interest, and I've already got AES's web site up and plan to buy a set of these OT's. I do have to ask: Did you by any chance 'verify' the PI on them? I'm just wondering if the stated 8 ohms and 5K worked out when and if you actually checked? Also, I had some advice from Paul (Kegger) about adding some additional capacity to the PS in these units......which I've done on a couple of them, and found that helped to some degree.
I'll end my post here by saying that I also did the swap of the 6EU7 wiring and reconfigured mine to 12AX wiring...which not only allowed the use of the 12AX7, but I've also tossed in other members of the family (ie the AT and AU versions), and found that the amp actually sounded a bit better with a 12AU in place vs the AX. I'll perhaps dig a couple of mine out and post a few shots!
Thanks Dave...... this stuff teaches us 'rookies' SOoooo much!!!
TSD
I too will join the 'great thread!' commentary. I have a couple of these little amps, as well as a few 'clones' that I fabricated along the way.....and some time back I secured the parts to build Ned's (Redboy) clone of a SE amp that he had fabricated some time back. His amp used some OTs from Edcore....which are significantly larger than the Maggie OTs..... but alas, I've yet to complete that project in order to do any listening comparisons. But, I definitely fall into the 'enamored with flea amps' category, and having a set of Cornwalls myself, I too can relate to the idea of making a small amp sound great.
To that end, I will be watching and reading this thread with great interest, and I've already got AES's web site up and plan to buy a set of these OT's. I do have to ask: Did you by any chance 'verify' the PI on them? I'm just wondering if the stated 8 ohms and 5K worked out when and if you actually checked? Also, I had some advice from Paul (Kegger) about adding some additional capacity to the PS in these units......which I've done on a couple of them, and found that helped to some degree.
I'll end my post here by saying that I also did the swap of the 6EU7 wiring and reconfigured mine to 12AX wiring...which not only allowed the use of the 12AX7, but I've also tossed in other members of the family (ie the AT and AU versions), and found that the amp actually sounded a bit better with a 12AU in place vs the AX. I'll perhaps dig a couple of mine out and post a few shots!
Thanks Dave...... this stuff teaches us 'rookies' SOoooo much!!!
TSD
Analog -- Thank-you so much for posting the schematic!
6D -- Yeah, it's weird. But Magnavox did some weird things every now and then. With the exception of the EQ section in the NFB loop which has already been deleted in that schematic, it exactly represents the original build of my amplifier. The only difference in that schematic and my unit (besides the removed EQ components) is that the tap into the NFB system for the external balance control was accomplished on my unit with flying leads ending in a Molex plug. On the schematic provided, those connection points appear at pins 6&9 of the tuner power plug.
Dave
6D -- Yeah, it's weird. But Magnavox did some weird things every now and then. With the exception of the EQ section in the NFB loop which has already been deleted in that schematic, it exactly represents the original build of my amplifier. The only difference in that schematic and my unit (besides the removed EQ components) is that the tap into the NFB system for the external balance control was accomplished on my unit with flying leads ending in a Molex plug. On the schematic provided, those connection points appear at pins 6&9 of the tuner power plug.
Dave
